Sign in with
Sign up | Sign in
Your question

BBC considering HDTV rollout

Last response: in Home Theatre
Share
Anonymous
April 27, 2004 9:07:59 AM

Archived from groups: alt.video.digital-tv (More info?)

<http://www.digitalspy.co.uk/article/ds14311.html&gt;

IIRC, 576p is defined as HDTV in Australia.

Kirk Bayne
alt.video.digital-tv Home Page
<http://www.geocities.com/lislislislis/avdtv.htm&gt;

More about : bbc hdtv rollout

Anonymous
April 27, 2004 6:18:24 PM

Archived from groups: alt.video.digital-tv (More info?)

K. B. wrote:

> <http://www.digitalspy.co.uk/article/ds14311.html&gt;
>
> IIRC, 576p is defined as HDTV in Australia.

Since the article states lack of available content is one of the
problematic issues, it's a no-brainer that switching to 60hz (1080i,
720p) would facilitate the change, enabling immediate HDTV telecasts of
increasing portions of current and ongoing American programing, which
I'm aware is very popular in the UK.

576p requires nothing more than display devices with line-doubling and
perhaps frame doubling (to 100hz) capacity. This has been available
from Faroudja for several years. I've heard mixed comments re the 100hz
TVs sold in Europe. Obviously they would eliminate the flicker
associated with 50hz; however, if the rescaling (especially for
interlace-video sourced material) is not done well the results might be
less than stellar.

And it's well known that modern TVs sold in Britain usually have
multi-scan capability (60hz for NTSC [or so-called PAL-60, aka
Pseudo-PAL] sources).

Even in the US and Canada, multi-scan capability is effectively standard
for HDTV receivers, accommodating 480i, 480p and 1080i sources. Some
HDTVs here now also accept 720p, though most (such as the current Sony
XBRs) convert it to 1080i. [Last I heard, Princeton Graphics was the
only US manufacturer to produce TVs which displayed native 720p.]

Obviously, Europe should have a higher refresh rate than 50hz for HDTV,
though future sets obviously will need to accommodate legacy 50hz sources.






C.
April 27, 2004 9:03:51 PM

Archived from groups: alt.video.digital-tv (More info?)

"manitou910" <manitou910@rogers.com> wrote in message
news:QYtjc.3220$EPt1.3179@news01.bloor.is.net.cable.rogers.com...
> K. B. wrote:
>
> > <http://www.digitalspy.co.uk/article/ds14311.html&gt;
> >
> > IIRC, 576p is defined as HDTV in Australia.
>
> Since the article states lack of available content is one of the
> problematic issues, it's a no-brainer that switching to 60hz (1080i,
> 720p) would facilitate the change, enabling immediate HDTV telecasts of
> increasing portions of current and ongoing American programing, which
> I'm aware is very popular in the UK.

That's a question of physically producing enough material that people want
to watch rather than technical problems in converting existing programming.

You'll find that 50Hz broadcasts out-populate 60Hz regions by quite degree,
that's quite a back library of programming to obsolete or standards convert
for the sake of imports from other regions.

Most decently produced US shows have replaced film with 24p HiDef so the
frame-rate issue isn't a problem, if you're watching CSI, '24' etc in HD in
the US then you're already watching converted material (your set will
hopefully pull-down). We already receive very nice down-conversions of HD
material for 25fps, it doesn't matter if this is 576i, 720p, 1080i.

I've watched Euro1080 (50Hz) on a plasma and it's not an issue, the set
wasn't actually scanning at 50Hz.

Az.
Related resources
Anonymous
April 29, 2004 6:24:23 AM

Archived from groups: alt.video.digital-tv (More info?)

manitou910 wrote:
> K. B. wrote:
>
>> <http://www.digitalspy.co.uk/article/ds14311.html&gt;
>>
>> IIRC, 576p is defined as HDTV in Australia.
>
> Since the article states lack of available content is one of the
> problematic issues, it's a no-brainer that switching to 60hz (1080i,
> 720p) would facilitate the change, enabling immediate HDTV telecasts
> of increasing portions of current and ongoing American programing,
> which
> I'm aware is very popular in the UK.

Err - the 50/60 issue is a non-issue. A lot of US drama is shot 1080/24p
isn't it - which we have no problems watching in 576/50i, 576/25p, 1080/50i
etc. with the 4% speed-up we are used to when watching 24fps film material.
We don't have the 24p/60i 3:2 pulldown problem either, as 24p/50i can be
done 2:2 with the speed-up.

Similarly the UK HDTV material shot 1080/25p is slowed down for the US
market and run as 1080/24p (or then 3:2ed to 1080/60i)

Where frame rate conversion is required (say from 60i/p to 50i/p) standards
convesion is improving now to the point where 50->60 and 60->50 conversion
is more transparent than MPEG2 encoding... (i.e. more motion artefacts are
introduced by the MPEG process than the conversion) Using phase correlation
techniques especially means it is not a major issue.

>
> 576p requires nothing more than display devices with line-doubling and
> perhaps frame doubling (to 100hz) capacity.

576/50p requires at worst 576/50p acquisition gear - which is rare. I think
it is more common to shoot 1080/50i and then downconvert to 576/50p, or
1080/25p frame doubled? Of course you can also TK 25fps film (sped up 24fps
film as well) to 50p by frame doubling. There is no advantage over 576/25p,
but an improvement over 576/50i if you don't vertically filter to reduce
interlace twitter.

> This has been available
> from Faroudja for several years.

Faroudja are more in the display conversion rather than standards conversion
business aren't they? Snell and Wilcox have made HD standards converters
(and some of THE best SD converters) for a while.

>I've heard mixed comments re the
> 100hz TVs sold in Europe.

This is because they are mainly used to convert 576/50i to 576/100i - with
no real progressive conversion. Some (Philips Natural Motion) even convert
material (like film) sourced 576/25p, but broadcast 576/50i, to 576/100i by
interpolating in-between fields, giving the whole film sequence a really
fluid "video" look. Very disconcerting - and the total opposite of what
many producers are now doing (shooting video at 576/50i but reducing the
temporal resolution in post-production to 576/25 to give a "film" look)...

One problem is that it is difficult to buy a direct view CRT set with
external 576/100i or 1080/50i inputs in the UK - so you are dependent on the
internal processing architecture - and can't chose your own.


> Obviously they would eliminate the flicker
> associated with 50hz; however, if the rescaling (especially for
> interlace-video sourced material) is not done well the results might
> be less than stellar.

The interpolation, motion compensation/tracking etc. used to generate twice
the number of fields is the real problem with European 100Hz sets. You get
smearing on fast motion, nasty overly vicious noise reduction, enhancement
of MPEG2 blocking and HF artefacts etc. introduced on DVD, DSat and DTT
transmissions. The RGB interconnects commonly used between DVD / Digital TV
set top boxes and European TVs means that some of the MPEG2 artefacts that
would be hidden by PAL (or NTSC) composite or S-video encoding aren't - so
you see more of the coding errors.

There have been some 576/50i to 576/50p sets sold but not many - the
benefits of progressive at 50Hz are less marked (and we don't have 3:2
issues) Some Sony DRC sets now offer 625/50i to 1250/50i or 625/100i (I've
moved to total lines from active here as this is what they are described as)
and the 1250 stuff looks better to me than the 100Hz...

>
> And it's well known that modern TVs sold in Britain usually have
> multi-scan capability (60hz for NTSC [or so-called PAL-60, aka
> Pseudo-PAL] sources).

Yep - though in many cases this is used for RGB-60 display from DVD players
these days - with no need for any PAL or NTSC composite or S-video
encoding/decoding to take place. Most TVs sold in Europe for years have
been able to lock to both 576/50i and 480/60i signals - though not all (but
most) have NTSC 3.58/4.43 or PAL-60 chroma decoding. (Meaning you get a
colour 480/60i picture in RGB but not in composite/s-video)

VHS machines do not record PAL-60 in the UK in the main, though some DVD
recorders will record NTSC 3.58 or RGB 480/60i.

>
> Even in the US and Canada, multi-scan capability is effectively
> standard for HDTV receivers, accommodating 480i, 480p and 1080i
> sources.

Isn't 480i not accommodated by multi-scan - instead the signal is
upconverted either to 480p or 1080i? It is quite a tall order to build a
tubed TV set capable of scanning both SDTV and HDTV rates. (525x60 and
1125ishx30 aren't that different, but 525x30 is quite a bit slower...)

> Some HDTVs here now also accept 720p, though most (such as
> the current Sony XBRs) convert it to 1080i. [Last I heard, Princeton
> Graphics was the only US manufacturer to produce TVs which displayed
> native 720p.]
>
> Obviously, Europe should have a higher refresh rate than 50hz for
> HDTV, though future sets obviously will need to accommodate legacy
> 50hz sources.

Hmm - 50 Hz is fine for temporal purposes (especially given many people
consider 24 Hz fine - as that is what film, and much US HD drama, runs at
for acquisition)

Just because you shoot at this rate there is no need to refresh your display
at this rate - and it is only a real issue with CRT based devices that scan
a flying spot. Film projectors double or triple expose each film frame on
projection (giving 48 or 72 Hz refresh from a 24Hz source)

Plasmas displays run at much higher sub-field rates than their source
material (they often run at 300Hz ish don't they to generate the grey
scale - along with dithering) - and LCDs and DLPs have a different refresh
dynamic, so don't appear to flicker in the same way CRTs do?

The flicker introduced by 50Hz scanning (rather than 50Hz acquisition) is a
function of CRTs - which are presumably less an issue for future HD
standards?

I think that the benefits of sticking with the same production format
(timecode compatibility, simultaneous downconversion to SDTV easily,
lighting flicker etc.) probably outweigh the benefits of moving to a
different system for the sake of it.
Anonymous
April 29, 2004 7:06:51 AM

Archived from groups: alt.video.digital-tv (More info?)

Stephen Neal wrote:
>>
>>>IIRC, 576p is defined as HDTV in Australia.
>>
>>Since the article states lack of available content is one of the
>>problematic issues, it's a no-brainer that switching to 60hz (1080i,
>>720p) would facilitate the change, enabling immediate HDTV telecasts
>>of increasing portions of current and ongoing American programing,
>>which I'm aware is very popular in the UK.
>
> Err - the 50/60 issue is a non-issue. A lot of US drama is shot 1080/24p
> isn't it - which we have no problems watching in 576/50i, 576/25p, 1080/50i
> etc. with the 4% speed-up we are used to when watching 24fps film material.
> We don't have the 24p/60i 3:2 pulldown problem either, as 24p/50i can be
> done 2:2 with the speed-up.

The speedup can work havoc with an elaborate soundtrack. While pitch is
no longer an issue for digital audio, a 4% speedup can drastically alter
the character of any music. It is correct, however, that most US dramas
(and comedies) are shot at 24fps and still more on film than video.

> Similarly the UK HDTV material shot 1080/25p is slowed down for the US
> market and run as 1080/24p (or then 3:2ed to 1080/60i)

This has been done for a few movies shot on digital PAL (eg, "The
Anniversary Party"), but I've never heard of UK TV shows being slowed
down for US telecasts.

> Where frame rate conversion is required (say from 60i/p to 50i/p) standards
> convesion is improving now to the point where 50->60 and 60->50 conversion
> is more transparent than MPEG2 encoding... (i.e. more motion artefacts are
> introduced by the MPEG process than the conversion) Using phase correlation
> techniques especially means it is not a major issue.

I'm very skeptical re standards conversions because I've seen so many
that looked atrocious.

After all the effort to restore "Brideshead Revisited" the NTSC
broadcasts (and DVDs) were literally unwatchable.

Ditto for the new "Forsyte Saga" (though the PAL DVDs for this were
pretty bad too).

>>576p requires nothing more than display devices with line-doubling and
>>perhaps frame doubling (to 100hz) capacity.
>
> 576/50p requires at worst 576/50p acquisition gear - which is rare. I think
> it is more common to shoot 1080/50i and then downconvert to 576/50p, or
> 1080/25p frame doubled? Of course you can also TK 25fps film (sped up 24fps
> film as well) to 50p by frame doubling. There is no advantage over 576/25p,
> but an improvement over 576/50i if you don't vertically filter to reduce
> interlace twitter.

I'd assume that with proper conversion to progressive scan, this should
work.

>>This has been available from Faroudja for several years.
>
> Faroudja are more in the display conversion rather than standards conversion
> business aren't they? Snell and Wilcox have made HD standards converters
> (and some of THE best SD converters) for a while.

Faroudja is best known for its high-end home scalers (I have one of
their NRS series feeding 720p/60 to an Ampro HD3600 CRT projector) and,
more recently, the FLI2310 chip used in certain progressive-scan DVD
players (with scaling to 720p and 1080i as well as 480p).

I'm aware that some US stations are using industrial Faroudja gear to
up-rez NTSC programs for HDTV. However I don't know exactly which
stations and/or networks are doing this, so I can't really comment.

The Faroudja NRS scaler can be set to convert PAL sources to 100hz
progressive scan. I haven't seen this, but I'd expect it would be very
good.

I'm well aware of Snell & Wilcox' reputation, but can mention that
PAL-to-NTSC conversions always look a lot better with Faroudja scaling
at the display end, than without.

>>I've heard mixed comments re the 100hz TVs sold in Europe.
>
> This is because they are mainly used to convert 576/50i to 576/100i - with
> no real progressive conversion. Some (Philips Natural Motion) even convert
> material (like film) sourced 576/25p, but broadcast 576/50i, to 576/100i by
> interpolating in-between fields, giving the whole film sequence a really
> fluid "video" look. Very disconcerting - and the total opposite of what
> many producers are now doing (shooting video at 576/50i but reducing the
> temporal resolution in post-production to 576/25 to give a "film" look)...
>
> One problem is that it is difficult to buy a direct view CRT set with
> external 576/100i or 1080/50i inputs in the UK - so you are dependent on the
> internal processing architecture - and can't chose your own.

Has their been much interest in progressive-scan DVD players in the UK?

I'm aware of them being marketed but, as you say, without compatible
displays their use would be limited.

>>Obviously they would eliminate the flicker
>>associated with 50hz; however, if the rescaling (especially for
>>interlace-video sourced material) is not done well the results might
>>be less than stellar.
>
> The interpolation, motion compensation/tracking etc. used to generate twice
> the number of fields is the real problem with European 100Hz sets. You get
> smearing on fast motion, nasty overly vicious noise reduction, enhancement
> of MPEG2 blocking and HF artefacts etc. introduced on DVD, DSat and DTT
> transmissions. The RGB interconnects commonly used between DVD / Digital TV
> set top boxes and European TVs means that some of the MPEG2 artefacts that
> would be hidden by PAL (or NTSC) composite or S-video encoding aren't - so
> you see more of the coding errors.

Rescaling is always a balancing act. If the source is technically poor
it will likely look worse on a high-resolution display.

> There have been some 576/50i to 576/50p sets sold but not many - the
> benefits of progressive at 50Hz are less marked (and we don't have 3:2
> issues) Some Sony DRC sets now offer 625/50i to 1250/50i or 625/100i (I've
> moved to total lines from active here as this is what they are described as)
> and the 1250 stuff looks better to me than the 100Hz...
>
>>And it's well known that modern TVs sold in Britain usually have
>>multi-scan capability (60hz for NTSC [or so-called PAL-60, aka
>>Pseudo-PAL] sources).
>
> Yep - though in many cases this is used for RGB-60 display from DVD players
> these days - with no need for any PAL or NTSC composite or S-video
> encoding/decoding to take place. Most TVs sold in Europe for years have
> been able to lock to both 576/50i and 480/60i signals - though not all (but
> most) have NTSC 3.58/4.43 or PAL-60 chroma decoding. (Meaning you get a
> colour 480/60i picture in RGB but not in composite/s-video)
>
> VHS machines do not record PAL-60 in the UK in the main, though some DVD
> recorders will record NTSC 3.58 or RGB 480/60i.
>
>>Even in the US and Canada, multi-scan capability is effectively
>>standard for HDTV receivers, accommodating 480i, 480p and 1080i
>>sources.
>
> Isn't 480i not accommodated by multi-scan - instead the signal is
> upconverted either to 480p or 1080i? It is quite a tall order to build a
> tubed TV set capable of scanning both SDTV and HDTV rates. (525x60 and
> 1125ishx30 aren't that different, but 525x30 is quite a bit slower...)

Yes. The HDTVs don't display native NTSC; they rescale to 480p or 960i
(a few may now do 1080i, or 720p). My point was that they could handle
a range of different standards.

>>Some HDTVs here now also accept 720p, though most (such as
>>the current Sony XBRs) convert it to 1080i. [Last I heard, Princeton
>>Graphics was the only US manufacturer to produce TVs which displayed
>>native 720p.]
>>
>>Obviously, Europe should have a higher refresh rate than 50hz for
>>HDTV, though future sets obviously will need to accommodate legacy
>>50hz sources.
>
> Hmm - 50 Hz is fine for temporal purposes (especially given many people
> consider 24 Hz fine - as that is what film, and much US HD drama, runs at
> for acquisition)
>
> Just because you shoot at this rate there is no need to refresh your display
> at this rate - and it is only a real issue with CRT based devices that scan
> a flying spot. Film projectors double or triple expose each film frame on
> projection (giving 48 or 72 Hz refresh from a 24Hz source)
>
> Plasmas displays run at much higher sub-field rates than their source
> material (they often run at 300Hz ish don't they to generate the grey
> scale - along with dithering) - and LCDs and DLPs have a different refresh
> dynamic, so don't appear to flicker in the same way CRTs do?
>
> The flicker introduced by 50Hz scanning (rather than 50Hz acquisition) is a
> function of CRTs - which are presumably less an issue for future HD
> standards?
>
> I think that the benefits of sticking with the same production format
> (timecode compatibility, simultaneous downconversion to SDTV easily,
> lighting flicker etc.) probably outweigh the benefits of moving to a
> different system for the sake of it.

If new display formats eliminate flicker associated with PAL (even NTSC)
and do so without compromising other technical aspects this is the way
to go IMO.

Someone has even suggested a 120hz standard for American HDTV which
would be fully compatible with both legacy 60hz NTSC video and all
movies and 24fps film-sourced TV series (for which each frame would be
shown five times, eliminating 3:2 pulldown artifacts).

My major issue with 50hz standards is the excess flicker for standard
displays, and the fact that all movies and film-sourced American TV
series must be speeded up which may cause serious audio compromise.






C.
Anonymous
April 29, 2004 3:29:08 PM

Archived from groups: alt.video.digital-tv (More info?)

manitou910 wrote:
> Stephen Neal wrote:
>>>
>>>> IIRC, 576p is defined as HDTV in Australia.
>>>
>>> Since the article states lack of available content is one of the
>>> problematic issues, it's a no-brainer that switching to 60hz (1080i,
>>> 720p) would facilitate the change, enabling immediate HDTV telecasts
>>> of increasing portions of current and ongoing American programing,
>>> which I'm aware is very popular in the UK.
>>
>> Err - the 50/60 issue is a non-issue. A lot of US drama is shot
>> 1080/24p isn't it - which we have no problems watching in 576/50i,
>> 576/25p, 1080/50i etc. with the 4% speed-up we are used to when
>> watching 24fps film material. We don't have the 24p/60i 3:2 pulldown
>> problem either, as 24p/50i can be done 2:2 with the speed-up.
>
> The speedup can work havoc with an elaborate soundtrack. While pitch
> is no longer an issue for digital audio, a 4% speedup can drastically
> alter the character of any music. It is correct, however, that most
> US dramas (and comedies) are shot at 24fps and still more on film
> than video.

Given that we've put up with the speed-up on 24p material since about 1936,
when we launched our 50i TV system (405/50i), I think we're used to it in
the UK. With digital pitch correction the 4% speed-up is normally not an
issue. I agree it changes the nature of music programmes - but these are
often shot 60i or 50i rather than 24p so there is no speed change (as a
standards conversion rather than speed change is used)

>
>> Similarly the UK HDTV material shot 1080/25p is slowed down for the
>> US market and run as 1080/24p (or then 3:2ed to 1080/60i)
>
> This has been done for a few movies shot on digital PAL (eg, "The
> Anniversary Party"), but I've never heard of UK TV shows being slowed
> down for US telecasts.

The BBC series Rockface was shot 1080/25p on HDCam and was co-produced with
a US broadcaster/studio. The 25p material was slowed to 24p for the US
transmissions. (Presumably it was then either broadcast 24p or 60i with 3:2
pull-down)

I imagine that UK drama shot 25p will follow this model of slow-down to 24p.
Of course 50i drama will still be converted to 60i with no speed change.

>
>> Where frame rate conversion is required (say from 60i/p to 50i/p)
>> standards convesion is improving now to the point where 50->60 and
>> 60->50 conversion is more transparent than MPEG2 encoding... (i.e.
>> more motion artefacts are introduced by the MPEG process than the
>> conversion) Using phase correlation techniques especially means it
>> is not a major issue.
>
> I'm very skeptical re standards conversions because I've seen so many
> that looked atrocious.

They have got massively better in recent years. You get what you pay for.
In the last 3 or 4 years the Snell and Wilcox Alchemist PhC and Platinum
conversions approach transparency - even on fast moving sports.

We also have the problem in the UK of having to un-pick 3:2 pull-down of US
film transfers, as the non-continuous movement confuses motion tracking,
instead the 3rd redundant field has to be detected and removed. So we go
from 60i 3:2 to 48i 2:2 using DEFT (Digital Electronic Film Transfer -
another S&W technique) and then speed up to 50i. It works really well on US
film drama edited on 480/60i tape. You only notice the slight softness
difference between 480 and 576 - which is far less obvious than the 3.58NTSC
vs 4.43PAL softness difference.

>
> After all the effort to restore "Brideshead Revisited" the NTSC
> broadcasts (and DVDs) were literally unwatchable.

Well the broadcasts will have been made on 1980s converters. I'm
disappointed that the DVDs were standards converted at all - AIUI Brideshead
was shot and edited on film so could have been telecined from the film
masters to 480/60i or 480/24p for DVD mastering. You'd have probably had a
slow-down if this were the case as I suspect the UK TV production will have
been shot 25fps film rather than 24fps (as in Europe film-for-TV is normally
shot at 25 rather than 24)

The chances are that if the DVD was mastered from the UK VT transfer it may
well have been the 1980s 1" C format PAL, and the telecine-ing would have
happened in the 80s as well, if it were mastered from the US broadcast
masters made in the 80s it will have had a pretty bad conversion as well I
imagine. Telecine and conversion technology has moved on a-pace since the
80s and even the mid 90s...

There are of course all the usual DVD mastering annoyances as well - too
much noise reduction, excessive compression and addition aperture correction
etc.

>
> Ditto for the new "Forsyte Saga" (though the PAL DVDs for this were
> pretty bad too).

Again - a commercial DVD release - presumably not optimised for quality.
There seem to be a massive range in quality of UK DVD releases - the good
ones are very good, the bad ones are terrible. The best feature film
transfers are sourced from 1080/24p HD masters - which are converted to
576/50i (576/25p) via scaling and speed-up.

For high quality R1 PAL video and film DVD remastering the Dr Who
Restoration Team have worked marvels. They use a BBC R&D PAL decoder to
transfer the D3 copies of the 2" or 1" video masters (which leaves almost no
composite artefacts), and where possible re-transfer and re-grade any film
inserts that remain as film in the BBC archive. The quality is often
stunningly good.

>
>>> 576p requires nothing more than display devices with line-doubling
>>> and perhaps frame doubling (to 100hz) capacity.
>>
>> 576/50p requires at worst 576/50p acquisition gear - which is rare.
>> I think it is more common to shoot 1080/50i and then downconvert to
>> 576/50p, or 1080/25p frame doubled? Of course you can also TK 25fps
>> film (sped up 24fps film as well) to 50p by frame doubling. There is
>> no advantage over 576/25p, but an improvement over 576/50i if you
>> don't vertically filter to reduce interlace twitter.
>
> I'd assume that with proper conversion to progressive scan, this
> should work.
>

576/50i to 576/50p should be OK if the de-interlacing processing is high
enough quality. 576/50i to 576/25p will look horrid though, as you are
chucking away half of the motion information. To reduce display flicker
then you would either have to move to 576/100p (quite a high scan rate) or
576/100i (which will have some interlace artefacts - but these will be at
50Hz not 25Hz as is the case with 50i)

>>> This has been available from Faroudja for several years.
>>
>> Faroudja are more in the display conversion rather than standards
>> conversion business aren't they? Snell and Wilcox have made HD
>> standards converters (and some of THE best SD converters) for a
>> while.
>
> Faroudja is best known for its high-end home scalers (I have one of
> their NRS series feeding 720p/60 to an Ampro HD3600 CRT projector)
> and, more recently, the FLI2310 chip used in certain progressive-scan
> DVD players (with scaling to 720p and 1080i as well as 480p).

Yep -scaling rather than conversion - i.e. the field/frame rates are not
changing? (Apart from 24 to 60 which is a 3:2 conversion rather than
anything more complex?)

In other words scalers used to get between 480i to 480p to 720p to 1080i at
either 30 or 60 (or 24 with 3:2)

[snip]

If we were to move to 60i, 60p or 30p production in Europe we'd need to use
field/frame rate tstandards conversion, rather than scaling, permanently to
convert to or from 576/50i for upconversion of SDTV material, and
downconversion of HD. Whilst this is now a high quality process it is also
a much more expensive process than scaling - especially if you have to do it
for all of your outlets, especially regional ones. Far better to only
field/frame rate convert the small amount of imported material rather than
your entire archive and continued SDTV production base?

Given that much of what the UK would import from the US would be 24p, rather
than 30p or 60i, that would be easily converted to 48i then sped up to 50i
for playout without conversion. We'd then only need to do a standards
conversion for the 30p, 60i or 60p material such as sport, gameshows etc.
which are far less common and popular imports than the 24p stuff.


>>> I've heard mixed comments re the 100hz TVs sold in Europe.
>>
>> This is because they are mainly used to convert 576/50i to 576/100i
>> - with no real progressive conversion. Some (Philips Natural Motion)
>> even convert material (like film) sourced 576/25p, but broadcast
>> 576/50i, to 576/100i by interpolating in-between fields, giving the
>> whole film sequence a really fluid "video" look. Very disconcerting
>> - and the total opposite of what many producers are now doing
>> (shooting video at 576/50i but reducing the temporal resolution in
>> post-production to 576/25 to give a "film" look)...
>>
>> One problem is that it is difficult to buy a direct view CRT set with
>> external 576/100i or 1080/50i inputs in the UK - so you are
>> dependent on the internal processing architecture - and can't chose
>> your own.
>
> Has their been much interest in progressive-scan DVD players in the
> UK?

Not huge amounts - though quite a few players are available. However there
are very few 50p CRT sets on the market. Prog scan DVDs are mainly used
with inherently progressive display devices that can't display interlace
native - plasmas, DLPs, LCDs etc - which have 576/50p input compatibility.

However the vast majority of TV sales in the UK are still direct-view CRTs -
predominantly in the sub 28" diagonal. However increasingly >21" sets sold
are 16:9 CRT shape - it is very difficult to buy a 4:3 set bigger than 21"
in most UK stores these days. 32" is the largest mass market 16:9 CRT -
though 36"ers are now on offer.

Plasmas are growing in popularity, though like projectors they are still
quite niche.

>
> I'm aware of them being marketed but, as you say, without compatible
> displays their use would be limited.

100i rather than 50p is the marketing thing for direct-view CRTs in the UK.
Annoying as I would rather watch an unprocessed 50i picture - but it is
difficult to buy a high-end TV without processing to 100i now... (And only
the higher end sets have multiple RGB inputs etc.)

>
>>> Obviously they would eliminate the flicker
>>> associated with 50hz; however, if the rescaling (especially for
>>> interlace-video sourced material) is not done well the results might
>>> be less than stellar.
>>
>> The interpolation, motion compensation/tracking etc. used to
>> generate twice the number of fields is the real problem with
>> European 100Hz sets. You get smearing on fast motion, nasty overly
>> vicious noise reduction, enhancement of MPEG2 blocking and HF
>> artefacts etc. introduced on DVD, DSat and DTT transmissions. The
>> RGB interconnects commonly used between DVD / Digital TV set top
>> boxes and European TVs means that some of the MPEG2 artefacts that
>> would be hidden by PAL (or NTSC) composite or S-video encoding
>> aren't - so you see more of the coding errors.
>
> Rescaling is always a balancing act. If the source is technically
> poor it will likely look worse on a high-resolution display.

Yep - though a lot of 100Hz introduces as many artefacts as it also exposes!

[snip]

> If new display formats eliminate flicker associated with PAL (even
> NTSC) and do so without compromising other technical aspects this is
> the way to go IMO.

Technically the flicker is nothing to do with PAL - that is only the chroma
standard!

I think you have to divorce your display rate and your broadcast rate don't
you?

I think that we'd be best suited eventually moving to 1080/25p and 1080/50p
broadcast standards - display on direct CRTs would be at 100i (with
interlace artefacts at 50Hz) - but display on plasma, DLP, LCD etc would be
at 50p without major flicker (as the flicker is more a CRT scanning raster
thing which isn't inherent in other display technologies?)

>
> Someone has even suggested a 120hz standard for American HDTV which
> would be fully compatible with both legacy 60hz NTSC video and all
> movies and 24fps film-sourced TV series (for which each frame would be
> shown five times, eliminating 3:2 pulldown artifacts).

Yep - presumably retaining 24p and 60i/p transmission systems and only
up-converting to 120p or i in the display device by frame repetition?

>
> My major issue with 50hz standards is the excess flicker for standard
> displays, and the fact that all movies and film-sourced American TV
> series must be speeded up which may cause serious audio compromise.
>

Have you watched 50i or 25p material on a non-CRT device? The audio
compromise is only an issue for the small amount of imported stuff we have -
our native stuff has none. The major UK networks show quite small amounts
of imported material - though feature films I agree are compromised
slightly, but I personally find 3:2 more annoying than the audio change!

Steve
Anonymous
April 29, 2004 4:15:15 PM

Archived from groups: alt.video.digital-tv (More info?)

Stephen Neal wrote:

>
> Where frame rate conversion is required (say from 60i/p to 50i/p) standards
> convesion is improving now to the point where 50->60 and 60->50 conversion
> is more transparent than MPEG2 encoding... (i.e. more motion artefacts are
> introduced by the MPEG process than the conversion) Using phase correlation
> techniques especially means it is not a major issue.

I don;t understand how this conversion can be done well. Consider
the typical scenario when I watch a 50->60 Hz conversion: A F1 car
is travelling down the track to 100 mph, with buildings and fences
with narrow highly visible poles moving rapidly, as the camera
pans to keep teh car centered. How do you fix up the 5:6 motion of
those highly visible discrete poles? Aboyt 90% of 50Hz originaled TV
is of this sort. It looks bad.



Doug McDonald
Anonymous
April 29, 2004 4:21:40 PM

Archived from groups: alt.video.digital-tv (More info?)

manitou910 wrote:

>>> Some HDTVs here now also accept 720p, though most (such as
>>> the current Sony XBRs) convert it to 1080i. [Last I heard, Princeton
>>> Graphics was the only US manufacturer to produce TVs which displayed
>>> native 720p.]

You are talking here only legacy CRT diplays. These will dominate
the HD and all but niche markest only a couple of more years. Then
the non-scanned displays such as plasma, LCD, and DLP will be the
"norm" for all but the most budget concious. These are currently
all either 720p or, disastrously (Sony, of course), 768p. Eventually
they will be 1080p.

Doug McDonald
Anonymous
April 29, 2004 4:24:52 PM

Archived from groups: alt.video.digital-tv (More info?)

Could somebody explain how you can change the pitch of music
without changing the speed, without totally screwing up the phase
relationships that control soundstage imaging?

Doug McDonald
Anonymous
April 29, 2004 6:31:57 PM

Archived from groups: alt.video.digital-tv (More info?)

Stephen Neal wrote:

>>>>>IIRC, 576p is defined as HDTV in Australia.
>>>>
>>>>Since the article states lack of available content is one of the
>>>>problematic issues, it's a no-brainer that switching to 60hz (1080i,
>>>>720p) would facilitate the change, enabling immediate HDTV telecasts
>>>>of increasing portions of current and ongoing American programing,
>>>>which I'm aware is very popular in the UK.
>>>
>>>Err - the 50/60 issue is a non-issue. A lot of US drama is shot
>>>1080/24p isn't it - which we have no problems watching in 576/50i,
>>>576/25p, 1080/50i etc. with the 4% speed-up we are used to when
>>>watching 24fps film material. We don't have the 24p/60i 3:2 pulldown
>>>problem either, as 24p/50i can be done 2:2 with the speed-up.
>>
>>The speedup can work havoc with an elaborate soundtrack. While pitch
>>is no longer an issue for digital audio, a 4% speedup can drastically
>>alter the character of any music. It is correct, however, that most
>>US dramas (and comedies) are shot at 24fps and still more on film
>>than video.
>
> Given that we've put up with the speed-up on 24p material since about 1936,
> when we launched our 50i TV system (405/50i), I think we're used to it in
> the UK. With digital pitch correction the 4% speed-up is normally not an
> issue. I agree it changes the nature of music programmes - but these are
> often shot 60i or 50i rather than 24p so there is no speed change (as a
> standards conversion rather than speed change is used)
>
>>>Similarly the UK HDTV material shot 1080/25p is slowed down for the
>>>US market and run as 1080/24p (or then 3:2ed to 1080/60i)
>>
>>This has been done for a few movies shot on digital PAL (eg, "The
>>Anniversary Party"), but I've never heard of UK TV shows being slowed
>>down for US telecasts.
>
> The BBC series Rockface was shot 1080/25p on HDCam and was co-produced with
> a US broadcaster/studio. The 25p material was slowed to 24p for the US
> transmissions. (Presumably it was then either broadcast 24p or 60i with 3:2
> pull-down)
>
> I imagine that UK drama shot 25p will follow this model of slow-down to 24p.
> Of course 50i drama will still be converted to 60i with no speed change.

Most of the standards concersions I've seen were very obviously done
with the field/frame interpolation which takes portions of one
field/frame and puts it into another, which always creates ghosting and
other artifacts. While this can be minimized by additional processing,
the result is a much softer picture, and usually (though I don't
understand why) desaturated and otherwise poor color.

>>>Where frame rate conversion is required (say from 60i/p to 50i/p)
>>>standards convesion is improving now to the point where 50->60 and
>>>60->50 conversion is more transparent than MPEG2 encoding... (i.e.
>>>more motion artefacts are introduced by the MPEG process than the
>>>conversion) Using phase correlation techniques especially means it
>>>is not a major issue.
>>
>>I'm very skeptical re standards conversions because I've seen so many
>>that looked atrocious.
>
> They have got massively better in recent years. You get what you pay for.
> In the last 3 or 4 years the Snell and Wilcox Alchemist PhC and Platinum
> conversions approach transparency - even on fast moving sports.
>
> We also have the problem in the UK of having to un-pick 3:2 pull-down of US
> film transfers, as the non-continuous movement confuses motion tracking,
> instead the 3rd redundant field has to be detected and removed. So we go
> from 60i 3:2 to 48i 2:2 using DEFT (Digital Electronic Film Transfer -
> another S&W technique) and then speed up to 50i. It works really well on US
> film drama edited on 480/60i tape. You only notice the slight softness
> difference between 480 and 576 - which is far less obvious than the 3.58NTSC
> vs 4.43PAL softness difference.
>
>>After all the effort to restore "Brideshead Revisited" the NTSC
>>broadcasts (and DVDs) were literally unwatchable.
>
> Well the broadcasts will have been made on 1980s converters. I'm
> disappointed that the DVDs were standards converted at all - AIUI Brideshead
> was shot and edited on film so could have been telecined from the film
> masters to 480/60i or 480/24p for DVD mastering. You'd have probably had a
> slow-down if this were the case as I suspect the UK TV production will have
> been shot 25fps film rather than 24fps (as in Europe film-for-TV is normally
> shot at 25 rather than 24)
>
> The chances are that if the DVD was mastered from the UK VT transfer it may
> well have been the 1980s 1" C format PAL, and the telecine-ing would have
> happened in the 80s as well, if it were mastered from the US broadcast
> masters made in the 80s it will have had a pretty bad conversion as well I
> imagine. Telecine and conversion technology has moved on a-pace since the
> 80s and even the mid 90s...
>
> There are of course all the usual DVD mastering annoyances as well - too
> much noise reduction, excessive compression and addition aperture correction
> etc.

I'm fairly sure the NTSC discs were new conversions (the color was
improved, and identical to the new PAL discs). For some reason they
were just done incredibly badly. I got the PAL discs when they became
available and they are fine.

The R1/NTSC discs, though, are appalling for the unending chop-chop
arising from sub-mediocre frame/field interpolation, especially for
moderate pans and rolling credits.

>>Ditto for the new "Forsyte Saga" (though the PAL DVDs for this were
>>pretty bad too).
>
> Again - a commercial DVD release - presumably not optimised for quality.
> There seem to be a massive range in quality of UK DVD releases - the good
> ones are very good, the bad ones are terrible. The best feature film
> transfers are sourced from 1080/24p HD masters - which are converted to
> 576/50i (576/25p) via scaling and speed-up.
>
> For high quality R1 PAL video and film DVD remastering the Dr Who
> Restoration Team have worked marvels. They use a BBC R&D PAL decoder to
> transfer the D3 copies of the 2" or 1" video masters (which leaves almost no
> composite artefacts), and where possible re-transfer and re-grade any film
> inserts that remain as film in the BBC archive. The quality is often
> stunningly good.

I've noted that the quality of PAL discs seems to vary a lot, and I've
been disappointed often by the discs for UK TV series.

As you say, most movies now are mastered at 1080/24, which makes for
easy downconversion to NTSC and PAL specs.

>>>>576p requires nothing more than display devices with line-doubling
>>>>and perhaps frame doubling (to 100hz) capacity.
>>>
>>>576/50p requires at worst 576/50p acquisition gear - which is rare.
>>>I think it is more common to shoot 1080/50i and then downconvert to
>>>576/50p, or 1080/25p frame doubled? Of course you can also TK 25fps
>>>film (sped up 24fps film as well) to 50p by frame doubling. There is
>>>no advantage over 576/25p, but an improvement over 576/50i if you
>>>don't vertically filter to reduce interlace twitter.
>>
>>I'd assume that with proper conversion to progressive scan, this
>>should work.
>
> 576/50i to 576/50p should be OK if the de-interlacing processing is high
> enough quality. 576/50i to 576/25p will look horrid though, as you are
> chucking away half of the motion information. To reduce display flicker
> then you would either have to move to 576/100p (quite a high scan rate) or
> 576/100i (which will have some interlace artefacts - but these will be at
> 50Hz not 25Hz as is the case with 50i)
>
>>>>This has been available from Faroudja for several years.
>>>
>>>Faroudja are more in the display conversion rather than standards
>>>conversion business aren't they? Snell and Wilcox have made HD
>>>standards converters (and some of THE best SD converters) for a
>>>while.
>>
>>Faroudja is best known for its high-end home scalers (I have one of
>>their NRS series feeding 720p/60 to an Ampro HD3600 CRT projector)
>>and, more recently, the FLI2310 chip used in certain progressive-scan
>>DVD players (with scaling to 720p and 1080i as well as 480p).
>
> Yep -scaling rather than conversion - i.e. the field/frame rates are not
> changing? (Apart from 24 to 60 which is a 3:2 conversion rather than
> anything more complex?)
>
> In other words scalers used to get between 480i to 480p to 720p to 1080i at
> either 30 or 60 (or 24 with 3:2)

Faroudja scalers rescale all NTSC sources to 60 frames. For 24fps
source material, the additional field generated by the 3:2 pulldown is
discarded, a progressive frame is created by combining the two fields
from each movie frame, and the result is output at 60fps, reinstating
the pulldown but eliminating all interlace artifacts.

> If we were to move to 60i, 60p or 30p production in Europe we'd need to use
> field/frame rate tstandards conversion, rather than scaling, permanently to
> convert to or from 576/50i for upconversion of SDTV material, and
> downconversion of HD. Whilst this is now a high quality process it is also
> a much more expensive process than scaling - especially if you have to do it
> for all of your outlets, especially regional ones. Far better to only
> field/frame rate convert the small amount of imported material rather than
> your entire archive and continued SDTV production base?
>
> Given that much of what the UK would import from the US would be 24p, rather
> than 30p or 60i, that would be easily converted to 48i then sped up to 50i
> for playout without conversion. We'd then only need to do a standards
> conversion for the 30p, 60i or 60p material such as sport, gameshows etc.
> which are far less common and popular imports than the 24p stuff.

I realize we can't expect broadcast networks to shange their ways, but
there is no reason for serious home theater people to be short-changed.

Ideally all video sources should be displayed in their original format.

>>>>I've heard mixed comments re the 100hz TVs sold in Europe.
>>>
>>>This is because they are mainly used to convert 576/50i to 576/100i
>>>- with no real progressive conversion. Some (Philips Natural Motion)
>>>even convert material (like film) sourced 576/25p, but broadcast
>>>576/50i, to 576/100i by interpolating in-between fields, giving the
>>>whole film sequence a really fluid "video" look. Very disconcerting
>>>- and the total opposite of what many producers are now doing
>>>(shooting video at 576/50i but reducing the temporal resolution in
>>>post-production to 576/25 to give a "film" look)...
>>>
>>>One problem is that it is difficult to buy a direct view CRT set with
>>>external 576/100i or 1080/50i inputs in the UK - so you are
>>>dependent on the internal processing architecture - and can't chose
>>>your own.
>>
>>Has their been much interest in progressive-scan DVD players in the
>>UK?
>
> Not huge amounts - though quite a few players are available. However there
> are very few 50p CRT sets on the market. Prog scan DVDs are mainly used
> with inherently progressive display devices that can't display interlace
> native - plasmas, DLPs, LCDs etc - which have 576/50p input compatibility.
>
> However the vast majority of TV sales in the UK are still direct-view CRTs -
> predominantly in the sub 28" diagonal. However increasingly >21" sets sold
> are 16:9 CRT shape - it is very difficult to buy a 4:3 set bigger than 21"
> in most UK stores these days. 32" is the largest mass market 16:9 CRT -
> though 36"ers are now on offer.
>
> Plasmas are growing in popularity, though like projectors they are still
> quite niche.
>
>>I'm aware of them being marketed but, as you say, without compatible
>>displays their use would be limited.
>
> 100i rather than 50p is the marketing thing for direct-view CRTs in the UK.
> Annoying as I would rather watch an unprocessed 50i picture - but it is
> difficult to buy a high-end TV without processing to 100i now... (And only
> the higher end sets have multiple RGB inputs etc.)
>
>>>>Obviously they would eliminate the flicker
>>>>associated with 50hz; however, if the rescaling (especially for
>>>>interlace-video sourced material) is not done well the results might
>>>>be less than stellar.
>>>
>>>The interpolation, motion compensation/tracking etc. used to
>>>generate twice the number of fields is the real problem with
>>>European 100Hz sets. You get smearing on fast motion, nasty overly
>>>vicious noise reduction, enhancement of MPEG2 blocking and HF
>>>artefacts etc. introduced on DVD, DSat and DTT transmissions. The
>>>RGB interconnects commonly used between DVD / Digital TV set top
>>>boxes and European TVs means that some of the MPEG2 artefacts that
>>>would be hidden by PAL (or NTSC) composite or S-video encoding
>>>aren't - so you see more of the coding errors.
>>
>>Rescaling is always a balancing act. If the source is technically
>>poor it will likely look worse on a high-resolution display.
>
> Yep - though a lot of 100Hz introduces as many artefacts as it also exposes!
>
>>If new display formats eliminate flicker associated with PAL (even
>>NTSC) and do so without compromising other technical aspects this is
>>the way to go IMO.
>
> Technically the flicker is nothing to do with PAL - that is only the chroma
> standard!
>
> I think you have to divorce your display rate and your broadcast rate don't
> you?
>
> I think that we'd be best suited eventually moving to 1080/25p and 1080/50p
> broadcast standards - display on direct CRTs would be at 100i (with
> interlace artefacts at 50Hz) - but display on plasma, DLP, LCD etc would be
> at 50p without major flicker (as the flicker is more a CRT scanning raster
> thing which isn't inherent in other display technologies?)
>
>>Someone has even suggested a 120hz standard for American HDTV which
>>would be fully compatible with both legacy 60hz NTSC video and all
>>movies and 24fps film-sourced TV series (for which each frame would be
>>shown five times, eliminating 3:2 pulldown artifacts).
>
> Yep - presumably retaining 24p and 60i/p transmission systems and only
> up-converting to 120p or i in the display device by frame repetition?

Exactly. Movies would be should with each frame repeated five times.
NTSC sources would be line doubled and frame doubled.

>>My major issue with 50hz standards is the excess flicker for standard
>>displays, and the fact that all movies and film-sourced American TV
>>series must be speeded up which may cause serious audio compromise.
>
> Have you watched 50i or 25p material on a non-CRT device?

No, only on my Sony PVM monitors, all of which are multistandard.

> The audio
> compromise is only an issue for the small amount of imported stuff we have -
> our native stuff has none. The major UK networks show quite small amounts
> of imported material - though feature films I agree are compromised
> slightly, but I personally find 3:2 more annoying than the audio change!

Feature films IMO are very seriously compromised, and this is my primary
objection re 50hz.

Assuming a multiscan display can be available, my ideal DVD player
would, first, rescale all NTSC and PAL movies to 720p using Faroudja
technology; second, discard the pulldown for NTSC and the speedup for
PAL and output a 72fps display; third, restore the audio for PAL discs
to their correct speed using a digital chip.







C.
Anonymous
April 29, 2004 9:30:29 PM

Archived from groups: alt.video.digital-tv (More info?)

Doug McDonald wrote:

> Could somebody explain how you can change the pitch of music
> without changing the speed, without totally screwing up the phase
> relationships that control soundstage imaging?

Analog recording changes the pitch of audio if the playback speed is
changed.

Digital does not, because digital recording increases the speed (ie,
tempo) but not the content (ie, pitch).

I can't comment on phase or soundstage.








C.
Anonymous
April 29, 2004 9:30:30 PM

Archived from groups: alt.video.digital-tv (More info?)

manitou910 wrote:
> Doug McDonald wrote:
>
>> Could somebody explain how you can change the pitch of music
>> without changing the speed, without totally screwing up the phase
>> relationships that control soundstage imaging?
>
>
> Analog recording changes the pitch of audio if the playback speed is
> changed.
>
> Digital does not, because digital recording increases the speed (ie,
> tempo) but not the content (ie, pitch).
>
> I can't comment on phase or soundstage.
>
>

You did'nt even try to answer my question. How do they do this?

Doug McDonald
Anonymous
April 29, 2004 9:36:39 PM

Archived from groups: alt.video.digital-tv (More info?)

Doug McDonald wrote:
>>
>> Where frame rate conversion is required (say from 60i/p to 50i/p)
>> standards
>> convesion is improving now to the point where 50->60 and 60->50
>> conversion
>> is more transparent than MPEG2 encoding... (i.e. more motion artefacts
>> are
>> introduced by the MPEG process than the conversion) Using phase
>> correlation
>> techniques especially means it is not a major issue.
>
> I don;t understand how this conversion can be done well. Consider the
> typical scenario when I watch a 50->60 Hz conversion: A F1 car
> is travelling down the track to 100 mph, with buildings and fences with
> narrow highly visible poles moving rapidly, as the camera
> pans to keep teh car centered. How do you fix up the 5:6 motion of those
> highly visible discrete poles? Aboyt 90% of 50Hz originaled TV
> is of this sort. It looks bad.

IMO it is never done well. Even the Snell & Wilcox systems still take
portions of each field or frame and put them in an adjacent one.

While motion artifact correction may minimize artifacts it cannot
eliminate them completely and, at best, these standards conversions from
PAL to NTSC always look soft and fuzzy.

It's important to understand that even the S&W Alchemist was likely
designed for interlace display which camouflages a multitude of sins.

With conversion to progressive scan, these conversions are brutally
unmasked, though they look least bad with Faroudja scaling.







C.
Anonymous
April 29, 2004 10:08:09 PM

Archived from groups: alt.video.digital-tv (More info?)

In article <VYakc.28745$huU.18029@news04.bloor.is.net.cable.rogers.com>,
manitou910 <manitou910@rogers.com> writes:
> Doug McDonald wrote:
>
>> Could somebody explain how you can change the pitch of music
>> without changing the speed, without totally screwing up the phase
>> relationships that control soundstage imaging?
>
> Analog recording changes the pitch of audio if the playback speed is
> changed.
>
> Digital does not, because digital recording increases the speed (ie,
> tempo) but not the content (ie, pitch).
>
> I can't comment on phase or soundstage.
>
A really good speed conversion will still likely cause
problems. Theoretically, one might split the modulation
from the tones (al Cepstrum), but that is messy and
impercise also.

There really is no magic bullet for digital audio speed
conversion. Think of it like this:

1) The tones have to be maintained. (maintains musicality.)
2) The modulation of the tones has to be changed. (changes the timing.)

This isn't easy (or even perfectly possible.) For speech or low-fi
(maybe FM radio quality), it isn't probably too bad. I used to be
an aggressive developer, and that is something that even I wouldn't
tackle.

John
Anonymous
April 29, 2004 10:08:10 PM

Archived from groups: alt.video.digital-tv (More info?)

John S. Dyson wrote:
=
>
> This isn't easy (or even perfectly possible.) For speech or low-fi
> (maybe FM radio quality), it isn't probably too bad.


FM radio is not low fi, at least not here in the US for the station
I listen to ... it's definately hi-fi.

Doug McDonald
Anonymous
April 30, 2004 1:04:31 AM

Archived from groups: alt.video.digital-tv (More info?)

In article <c6rih8$2vd$2@news.ks.uiuc.edu>,
Doug McDonald <mcdonald@scs.uiuc.edu> writes:
> John S. Dyson wrote:
> =
>>
>> This isn't easy (or even perfectly possible.) For speech or low-fi
>> (maybe FM radio quality), it isn't probably too bad.
>
>
> FM radio is not low fi, at least not here in the US for the station
> I listen to ... it's definately hi-fi.
>
I meant, in comparison with CD-like quality. When I was speaking
of FM quality, I meant TYPICAL FM quality. Theoretically, you might
get 80dBSNR with <0.1% distortion over 20-15KHz, but most of the
time, that isn't true :-).

John
Anonymous
April 30, 2004 1:04:32 AM

Archived from groups: alt.video.digital-tv (More info?)

John S. Dyson wrote:
>
>
> I meant, in comparison with CD-like quality. When I was speaking
> of FM quality, I meant TYPICAL FM quality. Theoretically, you might
> get 80dBSNR with <0.1% distortion over 20-15KHz, but most of the
> time, that isn't true :-).


Well, there IS the Optimud.

Doug McDonald
Anonymous
April 30, 2004 1:16:31 AM

Archived from groups: alt.video.digital-tv (More info?)

Doug McDonald wrote:
>>
>>> Could somebody explain how you can change the pitch of music
>>> without changing the speed, without totally screwing up the phase
>>> relationships that control soundstage imaging?
>>
>> Analog recording changes the pitch of audio if the playback speed is
>> changed.
>>
>> Digital does not, because digital recording increases the speed (ie,
>> tempo) but not the content (ie, pitch).
>
> You did'nt even try to answer my question. How do they do this?

Er......, why would anyone want to change the pitch of music, other
things being equal?

And, sorry, I don't know (or need to know) how this might be done.

The earlier portions of this thread were concerned with _avoiding_ pitch
distortions resulting from the 4% speedup which analog[audio] PAL videos
and telecasts of theatrical movies create.

Digital audio at least avoids increasing the pitch.






C.
Anonymous
April 30, 2004 1:16:32 AM

Archived from groups: alt.video.digital-tv (More info?)

manitou910 wrote:

>>
>>
>> You did'nt even try to answer my question. How do they do this?
>
>
> Er......, why would anyone want to change the pitch of music, other
> things being equal?
>
> And, sorry, I don't know (or need to know) how this might be done.
>
> The earlier portions of this thread were concerned with _avoiding_ pitch
> distortions resulting from the 4% speedup which analog[audio] PAL videos
> and telecasts of theatrical movies create.
>
> Digital audio at least avoids increasing the pitch.
>
>

My question is, how do they avoid changing the pitch
when speeding it up?

Doug McDonald
Anonymous
April 30, 2004 1:41:18 AM

Archived from groups: alt.video.digital-tv (More info?)

In article <Pgekc.318598$2oI1.21001@twister01.bloor.is.net.cable.rogers.com>,
manitou910 <manitou910@rogers.com> writes:
> Doug McDonald wrote:
>>>
>>>> Could somebody explain how you can change the pitch of music
>>>> without changing the speed, without totally screwing up the phase
>>>> relationships that control soundstage imaging?
>>>
>>> Analog recording changes the pitch of audio if the playback speed is
>>> changed.
>>>
>>> Digital does not, because digital recording increases the speed (ie,
>>> tempo) but not the content (ie, pitch).
>>
>> You did'nt even try to answer my question. How do they do this?
>
> Er......, why would anyone want to change the pitch of music, other
> things being equal?
>
> And, sorry, I don't know (or need to know) how this might be done.
>
> The earlier portions of this thread were concerned with _avoiding_ pitch
> distortions resulting from the 4% speedup which analog[audio] PAL videos
> and telecasts of theatrical movies create.
>
> Digital audio at least avoids increasing the pitch.
>
Note that we are speaking of the necessary transforms. If you
speed up the timing, then you have to do pitch correction. Digital
audio doesn't MAGICALLY allow for variable timing without pitch
change!!! You cannot just 'clock' the signal faster, because
everything will be faster/slower similar to good old magnetic
tape.

John
Anonymous
April 30, 2004 1:41:19 AM

Archived from groups: alt.video.digital-tv (More info?)

John S. Dyson wrote:

>
> Note that we are speaking of the necessary transforms. If you
> speed up the timing, then you have to do pitch correction. Digital
> audio doesn't MAGICALLY allow for variable timing without pitch
> change!!! You cannot just 'clock' the signal faster, because
> everything will be faster/slower similar to good old magnetic
> tape.


How is it done? This is non-obvious.

Doug McDonald
April 30, 2004 2:26:24 AM

Archived from groups: alt.video.digital-tv (More info?)

"manitou910" <manitou910@rogers.com> wrote in message news:H2bkc.28789
<
> IMO it is never done well. Even the Snell & Wilcox systems still take
> portions of each field or frame and put them in an adjacent one.
>
> While motion artifact correction may minimize artifacts it cannot
> eliminate them completely and, at best, these standards conversions from
> PAL to NTSC always look soft and fuzzy.
>
> It's important to understand that even the S&W Alchemist was likely
> designed for interlace display which camouflages a multitude of sins.
>
> With conversion to progressive scan, these conversions are brutally
> unmasked, though they look least bad with Faroudja scaling.

Do you find watching CSI, '24', Six Feet Under, Buffy, along with all other
modern HD production in the US, equally troublesome? These are all filmed in
24p and standards converted to 60Hz.

Az.
Anonymous
April 30, 2004 2:52:44 AM

Archived from groups: alt.video.digital-tv (More info?)

In article <c6rvd4$7pg$3@news.ks.uiuc.edu>,
Doug McDonald <mcdonald@scs.uiuc.edu> writes:
> John S. Dyson wrote:
>
>>
>> Note that we are speaking of the necessary transforms. If you
>> speed up the timing, then you have to do pitch correction. Digital
>> audio doesn't MAGICALLY allow for variable timing without pitch
>> change!!! You cannot just 'clock' the signal faster, because
>> everything will be faster/slower similar to good old magnetic
>> tape.
>
>
> How is it done? This is non-obvious.
>
(I know that you have already thought about this, but I am answering
for other people to read also. Unfortunately, I am not giving an
answer, because I probably don't know as many details as you do :-).)

It certainly is non-obvious. A first pass guess would be to seperate
the modulation envelope from the tonal information. (A Cepstrum would
provide some of the raw information, but not enough for full
reconstruction.) (Cepstrum techniques can be useful for seperating
modulation from 'carriers' in very limited and ad-hoc situations.)
(For those who don't know what a Cepstrum is -- think of it as a
way for a signal to be demodulated against itself, and the modulation
components and carrier can be divined.)

I am sure that processing that is partially based upon Cepstrum type
techniques would be useful. I doubt that my blind, first guess
method would work. It does probably require the decomposition
and artful reconstruction of the timing (modulation) and the tonal
info. The 'noise' part of musical signals can probably be adequately
faked.

There are old methods that tend to add a modulation noise to the signal,
and is definitely good enough for speech. Speech signals tend to
have an easier-to-predict tonality and modulation waveform (with
a dual mode characteristic.) Some speech compression schemes even
remove the central region of the spectrum.

The other bad thing about speech vs. music is the much wider frequency
ranges.

I am sure that a 'REASONABLE' job can be done, but it would likely
impart similar effects to mp3 with just a little too small datarate.
In fact, I might have even heard the results of some research from
AT&T Bell Labs (back when it really was Bell Labs when I worked there)
in the late 1980s. It is probably better now, but it is probably
very messy (as you had expected.)

John
Anonymous
April 30, 2004 4:27:10 AM

Archived from groups: alt.video.digital-tv (More info?)

Doug McDonald wrote:
> Stephen Neal wrote:
>
>>
>> Where frame rate conversion is required (say from 60i/p to 50i/p)
>> standards convesion is improving now to the point where 50->60 and
>> 60->50 conversion is more transparent than MPEG2 encoding... (i.e.
>> more motion artefacts are introduced by the MPEG process than the
>> conversion) Using phase correlation techniques especially means it
>> is not a major issue.
>
> I don;t understand how this conversion can be done well. Consider
> the typical scenario when I watch a 50->60 Hz conversion: A F1 car
> is travelling down the track to 100 mph, with buildings and fences
> with narrow highly visible poles moving rapidly, as the camera
> pans to keep teh car centered. How do you fix up the 5:6 motion of
> those highly visible discrete poles? Aboyt 90% of 50Hz originaled TV
> is of this sort. It looks bad.

I'm not an expert in current standards conversion - but I did to a degree
project 10 years ago on motion detection to remove film weave from archive
sequences - very crude stuff for my degree but I read around a lot and have
retained an interest. (I spent a while as a broadcast TV R&D Engineer before
a radical career change took me to a different career in TV... I know say
"Run VT" a lot...)

Early 50/60 standards conversions didn't do any motion detection much at
all, they either ditched or repeated fields / frames, and scaled, causing
horrid juddery movement - this is often still used on consumer grade kit I
believe (such as digital conversion VHS VCRs etc.).

Then 4-field/4-line adaptive conversion - which was a mainstay for many
years (and is still widely used in cheaper, news-grade scenarios) was
introduced. This uses varying amounts of picture information from 4
successive fields, across 4 lines to generate an intermediate picture. The
weighting between the 4-fields contribution is altered depending upon the
amount of total movement within a frame detected. This crudely decides
whether to average a lot across 4 fields or not - depending upon the total
amount of motion between the fields. It is a kind of blanket motion
detection - not done on an area, by area basis.

More sophisticated block-based motion detection was then introduced (similar
to the motion detection used to remove temporal redundancy in MPEG2
encoding) - allowing small areas of the picture to be compared
field-by-field/frame-by-frame to see how much they have moved in between the
source fields. This movement detection then allows the block movement to be
interpolated to create wholly new frames for the standards converted
output - estimated where the blocks of picture would have been if they had
been captured in the output field/frame rate. This is commonly done using
block matching, where an 8x8 or 16x16 picture area is compared with a
similar area in frames/fields either side of the current field/frame - and
the nearest match position is decided to be where that area was previously /
will move to consequently. This works for a lot of movement - but only
allows one motion vector per block (or even per pixel) - which falls over if
you have semi-transparency (say a transparent rolling caption over a panning
shot - where there is two lots of movement in the same block.) It also
suffers with detecting rotational movement - say a spinning wheel.

Snell & Wilcox latest conversion (based on BBC R&D, and used for their
high-end converters - the Platinum and the fully loaded Alchemist - not all
Alchemists have it?) is Phase Correlation. This uses maths which I'm not
sure I have ever understood to provide a far better motion analysis, and
provide far better interpolation of intermediate frames/fields. It copes
better with multiple layers of motion, as well as improving on rotational
motion detection. It is the best conversion available I am advised.

From your example - a high end converter will isolate the poles and the car,
detect that they are moving at different amounts between frames, as well as
detecting the overall motion of the camera. It will then do its best to
create intermediate frames based on the source picture information and the
motion vectors - so will detect the motion of the poles, and create frames
with them in positions they never were in in the source material. This only
works if the poles are correctly tracked - and can go wrong on odd movement
of areas of uniform texture - though in these cases the motion tracking
algorithms usually give up gracefully and the older blending of fields takes
place (which is less of a problem on uniform stuff)

Fomula One material may not be a good choice - I don't know how much control
broadcasters have over the conversion of this stuff - as much of the
coverage is now provided by F1 themselves?

In comparison an older, or current cheaper, converter, characterised by
blurred/juddery output, would just mix varying amounts of the source fields
together to create the output frames, and you'd get multiple imaging etc.

The easiest way to tell which kind of converter is in use in 50Hz land is to
see how much clean field-rate motion is present. The older converters blend
and blur so much that most of the field-based motion is averaged out to a
kind of blurry mush, whereas the newer ones (especially PhC based stuff)
gives very clear and fluid field-based motion.

I am approaching this as a 50Hz viewer watching material converted from
60Hz - so we are throwing away temporal information rather than making it up
though - so we probably benefit in that way, though when 576/50 stuff is
converted to 480/60 you have the benefit of sharper source material?

Steve
Anonymous
April 30, 2004 4:30:11 AM

Archived from groups: alt.video.digital-tv (More info?)

Doug McDonald wrote:
> manitou910 wrote:
>
>>>
>>>
>>> You did'nt even try to answer my question. How do they do this?
>>
>>
>> Er......, why would anyone want to change the pitch of music, other
>> things being equal?
>>
>> And, sorry, I don't know (or need to know) how this might be done.
>>
>> The earlier portions of this thread were concerned with _avoiding_
>> pitch distortions resulting from the 4% speedup which analog[audio]
>> PAL videos and telecasts of theatrical movies create.
>>
>> Digital audio at least avoids increasing the pitch.
>>
>>
>
> My question is, how do they avoid changing the pitch
> when speeding it up?

I believe originally harmonisers or similar were used in the analogue-y
domain. Nowadays I think there are different digital algorithms, the is
certainly an option in many digital audio editing packages to alter the
pitch of a digital audio sample without changing its length, and conversely
to alter the length without changing the pitch.

What this does to embedded phase information - especially with material
already compressed using AC3 etc I have no idea. ISTR that this may be one
reason that a lot of broadcast audio is stored uncompressed on VTRs etc.

Steve
Anonymous
April 30, 2004 4:32:33 AM

Archived from groups: alt.video.digital-tv (More info?)

Aztech wrote:
> "manitou910" <manitou910@rogers.com> wrote in message news:H2bkc.28789
> <
>> IMO it is never done well. Even the Snell & Wilcox systems still
>> take portions of each field or frame and put them in an adjacent one.
>>
>> While motion artifact correction may minimize artifacts it cannot
>> eliminate them completely and, at best, these standards conversions
>> from PAL to NTSC always look soft and fuzzy.
>>
>> It's important to understand that even the S&W Alchemist was likely
>> designed for interlace display which camouflages a multitude of sins.
>>
>> With conversion to progressive scan, these conversions are brutally
>> unmasked, though they look least bad with Faroudja scaling.
>
> Do you find watching CSI, '24', Six Feet Under, Buffy, along with all
> other modern HD production in the US, equally troublesome? These are
> all filmed in 24p and standards converted to 60Hz.

Yes - though with a very crude conversion with no interpolation to create
intermediate fields? Good news for 50Hz viewers though, as we can un-pick
the 3:2 to 2:2 and get a 48Hz version with no field repetition.

Similar techniques are now being developed to "unpick" early 50/60 standards
conversions, to remove the artefacts introduced by early converters when
only a standards converted copy of an important programme exists, with the
master being lost. (Say for an early colour Dr Who...)

Steve
April 30, 2004 4:32:34 AM

Archived from groups: alt.video.digital-tv (More info?)

"Stephen Neal" <stephen.neal@nospam.please.as-directed.com> wrote in message
news:c6s3ah$eb8$1$8302bc10@news.demon.co.uk...
> Aztech wrote:
> > "manitou910" <manitou910@rogers.com> wrote in message news:H2bkc.28789
> > <
> >> IMO it is never done well. Even the Snell & Wilcox systems still
> >> take portions of each field or frame and put them in an adjacent one.
> >>
> >> While motion artifact correction may minimize artifacts it cannot
> >> eliminate them completely and, at best, these standards conversions
> >> from PAL to NTSC always look soft and fuzzy.
> >>
> >> It's important to understand that even the S&W Alchemist was likely
> >> designed for interlace display which camouflages a multitude of sins.
> >>
> >> With conversion to progressive scan, these conversions are brutally
> >> unmasked, though they look least bad with Faroudja scaling.
> >
> > Do you find watching CSI, '24', Six Feet Under, Buffy, along with all
> > other modern HD production in the US, equally troublesome? These are
> > all filmed in 24p and standards converted to 60Hz.
>
> Yes - though with a very crude conversion with no interpolation to create
> intermediate fields? Good news for 50Hz viewers though, as we can un-pick
> the 3:2 to 2:2 and get a 48Hz version with no field repetition.

If it were really such an issue wouldn't we have sets that remove the speed
up and produce a multiple thereof, similar to how decent US tv's are now
able to pull-down 3:2 material.

It' just that I find a lot of this criticism a bit rich when a lower field
rate is used as the main acquisition format in the US enough through HD or
film, this 'horrendous problem' doesn't seem to cause such consternation.

Az.
Anonymous
April 30, 2004 4:32:35 AM

Archived from groups: alt.video.digital-tv (More info?)

Aztech wrote:
>
>
> It' just that I find a lot of this criticism a bit rich when a lower field
> rate is used as the main acquisition format in the US enough through HD or
> film, this 'horrendous problem' doesn't seem to cause such consternation.
>

There is no horrendous problem with the vast majority of 24p
film or HD video programs in the US .... they are sitcoms or talking
heards with slow enough motion that just about anybody can get a
smooth conversion. Car chase scenes, of ccourse, look like film
at 24 HZ even if done to 24 Hz video, and this is not bad, just very
blurry.

I was serious about the F1 auto races: that's essentially all
I ever watch that originated at 50 HZ except news footage.

Doug McDonald
Anonymous
April 30, 2004 5:05:35 AM

Archived from groups: alt.video.digital-tv (More info?)

In article <IPgkc.3429003$iA2.403617@news.easynews.com>,
"Aztech" <az@tech.com> writes:
> "Stephen Neal" <stephen.neal@nospam.please.as-directed.com> wrote in message
> news:c6s3ah$eb8$1$8302bc10@news.demon.co.uk...
>> Aztech wrote:
>> > "manitou910" <manitou910@rogers.com> wrote in message news:H2bkc.28789
>> > <
>> >> IMO it is never done well. Even the Snell & Wilcox systems still
>> >> take portions of each field or frame and put them in an adjacent one.
>> >>
>> >> While motion artifact correction may minimize artifacts it cannot
>> >> eliminate them completely and, at best, these standards conversions
>> >> from PAL to NTSC always look soft and fuzzy.
>> >>
>> >> It's important to understand that even the S&W Alchemist was likely
>> >> designed for interlace display which camouflages a multitude of sins.
>> >>
>> >> With conversion to progressive scan, these conversions are brutally
>> >> unmasked, though they look least bad with Faroudja scaling.
>> >
>> > Do you find watching CSI, '24', Six Feet Under, Buffy, along with all
>> > other modern HD production in the US, equally troublesome? These are
>> > all filmed in 24p and standards converted to 60Hz.
>>
>> Yes - though with a very crude conversion with no interpolation to create
>> intermediate fields? Good news for 50Hz viewers though, as we can un-pick
>> the 3:2 to 2:2 and get a 48Hz version with no field repetition.
>
> If it were really such an issue wouldn't we have sets that remove the speed
> up and produce a multiple thereof, similar to how decent US tv's are now
> able to pull-down 3:2 material.
>
> It' just that I find a lot of this criticism a bit rich when a lower field
> rate is used as the main acquisition format in the US enough through HD or
> film, this 'horrendous problem' doesn't seem to cause such consternation.
>
60Hz TVs don't flicker as much as 50Hz TVs. That is the horrendous
flickerfest problem.

John
Anonymous
April 30, 2004 5:11:57 AM

Archived from groups: alt.video.digital-tv (More info?)

In article <c6s30i$e1e$1$8302bc10@news.demon.co.uk>,
"Stephen Neal" <stephen.neal@nospam.please.as-directed.com> writes:
>
> I am approaching this as a 50Hz viewer watching material converted from
> 60Hz - so we are throwing away temporal information rather than making it up
> though - so we probably benefit in that way, though when 576/50 stuff is
> converted to 480/60 you have the benefit of sharper source material?
>
Well -- NTSC has the same horizontal detail, and high end NTSC
cameras have had double the number of vertical pixels (giving a
wider/flatter vertical response than expected.) So, given
COMPONENT signals, the raw data in NTSC signals isn't necessarily
that much less detailed than PAL originated signals. The
differences in the amount of interline averaging from different
camera designs is probably a similar order of magnitude.

(My own camera has 960+Vertical pixels, which are DSPed
down to 480 with more detail in combo with the necessary
interline flicker reduction than is commonly assumed from
people used to 480V (or 576V) pixels.) AFAIR, the equivalent
PAL camera does NOT have 1140V pixels. Later PAL cameras might
have the larger number of V pixels -- but AFAIR not the
middle 1990s model that I have.

Given composite encoding, that tends to limit the available
vertical resolution without interfering with the chroma. Composite
encoding is mostly legacy (or commercials) at the studio level
nowadays.

John
Anonymous
April 30, 2004 5:53:39 AM

Archived from groups: alt.video.digital-tv (More info?)

Aztech wrote:
>
> Do you find watching CSI, '24', Six Feet Under, Buffy, along with all other
> modern HD production in the US, equally troublesome? These are all filmed in
> 24p and standards converted to 60Hz.

"Six Feet Under" is spectacular in HDTV, 1080i.

The "standards conversion" to which you refer is the same 3:2 pulldown
used for decades to convert movies and US-filmed TV series to 60hz for NTSC.

This, as you well know, is a completely different process from
converting PAL to NTSC and/or vice-versa by chopping up each field or
frame into tiny little bits then reassembling the little bits to fit the
Procrustes Bed of an incompatible frame rate and scanline structure.









C.
April 30, 2004 7:34:40 PM

Archived from groups: alt.video.digital-tv (More info?)

"Doug McDonald" <mcdonald@scs.uiuc.edu> wrote in message
<
> I was serious about the F1 auto races: that's essentially all
> I ever watch that originated at 50 HZ except news footage.

So that's not 90% of all 50Hz originated TV. Of course, if you personally
spend 90% of your time watching overseas F1 then you don't see 99.9% of all
other output.

As for racing, MPEG artefacts are usually the most apparent problem.

Az.
April 30, 2004 7:44:39 PM

Archived from groups: alt.video.digital-tv (More info?)

"John S. Dyson" <toor@iquest.net> wrote in message
<
> > It' just that I find a lot of this criticism a bit rich when a lower
field
> > rate is used as the main acquisition format in the US enough through HD
or
> > film, this 'horrendous problem' doesn't seem to cause such
consternation.
> >
> 60Hz TVs don't flicker as much as 50Hz TVs. That is the horrendous
> flickerfest problem.

Very good of you to remind us, unfortunately this thread isn't about display
devices, a lot of TV don't lock to native resolutions anymore, or in terms
of HD on plasmas or LCD's becoming mainstream it becomes a non-issue because
they don't scan at those rates.

From looking above you'll see people discussing the merits of particular
conversions, and whether this is such a problem as to shift to a common
frame-rate. As you can gather, if things are being produced in 24/25p HiDef
they can be adequately converted to play in any region without any major
detriment. Native HiDef productions in the US have to "suffer" the same 3:2
pull-down conversion as foreign material, other countries have to suffer the
speed-up instead.

Az.
April 30, 2004 7:56:29 PM

Archived from groups: alt.video.digital-tv (More info?)

"manitou910" <manitou910@rogers.com> wrote in message
news:D kikc.36649$huU.26335@news04.bloor.is.net.cable.rogers.com...
> Aztech wrote:
> >
> > Do you find watching CSI, '24', Six Feet Under, Buffy, along with all
other
> > modern HD production in the US, equally troublesome? These are all
filmed in
> > 24p and standards converted to 60Hz.
>
> "Six Feet Under" is spectacular in HDTV, 1080i.
>
> The "standards conversion" to which you refer is the same 3:2 pulldown
> used for decades to convert movies and US-filmed TV series to 60hz for
NTSC.
>
> This, as you well know, is a completely different process from
> converting PAL to NTSC and/or vice-versa by chopping up each field or
> frame into tiny little bits then reassembling the little bits to fit the
> Procrustes Bed of an incompatible frame rate and scanline structure.

NTSC to PAL is quite a bit different.

Any decent production of note shouldn't be using native interlaced NTSC or
PAL, take co-productions like the Band of Brothers, that was done at 24fps
HiDef and converted as required for each region. I believe BBC HD production
is at 25fps, this may be slowed to 24fps for conversion to NTSC regions.

It's no longer about throwing native material through a box and hoping the
end result is acceptable.

Az.
Anonymous
April 30, 2004 8:48:18 PM

Archived from groups: alt.video.digital-tv (More info?)

In article <Hvukc.14335165$Id.2378768@news.easynews.com>,
"Aztech" <az@tech.com> writes:
> "John S. Dyson" <toor@iquest.net> wrote in message
> <
>> > It' just that I find a lot of this criticism a bit rich when a lower
> field
>> > rate is used as the main acquisition format in the US enough through HD
> or
>> > film, this 'horrendous problem' doesn't seem to cause such
> consternation.
>> >
>> 60Hz TVs don't flicker as much as 50Hz TVs. That is the horrendous
>> flickerfest problem.
>
> Very good of you to remind us, unfortunately this thread isn't about display
> devices,
>
Note that there are alot of CRT HDTVs (quite common.) Also, you brought
it up... The data update rate is quite different from the scan rate,
where 50Hz would look much worse with a wider field of view.

John
Anonymous
May 1, 2004 12:09:46 AM

Archived from groups: alt.video.digital-tv (More info?)

In article <c6u9cp$rer$1$8300dec7@news.demon.co.uk>,
"Stephen Neal" <stephen.neal@nospam.please.as-directed.com> writes:
> John S. Dyson wrote:
>> In article <c6s30i$e1e$1$8302bc10@news.demon.co.uk>,
>> "Stephen Neal" <stephen.neal@nospam.please.as-directed.com> writes:
>>>
>>> I am approaching this as a 50Hz viewer watching material converted
>>> from 60Hz - so we are throwing away temporal information rather than
>>> making it up though - so we probably benefit in that way, though
>>> when 576/50 stuff is converted to 480/60 you have the benefit of
>>> sharper source material?
>>>
>> Well -- NTSC has the same horizontal detail, and high end NTSC
>> cameras have had double the number of vertical pixels (giving a
>> wider/flatter vertical response than expected.)
>
> Yep - though in this case you are referring to NTSC as the line rather than
> chroma standard? (i.e. Component NTSC aka 601/656 video - 720x480.)
>
> The Philips/BTS/Thomson LDK DPMS cameras used as 16:9/4:3 switchable cameras
> have a 4:3 sensor, with 4x the number of horizontal scanlines as the line
> standard. 4 lines are integrated together for 4:3 images, 3 are integrated
> for 16:9 - within the central 16:9 area. (i.e. the image WIDTH remains the
> same in 4:3 and 16:9, with 16:9 images having a reduced vertical HEIGHT -
> this is different to most other 16:9 cameras where a 16:9 sensor is used
> with a 4:3 centre cut out, and the height remains fixed with the width
> changing.)
>
Those are definitely newer units. When speaking of current material, it
is much much better than the PALies might suggest (in fact, NTSC component
can essentially provide the vertical resolution of PAL, within the
bounds of normal variations in cameras.) No matter what you do, if
you are ever targeting PAL composite, you have to filter like hell
to avoid artifacts.

My own camera is a moderate end ENG unit, nothing special and not a studio
unit. De-interlacing (much more seamless than high scanrates) is common
fare in the US, and full frame combs are also common (which is still
BBC research material for PAL.)

Bottom line: denegrating NTSC (and trying to discount the full superiority
of 60Hz update or faster without artifacting) is specious based upon the
resolution numbers (well, except for OTA pal, where UK allows very
generous transmission bandwidths :-)). You cannot even perceive significant
improved resolution given a 50Hz confusion flicker rate.


John
Anonymous
May 1, 2004 12:28:25 AM

Archived from groups: alt.video.digital-tv (More info?)

John S. Dyson wrote:
> In article <c6s30i$e1e$1$8302bc10@news.demon.co.uk>,
> "Stephen Neal" <stephen.neal@nospam.please.as-directed.com> writes:
>>
>> I am approaching this as a 50Hz viewer watching material converted
>> from 60Hz - so we are throwing away temporal information rather than
>> making it up though - so we probably benefit in that way, though
>> when 576/50 stuff is converted to 480/60 you have the benefit of
>> sharper source material?
>>
> Well -- NTSC has the same horizontal detail, and high end NTSC
> cameras have had double the number of vertical pixels (giving a
> wider/flatter vertical response than expected.)

Yep - though in this case you are referring to NTSC as the line rather than
chroma standard? (i.e. Component NTSC aka 601/656 video - 720x480.)

The Philips/BTS/Thomson LDK DPMS cameras used as 16:9/4:3 switchable cameras
have a 4:3 sensor, with 4x the number of horizontal scanlines as the line
standard. 4 lines are integrated together for 4:3 images, 3 are integrated
for 16:9 - within the central 16:9 area. (i.e. the image WIDTH remains the
same in 4:3 and 16:9, with 16:9 images having a reduced vertical HEIGHT -
this is different to most other 16:9 cameras where a 16:9 sensor is used
with a 4:3 centre cut out, and the height remains fixed with the width
changing.)

These cameras are available in 576/50i - they are widely used in Europe. I
would expect they are also available in 480/60i models as well.

> So, given
> COMPONENT signals, the raw data in NTSC signals isn't necessarily
> that much less detailed than PAL originated signals.

Yep - though in this case presumably you mean 601/656 720x576 component
signals (PAL technically only refers to the chroma system - NTSC refers to a
line standard and a chroma system?) - you can't have PAL component signals
technically (PAL component is an oxymoron?)

Modern cameras are pretty similar between 480/60 and 576/50 modes - so given
the same model of camera there is still surely more vertical resolution in
the 576/50i signal?

> The
> differences in the amount of interline averaging from different
> camera designs is probably a similar order of magnitude.
>
> (My own camera has 960+Vertical pixels, which are DSPed
> down to 480 with more detail in combo with the necessary
> interline flicker reduction than is commonly assumed from
> people used to 480V (or 576V) pixels.)

Yep - you are making a specific camera comparison though - not a general

> AFAIR, the equivalent
> PAL camera does NOT have 1140V pixels. Later PAL cameras might
> have the larger number of V pixels -- but AFAIR not the
> middle 1990s model that I have.

Yep - though DPMS cameras with their 3x or 4x the number of vertical sensor
lines are the same in both standards? We are comparing standards not
capture technology. If one assumes common source material - say 1080/24p
HDTV material, the 576/50i signal will have a greater vertical resolution
than the 480/60i version surely - everything else being equal.

>
> Given composite encoding, that tends to limit the available
> vertical resolution without interfering with the chroma. Composite
> encoding is mostly legacy (or commercials) at the studio level
> nowadays.

Composite is all but dead surely - only old regional stations and the final
stage of the analogue broadcast chain is now PAL in the UK. (Certainly all
UK commercials are delivered in 16:9 component - with only very occasional
composite issues)

Digital TV in both the US and UK mainly uses 4:2:0 sampling for MPEG2 -
chroma resolution subsampled vertically to match the horizontal
sub-sampling? This is the same in both line standards - though DVCam is
different I think (4:1:1 in 480/60, 4:2:0 in 576/50)?

Not sure what point you were trying to make there?

Incidentally - AIUI 50i and 25p signals MPEG2 compress slightly better with
60i and 30p stuff - as there is more temporal redundancy to benefit from at
a lower frame rate? This may offset the larger image area?


Steve
Anonymous
May 1, 2004 12:31:37 AM

Archived from groups: alt.video.digital-tv (More info?)

Aztech wrote:
[snip]
>>> Do you find watching CSI, '24', Six Feet Under, Buffy, along with
>>> all other modern HD production in the US, equally troublesome?
>>> These are all filmed in 24p and standards converted to 60Hz.
>>
>> Yes - though with a very crude conversion with no interpolation to
>> create intermediate fields? Good news for 50Hz viewers though, as
>> we can un-pick the 3:2 to 2:2 and get a 48Hz version with no field
>> repetition.
>
> If it were really such an issue wouldn't we have sets that remove the
> speed up and produce a multiple thereof, similar to how decent US
> tv's are now able to pull-down 3:2 material.

Yep - though they'd need quite a big hard-drive wouldn't they? If you fed in
a 50i signal and replayed it at 48i you'd need to store 2 frames for every
second?

>
> It' just that I find a lot of this criticism a bit rich when a lower
> field rate is used as the main acquisition format in the US enough
> through HD or film, this 'horrendous problem' doesn't seem to cause
> such consternation.

Yep - and both intellectually and visually 3:2 pull-down is not a nice
solution to 24p display on 60i devices... I certainly chose to watch 24p
material as 50i rather than 60i when I have the choice (and there is quite a
lot of 24p material out there)

Steve
Anonymous
May 1, 2004 12:31:38 AM

Archived from groups: alt.video.digital-tv (More info?)

Stephen Neal wrote:
>
>
> Yep - and both intellectually and visually 3:2 pull-down is not a nice
> solution to 24p display on 60i devices... I certainly chose to watch 24p
> material as 50i rather than 60i when I have the choice (and there is quite a
> lot of 24p material out there)


2:3 pulldown is eessentially invisible and is never an issue.

50i, at that display rate, is simply, absolutely, utterly,
completely, UNWATCHABLE.

Doug McDonald
Anonymous
May 1, 2004 4:25:31 AM

Archived from groups: alt.video.digital-tv (More info?)

Doug McDonald wrote:
> Stephen Neal wrote:
>>
>>
>> Yep - and both intellectually and visually 3:2 pull-down is not a
>> nice solution to 24p display on 60i devices... I certainly chose to
>> watch 24p material as 50i rather than 60i when I have the choice
>> (and there is quite a lot of 24p material out there)
>
>
> 2:3 pulldown is eessentially invisible and is never an issue.
>
> 50i, at that display rate, is simply, absolutely, utterly,
> completely, UNWATCHABLE.

Interesting comments. I completely believe that that is your experience -
just as I believe that some people prefer 100Hz TVs over 50Hz. Personally I
have yet to see a 100Hz TV I could watch for any length of time.
Increasingly I believe that environment and familiarity influence what we
see as artefacts and defects in our system.

You say 3:2 pulldown is essentially invisible - yet when I watch 24p
material with 3:2 pulldown added in my replay device (and viewed 60i on my
display) I find it less fluid in motion terms than the same material sped up
to 25fps and viewed 2:2 50i (which is another option on my replay device).
I have no problem spotting it - I see it almost instantly and find it quite
unpleasant - but this may be because I have only recently started watching
material in this format in native 60i (since 2000 when I bought a R1
compatible DVD player) If you have been watching 3:2 stuff for a long
period I suspect you are used to it as a norm, and ignore the artefacts that
I find obvious.

Similarly I find 2:2 50i pulldown films on DVD and TV more watchable than
3:2 pulldown 60i films. I notice the judder. Because I have watched 50i TV
displays all my life I guess I just don't see the flicker as problem. I
guess I find 50Hz refresh rate normal...

However what is odd - as I find 60Hz refresh PC displays unwatchable (and
get a headache within minutes if using them). I suspect it is a
brightness/contrast and field of view issue - PC CRT monitors often display
bright white displays, and I suspect their phosphors decay quicker (to cope
with higher refresh rates?). On the other hand I spend my days with my
entire field of view filled with 50Hz broadcast monitors (I stare at a
production monitor stack for my living) and am used to assessing picture
quality at the point of broadcast. I don't get headaches in this
environment - but then again I am looking at them differently - flitting
between screens rather than looking at a single one intently.

I have no problems with 60i or 50i viewing - and would not suggest for a
moment that the US move to 50i, just as I see no reason for the UK to move
to a 60i HDTV system. There are perfectly valid reasons for both our
refresh rates - and both regions have made pragmatic choices for film
replay, HDTV production etc. IMHO our SDTV and legacy production systems
are too well established to allow a universal production/broadcast
field/frame rate to be adopted at this point.

With 24p and 25p perfectly valid as HDTV systems - and modern display
devices no longer tied to broadcast refresh rates, or less "refresh-related"
(i.e. many non-CRT displays no longer have a similar pulsing refresh
dynamic, or a raster scan and decay system in the way that a CRT system has)
there is less reason for the transmission system to run at the display
refresh rate? On the other hand I think that for non-drama production, 50
or 60 Hz motion rendition is important, so any system should not be limited
to 24/25/30 fps motion rendition - instead 50 or 60 Hz interlace, or ideally
(as non-CRT devices cope better with them) progressive systems should be
used?

Steve
May 1, 2004 4:25:32 AM

Archived from groups: alt.video.digital-tv (More info?)

"Stephen Neal" <stephen.neal@nospam.please.as-directed.com> wrote in message
<
> Interesting comments. I completely believe that that is your experience -
> just as I believe that some people prefer 100Hz TVs over 50Hz. Personally
I
> have yet to see a 100Hz TV I could watch for any length of time.
> Increasingly I believe that environment and familiarity influence what we
> see as artefacts and defects in our system.
>
> You say 3:2 pulldown is essentially invisible - yet when I watch 24p
> material with 3:2 pulldown added in my replay device (and viewed 60i on my
> display) I find it less fluid in motion terms than the same material sped
up
> to 25fps and viewed 2:2 50i (which is another option on my replay device).
> I have no problem spotting it

Season 2 of '24' comes to mind, *shudder* (or should that be *stutter*)

Az.
Anonymous
May 1, 2004 5:12:04 AM

Archived from groups: alt.video.digital-tv (More info?)

John S. Dyson wrote:
> In article <c6u9cp$rer$1$8300dec7@news.demon.co.uk>,
> "Stephen Neal" <stephen.neal@nospam.please.as-directed.com> writes:
>> John S. Dyson wrote:
>>> In article <c6s30i$e1e$1$8302bc10@news.demon.co.uk>,
>>> "Stephen Neal" <stephen.neal@nospam.please.as-directed.com> writes:
>>>>
>>>> I am approaching this as a 50Hz viewer watching material converted
>>>> from 60Hz - so we are throwing away temporal information rather
>>>> than making it up though - so we probably benefit in that way,
>>>> though when 576/50 stuff is converted to 480/60 you have the
>>>> benefit of sharper source material?
>>>>
>>> Well -- NTSC has the same horizontal detail, and high end NTSC
>>> cameras have had double the number of vertical pixels (giving a
>>> wider/flatter vertical response than expected.)
>>
>> Yep - though in this case you are referring to NTSC as the line
>> rather than chroma standard? (i.e. Component NTSC aka 601/656 video
>> - 720x480.)
>>
>> The Philips/BTS/Thomson LDK DPMS cameras used as 16:9/4:3 switchable
>> cameras have a 4:3 sensor, with 4x the number of horizontal
>> scanlines as the line standard. 4 lines are integrated together for
>> 4:3 images, 3 are integrated for 16:9 - within the central 16:9
>> area. (i.e. the image WIDTH remains the same in 4:3 and 16:9, with
>> 16:9 images having a reduced vertical HEIGHT - this is different to
>> most other 16:9 cameras where a 16:9 sensor is used with a 4:3
>> centre cut out, and the height remains fixed with the width
>> changing.)
>>
> Those are definitely newer units.

I first came across thse operationally in 1997, when I moved from working in
a PAL 4:3 to a 16:9 576/50i component digital gallery. The models I worked
with then are now actually obsolete.

Whilst LDK (Philips/BTS now ThomsonGrassValley) DPMS cameras are not
universal - they are very popular, and are capable of generating very nice
SDTV pictures if well looked after.

> When speaking of current material,
> it
> is much much better than the PALies might suggest (in fact, NTSC
> component can essentially provide the vertical resolution of PAL,
> within the
> bounds of normal variations in cameras.) No matter what you do, if
> you are ever targeting PAL composite, you have to filter like hell
> to avoid artifacts.

Depends what you mean by "targeting" PAL composite. All network production
in the UK is 576/50i 16:9 (apart from the 1080/25p or 50i stuff) and
therefore there is no real sense in targeting PAL - as depending on the
situation the vertical resolution will depend on the PAL display system.
Some PAL viewers will be 4:3 analogue network viewers - who will see 12F12
(4:3 full frame), 14L12 (14:9 letterbox in 4:3) or 16L12 (16:9 letterbox in
4:3) depending on the source material and its shoot/protect setting.
Similarly some other digital viewers watching via PAL or RGB interconnects
on 4:3 sets will see similar ratios, i.e. 12F12 (4:3 centre cut), 14L12
(14:9 letterbox), 16L12 (16:9 letterbox) - so the display resolution as well
as the composite chroma system plays a part.

As far as I am concerned all UK 625 (aka 576/50i) production is aimed at
16:9 digital RGB viewers (over 50% of UK homes now have at least one TV fed
from a 16:9 digital component source - though some of these will be fed PAL
not RGB) What processing do you believe is applied to 576/50i stuff that
isn't to 480/60i stuff in the component domain - I don't see how any "PAL"
optimisation can be made as the broadcaster can have no idea of the target
display aspect ratio and therefore no idea of the horizontal or vertical
scaling applied for 4:3 PAL viewers (Though with AFDs - Active Format
Descriptors - they can signal their suggested display ratio on some
platforms that support AFDs)

Also - surely both 480/60i and 576/50i digital TV systems are broadcast6
using the same 4:2:0 system with the same vertical chroma sub-sampling - so
the vertical resolution loss (in chroma terms) is as bad in NTSCland as in
PALland? Are you saying that 576/50i stuff is luminance filtered more than
480/60i vertically as 30Hz interline twitter is less obvious than 25Hz
twitter? Whilst this may be the case does it explain a 96 line resolution
difference alone?

Again - please don't think I'm saying 480/60i material is unwatchable - I
know it isn't. I have watched many a high-quality 480/60i D1 or DigiBeta VTR
replay of material - and it can look very nice. I don't think it looks
better than similar replays of 576/50i material though, it normally looks a
bit softer and has a more obvious line-structure in my experience...

If you want to be brutal though - NTSC 3.58 material compared to PAL 4.43
material normally looks much more different - whether from a 1" C format VTR
or off-air. (I've never seen 2" Quad NTSC so can't compare that with the 2"
Quad PAL I have seen)

>
> My own camera is a moderate end ENG unit, nothing special and not a
> studio unit. De-interlacing (much more seamless than high scanrates)
> is common fare in the US, and full frame combs are also common (which
> is still
> BBC research material for PAL.)
>

Yep - but no-one uses PAL for production any more do they? Certainly it is
almost dead as a production format outside news, and even within news it is
slowly dying? 576/50i component production (normally in 16:9)

DVCam, DVCPro and BetaSX (as well as the newer XDCam and IMX formats) are
all 4:1:1, 4:2:0 or 4:2:2 component. Most newer SNG trucks in the UK are
component digital internally with SDI or fibre links to the cameras for live
links (though increasingly component digital microwave links are also used
to get single person wireless cameras back to SNG trucks - especially in
crowded situations where cables couldn't safely be used) The MPEG uplinks
are normally component digital as well. Whilst I am not saying that PAL
analogue is dead - and it is still used for some applications (even in 16:9
production) - it is no longer as relevant as it once was. (Significant
numbers of PAL analogue microwave links are still in use for regional news
in the UK - and many regional news centres are still 4:3->14:9 PAL
analogue - but many of these centres are converting to 16:9 576/50i
production)

> Bottom line: denegrating NTSC (and trying to discount the full
> superiority of 60Hz update or faster without artifacting) is specious
> based upon the resolution numbers (well, except for OTA pal, where UK
> allows very
> generous transmission bandwidths :-)).

I'm not saying 480/60i is bad - though I have yet to be convinced of the
superiority of NTSC 3.58 over PAL 4.43 in the real-world. I am saying that
576/50i has advantages over 480/60i in some areas - specifically in carrying
24/25fps material.

I'd rather watch 576/24p material than 480/24p material everything else
being equal - and that is effectively the difference between 24fps material
carried 480/60i and the same material carried 576/50i. Why else would
576/24p 16:9 production (using 576/25p kit) have been used for post
production on a number of US series ? It allows downconversion to 480/24p
and 3:2 pulldown to 480/60i, as well as conversion to 576/25p or 576/50i
from a single edited master, and has benefits over post-production of the
same material in the 480/24 domain, especially for 576/50i viewers, and no
disadvantages for 480/60i viewers?

As for the PAL I transmission system - well it works quite nicely (and
allows enough space for a nice 728kbs digital audio system as well, though
B/G and SECAM L have also squeezed this in) - though I haven't watched PAL
off-air much since 1998 when I moved to RGB digital reception. Though there
are MPEG2 compression artefacts (increasingly so as the bit rates are
reduced on some output...) I find the lack of composite subcarrier
artefacts - such as the luminance resolution reduction, cross-luma and
cross-colour, and chroma resolution reduction - outweigh these in many
cases.

> You cannot even perceive
> significant improved resolution given a 50Hz confusion flicker rate.
>

Err - I can tell a 480/60i picture from a 576/50i picture with no problem at
all - the line-structure on a 480/60i picture is more visible, and picture
often looks a bit softer. I watch neither in NTSC or PAL, composite or
s-video, though, as I endeavour to use RGB interconnects wherever possible.
This is especially obvious on DVD releases - though often the 480 stuff has
quite nasty aperture correction/edge enhancement/artificial sharpness added
(though this is also annoyingly visible on some 576 material as well...)

I'm not saying I am a 480/60-phobe (or NTSC-phobe) - I have a number of R1
480 line DVDs - the content after all is why I am watching them. However
given a choice between a 480 and 576 line DVD of 24fps material (and
mastering issues being equal) - I'd chose the 576 line version. I'm not
saying you should - just that I would.

Steve
Anonymous
May 1, 2004 5:16:22 AM

Archived from groups: alt.video.digital-tv (More info?)

Aztech wrote:
> "Stephen Neal" <stephen.neal@nospam.please.as-directed.com> wrote in
> message <
>> Interesting comments. I completely believe that that is your
>> experience - just as I believe that some people prefer 100Hz TVs
>> over 50Hz. Personally I have yet to see a 100Hz TV I could watch
>> for any length of time. Increasingly I believe that environment and
>> familiarity influence what we see as artefacts and defects in our
>> system.
>>
>> You say 3:2 pulldown is essentially invisible - yet when I watch 24p
>> material with 3:2 pulldown added in my replay device (and viewed 60i
>> on my display) I find it less fluid in motion terms than the same
>> material sped up to 25fps and viewed 2:2 50i (which is another
>> option on my replay device). I have no problem spotting it
>
> Season 2 of '24' comes to mind, *shudder* (or should that be
> *stutter*)

Are you making this observation from watching a R1 DVD release of "24"
480/60i at home, or watching a really bad non-DEFT 480/60i to 576/50i
conversion where no 3:2 pull-down removal was done, so the motion
compensation/interpolation algorithms were trashed because the 3:2 pulldown
artefacts warped the motion vectors? (AIUI no 16:9 576/50i master was made
available with DEFT conversion? Or that was the reason quoted elsewhere for
the poor quality of the version shown on BBC Three?)

Steve
Anonymous
May 1, 2004 7:14:10 AM

Archived from groups: alt.video.digital-tv (More info?)

Aztech wrote:
>
>>to 25fps and viewed 2:2 50i (which is another option on my replay device).
>>I have no problem spotting it
>
> Season 2 of '24' comes to mind, *shudder* (or should that be *stutter*)

A _lot_ of DVDs (especially for TV series, American and British) have
incorrect flagging for pulldown (including 2:2 for PAL discs) and this
will cause major problems for any display.

The best progressive scan DVD players have circuits which (usually)
identify incorrect flagging and correct it as part of the de-interlacing
process.








C.
Anonymous
May 1, 2004 12:36:59 PM

Archived from groups: alt.video.digital-tv (More info?)

Stephen Neal wrote:

>>
>>50i, at that display rate, is simply, absolutely, utterly,
>>completely, UNWATCHABLE.
>
>
> Interesting comments. I completely believe that that is your experience -
> just as I believe that some people prefer 100Hz TVs over 50Hz.

Of course they are going to prefer 100HZ frame doubled! It doesn't
have the unwatchable flicker.


>
Personally I
> have yet to see a 100Hz TV I could watch for any length of time.
> Increasingly I believe that environment and familiarity influence what we
> see as artefacts and defects in our system.
>

50 Hz is simply to slow: it has a terrible, completely unwatchabble,
flicker.


> However what is odd - as I find 60Hz refresh PC displays unwatchable (and
> get a headache within minutes if using them).

Yes, and 50 Hz is FAR worse.


> I suspect it is a
> brightness/contrast and field of view issue - PC CRT monitors often display
> bright white displays, and I suspect their phosphors decay quicker (to cope
> with higher refresh rates?).

No, actually, they are all much faster than frame rate, of the order
of line rate.


>
> I have no problems with 60i or 50i viewing - and would not suggest for a
> moment that the US move to 50i, just as I see no reason for the UK to move
> to a 60i HDTV system. There are perfectly valid reasons for both our
> refresh rates

There is NO valid reason for 50 HZ ... it is simply too slow. The
poor folks saddled with 50 Hz line frequence should have used some
system to allow a faster refresh rate, at least after the UK dumped
405 lines .... there is plenty of space in your 8 MHz, and plenty
of things like three phase power for lighting.



> With 24p and 25p perfectly valid as HDTV systems - and modern display
> devices no longer tied to broadcast refresh rates, or less "refresh-related"
> (i.e. many non-CRT displays no longer have a similar pulsing refresh
> dynamic, or a raster scan and decay system in the way that a CRT system has)
> there is less reason for the transmission system to run at the display
> refresh rate? On the other hand I think that for non-drama production, 50
> or 60 Hz motion rendition is important, so any system should not be limited
> to 24/25/30 fps motion rendition - instead 50 or 60 Hz interlace, or ideally
> (as non-CRT devices cope better with them) progressive systems should be
> used?


That's true ... I would not in the least object to 50Hz
progressive on an "always on" display even for sports. It would
be fine. As a TRANSMISSION format, now that transmission and
display are becoming uncoupled, it is OK.

Doug McDonald
Anonymous
May 1, 2004 2:58:22 PM

Archived from groups: alt.video.digital-tv (More info?)

manitou910 wrote:
> Aztech wrote:
>>
>>> to 25fps and viewed 2:2 50i (which is another option on my replay
>>> device). I have no problem spotting it
>>
>> Season 2 of '24' comes to mind, *shudder* (or should that be
>> *stutter*)
>
> A _lot_ of DVDs (especially for TV series, American and British) have
> incorrect flagging for pulldown (including 2:2 for PAL discs) and this
> will cause major problems for any display.

Do you mean that the 25p material on a DVD is not indicated as progressive,
but interlaced, or that the material is indicated as progressive, but each
25p frame contains a field from each consecutive film frame and this is not
indicated in the DVD mastering? (i.e. the pull-down is a field out so if the
film frames went ZABC, the video fields would go Z2A1-A2B1-B2C1-C2). The
latter would look OK on a 50i display, but obviously would confuse the hell
out of any DVD player that based its progressive to interlace conversion on
the DVD flags rather than detecting the 2:2 (or 3:2) cadence itself.

Incidentally a lot of UK TV studio series containing film inserts (studio
video, location film was a common technique over here until the 90s) have
this out-of-phase telecine issue - not a problem for the viewer at the time,
though it did cause problems when such series were telerecorded to film for
foreign sales (rather than being standards converted and sold on VT, some
video series were sold on 16mm or 35mm film) as each telerecorded fillm
frame would contain a field from two film source frames, rather than two
fields from the same frame, causing nasty blurring effects etc.

>
> The best progressive scan DVD players have circuits which (usually)
> identify incorrect flagging and correct it as part of the
> de-interlacing process.
>

Yep - trusting any mastering information doesn't seem like the best idea to
me - given how often material is coded in a less than ideal manner.

Steve
May 1, 2004 4:08:04 PM

Archived from groups: alt.video.digital-tv (More info?)

"Stephen Neal" <stephen.neal@nospam.please.as-directed.com> wrote in message
<
> >> You say 3:2 pulldown is essentially invisible - yet when I watch 24p
> >> material with 3:2 pulldown added in my replay device (and viewed 60i
> >> on my display) I find it less fluid in motion terms than the same
> >> material sped up to 25fps and viewed 2:2 50i (which is another
> >> option on my replay device). I have no problem spotting it
> >
> > Season 2 of '24' comes to mind, *shudder* (or should that be
> > *stutter*)
>
> Are you making this observation from watching a R1 DVD release of "24"
> 480/60i at home, or watching a really bad non-DEFT 480/60i to 576/50i
> conversion where no 3:2 pull-down removal was done, so the motion
> compensation/interpolation algorithms were trashed because the 3:2
pulldown
> artefacts warped the motion vectors? (AIUI no 16:9 576/50i master was made
> available with DEFT conversion? Or that was the reason quoted elsewhere
for
> the poor quality of the version shown on BBC Three?)

The BBC version was really poor, but considering it probably went through
24p > 60i > 24p > 50i with bad conversions that's no surprise. Anyone
confirmed the rumour this was intentional as to not pre-empt DVD sales?

This was a DVD on a set that scans at 60Hz, there's quite a bit of movement
(and camera jitter) on '24' and I notice repeat frames and less than smooth
movement on pans. Personally, this irritates me more than the apparent
benefit of 60Hz, but maybe that's down to what we're used to. Same goes for
some peoples irritation concerning artefacts generated by some less than
wonderful 100Hz sets, though I think DRC-1250 looks pretty nice.

I will have to check Season 3, I can no-longer tolerate watching the show on
SkyOne, and looking at the viewing figures everyone but 500k people agree
with me. It's even more crass and aggravating than BBC Three ;) 

Az.
Anonymous
May 1, 2004 7:02:24 PM

Archived from groups: alt.video.digital-tv (More info?)

Stephen Neal wrote:
>
> Do you mean that the 25p material on a DVD is not indicated as progressive,
> but interlaced, or that the material is indicated as progressive, but each
> 25p frame contains a field from each consecutive film frame and this is not
> indicated in the DVD mastering? (i.e. the pull-down is a field out so if the
> film frames went ZABC, the video fields would go Z2A1-A2B1-B2C1-C2). The
> latter would look OK on a 50i display, but obviously would confuse the hell
> out of any DVD player that based its progressive to interlace conversion on
> the DVD flags rather than detecting the 2:2 (or 3:2) cadence itself.

IINM all DVDs at present are encoded as interlace. The flagging is
required to tell the DVD player the correct output order for the fields,
including indicating 3:2 pulldown for movies on NTSC discs.

If the 2:2 pulldown for PAL discs is incorrect it likely won't create a
problem for standard interlace displays, but will definitely cause
problems for progressive-scan players which fail to identify the error.

I have the Italian R2 DVD of Visconti's "The Damned". Play it in slo-mo
or freeze frame and all the wrong fields are combined -- field 2 from
one frame with field 1 from the next!

The new R1 disc from Warner is way better.

For more comprehensice info on this I'd recommend a visit to Jim
Taylor's DVD FAQ. I think the URL still is: http://www.dvddemystified.com






C.
Anonymous
May 2, 2004 2:15:54 AM

Archived from groups: alt.video.digital-tv (More info?)

"Aztech" <az@tech.com> wrote in message
news:EqMkc.14391668$Id.2386980@news.easynews.com...
> "Stephen Neal" <stephen.neal@nospam.please.as-directed.com> wrote in
message
> <
> > >> You say 3:2 pulldown is essentially invisible - yet when I watch 24p
> > >> material with 3:2 pulldown added in my replay device (and viewed 60i
> > >> on my display) I find it less fluid in motion terms than the same
> > >> material sped up to 25fps and viewed 2:2 50i (which is another
> > >> option on my replay device). I have no problem spotting it
> > >
> > > Season 2 of '24' comes to mind, *shudder* (or should that be
> > > *stutter*)
> >
> > Are you making this observation from watching a R1 DVD release of "24"
> > 480/60i at home, or watching a really bad non-DEFT 480/60i to 576/50i
> > conversion where no 3:2 pull-down removal was done, so the motion
> > compensation/interpolation algorithms were trashed because the 3:2
> pulldown
> > artefacts warped the motion vectors? (AIUI no 16:9 576/50i master was
made
> > available with DEFT conversion? Or that was the reason quoted elsewhere
> for
> > the poor quality of the version shown on BBC Three?)
>
> The BBC version was really poor, but considering it probably went through
> 24p > 60i > 24p > 50i with bad conversions that's no surprise. Anyone
> confirmed the rumour this was intentional as to not pre-empt DVD sales?
>

I think the point is that it definitely didn't follow that route :

If it had gone 24p>60i>24p(>48i)>50i it would have looked pretty good.
Instead I believe it went 24p>60i>50i - i.e. no attempt was made to unpick
the 3:2 pulldown so rogue motion artefacts were present and the 60i master
was effectively treated as 60i video in the conversion. AIUI the BBC were
delivered a 50i master - so they had no option to convert themselves using a
DEFT-style converter (where the 60i master would have had 60i>24p>48i>50i
conversion)

> This was a DVD on a set that scans at 60Hz, there's quite a bit of
movement
> (and camera jitter) on '24' and I notice repeat frames and less than
smooth
> movement on pans.

Yep - if you are viewing an R1 DVD on a standard DVD player you will get 3:2
pulldown 60i.

> Personally, this irritates me more than the apparent
> benefit of 60Hz, but maybe that's down to what we're used to.

Well with 24p material the only benefit possible is the 60Hz refresh rate. I
agree that the 3:2 pulldown motion artefacts kind of outweigh this - though
this may be because we are used to 2:2 50Hz film material - where there is
more refresh flicker, but no disription in the temporal domain.

> Same goes for
> some peoples irritation concerning artefacts generated by some less than
> wonderful 100Hz sets, though I think DRC-1250 looks pretty nice.
>

Yep - though DRC-1250 is 50 Hz isn't it? Instead of field-doubling it
line-doubles - delivering a 625/50i->1250/50i upconversion, instead of a
625/50i->625/100i conversion that DRC 100 delivers...

> I will have to check Season 3, I can no-longer tolerate watching the show
on
> SkyOne, and looking at the viewing figures everyone but 500k people agree
> with me. It's even more crass and aggravating than BBC Three ;) 
>

Yep - BBC Three may have dogged it but they didn't put ad-breaks in. (Though
they did bung trails into Liquid News when it was on Choice...)

Steve
Anonymous
May 2, 2004 2:20:28 AM

Archived from groups: alt.video.digital-tv (More info?)

"manitou910" <manitou910@rogers.com> wrote in message
news:4_Okc.345592$2oI1.286012@twister01.bloor.is.net.cable.rogers.com...
[snip]
> I have the Italian R2 DVD of Visconti's "The Damned". Play it in slo-mo
> or freeze frame and all the wrong fields are combined -- field 2 from
> one frame with field 1 from the next!

Yep - presumably the R2 DVD was mastered poorly - and before the 576
progressive standard was agreed, and thus the 2:2 pulldown cadence was less
of an issue. AIUI there were Macrovision issues that prevented 576
progressive players being produced for quite a while after 480 progressive
versions.

>
> The new R1 disc from Warner is way better.
>

Yep - though is this is only because the mastering is better (presumably as
480 progressive has been around for longer?) - nothing to do inherently with
the 480 / 576 difference?

Steve
!