Topic: 24-Bit / 48kHz or 96kHz

[page]

Since it's been a few years since I last checked with different gear, I did a 24/28 - 24/94 recording comparison last night on a couple pieces of avantgarde classical material in a pristine acoustic.  Checking this afternoon on the big system and on my best phones (Mytek DAC>Senn HD650) I might hear a subtle difference, but can't really be sure.  Need to do some more listening and file analysis. They both sound really good.  Unfortunately it isn't something I can post.

This is something of "cheating yourself" i think. Such comparison makes sense when you sit under your phones and another person chooses the samples and you don'T know if a 96 kHZ file is running to your ears or not. I bet you will hear no difference under that scheme...

Yeah, ABX testing is a wonderful revealer.

[slowburn]

I disregard data storage as a problem (try doing some HD video and the 'waste' of 24/96 will be insignificant). I certainly can't hear the difference between 24/48 and 24/96 on most recordings. On a few I can though.

But the main reason I've used 24/96 for years and always save those raw master files is that you never know what willl happen in the future. It's kind of the same reason you should never throw out an analog master. Cause you never know what new transfer metod will arrive in the future.

Who knows what will be possible to pull out of an audio signal in the future. Who knows if those signals we can't hear today will make a difference in whatever that process might be. Maybe in 50 years everyone will be able to hear better through some ear software. Maybe they will be able to pull some kind of instrument direction out of any audio in a way we can't even think of today. Or something else that makes no sense today. Try running some of what has happened in the audio/video world in the last 25 years by someone still living in 1985.

I don't really imagine many of my recordings will be interesting to anyone 50-100 years from now but maybe there is the one. Why not give all of them the best odds.

I actually recorded almost all my DAT masters in 16/48 instead of 16/44 back in the day. While not a major difference now I'm glad I did. When I started with DATs it just made sense to me to use the best possible since I only copied to analog cassettes then. When computerburned cds came along it became a bit of a hassle but I stuck with it and eventually it stopped being a problem again.
 

[LikeASong]

Thanks to all who answered to my battery-related question
I think I will stop recording at 96kHz (keeping the 24-bit setting). From what I've read -here and in other places- it really makes not much sense recording in 96kHz when there's a million steps (poor PAs, poor cables, poor connections, poor recorder, poor transfer, poor final-listening device or conditions...too many possible failures) that can ruin the extra bit of quality. And reocrding in a lower sample rate allows for more storage and less battery consumption. So... Bye bye 96kHz

I think somebody mentioned that _if_ you were in a good studio setting where you could control things, 96/24 might make sense.  I know when I recently transferred my vinyl collection, I couldn't really hear _THAT MUCH_ of a difference between 96/24 and 48/24.  I just wasted the bandwidth and space because I could.  Last night I ran a show where I had board access and just kept everything at 48/24 - there seems to be little benefit going above that (especially with the equipment many of us run).

That's as opposite as recording in a packed venue and in ste*lth (or at least not as conveniently as on a studio) mode...

[aaronji]

I disregard data storage as a problem (try doing some HD video and the 'waste' of 24/96 will be insignificant). I certainly can't hear the difference between 24/48 and 24/96 on most recordings. On a few I can though.

Can you describe the differences you hear?  Just curious: I can't hear a difference myself.

But the main reason I've used 24/96 for years and always save those raw master files is that you never know what willl happen in the future. It's kind of the same reason you should never throw out an analog master. Cause you never know what new transfer metod will arrive in the future.

Who knows what will be possible to pull out of an audio signal in the future. Who knows if those signals we can't hear today will make a difference in whatever that process might be. Maybe in 50 years everyone will be able to hear better through some ear software. Maybe they will be able to pull some kind of instrument direction out of any audio in a way we can't even think of today. Or something else that makes no sense today. Try running some of what has happened in the audio/video world in the last 25 years by someone still living in 1985.

If you want to record at 96 kHz, by all means do it.  But don't do it because you will somehow "future-proof" your recordings.  The laws of physics and the physiology of hearing aren't going to change (well, in the case of the physiology, at least not in a meaningful time frame).  On top of that, most instruments aren't generating any energy at very high frequencies, and most mics aren't capturing what's up there anyway...

And if that is the argument, why not DSD or 192 kHz? 

Just my two cents, YMMV, etc.

[Gutbucket]

Since it's been a few years since I last checked with different gear, I did a 24/28 - 24/94 recording comparison last night on a couple pieces of avantgarde classical material in a pristine acoustic.  Checking this afternoon on the big system and on my best phones (Mytek DAC>Senn HD650) I might hear a subtle difference, but can't really be sure.  Need to do some more listening and file analysis. They both sound really good.  Unfortunately it isn't something I can post.

This is something of "cheating yourself" i think. Such comparison makes sense when you sit under your phones and another person chooses the samples and you don'T know if a 96 kHZ file is running to your ears or not. I bet you will hear no difference under that scheme...

Yeah, ABX testing is a wonderful revealer.

Agreed.  Yet my casual listening tests of the material I recorded this weekend were less about the existence of any perceptual difference at all in and of itself, and more about confirming my doubts about the significance of any difference when weighed against a 200% storage space burden increase for a near best-case scenario.  If I had been convinced I heard much of a significant difference, I'd be more motivated to make the effort to set up an ABX test to confirm or refute it.

I also say that now that my playback is to the point where I doubt it is the limiting factor in the perception of the potential differences.  The limiting factors in this case are likely to be either the limits of the recording gear and setup I used, the loss of sensitivity of my aging ears, or the hard limits of even youthful human hearing perception.  The first is easily checked in a partial sense at least by looking to see if there is information captured above ~18-20kHz in the at the 96kHz files, but I haven’t gotten around to that yet. 

One thing I don't doubt is that there was most certainly ultrasonic energy produced by the instrumentation I recorded for my tests this weekend on Friday and Saturday in the form of inharmonic high frequency energy from things like cymbals, triangles, chimes other struck metallic percussion, and harmonics from horns, possibly violins and other instruments.  And compared to this classical percussion consort type material, there is probably even more ultrasonic energy present from the cymbals and rim shots when I make on-stage recordings in close proximity to jazz drum kits, and from the often harmonically brighter timbre of jazz brass pointed mic-wards, but in both cases I’m talking about direct radiation of acoustic instruments known to produce significant enegry above 20kHz, not PA amplified/re-amplified sound.

[F.O.Bean]

I love this thread

I used to run 24/44.1, and believe it or not, got bitched at it by someone downloading my source, for not recording in 24/48

I do now record in 24/48 and will probably do so forever Besides, 24/48 is a standard for BOTH DVD-Audio and DVD-Video

[Gutbucket]

I did a 24/28 - 24/94 recording comparison last night

How typo was that? yanowatimeen. thanks for the slide. 

[TimeBandit]

I’m talking about direct radiation of acoustic instruments known to produce significant enegry above 20kHz, not PA amplified/re-amplified sound.

Thats why i mentioned on first post in that thread:

"Recording full acoustic music or nature sounds, 96 kHz makes sense, or running SBD pull if the deck if full 96 khz in the chain", or Rendering music on VST instruments.

For stack taping far away that the PA sound simply overdoes the instruments, 96 KHz makes no sense (for me)

[aaronji]

One thing I don't doubt is that there was most certainly ultrasonic energy produced by the instrumentation I recorded for my tests this weekend on Friday and Saturday in the form of inharmonic high frequency energy from things like cymbals, triangles, chimes other struck metallic percussion, and harmonics from horns, possibly violins and other instruments.  And compared to this classical percussion consort type material, there is probably even more ultrasonic energy present from the cymbals and rim shots when I make on-stage recordings in close proximity to jazz drum kits, and from the often harmonically brighter timbre of jazz brass pointed mic-wards, but in both cases I’m talking about direct radiation of acoustic instruments known to produce significant enegry above 20kHz, not PA amplified/re-amplified sound.

The ultrasonics are definitely there, but, for most instruments, only a few tenths or hundredths of a percent of their total power is above 20 kHz.  And that's generally measured quite close (a few feet in the articles I have read) and at low levels.  Cymbals, like you mentioned, are the big exception.  Rimshots and some of the percussion have a moderate proportion of their power above 20 kHz (5 - 6% for rimshots, maybe 1% for a triangle).  Horns are also up there, but still not usually more than 1%. 

Moving away from the musicians, the levels of those ultrasonics will drop pretty quickly.  If your mics, and the rest of your chain, can reproduce those frequencies with any fidelity, on-stage would be the place to record them.  Further away, a lot of that stuff probably drops down to the level of system noise.  And, as you noted, it's all moot if you're taping the PA...

Of course, nobody can hear it anyway! 

[Gutbucket]

Of course, nobody can hear it anyway!

Most likely not.  But it should be taken into consideration in the the engineering and design of recording equipment and the way it is used to best effect.  In any case, as much as I understand the arguments and theory, nothing beats testing and confirming things for oneself. And even then I still consider it a practical rather than an absolute conclusion.

For anyone interested, here is a good resource on the measured output of instruments >20kHz-

"There's Life Above 20 Kilohertz: A Survey of Musical-Instrument Spectra to 102.4 kHz"
 
At least one member of each instrument family (strings, woodwinds, brass and percussion) produces energy to 40 kHz or above, and the spectra of some instruments reach the measurement limit of 102.4 kHz. Survey includes French horn, trumpet, violin, oboe, crash cymbals, sibilant speech, claves, a drum rimshot, triangle, jangling keys, and piano. Includes short description of others' work on perception of air- and bone-conducted ultrasound, and points out that even if ultrasound be taken as having no effect on perception of live sound, yet its presence may still pose a problem to the audio equipment designer and recording engineer.

[Gutbucket]

I consider it a positive indication when the pets start turning their ears and looking around with interst when I playback some of my recordings.  I should figure out a way to ABX test the cat. 

[achalsey]

I have no stake in this and don't understand the physics in the least so its only barely comprehensible, but am curious what people have to say about this souce (http://people.xiph.org/~xiphmont/demo/neil-young.html) posted in another thread but didn't seem to get much attention.

It says recording the ultrasonics at the higher frequency is actually very slightly detrimental to average playback systems.  Though, I didn't try to really  understand what intermodulation distortion is and how its reproduced on normal playback above 48khz.

One thing I don't doubt is that there was most certainly ultrasonic energy produced by the instrumentation I recorded for my tests this weekend on Friday and Saturday in the form of inharmonic high frequency energy from things like cymbals, triangles, chimes other struck metallic percussion, and harmonics from horns, possibly violins and other instruments.  And compared to this classical percussion consort type material, there is probably even more ultrasonic energy present from the cymbals and rim shots when I make on-stage recordings in close proximity to jazz drum kits, and from the often harmonically brighter timbre of jazz brass pointed mic-wards, but in both cases I’m talking about direct radiation of acoustic instruments known to produce significant enegry above 20kHz, not PA amplified/re-amplified sound.

The ultrasonics are definitely there, but, for most instruments, only a few tenths or hundredths of a percent of their total power is above 20 kHz.  And that's generally measured quite close (a few feet in the articles I have read) and at low levels.  Cymbals, like you mentioned, are the big exception.  Rimshots and some of the percussion have a moderate proportion of their power above 20 kHz (5 - 6% for rimshots, maybe 1% for a triangle).  Horns are also up there, but still not usually more than 1%. 

Moving away from the musicians, the levels of those ultrasonics will drop pretty quickly.  If your mics, and the rest of your chain, can reproduce those frequencies with any fidelity, on-stage would be the place to record them.  Further away, a lot of that stuff probably drops down to the level of system noise.  And, as you noted, it's all moot if you're taping the PA...

Of course, nobody can hear it anyway!

[hi and lo]

I have no stake in this and don't understand the physics in the least so its only barely comprehensible, but am curious what people have to say about this souce (http://people.xiph.org/~xiphmont/demo/neil-young.html) posted in another thread but didn't seem to get much attention.

It says recording the ultrasonics at the higher frequency is actually very slightly detrimental to average playback systems.  Though, I didn't try to really  understand what intermodulation distortion is and how its reproduced on normal playback above 48khz.

Thanks for the article! I will read it later in depth, but took a quick glance and it's definitely interesting. I think I already pointed this out in this thread, but...

The 24/192kHz format has been universally panned by critics. Many will tell you that it is clearly inferior to 24/96kHz and there is some science (or at least intarwebz science) that supports the claim that 192kHz is actually pushing and exceeding the theoretical limits of the modern ADC. Metaphorically, we're red-lining the RPMs and the engine is starting to overheat thus creating errors in the audio data.

24/96kHz does not suffer from this real-world performance limitation and is thought to be the superior format for anything that can hear issues with a 192kHz sampling rate.

[achalsey]

Thanks for the article! I will read it later in depth, but took a quick glance and it's definitely interesting. I think I already pointed this out in this thread, but...

The 24/192kHz format has been universally panned by critics. Many will tell you that it is clearly inferior to 24/96kHz and there is some science (or at least intarwebz science) that supports the claim that 192kHz is actually pushing and exceeding the theoretical limits of the modern ADC. Metaphorically, we're red-lining the RPMs and the engine is starting to overheat thus creating errors in the audio data.

24/96kHz does not suffer from this real-world performance limitation and is thought to be the superior format for anything that can hear issues with a 192kHz sampling rate.

Okay, thanks.

Someone else posted the article, but I didn't realize the issues brought up were more characteristic of 24/192 and not the difference between 48khz and 96khz.

[Gutbucket]

I've read several convincing (to me) arguments that a final delivery format of say 18 or 20 bits and a sample rate of about 60kHz effectively addresses all the real world applied engineering issues. 24/96 is pretty close to that without much additional wasted overhead and is an accepted standard. 16/44.1 as a delivery format is good enough for most uses as I see it.  I think the differences people hear between CD and higher-res formats is due to differences in mastering and more attention payed to the details of getting it right for hi-res, and partly the legacy of some older CD recordings that are not up to the quality of what can be done with modern equipment.

Recording and processing is a different thing though and has seperate issues than delivery format.  We all know the argument for more than 16 bits offering useful recording and processing headroom, and oversampling to higher rates is important in convesion and processing for filtering.  Less compelling to me is the question of if it makes sense to record at 96kHz or not if storage space isn't an issue, given my gear, the recording environment, the processing I'll do, etc.  I'm getting more motivated to setup a proper ABX test, but that's somewhat challenging to do as I'd need to run identical recorders at seperate rates.