De-Interlace Yes or No??

[Guest]

Archived from groups: rec.video.desktop (More info?)

If I plan to burn my mpegs to DVD is there any advantage to
deinterlacing my captured video?

I have been capturing old VHS to my PC, filtering it, then burning to
DVD.

Thanks,

JT

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

Jan Panteltje wrote:
> On a sunny day (15 Feb 2005 12:43:12 -0800) it happened "Goodner79"
> <jtg30@hotmail.com> wrote in
> <1108500192.114076.271280@c13g2000cwb.googlegroups.com>:
>
> >If I plan to burn my mpegs to DVD is there any advantage to
> >deinterlacing my captured video?
> No, DVD output (from player to TV) is interlaced.
> Only de-interlace if you only want to play on the PC.
> Even then one could use a player that can deinterlace on the fly,
> like xine or mplayer in Linux.

So DVD's are not interlaced? Only VHS and regular TV?

Thanks..JT

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

"Goodner79" <jtg30@hotmail.com> wrote in message
news:1108500192.114076.271280@c13g2000cwb.googlegroups.com...
> If I plan to burn my mpegs to DVD is there any advantage to
> deinterlacing my captured video?
>
> I have been capturing old VHS to my PC, filtering it, then burning to
> DVD.
>
> Thanks,
>
> JT
>

Check out this at www.digitalfaq.com

Short answer; if you plan on viewing it on an interlaced display,
keep the interlace. If your source had no interlace to keep, that's
no problem, don't try to add any.

If you use a progressive display then it is a good idea to
de-interlace.

Luck;
Ken

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

Goodner79 wrote:
> Jan Panteltje wrote:
>> On a sunny day (15 Feb 2005 12:43:12 -0800) it happened "Goodner79"
>> <jtg30@hotmail.com> wrote in
>> <1108500192.114076.271280@c13g2000cwb.googlegroups.com>:
>>
>>> If I plan to burn my mpegs to DVD is there any advantage to
>>> deinterlacing my captured video?
>> No, DVD output (from player to TV) is interlaced.
>> Only de-interlace if you only want to play on the PC.
>> Even then one could use a player that can deinterlace on the fly,
>> like xine or mplayer in Linux.
>
> So DVD's are not interlaced? Only VHS and regular TV?
>
> Thanks..JT


They are. Read above where it says " DVD output (from player to TV) is
interlaced". Pretty much any video signal that goes to a standard
definition NTSC tv is interlaced.

Mike

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

On a sunny day (15 Feb 2005 12:43:12 -0800) it happened "Goodner79"
<jtg30@hotmail.com> wrote in
<1108500192.114076.271280@c13g2000cwb.googlegroups.com>:

>If I plan to burn my mpegs to DVD is there any advantage to
>deinterlacing my captured video?
No, DVD output (from player to TV) is interlaced.
Only de-interlace if you only want to play on the PC.
Even then one could use a player that can deinterlace on the fly,
like xine or mplayer in Linux.

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

Goodner79 wrote:
> If I plan to burn my mpegs to DVD is there any advantage to
> deinterlacing my captured video?
>
> I have been capturing old VHS to my PC, filtering it, then burning to
> DVD.
>
> Thanks,
>
> JT

Deinterlacing will blur the original picture. There is no advantage.
Even if the original was shot using a progressive scan camera it would
not be wise to deinterlace the video if you don't know that the
deinterlacing algorithm is wise to the format of the source image.

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

On a sunny day (Wed, 16 Feb 2005 18:48:47 +0100) it happened Funprice
<jhbusscherNOSPAMWHATSOEVER@freemail.nl> wrote in
<qr1711tb2jlafth1qvh92vvp30npjre4f4@4ax.com>:

>So what you see on a normal tv-screen is just one field at a time?
>Does the first field completely disappear before the second is
>written? I mean: is the phosphor that fast?
Yes the phospor is that fast.
The ones who remember the 'lightpen'... it was just a photo diode
that picked up the signal of the spot passing under its lens.
Your eye may have some more 'persistence', it sort of integrates
over time.

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

On a sunny day (Wed, 16 Feb 2005 16:38:29 +0000 (UTC)) it happened Alan
Pollock <nex@nopanix.com> wrote in <cuvsu4$o97$1@reader2.panix.com>:

>So if you encode something into progressive and display it on a standard tv,
>will the picture be worse than if you'd encoded into interlaced?
>
>And if so, what difference would you see and how bad would it be? Nex

What happens is that for example in PAL the TV displays 50 fields
per second, 25 complete 'frames'.
In NTSC it is 60 and 30 respectively.
Since the normal analog (*) TV cannot adapt horizontal frequency in any way to
'double' the value, you end up with half the number of lines written over each
other. So 1,3,5 are written over 2,4,6.
The fields cover each other.
So you lose half the vertical resolution. (and have only 312 lines in PAL,
a little more then 200 in NTSC).
To get the normal number of lines you'd need to use double the line frequency.
And that would require twice the bandwidth for the videio, that a normal TV
does not have either.
Normally in PAL and NTSC one line is 64 us (micro second), in PAL 625 lines
= 40mS = 25 Hz! So a line duration would have to become 32 uS,
to get 625 lines in 1/50 second (20 ms).
1/50 = 0.02 seconds. 0.02 / 625 = 32 x 10 **-6 (32 micro seconds).


An other interesting point is that if you capture analog at half the vertical
size, so say PAL in 352x288, say from VHS, you in fact only use the even (or
odd) field.
In such a case de-interlace in not needed.
I recorded a lot of VHS that way long ago.

PC media players usually give you a choice to either display odd, even, or
some synthesis of both fileds as 'deinterlace' option.
IIRC my Power DVD does this (some menu option).

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

"Funprice" <jhbusscherNOSPAMWHATSOEVER@freemail.nl> wrote in message
news:qr1711tb2jlafth1qvh92vvp30npjre4f4@4ax.com...

> Thank you.
>
> So what you see on a normal tv-screen is just one field at a time?
> Does the first field completely disappear before the second is
> written? I mean: is the phosphor that fast?


The phospher has some persistence but it is the persistence of the eye that
makes this work. Annoying as it certainly can be.

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

> ----------
> --------
> ----------
> --------
> ----------

If you use a 'BOB' deinterlace you don't get this effect on a progressive
display. Basically it duplicates each field so that get a very good image
at 60fps. From memory, this is how expensive line doublers work.

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

On a sunny day (Thu, 17 Feb 2005 17:42:31 +1000) it happened "Steve Simpson"
<guest@anon.com> wrote in <37j07fF5bquafU1@individual.net>:

>> ----------
>> --------
>> ----------
>> --------
>> ----------
>
>If you use a 'BOB' deinterlace you don't get this effect on a progressive
>display. Basically it duplicates each field so that get a very good image
>at 60fps. From memory, this is how expensive line doublers work.
I dunno how 'expensive' line doublers work.
But I designed my own with a Xilinx FPGA (programmable gate array chip) and
some Philips AD converters, some R2R DA.
This was to play my (interlaced) DVD player output (YUV) on a 32Khz VGA monitor
(at 50 Hz).
Now I have to think really hard, but there were 2 issues.
First you have to double the H frequency (I used a PLL (= Phase Locked Loop) ).
Then I stored the digitized values of a line in the FPGA internal RAM.
These dual port RAMs can be written to and read from at the same time.
Then I did read the line stored in that RAM twice in double the speed.
(Generate 2 lines of 32 us (micro seconds) ).
This did what it was supposed to do, double the number of lines in the current
time.
I will not go into all the color issues here....
Finally I wanted to make it into a commercial product, but it did not meet my
quality standard, as playing the DVD on the PC was simply better....
Also it was expensive in small quantity to make.
Also there were better ones available for less I am sure.
Note that it does NOT have a field memory, but only some line memories!
Tried playing VHS time base corrector with it too... unfinished project,
likely never will be finished as VHS is out.
But it sure worked (or works as I still have the thing), and gave me some
insight in how to design a line doubler.
Exercise





>
>

 

[Guest]

Archived from groups: rec.video.desktop (More info?)

I should add that most software deinterlacing simply blends the fields and
then shows the result at 30fps which looks extremely poor IMO.