Sign in with
Sign up | Sign in
Your question

The Difference Betweeen 96khz & 192khz

Last response: in Home Audio
Share
Anonymous
September 23, 2005 3:24:28 PM

Archived from groups: rec.audio.pro (More info?)

CAn anyone tell the differnce?
kevin
September 23, 2005 4:34:03 PM

Archived from groups: rec.audio.pro (More info?)

92 kHz.

--RY
September 23, 2005 4:34:42 PM

Archived from groups: rec.audio.pro (More info?)

96 kHz?

(OK so that's what I get for being a smart$$$!)
Related resources
Anonymous
September 23, 2005 7:52:52 PM

Archived from groups: rec.audio.pro (More info?)

In article <1127499868.898009.290530@f14g2000cwb.googlegroups.com>,
"Matrixmusic" <kevindoylemusic@rogers.com> wrote:

> CAn anyone tell the differnce?
> kevin


I did a days worth of tests - the same program recorded at 48, 96, and
192k at the same time. We recorded piano, acoustic guitar, percussion,
drums. Not a scientific A/B/C test, but as blind as we could make it.
Everyone (6 people - musicians, engineers, bystander) picked the 192k.
Most telling, the piano player ran into the room after hearing the 192k
from outside the control room saying "I never heard it sound like I hear
it while I'm playing."

You never realize how bad 48k sounds until you do this test. 192k is
pure and airy, 96k has a mid-range grunge that appears, and 48k really
has a lot of the mid-range hardness.

I buy Dan Lavry's argument about poor converter implementation. I buy
all the other audio engineering guru's finding problems with 192k too.
They all make some good points. But I've heard it and I'm convinced it
sounds better.

That being said, I don't go out of my way to record anything at 96k
anymore. It's too much effort for not a lot of advantage in the end for
most types of recording. But if I had to make a real "audiophile"
recording, I'd do it at 192k in a flash.

--
Bobby Owsinski
Surround Associates
http://www.surroundassociates.com
Anonymous
September 23, 2005 8:16:49 PM

Archived from groups: rec.audio.pro (More info?)

Matrixmusic <kevindoylemusic@rogers.com> wrote:
>CAn anyone tell the differnce?

Sure, one gives you half the running time.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Anonymous
September 23, 2005 11:43:33 PM

Archived from groups: rec.audio.pro (More info?)

Matrixmusic wrote:

> CAn anyone tell the differnce?

It makes a really big difference in marketing. Those folks can tell
right away. <g>

--
ha
Anonymous
September 24, 2005 12:56:12 AM

Archived from groups: rec.audio.pro (More info?)

"Matrixmusic" <kevindoylemusic@rogers.com> wrote in message
news:1127499868.898009.290530@f14g2000cwb.googlegroups.com...
> CAn anyone tell the differnce?

According to Dan Lavry, 192kHz is harder to make work right, and is less
likely to sound good as a result. I haven't done the comparison, so I don't
know if he's right, but it's an interesting suggestion.

Peace,
Paul
Anonymous
September 24, 2005 2:51:16 AM

Archived from groups: rec.audio.pro (More info?)

On 23 Sep 2005 11:24:28 -0700, in rec.audio.pro "Matrixmusic"
<kevindoylemusic@rogers.com> wrote:

>CAn anyone tell the differnce?
>kevin
depends on which way the arrows on the speaker cable are pointing


martin
Anonymous
September 24, 2005 5:39:35 AM

Archived from groups: rec.audio.pro (More info?)

martin griffith schreef:
> On 23 Sep 2005 11:24:28 -0700, in rec.audio.pro "Matrixmusic"
> <kevindoylemusic@rogers.com> wrote:
>
>
>>CAn anyone tell the differnce?
>>kevin
>
> depends on which way the arrows on the speaker cable are pointing
>
>
> martin

No, but you do have a LOT of headroom for effects and such...
Anonymous
September 24, 2005 2:03:33 PM

Archived from groups: rec.audio.pro (More info?)

"Matrixmusic" <kevindoylemusic@rogers.com> wrote in message
news:1127499868.898009.290530@f14g2000cwb.googlegroups.com

> CAn anyone tell the difference?

The interesting question is 44 versus 96.

192 versus 96 is so deep into diminishing returns that it is
rediculous.

Listen for yourself with these samples, or make your own
samples with these samples as models:

http://www.pcabx.com/technical/low_pass/index.htm

http://www.pcabx.com/technical/sample_rates/index.htm
Anonymous
September 24, 2005 7:37:59 PM

Archived from groups: rec.audio.pro (More info?)

"Anonymouse" <yeah@rig.ht> wrote:
>
> No, but you do have a LOT of headroom for effects and such...



"Headroom?" By increasing the sample rate?

Gabe is spinning in his grave.

--
"It CAN'T be too loud... some of the red lights aren't even on yet!"
- Lorin David Schultz
in the control room
making even bad news sound good

(Remove spamblock to reply)
Anonymous
September 24, 2005 7:43:34 PM

Archived from groups: rec.audio.pro (More info?)

"Bobby Owsinski" <polymedia@earthlink.net> wrote:
>
> I did a days worth of tests - the same program recorded at 48, 96,
> and 192k at the same time.
[...]
> Everyone (6 people - musicians, engineers, bystander) picked the
> 192k.




Whereas my tests between 44.1 and 96 showed so little difference as to
be utterly insignificant.

Either the equipment I used is very, very different than Bobby's, or his
group of listeners is WAY more picky than me, or one of us ran a really
defective test.

My conclusion was that sample rates over 48K are a total waste of disk
space.

--
"It CAN'T be too loud... some of the red lights aren't even on yet!"
- Lorin David Schultz
in the control room
making even bad news sound good

(Remove spamblock to reply)
Anonymous
September 24, 2005 7:43:35 PM

Archived from groups: rec.audio.pro (More info?)

"Lorin David Schultz" <Lorin@DAMNSPAM!v5v.ca> wrote in
message news:GueZe.274200$tt5.235787@edtnps90
> "Bobby Owsinski" <polymedia@earthlink.net> wrote:
>>
>> I did a days worth of tests - the same program recorded
>> at 48, 96, and 192k at the same time.
> [...]
>> Everyone (6 people - musicians, engineers, bystander)
>> picked the 192k.

> Whereas my tests between 44.1 and 96 showed so little
> difference as to be utterly insignificant.

I read Bobby's post as hedging on the highly important
issues of level-matching, time-synching, and bias control.
It's like the availability of effective software for doing
listening tests based on file comparisons costs a lot of
money (it's free) or hard to find (see www.pcabx.com or the
Hydrogen Audio forums).

I would also be concerned about tests comparing sample rates
that involved different converters, or to a lesser degree
nominally identical converters running at different rates.

If you want to compare different sample rates, that's all
that should change. In the real world the best alternative
seems to be to resample the files. IME a good software
resampler can be more bullet-proof than a hardware
resampler, or converters running at different rates.

> Either the equipment I used is very, very different than
> Bobby's, or his group of listeners is WAY more picky than
> me, or one of us ran a really defective test.

In Bobby's tests, I see a number of relevant parameters that
were not as well-controlled as one might hope for.

If we want to play dueling authories, we can compare Bobby's
comments to those made here some months back by George
Massenburg and Bob Katz. As i recall, they came down on the
"no differences" side of the controversy.

> My conclusion was that sample rates over 48K are a total
> waste of disk space.

I'll see your 48 and *raise* you to 44.1! ;-)
Anonymous
September 24, 2005 7:53:24 PM

Archived from groups: rec.audio.pro (More info?)

Lorin David Schultz wrote:

> "Anonymouse" wrote:

> > No, but you do have a LOT of headroom for effects and such...

> "Headroom?" By increasing the sample rate?

> Gabe is spinning in his grave.

Since I'm not terrific at multitasking, Lanis increases my headroom by
turning up my simple rate.

I'm pretty sure Gabe would understand.

--
ha
Anonymous
September 24, 2005 10:07:03 PM

Archived from groups: rec.audio.pro (More info?)

"Lorin David Schultz" <Lorin@DAMNSPAM!v5v.ca> wrote in message news:GueZe.274200$tt5.235787@edtnps90...
> "Bobby Owsinski" <polymedia@earthlink.net> wrote:
> >
> > I did a days worth of tests - the same program recorded at 48, 96,
> > and 192k at the same time.
> [...]
> > Everyone (6 people - musicians, engineers, bystander) picked the
> > 192k.
>
>
>
>
> Whereas my tests between 44.1 and 96 showed so little difference as to
> be utterly insignificant.
>
> Either the equipment I used is very, very different than Bobby's, or his
> group of listeners is WAY more picky than me, or one of us ran a really
> defective test.
>
> My conclusion was that sample rates over 48K are a total waste of disk
> space.


It almost sounds as if there were a perfectly informed group of listeners
here, who decidedly believed that 192 should be better, and therefore
it was better. At any rate, there are a multitude of details missing.

The best case scenario leaves many questions to be asked: what sets
of gear were used to both record and playback in three sample rates
simultaneously?; what did they impart on the signal?; Playback paths
for each?; Time delay in playback switching? Et Al...

I've yet to meet an engineer that could detect 192 from 96 when applied
to the playback of a single, unprocessed audio track.
Anonymous
September 24, 2005 10:08:42 PM

Archived from groups: rec.audio.pro (More info?)

"Matrixmusic" <kevindoylemusic@rogers.com> wrote in message news:1127499868.898009.290530@f14g2000cwb.googlegroups.com...
> CAn anyone tell the differnce?
> kevin
>


Other than storage space? Not here. Since I mix in a hardware
world, I'm still quite happy with 44.1 or 48K.


--
David Morgan (MAMS)
http://www.m-a-m-s DOT com
Morgan Audio Media Service
Dallas, Texas (214) 662-9901
_______________________________________
http://www.artisan-recordingstudio.com
Anonymous
September 25, 2005 2:20:30 AM

Archived from groups: rec.audio.pro (More info?)

"Arny Krueger" <arnyk@hotpop.com> wrote in message
news:Fuadnfpd4_wEDKjenZ2dnUVZ_smdnZ2d@comcast.com...

> If you want to compare different sample rates, that's all
> that should change.

I think I disagree.

Suppose you want to digitize audio in the frequency range up to 20kHz.

If you digitize with a sampling rate of 48000, you need to use a filter
which stops all frequencies above 24000Hz, but allows through all
frequencies below 20000Hz.

However, if you digitize with a sampling rate of 192000Hz, you meed a
frequency which stops all frequencies above 96000Hz, but allows through all
frequencies below 20000Hz.

This is, in principle, an easier filter to build; and it's possible that
may result in the quality of the filtered analog (pre-digiization) audio
being better for the 192000Hz sampling than it is for the 48000Hz sampling.

So I think there are two separate questions:

1) do devices sampling at 192000Hz result in better digital files than
devices sampling at 48000Hz? To answer this question, you need to use
different sampling devices, as well as different sampling rates.

2) if you sample at 192000Hz, does it sound better if you store and play
back the data at 192000Hz compared with downsampling to 48000Hz and storing
and playing back at 48000Hz? To answer ths question, only one device is
needed.

I think in practice your comment addresses the second question, but not the
first.

Tim
Anonymous
September 25, 2005 2:20:31 AM

Archived from groups: rec.audio.pro (More info?)

"Tim Martin" <tim2718281@ntlworld.com> wrote in message
news:o ikZe.14506$wm3.2555@newsfe6-win.ntli.net
> "Arny Krueger" <arnyk@hotpop.com> wrote in message
> news:Fuadnfpd4_wEDKjenZ2dnUVZ_smdnZ2d@comcast.com...
>
>> If you want to compare different sample rates, that's all
>> that should change.
>
> I think I disagree.

> Suppose you want to digitize audio in the frequency range
> up to 20kHz.

> If you digitize with a sampling rate of 48000, you need
> to use a filter which stops all frequencies above
> 24000Hz, but allows through all frequencies below 20000Hz.

> However, if you digitize with a sampling rate of
> 192000Hz, you meed a frequency which stops all
> frequencies above 96000Hz, but allows through all
> frequencies below 20000Hz.

> This is, in principle, an easier filter to build; and
> it's possible that may result in the quality of the
> filtered analog (pre-digiization) audio being better for
> the 192000Hz sampling than it is for the 48000Hz
> sampling.

No, the low-pass filter for 96 KHz sampling is generally no
easier nor harder to build than the one for 44 KHz sampling.
In fact its often the very same filter!

For about the last 10-15 years the filter in question is
*always* implemented as a digital filter. In most cases
this filter implemented as a filter that has very flat
frequency and ideal phase response up to about 0.95 of the
Nyquist frequency, and also has tons of rejection at the
Nyquist frequency and above. If you operate this filter at
any reasonble frequency it does what is needed.

It is true that some people have thought that a more gentle
slope from 20 KHz to 0.95 of the Nyquist frequency for
higher frequencies might be a neat thing. This has been
implemented as a digital filter, as well. The design of
these filters is stored as sets of parameters on the
converter chip. It's up to the designer of the converter
product to pick which parameter set he wants to use when
multiple parameters are used.

IOW, its entirely feasible to design a converter that
operates very nicely at any sample rate from say 44,100 to
192,000 without any changes other than changing the clock
rate.

> So I think there are two separate questions:

> 1) do devices sampling at 192000Hz result in better
> digital files than devices sampling at 48000Hz? To
> answer this question, you need to use different sampling
> devices, as well as different sampling rates.

This hinges on how you define *better*. I don't know a
global definition for *better* but I do know a globally
applicable means for testing equipment to see if it sounds
different.

The question at hand is - does a good job of sampling at
192,000 Hz sound any different doing a good job of sampling
at 44,100? Quite a bit of evidence has been carefully
collected, and the answer seems to be: "No". Most if not all
contrary reports seem to be due to lack of appropriate care
while collecting data.

> 2) if you sample at 192000Hz, does it sound better if you
> store and play back the data at 192000Hz compared with
> downsampling to 48000Hz and storing and playing back at
> 48000Hz? To answer ths question, only one device is
> needed.

I think we can figure this one by inductive reasoning. If a
good job of sampling at 44,100 Hz sounds no different than
doing a good job of sampling at 96,000 Hz, increasing the
upper sample rate to 192,000 is probably not going to sound
any different.

> I think in practice your comment addresses the second
> question, but not the first.

If the converters are performing ideally or at least well
and consistently, then both situations should sound sound
the same.
Anonymous
September 25, 2005 2:57:01 AM

Archived from groups: rec.audio.pro (More info?)

no

Jonny Durango
Anonymous
September 25, 2005 3:04:01 AM

Archived from groups: rec.audio.pro (More info?)

Arny Krueger wrote:
> "Tim Martin" <tim2718281@ntlworld.com> wrote in message
> news:o ikZe.14506$wm3.2555@newsfe6-win.ntli.net
>
>>"Arny Krueger" <arnyk@hotpop.com> wrote in message
>>news:Fuadnfpd4_wEDKjenZ2dnUVZ_smdnZ2d@comcast.com...
>>
>>
>>>If you want to compare different sample rates, that's all
>>>that should change.
>>
>>I think I disagree.
>
>
>>Suppose you want to digitize audio in the frequency range
>>up to 20kHz.
>
>
>>If you digitize with a sampling rate of 48000, you need
>>to use a filter which stops all frequencies above
>>24000Hz, but allows through all frequencies below 20000Hz.
>
>
>>However, if you digitize with a sampling rate of
>>192000Hz, you meed a frequency which stops all
>>frequencies above 96000Hz, but allows through all
>>frequencies below 20000Hz.
>
>
>>This is, in principle, an easier filter to build; and
>>it's possible that may result in the quality of the
>>filtered analog (pre-digiization) audio being better for
>>the 192000Hz sampling than it is for the 48000Hz
>>sampling.
>
>
> No, the low-pass filter for 96 KHz sampling is generally no
> easier nor harder to build than the one for 44 KHz sampling.
> In fact its often the very same filter!
>
> For about the last 10-15 years the filter in question is
> *always* implemented as a digital filter. In most cases
> this filter implemented as a filter that has very flat
> frequency and ideal phase response up to about 0.95 of the
> Nyquist frequency, and also has tons of rejection at the
> Nyquist frequency and above. If you operate this filter at
> any reasonble frequency it does what is needed.
>
> It is true that some people have thought that a more gentle
> slope from 20 KHz to 0.95 of the Nyquist frequency for
> higher frequencies might be a neat thing. This has been
> implemented as a digital filter, as well. The design of
> these filters is stored as sets of parameters on the
> converter chip. It's up to the designer of the converter
> product to pick which parameter set he wants to use when
> multiple parameters are used.
>
> IOW, its entirely feasible to design a converter that
> operates very nicely at any sample rate from say 44,100 to
> 192,000 without any changes other than changing the clock
> rate.
>
>
>>So I think there are two separate questions:
>
>
>>1) do devices sampling at 192000Hz result in better
>>digital files than devices sampling at 48000Hz? To
>>answer this question, you need to use different sampling
>>devices, as well as different sampling rates.
>
>
> This hinges on how you define *better*. I don't know a
> global definition for *better* but I do know a globally
> applicable means for testing equipment to see if it sounds
> different.
>
> The question at hand is - does a good job of sampling at
> 192,000 Hz sound any different doing a good job of sampling
> at 44,100? Quite a bit of evidence has been carefully
> collected, and the answer seems to be: "No". Most if not all
> contrary reports seem to be due to lack of appropriate care
> while collecting data.
>
>
>>2) if you sample at 192000Hz, does it sound better if you
>>store and play back the data at 192000Hz compared with
>>downsampling to 48000Hz and storing and playing back at
>>48000Hz? To answer ths question, only one device is
>>needed.
>
>
> I think we can figure this one by inductive reasoning. If a
> good job of sampling at 44,100 Hz sounds no different than
> doing a good job of sampling at 96,000 Hz, increasing the
> upper sample rate to 192,000 is probably not going to sound
> any different.
>
>
>>I think in practice your comment addresses the second
>>question, but not the first.
>
>
> If the converters are performing ideally or at least well
> and consistently, then both situations should sound sound
> the same.
>
>

Are there any 96k or higher A/D's out there with an anti-alias filter
that starts rolling off @ around 30k for all the dog-eared audiophiles
out there using earthworks mics and such?

--

Jonny Durango

"If the key of C is the people's key, what is the key of the bourgeoisie?"
Anonymous
September 25, 2005 10:24:02 AM

Archived from groups: rec.audio.pro (More info?)

"Jonny Durango" <jonnydurango1BUSH_FROM_OFFICE@comcast.net>
wrote in message news:E76dnWHOveu3oKveRVn-uw@comcast.com

> Are there any 96k or higher A/D's out there with an
> anti-alias filter that starts rolling off @ around 30k
> for all the dog-eared audiophiles out there using
> earthworks mics and such?

Yes, I think that for example some or most of the M-Audio
Delta series interfaces have that feature. They inherited it
from the AKM converter chips that they use.

You can see what this looks like in this test report:

http://www.pcavtech.com/soundcards/delta-1010/2496.htm
Anonymous
September 25, 2005 2:32:48 PM

Archived from groups: rec.audio.pro (More info?)

Tim Martin wrote:

> So: the difference between 96KHz and 192KHz for this particular converter
> is that one samples at about 6.4MHz before downsampling, the other at about
> 12.8MHz ... :-)

And, according to Dan Lavry, sampling at 12.8 mHz is pretty hard to do
accurately, though we can do a decent job at half that rate.

Since there's no improvement other than wider bandwidth, at least for
recording audio, the tradeoff of greater distortion for higher sammple
rate isn't worth it.
Anonymous
September 25, 2005 4:25:23 PM

Archived from groups: rec.audio.pro (More info?)

"Arny Krueger" <arnyk@hotpop.com> wrote in message
news:o vqdnfPgp7dkQKjeRVn-vQ@comcast.com...

> For about the last 10-15 years the filter in question is
> *always* implemented as a digital filter.

How can the filter be implemented as a digital filter? The purpose of the
filter is to eliminate high frequencies from the analog signal before
conversion to digital, to prevent their producing aliases in the digital
representation.

You can't eliminate them with a digital filter, because it's too late - once
the data has been converted to digital, the aliases are already there, and
cannot be distinguished from analog signals within the passband.

It's easy to see this with a spreadsheet; if you're sampling at 48000Hz,
you can see that a sine wave at 18000Hz generates the same series of digital
values as a sine wave at 30000Hz offset by 180 degrees. That is, a 30000Hz
signal sampled at 48000hz creates an alias at 18000Hz.

The digital filter can't remove this alias, because it doesn't know whether
it's a wanted 18000hz signal, or an unwanted alias of a 30000hz signal.

Tim
Anonymous
September 25, 2005 4:25:24 PM

Archived from groups: rec.audio.pro (More info?)

Tim Martin <tim2718281@ntlworld.com> wrote:
>"Arny Krueger" <arnyk@hotpop.com> wrote in message
>
>> For about the last 10-15 years the filter in question is
>> *always* implemented as a digital filter.
>
>How can the filter be implemented as a digital filter? The purpose of the
>filter is to eliminate high frequencies from the analog signal before
>conversion to digital, to prevent their producing aliases in the digital
>representation.
>
>You can't eliminate them with a digital filter, because it's too late - once
>the data has been converted to digital, the aliases are already there, and
>cannot be distinguished from analog signals within the passband.

The way oversampling works is that you run a converter at a very fast
rate, so that the analogue filter required is up in the MHz region. This
allows you to use a first-order filter with a very high turnover frequency
and still assure there's nothing higher than the Nyquist rate, while really
having no effect in the audible range.

THEN you downsample that high rate data in the digital domain, with a
digital filter. The digital filter is as good as you want it to be or
at least as good as you're willing to pay for. All this stuff goes on
inside the box.

Oversampling has pretty much completely transformed converter design, and
made it possible to build digital systems that actually sound decent.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Anonymous
September 25, 2005 4:25:24 PM

Archived from groups: rec.audio.pro (More info?)

"Tim Martin" <tim2718281@ntlworld.com> wrote in message
news:TGwZe.15516$_56.12512@newsfe1-win.ntli.net

> "Arny Krueger" <arnyk@hotpop.com> wrote in message
> news:o vqdnfPgp7dkQKjeRVn-vQ@comcast.com...
>
>> For about the last 10-15 years the filter in question is
>> *always* implemented as a digital filter.

> How can the filter be implemented as a digital filter?

The digital filter runs at a far higher clock frequency than
the sample clock, and it is backed with a simple analog
filter based on that far higher clock frequency.

This is a benefit of the "oversampling" that is common on
converter chip spec sheets.

The oversampling is usually at 4x the clock frequency, or
higher. 8x is not unusual.

> The purpose of the filter is to eliminate high
> frequencies from the analog signal before conversion to
> digital, to prevent their producing aliases in the
> digital representation.

Agreed.

> You can't eliminate them with a digital filter, because
> it's too late - once the data has been converted to
> digital, the aliases are already there, and cannot be
> distinguished from analog signals within the passband.

See clever oversampling trickery, outlined above.
Anonymous
September 25, 2005 6:32:20 PM

Archived from groups: rec.audio.pro (More info?)

On Sun, 25 Sep 2005 08:37:50 -0400, Scott Dorsey wrote:

> The way oversampling works is that you run a converter at a very fast
> rate, so that the analogue filter required is up in the MHz region.
> This allows you to use a first-order filter with a very high turnover
> frequency and still assure there's nothing higher than the Nyquist rate,
> while really having no effect in the audible range.
>
> THEN you downsample that high rate data in the digital domain, with a
> digital filter. The digital filter is as good as you want it to be or
> at least as good as you're willing to pay for. All this stuff goes on
> inside the box.

That all makes sense, yet I've heard a lot of people voice the opinion
that most digital EQs don't sound very good. At the lowest level, aren't
EQs just filters & volume knobs? If it's so easy to make a digital
brickwall filter, why would it be so difficult to make a good sounding
digital EQ?
Anonymous
September 25, 2005 6:52:59 PM

Archived from groups: rec.audio.pro (More info?)

"Scott Dorsey" <kludge@panix.com> wrote in message
news:D h65mu$h76$1@panix2.panix.com...
>
> The way oversampling works is that you run a converter at a very fast
> rate, so that the analogue filter required is up in the MHz region. This
> allows you to use a first-order filter with a very high turnover frequency
> and still assure there's nothing higher than the Nyquist rate, while
really
> having no effect in the audible range.
>
> THEN you downsample that high rate data in the digital domain, with a
> digital filter. The digital filter is as good as you want it to be or
> at least as good as you're willing to pay for. All this stuff goes on
> inside the box.
>
> Oversampling has pretty much completely transformed converter design, and
> made it possible to build digital systems that actually sound decent.
> --scott

Thanks for that; I'd been wondering what Arny was telling me, and reading
the spec sheet for an ADC (an AK4584 was the first one picked by Google) and
come to that conclusion. It samples at 64 times the Fs (which would be
96khz or 192khz or whatever).

So, because the sampling frequency is very high - far higher than any audio
component present in the audio signal - there's no need to filter the analog
signal before conversion to digital. You can filter it digitally, then
downsample.

So: the difference between 96KHz and 192KHz for this particular converter
is that one samples at about 6.4MHz before downsampling, the other at about
12.8MHz ... :-)

Tim
Anonymous
September 25, 2005 8:34:04 PM

Archived from groups: rec.audio.pro (More info?)

> Tim Martin wrote:
>
> > So: the difference between 96KHz and 192KHz for this particular converter
> > is that one samples at about 6.4MHz before downsampling, the other at about
> > 12.8MHz ... :-)

Mike Rivers <mrivers@d-and-d.com> wrote:

> And, according to Dan Lavry, sampling at 12.8 mHz is pretty hard to do
> accurately, though we can do a decent job at half that rate.


In the case of a specific converter that can run at all of the standard
sample rates, wouldn't the front end clock be constant, such that the
oversampling at 96k might be 8x while the oversampling at 192k would be
4x? This makes much more sense.

ulysses
Anonymous
September 25, 2005 10:02:26 PM

Archived from groups: rec.audio.pro (More info?)

Agent 86 <maxwellsmart@control.gov> wrote:
>On Sun, 25 Sep 2005 08:37:50 -0400, Scott Dorsey wrote:
>
>> The way oversampling works is that you run a converter at a very fast
>> rate, so that the analogue filter required is up in the MHz region.
>> This allows you to use a first-order filter with a very high turnover
>> frequency and still assure there's nothing higher than the Nyquist rate,
>> while really having no effect in the audible range.
>>
>> THEN you downsample that high rate data in the digital domain, with a
>> digital filter. The digital filter is as good as you want it to be or
>> at least as good as you're willing to pay for. All this stuff goes on
>> inside the box.
>
>That all makes sense, yet I've heard a lot of people voice the opinion
>that most digital EQs don't sound very good. At the lowest level, aren't
>EQs just filters & volume knobs? If it's so easy to make a digital
>brickwall filter, why would it be so difficult to make a good sounding
>digital EQ?

Digital equalizers don't sound like analogue equalizers. Some of what people
like about analogue equalization are some of the side effects. I find with
an IIR digital filter, I have to use a lot more cut than I do to get a similar
sound with an analogue filter.

But I'll take digital equalization ANY day for things like notch filters
and brickwalls.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Anonymous
September 25, 2005 10:06:11 PM

Archived from groups: rec.audio.pro (More info?)

Tim Martin <tim2718281@ntlworld.com> wrote:
>
>So, because the sampling frequency is very high - far higher than any audio
>component present in the audio signal - there's no need to filter the analog
>signal before conversion to digital. You can filter it digitally, then
>downsample.
>
>So: the difference between 96KHz and 192KHz for this particular converter
>is that one samples at about 6.4MHz before downsampling, the other at about
>12.8MHz ... :-)

Maybe. Or MAYBE they both sample at the same rate, no matter what output
rate they are configured more, meaning the output filtering is more effective
when you use the lower sampling rate.

The thing is that oversampling gets you everything that an increased sample
rate does, EXCEPT extended ultrasonic frequency response. So, the only reason
you'd want to use the increased sample rate is if you really did need to
reproduce above 20 KHz. And I think the real verdict on whether that is useful
isn't even close to coming in yet.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Anonymous
September 26, 2005 12:43:42 AM

Archived from groups: rec.audio.pro (More info?)

In article <dh76pi$ens$1@panix2.panix.com>,
kludge@panix.com (Scott Dorsey) wrote:

> Digital equalizers don't sound like analogue equalizers. Some of what people
> like about analogue equalization are some of the side effects. I find with
> an IIR digital filter, I have to use a lot more cut than I do to get a similar
> sound with an analogue filter.
>
> But I'll take digital equalization ANY day for things like notch filters
> and brickwalls.
> --scott




Scott, at AES listen to the URS Neve plugs. They are the most analog
sounding tdm eq's I've heard.

The guy instrumental in creating them (Dan Zellman, who's also a master
tech) spent yesterday afternoon modding my console's master section.
Yumm.




David Correia
www.Celebrationsound.com
Anonymous
September 26, 2005 2:12:26 AM

Archived from groups: rec.audio.pro (More info?)

>Matrixmusic <kevindoylemusic@rogers.com> wrote:
>CAn anyone tell the differnce?
>

I don't know, but I record at the highest resolution possible with the gear I
have. Unfortunately, I can't go higher than 44.1. If I could I would go to
96, and if I could I would go to 192. I think some of the people who don't own
hardware capable of the higher sample rates and bitrates experience cognitive
dissonance which causes them to claim that anything above whatever sample
an/or bit rate they can do is superfulous. I think even some people who
eventually buy interfaces capable of higher rates still record at the lower
rate they're used to because the cognitive dissonance that made them decide it
was as good as it gets back when they only had an interface capable of it, is
so strong.

I think if everyone were capable of thinking rationally, everyone would agree
that recording at the highest rate possible is preferable. Afterall, how do we
know that 20 years from now we will meet an alien race of UFO travellers who
can hear way beyond human hearing?? And what if they want to rock???
Anonymous
September 26, 2005 5:19:13 AM

Archived from groups: rec.audio.pro (More info?)

Chevdo wrote:

> I think if everyone were capable of thinking rationally, everyone would agree
> that recording at the highest rate possible is preferable.

I think you owe it to yourself to find Dan Lavry's comments about that,
especially if you're suggesting rational thinking.

http://www.google.com/advanced_group_search?hl=en

--
ha
Anonymous
September 26, 2005 7:57:22 AM

Archived from groups: rec.audio.pro (More info?)

.....except, of course, that the signals that appear are confused with
those with which they are aliased. To make this a useful approach for
recording e.g., 160 kHz signals at a 44.1 kHz (typo above of '441 kHz',
yet another possibility!), one must bandpass around 160kHz with a
bandwidth of ~20 kHz to remove all spectral content from the
frequencies below and above the band of interest. But seems like that
might work well for bats.

dhs
Anonymous
September 26, 2005 9:19:16 AM

Archived from groups: rec.audio.pro (More info?)

Chevdo wrote:

> but I record at the highest resolution possible with the gear I
> have.

Marketing departments love people like you.

> Unfortunately, I can't go higher than 44.1. If I could I would go to
> 96, and if I could I would go to 192.

Fortunately, you're very lucky. You aren't wasting money on things that
you haven't heard and don't know will give you any real advantage.
However, I really think you should buy a 192 kHz A/D and D/A converter
and start doing some experimenting at 192 kHz. I suspect you'll hear a
difference and then you'll fee comfortable with your decision. But then
try 96, and even 44.1 again.

> I think some of the people who don't own
> hardware capable of the higher sample rates and bitrates experience cognitive
> dissonance which causes them to claim that anything above whatever sample
> an/or bit rate they can do is superfulous.

I don't own a car that can go 140 mph. I don't think it's suprerfluous
that I believe it's not necessary for a person, unless he's going to
race, to have a car that can go 140 mph. Are you planning to race
anyone at sample rates?

> I think even some people who
> eventually buy interfaces capable of higher rates still record at the lower
> rate they're used to because the cognitive dissonance that made them decide
> it
> was as good as it gets back when they only had an interface capable of it, is
> so strong.

Or maybe it's because they don't hear any practical advantage to using
more disk space and other resources. There may not be a dissonance at
all. Are you now an expert in pseudopsychoacoustics?

> I think if everyone were capable of thinking rationally, everyone would agree
> that recording at the highest rate possible is preferable. Afterall, how do
> we
> know that 20 years from now we will meet an alien race of UFO travellers who
> can hear way beyond human hearing?? And what if they want to rock???

What difference will that make? In fact where will those frequencies
come from? What mics are you using that are capable of usable frequency
response above 25 kHz?
Anonymous
September 26, 2005 11:57:11 AM

Archived from groups: rec.audio.pro (More info?)

"Chevdo" <chev@dont.com> wrote:
>
> I think if everyone were capable of thinking rationally

Yeah, yeah, yeah... you're the only one in the world who can think
straight, and everyone else is an idiot. We get it already. Can you
drop the pretentious, self-indulgent attitude soon? It muddies up your
posts and makes it too easy to miss points that actually make an ounce
of sense, like the following:



> everyone would agree that recording at the highest rate possible is
> preferable.

That's a reasonable position. In certain cases, like archiving, I might
agree, *assuming* that the converter was as accurate at high rates as at
lower ones. Like Mike illustrated, some converters are actually more
accurate at 48 than 96 or 192. I'd also rather convert at 48K with a
Lucid converter than at 192K with an M-Audio card.

In some cases, practical realities outweigh intangible advantages.
Doubling the sample rate doubles hard drive throughput requirements, and
increases demands on the DSP system for signal processing. In a busy
mix, high sample rates might limit processing capability or track
capacity. In that case, reducing the sample rate may increase creative
possibilities. Then it comes back to whether the difference can even be
heard, and if so, is it better enough to warrant the trade-offs?

So, there *are* reasons not to just automatically dial up the highest
rate on the box.

--
"It CAN'T be too loud... some of the red lights aren't even on yet!"
- Lorin David Schultz
in the control room
making even bad news sound good

(Remove spamblock to reply)
Anonymous
September 26, 2005 12:01:45 PM

Archived from groups: rec.audio.pro (More info?)

Don Pearce wrote:
> On 26 Sep 2005 03:57:22 -0700, dhs wrote:
>
> > ....except, of course, that the signals that appear are confused with
> > those with which they are aliased. To make this a useful approach for
> > recording e.g., 160 kHz signals at a 44.1 kHz (typo above of '441 kHz',
> > yet another possibility!), one must bandpass around 160kHz with a
> > bandwidth of ~20 kHz to remove all spectral content from the
> > frequencies below and above the band of interest. But seems like that
> > might work well for bats.
> >
> > dhs
>
> No, there is no reason for any confusion. As long as the correct filtering
> is present to select the correct alias (the baseband one for normal audio)
> then the exact frequencies are uniquely defined. As you say, in this case
> that would be a bandpass filter around 160kHz.
>

We agree, I believe -- my note was to underscore the need for the
'correct filtering' you mention.

dhs

> Signals don't get alised with each other. That isn't how it works.
>
> d
Anonymous
September 26, 2005 1:33:40 PM

Archived from groups: rec.audio.pro (More info?)

david correia <cassette1@comcast.net> wrote:
>
>Scott, at AES listen to the URS Neve plugs. They are the most analog
>sounding tdm eq's I've heard.

I will. The Oxford EQ comes pretty close, too, but I will give the stuff
a listen.
--scott


--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Anonymous
September 26, 2005 1:47:47 PM

Archived from groups: rec.audio.pro (More info?)

Matrixmusic wrote:
> CAn anyone tell the differnce?

For music, no.

As Arny pointed out, recording music at a sample rate of 192KHz is
firmly in the realms of diminishing returns.

However, for specialist applications that require recording sounds above
the limit of human hearing, then 192KHz is not only beneficial, but
essential.

There was a recent discussion on alt.sci.physics.acoustics about
recording the echo location signals of bats and then using time
stretching on playback to bring the sound down into the audible human
range. It is thought that certain species of bat can emit sounds in
excess of 160KHz, therefore not even a sample rate of 192KHz will be enough.

£0.025

Chris W

--
The voice of ignorance speaks loud and long,
But the words of the wise are quiet and few.
---
Anonymous
September 26, 2005 1:55:23 PM

Archived from groups: rec.audio.pro (More info?)

On Mon, 26 Sep 2005 09:47:47 +0100, Chris Whealy wrote:

> Matrixmusic wrote:
>> CAn anyone tell the differnce?
>
> For music, no.
>
> As Arny pointed out, recording music at a sample rate of 192KHz is
> firmly in the realms of diminishing returns.
>
> However, for specialist applications that require recording sounds above
> the limit of human hearing, then 192KHz is not only beneficial, but
> essential.
>
> There was a recent discussion on alt.sci.physics.acoustics about
> recording the echo location signals of bats and then using time
> stretching on playback to bring the sound down into the audible human
> range. It is thought that certain species of bat can emit sounds in
> excess of 160KHz, therefore not even a sample rate of 192KHz will be enough.
>
> £0.025
>
> Chris W

A bit light on sampling theory here. The unambiguous bandwidth you can use
is limited to less than half the sampling rate (or the Nyquist rate). But
that bandwidth can be at any multiple of the Nyquist rate you want - you
just need to filter appropriately.

So it is perfectly possible to record frequencies 0f 160kHz with a sampling
rate of 441.kHz. It is called an alias response, and is just as good and
valid as the baseband response.

d
Anonymous
September 26, 2005 2:43:31 PM

Archived from groups: rec.audio.pro (More info?)

Lorin David Schultz <Lorin@DAMNSPAM!v5v.ca> wrote:

> In some cases, practical realities outweigh intangible advantages.
> Doubling the sample rate doubles hard drive throughput requirements, and
> increases demands on the DSP system for signal processing. In a busy
> mix, high sample rates might limit processing capability or track
> capacity. In that case, reducing the sample rate may increase creative
> possibilities. Then it comes back to whether the difference can even be
> heard, and if so, is it better enough to warrant the trade-offs?

So, if tracking at 192k limits the total number of tracks you can
record, and thereby prevents the arhythmic lead singer from recording
that ill-advised tambourine overdub, then the higher sampling rate will
offer a tangible sonic improvement over the lower sample rate.

ulysses
Anonymous
September 26, 2005 3:16:43 PM

Archived from groups: rec.audio.pro (More info?)

"Chevdo" <chev@dont.com> wrote in message
news:ehFZe.249918$9A2.192816@edtnps89...

>. I think some of the people who don't own
> hardware capable of the higher sample rates and bitrates experience
> cognitive
> dissonance which causes them to claim that anything above whatever sample
> an/or bit rate they can do is superfulous.

And vice-versa.

I can't tell any difference with full-range acoustic music on my Audiophile
2496 between 44K1 and 96K.

geoff
Anonymous
September 26, 2005 3:16:44 PM

Archived from groups: rec.audio.pro (More info?)

In article <wdGZe.14301$iM2.1183580@news.xtra.co.nz>,
gwood@nospam-audioproducts.co.nz says...
>
>
>"Chevdo" <chev@dont.com> wrote in message
>news:ehFZe.249918$9A2.192816@edtnps89...
>
>>. I think some of the people who don't own
>> hardware capable of the higher sample rates and bitrates experience
>> cognitive
>> dissonance which causes them to claim that anything above whatever sample
>> an/or bit rate they can do is superfulous.
>
>And vice-versa.
>
>I can't tell any difference with full-range acoustic music on my Audiophile
>2496 between 44K1 and 96K.
>

except the cognitive dissonance I was describing has nothing to do with hearing
a difference or not, so vice-versa doesn't really apply.

Anyway as I've pointed out before, anyone who has been around music for any
significant length of time (ie all of us) probably has destroyed their hearing
by now. So, depending on whether you are recording for playback to yourself
only, or to other people, whether you can hear a difference may make a
difference or it may not. Since one doesn't know whether anyone else can hear
a difference, it would be rational to err on the side of caution and use the
highest rates possible, just in case.

Besides doesn't this also have a lot to do with the interface being used?
192khz and 44.1khz are going to sound a lot different depending on the quality
of the playback device. Some devices may be able to sound better at higher
rates than others can.
Anonymous
September 26, 2005 3:16:44 PM

Archived from groups: rec.audio.pro (More info?)

Geoff@work <gwood@nospam-audioproducts.co.nz> wrote:
>
>I can't tell any difference with full-range acoustic music on my Audiophile
>2496 between 44K1 and 96K.

Since my monitors are down 3 dB at 16 KHz and start dropping pretty fast
above that, I probably wouldn't hear a difference even if there was one.

Then again, the ATR-100 is flat to 35 KHz....
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Anonymous
September 26, 2005 3:16:45 PM

Archived from groups: rec.audio.pro (More info?)

>>Anyway as I've pointed out before, anyone who has been around music for
>>any significant length of time (ie all of us) probably has destroyed their
>>hearing by now.

Depends on the kind of music. Most of the varieties I listen to, no.
Heavy metal, definitely.
Anonymous
September 26, 2005 3:16:45 PM

Archived from groups: rec.audio.pro (More info?)

Chevdo <chevdo@chevdont.com> wrote:
>
>Besides doesn't this also have a lot to do with the interface being used?
>192khz and 44.1khz are going to sound a lot different depending on the quality
>of the playback device. Some devices may be able to sound better at higher
>rates than others can.

Right, this is why you can't make blanket statements about one rate
necessarily being better than another. The implementation issues swamp
everything else right now.

Hell, on the SV3700, you can tell the difference between 44.1 and 48
ksamp/sec rates, which should definitely be inaudible. Clearly a severe
implementation problem, and by no means the only one out there.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Anonymous
September 26, 2005 3:16:45 PM

Archived from groups: rec.audio.pro (More info?)

<chevdo@chevdont.com> wrote:

> Besides doesn't this also have a lot to do with the interface being used?
> 192khz and 44.1khz are going to sound a lot different depending on the
> quality
> of the playback device. Some devices may be able to sound better at higher
> rates than others can.

And some devices (both record and playback) may be able to sound better
at lower sample rates than at higher sample rates. If the system works
better at a lower sample rate, which if you've been following this
thread you might start to believe, then the only reason to use the
higher rates would be for the additional HF information. But the rest
of your recording chain more than likely has already excluded all that
HF information, so you're not recording it anyway. Therefore, the best
sample rate to use will be the one that your equipment executes most
successfully.

ulysses
Anonymous
September 26, 2005 4:01:15 PM

Archived from groups: rec.audio.pro (More info?)

On 26 Sep 2005 03:57:22 -0700, dhs wrote:

> ....except, of course, that the signals that appear are confused with
> those with which they are aliased. To make this a useful approach for
> recording e.g., 160 kHz signals at a 44.1 kHz (typo above of '441 kHz',
> yet another possibility!), one must bandpass around 160kHz with a
> bandwidth of ~20 kHz to remove all spectral content from the
> frequencies below and above the band of interest. But seems like that
> might work well for bats.
>
> dhs

No, there is no reason for any confusion. As long as the correct filtering
is present to select the correct alias (the baseband one for normal audio)
then the exact frequencies are uniquely defined. As you say, in this case
that would be a bandpass filter around 160kHz.

Signals don't get alised with each other. That isn't how it works.

d
Anonymous
September 26, 2005 7:37:08 PM

Archived from groups: rec.audio.pro (More info?)

Chevdo wrote:

> I rarely use mics at all, and often I am generating my signals from softsynths,
> so they never enter an A/D stage.

Plop! So what frequencies are you generating with these synths,
anyway? And what are you listening to them on?

> Besides, the fact that increasing
> sample-rate allows one to record higher frequencies is an incidental benefit to
> recording at higher sample-rates. A higher rate captures a more accurate
> waveform, and a finer resolution between samples means that the waveform isn't
> subject to as much interpolation.

My, you have so much to learn. I won't tutor you though since you
apparently don't care about facts. The short answer is that you're
quite wrong about this perceived more accurate waveform and finer
resolution. You get that with more bits, not a higher sample rate. I
know that what you think is true is intuitive, but it's false.

I'd suggest that you investigate 32-bit recording if you want better
accuracy and resolution.

> I assume this is why people notice an
> audible difference between higher and lower sample rates, not because of any
> difference in frequency content.

And your assumption would be wrong. As would your belief that "people"
notice an audible difference between higher and lower sample rates,
unless you're talking about 12 kHz vs. 44.1 kHz.
Anonymous
September 26, 2005 8:09:36 PM

Archived from groups: rec.audio.pro (More info?)

On 26 Sep 2005 08:01:45 -0700, dhs wrote:

> Don Pearce wrote:
>> On 26 Sep 2005 03:57:22 -0700, dhs wrote:
>>
>>> ....except, of course, that the signals that appear are confused with
>>> those with which they are aliased. To make this a useful approach for
>>> recording e.g., 160 kHz signals at a 44.1 kHz (typo above of '441 kHz',
>>> yet another possibility!), one must bandpass around 160kHz with a
>>> bandwidth of ~20 kHz to remove all spectral content from the
>>> frequencies below and above the band of interest. But seems like that
>>> might work well for bats.
>>>
>>> dhs
>>
>> No, there is no reason for any confusion. As long as the correct filtering
>> is present to select the correct alias (the baseband one for normal audio)
>> then the exact frequencies are uniquely defined. As you say, in this case
>> that would be a bandpass filter around 160kHz.
>>
>
> We agree, I believe -- my note was to underscore the need for the
> 'correct filtering' you mention.
>

Ah - OK. I misunderstodd you.

d
!