What is the preferred recording format, WAV or AIFF?

[bayswater]

I'm confident someone could design a test miminizing the issues you pointed out... including what you recommended in your last paragraph.

Such tests are common practise and have been around for well over 100 years, pioneered by Helmholtz, and used to establish loudness scales by measuring the minimum difference in loudness we can detect. The tests are used to measure a just noticeable difference (JND) for all sorts of stimuli. They are also used to test for telepathy, clairvoyance, and placebos.

[EMRR]

Wow, it got all rabid up in here.

Labels like their high sample rate archiving for some reason. And archivists. Maybe they are stupid? It's like 4x5 film versus 35mm, more to work with; again important for anyone who sees analog processing as part of their workflow - like any large studio working on projects for record labels.

I use high rates because there is an obvious sonic difference in my work flow. No one pressured me, on the contrary, it was a pain in the ass to adopt, but it clearly sound better to my ears. The MOTU 24i/o sounds pretty trashy at 48K compared to 88K2.

I find the disc space and processing power argument questionable. This stuff is cheap now. I don't have a problem with either, and I'm not running any sort of smokin' system here.

It's premature to suggest apple will sell hi-res files in the future....they're not selling 'standard' res now! I think there's no conspiracy.

I don't really care what you use, ok? I don't see a reason you should care what I use either.

[bayswater]

The MOTU 24i/o sounds pretty trashy at 48K compared to 88K2.

I've read that some converters work better at some speeds rather than other.

I don't think anyone cares whether you work at 44.1, 48, 88.2 or 96. That wasn't the point. The question was whether speeds higher than 48 necessarily produce better sound.

There is also the question of whether it is better to work at the rate of the final product, or whether the imperfection of sample rate conversion offsets any advantage.

[mikehalloran]

It's premature to suggest apple will sell hi-res files in the future....they're not selling 'standard' res now! I think there's no conspiracy.

Agreed.

[FMiguelez]

I find the disc space and processing power argument questionable. This stuff is cheap now. I don't have a problem with either, and I'm not running any sort of smokin' system here.

It is just a small part of a bigger argument.

I probably could not run what I currently run at 192 KHz with my current setup.
Like I said, if space and processing power are not an issue anymore, why stop at a mere 192 KHz SR? Why not 500 GHz?

Why would I record my audio tracks at 192 KHz if the libraries I use were recorded at 44,1???

It's premature to suggest apple will sell hi-res files in the future....they're not selling 'standard' res now! I think there's no conspiracy.

If that was directed at me, I never said anything about a conspiracy.
I simply predicted that Apple will re-sell their catalogue at higher resolutions in the future. Why not sell a thing twice if you can? Hey, they are high-res and sound better!

Haven't they already done something similar by giving you the choice now to buy "high quality" AACs instead of the old/past mp3s rendered at 128 Kbps?

My prediction might not be correct and I'll end up looking like an arse. Tough! We'll see about that in a few years, I suppose.

I don't really care what you use, ok? I don't see a reason you should care what I use either.

I suppose you are joking, and I'm not sure who that comment is aimed at.

So you joined a discussion about SR in a public discussion forum because you don't care what other people use and what other people think??

And I don't really care about what people use. I care more about what people think and why they think what they do.

[EMRR]

I'm addressing what I took as a general tone of anti-high SR I read through multiple pages of before commenting.

And back at ya: The "I don't really care what you use" comment came about precisely because previous statements about the use of higher SR's appeared to have partially generated an anti-high SR rant for multiple pages.

I don't hear a point beyond 88K2 myself with the work I do, though I've heard 384kHz operatic recordings in a mastering room on Pass amps with Dunlavy speakers at 384, 192, 96, and 48K, stepping down through the rates, and each step down lost an audible bit of fine resolution.

If you are using Plangent process to restore analog tape, I believe you need to work at at least 192 to recover enough bias data, so anything analog archived at a lower rate is potentially 'unrestorable' with regards to absolute time base.

[MIDI Life Crisis]

It's like the old trick audio engineers play with mixes. Give a producer 5 identical mixes and they will always pick the one that sounds best. "Mix #5; that's the one, such clarity!"

Audio technology does have very specific parameters and technical specifications that go well into and beyond the areas of sound that dogs and bats can perceive. But as far as archiving music for human ears, it'll be a few million years (if ever) that our hearing will evolve sufficiently to take advantage of those extreme frequencies.

That is not to say that extreme frequencies aren't useful, especially as weapons and in other hi-tech application, but we're talking about music here and in that regard I seriously doubt ANYONE can hear the difference between a resolution of 48k v. 96k, regardless of their anecdotal "evidence."

They're simply listening to "mix #5."

[EMRR]

Again: it's the processing, not the end result. Trickle-down. Etc. IF you really don't use any hardware on playback, it's a much smaller issue.

And again: plenty of people experience audible differences at different SR settings of certain converters, and not with others. If you don't hear one, don't worry about it.

[FMiguelez]

Again: it's the processing, not the end result. Trickle-down. Etc. IF you really don't use any hardware on playback, it's a much smaller issue.

And again: plenty of people experience audible differences at different SR settings of certain converters, and not with others. If you don't hear one, don't worry about it.

Will you, or any user of high SRs, AT LEAST be willing to accept that you may be wrong and you are not really hearing what you think you are?
Emotional investments can become quite expensive at MANY LEVELS...

Do you at least acknowledge that possibility, even if you think it's quite unlikely???

[FMiguelez]

But as far as archiving music for human ears, it'll be a few million years (if ever) that our hearing will evolve sufficiently to take advantage of those extreme frequencies.

For that to happen, girls would have to find it attractive to mate with bat-eared guys (and viceversa)

I wonder how modern medicine is impacting evolution by natural selection? It seems we're close to almost controlling certain aspects of it...

[stubbsonic]

There's a podcast called "You Are Not So Smart" that talks about various fallibilities of human cognition. And I think the audio placebo effect could easily be an example of several of these.

I have played swept sine waves for many of my classes, as we listen to the rising tone we each signal the frequency goes too high for us to hear it. I'm always the first to signal, though I won't hold those young musicians entirely responsible for shaving off half of my top octave.

My point is, just because I can't hear it, doesn't mean it isn't there for others or doesn't matter. That is a strong argument for a little bit of overkill. And even though my hearing doesn't do anything about 15K (maybe 14K at this point), that doesn't mean I need to hang up my hat. I do need to be more careful though. It doesn't make me hype high-frequencies (but that's another topic).

A carefully designed double-blind test with very experienced subjects and randomly ordered examples could reveal what people can and can't distinguish. Furthermore, it could show that people will SWEAR they hear a difference when there is none.

I've listened to plenty of beautiful music at 44.1/16. The audio quality is a little bit compromised, but that is relative to the superior sound of 24 bit. My compromised ultra-high hearing disqualifies me from commenting about high rates. But it is fair to say that whether or not differences can be heard, there is not much harm in working at well beyond our range of hearing, so that our cats and dogs can enjoy higher fidelity recordings.

[MIDI Life Crisis]

For that to happen, girls would have to find it attractive to mate with bat-eared guys (and viceversa) ...

Now you;re just Bat-baiting me... LOL

[EMRR]

It's situation specific. You are asking an absolutist question. I can take converters I've lived with for years like the 24i/o and 2408mkIII and arrive at a consistent repeatable sonic result at varying rates. Cymbals take on a trashy indistinct quality at lower rates, they sound much more clear at 88K2 and higher. My 16A, not so much difference, much harder to tell. As an aside, the noise floor of the 16A is vanishingly low compared to the PCIe line, which is not so great in comparison. They're both '24 bit' (16A sends/receives 32 bit), so the difference is in the power supplies and analog side implementation.

Take a look at this, you can see the phase results of varying rates. Take multiple passes though converted at lower rates, you can warp the top end 180 degrees. You can see the 192K native rate of the 16A converters have the least and smoothest filter shape: those filter artifacts can and do reflect down into audible frequencies, so the further away they are from audible, the less impact they have. Chasing 'frequencies people hear' is total red herring in my mind. I'm much more inclined to believe phase and filter distortion components can define audible differences. Note also while the filter shapes are reasonably similar in phase angle for corresponding sample rates, the sonic differences between converters are totally different when comparing rates within one converter. Noise shape of dither is interesting to look at too, with the 16A it's much higher in frequency than the PCIe line products, whatever that's worth.

Full size here

[FMiguelez]

Sorry to repeat myself, but I would REALLY like to know the answer to this question from some high SR users here...

Will you, or any user of high SRs, AT LEAST be willing to accept that you may be wrong and you are not really hearing what you think you are?
Emotional investments can become quite expensive at MANY LEVELS...

Do you at least acknowledge that possibility, even if you think it's quite unlikely???

[FMiguelez]

There's a podcast called "You Are Not So Smart" that talks about various fallibilities of human cognition. And I think the audio placebo effect could easily be an example of several of these.

Sounds like something I'd enjoy listening to. Thanks

My point is, just because I can't hear it, doesn't mean it isn't there for others or doesn't matter. That is a strong argument for a little bit of overkill.

That's a fair point.
However, and by the same token, just because it's theoretically plausible, or because some people CLAIM to hear a difference, it does not mean it's actually real or there's anything up there...

I think the weakness of that concept is that it's similar to the Pascal's Wager approach (the "just in case" part).


People are notorious for downright lying (everybody lies), getting confused, selective bias, etc. Even conditions like pareidolia, in the worst cases, are extremely common (and quite amusing for cynics like I am sometimes).

I think something like that has more explanatory power and is more parsimonious than the convoluted explanations and evidence-free assertions of some people who claim having bat-grade hearing.

But I could be proven wrong on all counts. And it would be fine for me.

There ARE lots of things that are apparently not there but they are (bacteria, subatomic particles, Dark Energy, etc). There are more things that apparently are there but they are really not (I better not go there, ).

To be sure, all this people (who think they hear differences) need to do is to subject themselves to a well-designed double blind listening test and be willing to accept the results, even if they are disappointing. IMHE, not everybody is capable of being totally honest with themselves like that.