The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Gear for sale     Latest  Trending
192khz, 96khz, 48khz. I hear the difference.
Old 4 weeks ago
  #511
Lives for gear
 
norfolk martin's Avatar
 

Quote:
Originally Posted by dinococcus View Post
No one for 176 kHz?
I found the sweet spot to be around 184.3 KHz Everything else sound terrible.
Old 4 weeks ago
  #512
Gear Guru
An engineer I'm working with, basically said good luck hearing any real difference at a higher sample rate recorded and played back. It might make your plug ins work better, but unless you're doing high level stuff, you probably don't need the file size. Yes for 192K for efx so you can play at half speed. FWIW we produce broadcast and do a lot of radio and mix to pix......

The only reason I'm pointing this out, is for bedroom guys that may feel compelled to buy gear to allow higher rates. There are a lot of other areas that will have more impact on sound. Totally my opinion, but getting into the rarefied kind of territory like tube rolling, cabling, and top class conversion..... handling big files can be cumbersome....
Old 4 weeks ago
  #513
Lives for gear
 
IanBSC's Avatar
Quote:
Originally Posted by sax512 View Post
But the fact that he also happens to be a designer of actual converters is what sets him apart from the hordes of engineers that also know the theory.
For example, I was very curious to see why he said that oversampling to x2 makes things easier, but taking it to x4 makes them harder.
Here is an example: I couldn't find the specific passage so I'm assuming he is talking about oversampling in a converter and not a plugin. Virtually all modern converters oversample at 128 or 256 times, WAY beyond x2. Lavry doesn't like how 192khz effects HIS converters. And it's a whole different question whether or not his converters sound the best...I've found them to be very good, but I don't favor them above all others.

Ironically, Lavry makes a case for 96khz, but he tends to get cited most often by advocates of 44.1.

Quote:
Originally Posted by sax512 View Post
This sea of meaningless opinions is, I believe, the reason why guys like Lavry don't post on GS anymore. Congratulations, I guess?
This is true AND its also why Greg Wells doesn't post here. He started a thread about how claimed his Antelope clock improved the sound of his mix and people came out of the woodwork calling him a fraud and a shill, or simply deluded by his gear purchase.

Dave Derr made a large contribution to this thread early on, and he isn't around anymore either.

Quote:
Originally Posted by sax512 View Post
Back to latency. It doesn't change with higher sample rates, because at say double the rate, one needs double the filter taps to maintain the same quality of filter.
So the real thing to get hung up on is the -3dB at 20 kHz, with 44.1 kHz sampling rate.
I think you'll find most converters have lower latency with higher sample rates.


Quote:
Originally Posted by sax512 View Post
Sure, if you move the filter up in frequency you get flat at 20 kHz, but you also make the system deal with more content that WILL cause distortion in the audio band because of intermodulation products caused by the fact that your DAC can be made in China, the US, the UK or whatever, but it's not made in Wakanda (that movie sucked, by the way).
And yet nobody has definitively encountered this distortion in 20 years. I concede that it is possible that your analog chain could have problems if you have tons of ultrasonic junk, but where is the actual evidence that this has been a problem? I've done a number of projects at 192khz and I haven't run across it. The only real hazard I found was older converters with less precise clocks are a bit less defined at these rates, which seems to be a jitter issue not intermodulation.
Old 4 weeks ago
  #514
Gear Maniac
 

Quote:
Originally Posted by IanBSC View Post
Here is an example: I couldn't find the specific passage so I'm assuming he is talking about oversampling in a converter and not a plugin. Virtually all modern converters oversample at 128 or 256 times, WAY beyond x2. Lavry doesn't like how 192khz effects HIS converters. And it's a whole different question whether or not his converters sound the best...I've found them to be very good, but I don't favor them above all others.

Ironically, Lavry makes a case for 96khz, but he tends to get cited most often by advocates of 44.1.



This is true AND its also why Greg Wells doesn't post here. He started a thread about how claimed his Antelope clock improved the sound of his mix and people came out of the woodwork calling him a fraud and a shill, or simply deluded by his gear purchase.

Dave Derr made a large contribution to this thread early on, and he isn't around anymore either.



I think you'll find most converters have lower latency with higher sample rates.

Let's differentiate between DACs and ADCs.

ADCs:
The MHz input stage of a AD converter works at low bit resolution. It is just an intermediate step to make it so that the analog anti-aliasing filter that precedes it doesn't need to have a steep roll-off.
That being said, while it is true that (in most cases at least) all sample rates that are multiples of each other, 44.1-88.2-etc vs. 96-192-etc, have the same 'father', it is also true that they start to differ depending on the filter on the decimation stage that brings the digital signal to the final sample frequency required. That filter is at the father's sample rate, but it has (or should have) different cut off frequency points (half of the final sample rate required).
To filter at lower cut off frequency one needs more taps (because of less wide transition band), so more latency.
I guess if you need to do stuff on the fly like in the case of sound reinforcement every bit of latency saved counts. That's hardly a situation where the most accuracy of sound is the main goal, but fine, let's save that fraction of ms. Yay!

Where people get hung up is the mixing phase (as in studio mix, not live mixing console). Higher sample rate means more space for the harmonics created by the plug in to 'live' without getting in the aliasing zone. Ok... Oversample, then! You can do that just fine with the addition of 0s in between samples. The low pass filter in that stage (that I previously forgot to specifically mention) is the one that can be made easy on the CPU, because of the signal having 0s in between the original sample frequency values. After that, you have even more space for the harmonics to live in, and you will have generated less of them.

BUT... it does depend on the design of the plug in. If it's not done right, it might very well be that lower sample rates will sound like crap, after you put a few plug ins in series.

DACs:
Same oversampling principles.
You start with low sample frequency, you oversample, you filter, you feed it to the DAC stage that can have a relaxed analog filter at the output stage.
Again, the filter can be made easy on the DACs CPU because it has samples known to be 0 in between 'real' samples, so it can be done faster than with a random signal at the oversampled rate.
Not that it matters, because when you are listening, what's the hurry that you can't wait 0.5 sec (which is a loooong time, practically overkill) to make sure the signal is filtered right? Unless you are in the sound reinforcement situation above, which is not one that warrants accuracy above all else.

So what's the point of going above 44.1 (or 48)? Only in the case one NEEDS low latency. In that case the digital filter can be made more 'relaxed', which means less taps, which means less latency.
I can't think of any situation where that need coexists with maximum transparency and purity of sound. And for sure none of the people that are saying that 92 kHz sounds better than 44.1 kHz have done their tests in such a situation.

To sum this up:
I'm not saying most consumer or even some pro products actually take the time to let the digital filters do things optimally.
I'm willing to bet there are DACs (and ADCs, and plug ins) that, for the sake of low latency, will do it less than egregiously.

But is it an INTRINSIC flaw of the lower sample rate? NO!
Could they take their f***ing time and do it right at the lower sample rate in most cases, and absolutely in EVERY SINGLE ONE of the cases that people posting on this thread have employed them? YES!


Quote:
Originally Posted by IanBSC View Post
And yet nobody has definitively encountered this distortion in 20 years. I concede that it is possible that your analog chain could have problems if you have tons of ultrasonic junk, but where is the actual evidence that this has been a problem? I've done a number of projects at 192khz and I haven't run across it. The only real hazard I found was older converters with less precise clocks are a bit less defined at these rates, which seems to be a jitter issue not intermodulation.
Of course they have! Non linearities in DACs and amps (and don't even get me started on speakers) are measured all the time. Mostly they are bundled up in figures that are somewhat indirect, like THD and the likes, but non linearity is absolutely measured for audio components. I'm surprised you didn't know about that, to be honest.
You have definitely run into it in your 192 kHz projects. It could have been measured, although it's program specific and quite hard to do. Whether you might have actually heard it, or even preferred it, is a whole different subject.
Old 4 weeks ago
  #515
Lives for gear
Quote:
Originally Posted by sax512 View Post
"So what's the point of going above 44.1 (or 48)? Only in the case one NEEDS low latency. In that case the digital filter can be made more 'relaxed', which means less taps, which means less latency.
I can't think of any situation where that need coexists with maximum transparency and purity of sound. And for sure none of the people that are saying that 92 kHz sounds better than 44.1 kHz haven't done their tests in such a situation."
I am the case that needs that. I am trying to blend together recorded and triggered drums in real time (While recording the analog drums). Have to have it all at once to get it done.
Old 4 weeks ago
  #516
Lives for gear
 
IanBSC's Avatar
Quote:
Originally Posted by sax512 View Post
Let's differentiate between DACs and ADCs.

ADCs:
The MHz input stage of a AD converter works at low bit resolution. It is just an intermediate step to make it so that the analog anti-aliasing filter that precedes it doesn't need to have a steep roll-off.
That being said, while it is true that (in most cases at least) all sample rates that are multiples of each other, 44.1-88.2-etc vs. 96-192-etc, have the same 'father', it is also true that they start to differ depending on the filter on the decimation stage that brings the digital signal to the final sample frequency required. That filter is at the father's sample rate, but it has (or should have) different cut off frequency points (half of the final sample rate required).
To filter at lower cut off frequency one needs more taps (because of less wide transition band), so more latency.
I guess if you need to do stuff on the fly like in the case of sound reinforcement every bit of latency saved counts. That's hardly a situation where the most accuracy of sound is the main goal, but fine, let's save that fraction of ms. Yay!

Where people get hung up is the mixing phase (as in studio mix, not live mixing console). Higher sample rate means more space for the harmonics created by the plug in to 'live' without getting in the aliasing zone. Ok... Oversample, then! You can do that just fine with the addition of 0s in between samples. The low pass filter in that stage (that I previously forgot to specifically mention) is the one that can be made easy on the CPU, because of the signal having 0s in between the original sample frequency values. After that, you have even more space for the harmonics to live in, and you will have generated less of them.

BUT... it does depend on the design of the plug in. If it's not done right, it might very well be that lower sample rates will sound like crap, after you put a few plug ins in series.

DACs:
Same oversampling principles.
You start with low sample frequency, you oversample, you filter, you feed it to the DAC stage that can have a relaxed analog filter at the output stage.
Again, the filter can be made easy on the DACs CPU because it has samples known to be 0 in between 'real' samples, so it can be done faster than with a random signal at the oversampled rate.
Not that it matters, because when you are listening, what's the hurry that you can't wait 0.5 sec (which is a loooong time, practically overkill) to make sure the signal is filtered right? Unless you are in the sound reinforcement situation above, which is not one that warrants accuracy above all else.

So what's the point of going above 44.1 (or 48)? Only in the case one NEEDS low latency. In that case the digital filter can be made more 'relaxed', which means less taps, which means less latency.
I can't think of any situation where that need coexists with maximum transparency and purity of sound. And for sure none of the people that are saying that 92 kHz sounds better than 44.1 kHz haven't done their tests in such a situation.

To sum this up:
I'm not saying most consumer or even some pro products actually take the time to let the digital filters do things optimally.
I'm willing to bet there are DACs (and ADCs, and plug ins) that, for the sake of low latency, will do it less than egregiously.

But is it an INTRINSIC flaw of the lower sample rate? NO!
Could they take their f***ing time and do it right at the lower sample rate in most cases, and absolutely in EVERY SINGLE ONE of the cases that people posting on this thread have employed them? YES!




Of course they have! Non linearities in DACs and amps (and don't even get me started on speakers) are measured all the time. Mostly they are bundled up in figures that are somewhat indirect, like THD and the likes, but non linearity is absolutely measured for audio components. I'm surprised you didn't know about that, to be honest.
You have definitely run into it in your 192 kHz projects. It could have been measured, although it's program specific and quite hard to do. Whether you might have actually heard it, or even preferred it, is a whole different subject.
I agree almost completely with your first response. However, I also suspect that Dave Derr and others may be right that transient edges shorter than the 250us (or so) you get with 44.1 sampling/filtering may be important to reproduce room reflections, spacial cues and other things accurately.

For the second point, if a distortion doesn't show up in a device's specs, doesn't show up in my metering, doesn't effect my playback equipment, and can't be heard identified audibly, doesn't sound bad, how would anyone know it exists, and what's the fuss?

Last edited by IanBSC; 4 weeks ago at 01:55 AM.. Reason: Including ringing 44.1 impulse length is about 250us on each side with FIR
Old 4 weeks ago
  #517
Motown legend
 
Bob Olhsson's Avatar
 

Professionals work with the best of what's available rather than waiting around for better products that are theoretically possible. I certainly didn't want to buy a new computer so that I could work at 96k but my results in final 44.1 files and even MP3s demonstrated to me that it sounded better. Quality upsampling within plug-ins is going to use at least as much computer power so the only thing being saved is disk space which is dirt cheap at this point.
Old 4 weeks ago
  #518
Gear Maniac
 

Quote:
Originally Posted by IanBSC View Post
I agree almost completely with your first response. However, I also suspect that Dave Derr and others may be right that transient edges shorter than the 100us (or so) you get with 44.1 sampling/filtering may be important to reproduce room reflections, spacial cues and other things accurately.

For the second point, if a distortion doesn't show up in a device's specs, doesn't show up in my metering, doesn't effect my playback equipment, and can't be heard identified audibly, doesn't sound bad, how would anyone know it exists, and what's the fuss?
Transient edges and frequency band are related. If you can't hear above 20 kHz, you only hear the frequency content under 20 kHz that makes up for those transient edges. Just because there is a square wave floating around in the air, for example, which there isn't because it doesn't exist in nature, doesn't mean that that's what your eardrums perceive. They only perceive its harmonic content up to 20 kHz (even less than that, to tell you the truth).

For the second point, the distortion does show up in a device's specs. THD. Total Harmonic Distortion. It's something provided by virtually every half decent manufacturer. Why do you keep saying this distortion doesn't exist? That is just silly.
If you can't hear it that's another matter. But then again you are the one pushing for recording frequency content that nobody can possibly hear.. So which is it? At least the THD distortion can absolutely be heard.
Old 4 weeks ago
  #519
Gear Maniac
 

Quote:
Originally Posted by Bob Olhsson View Post
Professionals work with the best of what's available rather than waiting around for better products that are theoretically possible. I certainly didn't want to buy a new computer so that I could work at 96k but my results in final 44.1 files and even MP3s demonstrated to me that it sounded better. Quality upsampling within plug-ins is going to use at least as much computer power so the only thing being saved is disk space which is dirt cheap at this point.
They are not just theoretically possible. They are out there and have been for a long time.

While I have no problem believing that some plug ins don't do things right when it comes to take care of all sampling rate at their input, I find it extremely hard to believe that none of them does. Or even that the majority doesn't.

When you say better sound, that is not a scientific statement. First of all, you would need to do ABX tests. Second, just because it sounds different, or even more pleasing, doesn't necessarily mean that it sounds more accurate.
I get that you are only interested in what's pleasing, given your job. But in a conversation about the merit of higher sample rates vs. lower ones, I think we should stick to non subjective stuff.
Old 4 weeks ago
  #520
Gear Maniac
 

Quote:
Originally Posted by elegentdrum View Post
I am the case that needs that. I am trying to blend together recorded and triggered drums in real time (While recording the analog drums). Have to have it all at once to get it done.
Great. What S/N do you need? That will tell you the filter's rejection and how many taps the filter needs to achieve that.
After some math, you will come up with the lowest possible sample rate you can use.
Which might seem like an exercise in futility, IF you don't take into account the draw backs of having extra, unnecessary frequency content being processed in your audio chain. That content is not just a nice, innocuous gravy on top as many here seem to think. It better be filtered out, or it will create intermodulation products in the audio band.

Just to be clear, 44.1 is not a magic sample rate above which intermodulation products happen. Those happen all the time among all frequencies being fed to the audio chain.
While you can't remove the ones caused by the frequency content in the audio band without messing with what you can and want to hear, you absolutely should remove all the extra unnecessary stuff above the hearing range (and below).
Or don't, but know that there will be added distortion.
Sounds still OK to you? Sounds even better? Go for it, then. That is part of the artistic process of creating music.

I am talking about the accurate recording and reproduction of sound waves, which requires a more scientific approach.
Old 4 weeks ago
  #521
Lives for gear
 
IanBSC's Avatar
Quote:
Originally Posted by sax512 View Post
For the second point, the distortion does show up in a device's specs. THD. Total Harmonic Distortion. It's something provided by virtually every half decent manufacturer. Why do you keep saying this distortion doesn't exist? That is just silly.
If you can't hear it that's another matter. But then again you are the one pushing for recording frequency content that nobody can possibly hear.. So which is it? At least the THD distortion can absolutely be heard.
All my converters record at 96 and 192, and one even records at 384khz. The highest distortion converter is -105db THD and the lowest is -122dbTHD+N. If I am recording 384khz at spec'd -122db THD+N, better than most converters DNR, where is the distortion?

As I've explained before, the issue is not frequency content as it is temporal accuracy.
Old 4 weeks ago
  #522
Lives for gear
Here is the math that's my problem.

Triggers want 3ms to decide how loud they have been hit.
Every 1ms is about 1' of distance precieved.
So at 3ms, the drums sound 3' away....ok

At 4.5ms, this is the furthest away that feels like a real kit. That leaves 1.5 ms to send MIDI into BFD and get the sound out to outbaord gear like the mic's and back into the DAW for blending with the mic's comming in.

So I want 2 round trips of audio in 1.5ms. not possible, but at 96K and RME PCI drivers/MADI I can get close.

Past 96K, the comuputer has to work to hard for 32 softsynth BFD tracks and 32 mic's going at once for a blend of electronic and acoustic drums along with the rest of the band.
Old 4 weeks ago
  #523
Gear Maniac
 

Quote:
Originally Posted by IanBSC View Post
All my converters record at 96 and 192, and one even records at 384khz. The highest distortion converter is -105db THD and the lowest is -122dbTHD+N. If I am recording 384khz at spec'd -122db THD+N, better than most converters DNR, where is the distortion?

As I've explained before, the issue is not frequency content as it is temporal accuracy.
THD is just INDICATIVE of the distortion power you actually end up with, which is signal dependent.
It is usually calculated when the signal being processed is 2 single sinewaves (or even 1), but there is no standard and every manufacturer will perform the test under the conditions that will make that figure show up as low as possible.
Nothing wrong with that. Real intermodulation distortion is probably impossible to measure directly, because it varies with the signal (that's not the same thing as saying it doesn't exist, goes without saying). But one just needs to be aware of this fact. It is just a single number, that bundles up all distortion under very specific test conditions that are NEVER experienced in real life applications. The lower the THD, the better the component... Probably. It depends on what really happens under real use of the component.
But as I said, you need to worry A LOT more about non linearities of your speakers.

I have no idea what temporal accuracy you are referring to. Can you explain?
Old 4 weeks ago
  #524
Gear Maniac
 

Quote:
Originally Posted by elegentdrum View Post
Here is the math that's my problem.

Triggers want 3ms to decide how loud they have been hit.
Every 1ms is about 1' of distance precieved.
So at 3ms, the drums sound 3' away....ok

At 4.5ms, this is the furthest away that feels like a real kit. That leaves 1.5 ms to send MIDI into BFD and get the sound out to outbaord gear like the mic's and back into the DAW for blending with the mic's comming in.

So I want 2 round trips of audio in 1.5ms. not possible, but at 96K and RME PCI drivers/MADI I can get close.

Past 96K, the comuputer has to work to hard for 32 softsynth BFD tracks and 32 mic's going at once for a blend of electronic and acoustic drums along with the rest of the band.
Provided that the S/N required can't be achieved at lower sample rates, sounds like you have your min sample rate figured out, then. No reason to go above that, even if you had less mics, is my point.
Old 4 weeks ago
  #525
Motown legend
 
Bob Olhsson's Avatar
 

The theory is math, practice is entirely subjective!

Again, this is all about the minimizing artifacts of filters within the audible range below 20k. Aliasing is especially troubling because it masks detail in the upper midrange as opposed to being an identifiable distortion. I've literally heard brushes on a snare drum disappear.
Old 4 weeks ago
  #526
Gear Maniac
 

Quote:
Originally Posted by Bob Olhsson View Post
The theory is math, practice is entirely subjective!

Again, this is all about the minimizing artifacts of filters within the audible range below 20k. Aliasing is especially troubling because it masks detail in the upper midrange as opposed to being an identifiable distortion. I've literally heard brushes on a snare drum disappear.
Get a better plug in..
And, again, if this is all about minimizing artifacts below 20 kHz, why is intermodulation not a concern then?
You mean to tell me intermodulation is less audible than aliasing? Fine. On YOUR rig, maybe it is (I strongly doubt it). But it's not because 44.1 is inferior to any higher sample rate.
I really am running out of different ways to try to deliver this concept.
Old 4 weeks ago
  #527
Lives for gear
Quote:
Originally Posted by Bob Olhsson View Post
The theory is math, practice is entirely subjective!

Again, this is all about the minimizing artifacts of filters within the audible range below 20k. Aliasing is especially troubling because it masks detail in the upper midrange as opposed to being an identifiable distortion. I've literally heard brushes on a snare drum disappear.
New ice cream, the aliasing.
Old 4 weeks ago
  #528
Lives for gear
Quote:
Originally Posted by sax512 View Post
... But it's not because 44.1 is inferior to any higher sample rate.
He never said it was. It's the bandwidth for the anti-aliasing filters to work. When so restricted, to 1/10 of an octave in the case of 44.1KHz, it becomes very difficult to cut everything above the Nyquist and leave <20KHz untouched.

If the perfect(no latency, no ringing, no phase shift, no frequency bumps or dips), real-time, brickwall filter could be invented, 44.1 would be just fine for recording.
Old 4 weeks ago
  #529
Gear Maniac
 

Quote:
Originally Posted by dinococcus View Post
New ice cream, the aliasing.
Come on, man. You're still talking to a living legend.
I mean.. sure I may come up with the occasional s**t and f**k, but there is no reason for mocking each other.
I honestly believe the misunderstanding lies, when you get down to it, to simply looking at the issue from a different angle.
Our angle is the theory, and the fact that 44.1 kHz already has ALL the info one needs.
His angle is that sometimes to extract it it's easier to go higher in sampling rate, even at the expense of higher distortion.
To which I reply that one needs ABX to really extract any value from subjective tests.
Old 4 weeks ago
  #530
Lives for gear
I wait the jitter's lunar landing with great impatience.
Old 4 weeks ago
  #531
Gear Maniac
 

Quote:
Originally Posted by u87allen View Post
He never said it was. It's the bandwidth for the anti-aliasing filters to work. When so restricted, to 1/10 of an octave in the case of 44.1KHz, it becomes very difficult to cut everything above the Nyquist and leave <20KHz untouched.

If the perfect(no latency, no ringing, no phase shift, no frequency bumps or dips), real-time, brickwall filter could be invented, 44.1 would be just fine for recording.
If the perfect linear DAC, but MOSTLY speakers, could be invented, I wouldn't bust your balls and keep telling you that your assumption that higher sample rates will make things better for sure is not a good one.
Old 4 weeks ago
  #532
Lives for gear
Quote:
Originally Posted by sax512 View Post
If the perfect linear DAC, but MOSTLY speakers, could be invented, I wouldn't bust your balls and keep telling you that your assumption that higher sample rates will make things better for sure is not a good one.
I think the DAC side is not where the problem is. It's the ADC side. And even here, of course a higher sample rate is not automatically better. As Bob has pointed out, he's heard every sample rate sound better than another, including lower sample rates sounding better than higher ones. There's just more room for the anti-aliasing filter to work at a higher sample rate. But if it's not a well designed filter, if other components of the ADC aren't good or designed well, then sure, the higher sample rate could sound worse. But all things being equal, more room for the filter to work equals a better filter and better sound.

But I suppose you're arguing no one can hear these differences without a perfect DAC, amp and speakers. So then the remaining question is what is responsible for the more open top end at higher sample rates? I've heard the more open top end. And no one's telling me I didn't. I used to think it was the higher sample rates catching things "between the samples" of lower sample rates. I no longer believe that to be true. So what are we left with?
Old 4 weeks ago
  #533
Lives for gear
 

Quote:
Originally Posted by Bob Olhsson View Post
The theory is math, practice is entirely subjective!

Again, this is all about the minimizing artifacts of filters within the audible range below 20k. Aliasing is especially troubling because it masks detail in the upper midrange as opposed to being an identifiable distortion. I've literally heard brushes on a snare drum disappear.
Hi Bob,

If I might add to this, an often neglected part of the sampling theorem is that it relies on continuous signals, they are band limited but also infinite in length.

In practice you need a sufficient number of samples for accurate time domain reconstruction. The consequence is a sine wave can get by with lower sample rates, transient events require higher rates.

So it's not entirely about the filters, though they certainly are crucially important.

Last edited by johnnyc; 3 weeks ago at 04:07 AM..
Old 4 weeks ago
  #534
Lives for gear
Quote:
Originally Posted by u87allen View Post
... And no one's telling me I didn't...
6 o'clock am on Mission road, tomorrow. At 12 feet with tree shots cream pies.
Old 4 weeks ago
  #535
Motown legend
 
Bob Olhsson's Avatar
 

Quote:
Originally Posted by sax512 View Post
...You mean to tell me intermodulation is less audible than aliasing? Fine. On YOUR rig, maybe it is (I strongly doubt it). But it's not because 44.1 is inferior to any higher sample rate.
I really am running out of different ways to try to deliver this concept.
Are are you claiming 96k must have more intermodulation distortion than 44.1? I'm afraid I'm not understanding your concept.
Old 4 weeks ago
  #536
Gear Maniac
 

Quote:
Originally Posted by Bob Olhsson View Post
Are are you claiming 96k must have more intermodulation distortion than 44.1? I'm afraid I'm not understanding your concept.
Exactly!
Old 4 weeks ago
  #537
Gear Maniac
 

Quote:
Originally Posted by u87allen View Post
I think the DAC side is not where the problem is. It's the ADC side.
It's actually not the DAC nor the ADC. It's mostly the speakers. Those are the most non linear components of the whole audio chain. By far.

Quote:
Originally Posted by u87allen View Post
And even here, of course a higher sample rate is not automatically better. As Bob has pointed out, he's heard every sample rate sound better than another, including lower sample rates sounding better than higher ones.
Better doesn't mean more accurate, which is what a conversation about sample rate comparison should be about. The rest is subjective (and mostly unreliable) personal opinions.

Quote:
Originally Posted by u87allen View Post
There's just more room for the anti-aliasing filter to work at a higher sample rate. But if it's not a well designed filter, if other components of the ADC aren't good or designed well, then sure, the higher sample rate could sound worse. But all things being equal, more room for the filter to work equals a better filter and better sound.
That's not true at all. The higher sample rate could absolutely sound worse even if the filter and the whole ADC are an amazing design and components.
Forget the ADC (or DAC) intrinsic non linearities. I'll go with the case they are perfect.
With 44.1 they feed content up to 22kHz to the amp, with 96 kHz they feed extra stuff between 22 kHz and 48. That's a whole extra audio band of content that nobody can possibly hear, but will create problems much bigger than aliasing once it goes through the amp and (above all else) the speakers.
Think of frequency content as a mortgage. If you stop at the mere first step, you may think that getting more money from the bank is better than less.
Then you quickly realize that you have to pay interests on the money borrowed (that's the intermodulation distortion).
You want to borrow enough money so that you can buy your house (the audio band), but anything extra will get you in a worse position. Less is more, in this case.

Quote:
Originally Posted by u87allen View Post
But I suppose you're arguing no one can hear these differences without a perfect DAC, amp and speakers.
Incorrect. I'm saying you can absolutely hear the difference. You may actually even prefer the more inaccurate sound. That would be the 96 kHz. Not as much because there is such a huge difference right after the DAC (or ADC) when filtering is done on the fly between the power of increased intermodulation distortion for 96 kHz and increased aliasing for 44.1 kHz, but because in the end all that extra content ends up at the terminals of a pair of speakers.

Quote:
Originally Posted by u87allen View Post
So then the remaining question is what is responsible for the more open top end at higher sample rates?
I don't know. A few culprits are:
1. Small amounts of distortions are often perceived as more detail
2. You looking for a certain quality in the sound you're listening to will make your brain find it (self conditioning)
3. Comparison with lower sample rates done badly with specific hardware/software and following erroneous generalization

Quote:
Originally Posted by u87allen View Post
I've heard the more open top end. And no one's telling me I didn't. I used to think it was the higher sample rates catching things "between the samples" of lower sample rates. I no longer believe that to be true. So what are we left with?
See above
Old 4 weeks ago
  #538
Motown legend
 
Bob Olhsson's Avatar
 

So a converter downsampled to 96k has more IM distortion than one downsampled to 44.1? (To be clear, there has been no such thing as sampling at lower than 176k since the 1980s.) That's a bit hard to believe.
Old 4 weeks ago
  #539
Gear Maniac
 

Quote:
Originally Posted by Bob Olhsson View Post
So a converter downsampled to 96k has more IM distortion than one downsampled to 44.1? (To be clear, there has been no such thing as sampling at lower than 176k since the 1980s.) That's a bit hard to believe.
Yes. More harmonic content to interact within itself to create intermodulation products. Lavry wrote a post with a nice example in one of the first pages of this thread. You don't have to believe it. It can be demonstrated mathematically.
When the processing is non linear, feeding two sinewaves at its input of frequency A and B will result in an output that also has signals at (nA+mB) frequencies, with n and m integers from - to + infinity (obviously of diminishing amplitude, as n and m get bigger).
The more stuff at the input, the more extra signals at the output.

It's also the same principle that creates ultra sonic harmonics at the output of a compressor.
Old 4 weeks ago
  #540
Lives for gear
Quote:
Originally Posted by sax512 View Post
See above

Quote:
Better doesn't mean more accurate, which is what a conversation about sample rate comparison should be about. The rest is subjective (and mostly unreliable) personal opinions.
Yes, you're right. I should have clarified. When I say "better" I mean more accurate in this case.

Quote:
With 44.1 they feed content up to 22kHz to the amp, with 96 kHz they feed extra stuff between 22 kHz and 48. That's a whole extra audio band of content that nobody can possibly hear, but will create problems much bigger than aliasing once it goes through the amp and (above all else) the speakers.
Think of frequency content as a mortgage. If you stop at the mere first step, you may think that getting more money from the bank is better than less.
Then you quickly realize that you have to pay interests on the money borrowed (that's the intermodulation distortion).
If one records at 96 and then downsamples to 48 or 44.1 this problem would disappear, right?

Also, in order for these components to be a factor, they would have to be large enough to make the amp and/or speakers swing enough to make the non-linearities of the amp and speaker matter. I'm thinking these components will be quite small. Other components in the recording chain greatly attenuate these frequencies. And the anti-aliasing filter at 96KHz doesn't allow everything below 48KHz through. It's not a perfect, brickwall filter set at 48K.

Quote:
I don't know. A few culprits are:
1. Small amounts of distortions are often perceived as more detail
2. You looking for a certain quality in the sound you're listening to will make your brain find it (self conditioning)
3. Comparison with lower sample rates done badly with specific hardware/software and following erroneous generalization
The difference I heard was "Oh, that's how the top end is supposed to sound" So open and so much depth. The 44.1 sounded like everything above 12KHz was squashed into a tiny sonic space. Yeah, all the frequencies were there. But no definition. If I'm going to pick one of your options, I'd go with #3 . But the possibility of an anti-aliasing filter that's screwing up the top end is still on the table as far as I'm concerned.

Last edited by u87allen; 4 weeks ago at 02:52 AM..
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump