High resolution audio. The science, or lack of...?

[manicm]

The empirical data is very easily checked.

1) Find a good quality high resolution (24/96 or better) file.

2) Down sample to 16/44.1.

3) Compare the two in a properly set up (phase correct, level matched, etc) blind test.

So you reduce a hi-res file to 16/44.1, hear no difference, and conclude 24/96 is no better?? What's wrong with this picture?? The only accurate way to put an end to this issue is if Linn or someone else produce Studio Masters of the same recording i.e. one recorded in 16/44 and 24/96 and compare the two. Otherwise comparisons like this are just nonsense to me.

If I'm given 16/44 and 24/96+ of the same recording and the latter sounds better then it IS better, be it down to better mastering, or higher bit depth and sampling rate advantages or both.

 

[BigH]

The empirical data is very easily checked.

1) Find a good quality high resolution (24/96 or better) file.

2) Down sample to 16/44.1.

3) Compare the two in a properly set up (phase correct, level matched, etc) blind test.

So you reduce a hi-res file to 16/44.1, hear no difference, and conclude 24/96 is no better?? What's wrong with this picture?? The only accurate way to put an end to this issue is if Linn or someone else produce Studio Masters of the same recording i.e. one recorded in 16/44 and 24/96 and compare the two. Otherwise comparisons like this are just nonsense to me.

If I'm given 16/44 and 24/96+ of the same recording and the latter sounds better then it IS better, be it down to better mastering, or higher bit depth and sampling rate advantages or both.

Not sure I understand this but why do 2 recordings, will they be the same? It usual to record in 24 bit now and then reduce to 16bit for cd.

 

[manicm]

Not sure I understand this but why do 2 recordings, will they be the same? It usual to record in 24 bit now and then reduce to 16bit for cd.

For this (very long) argument's sake, the only sure way would be a simultaneous 16 and 24 bit recording/mastering of a studio performance, and then compare the two and decide which sounds better.

But I don't get Dave's and others' reduction of a high-res file to 16/44 and then deciding it sounds no different! It's this kind of obfuscation I can't stomach.

 

[fr0g]

Not sure I understand this but why do 2 recordings, will they be the same? It usual to record in 24 bit now and then reduce to 16bit for cd.

For this (very long) argument's sake, the only sure way would be a simultaneous 16 and 24 bit recording/mastering of a studio performance, and then compare the two and decide which sounds better.

But I don't get Dave's and others' reduction of a high-res file to 16/44 and then deciding it sounds no different! It's this kind of obfuscation I can't stomach.

You are being very illlogical.

If YOU get a 24/96 track and resuce it to 16/44.1yourself then you know it is the same recording. Not hearing any differences in this case will confirm to YOU that they are audibly the same.

Take many people doing the same test and you get a statistically larger result.

ALL recordings are done at the higher bit rate these days. When recording is makes sense as there is room for adjustments to be made.

 

[fr0g]

I agree. Although in a way, I don't. The "proof" is in the maths.

And so completely ignore our ears, or the fact we have any...

No, and I didnt say that either.

The "proof" is indeed in the maths.

I agree that for people who don''t understand the maths, pjysical proof is needed. The only real way we can do that is with ABX.

And the only way we can be sure we are ABXing the same music is if we downsample the high res track ourselves...see previous post.

Itls similar to the "debate" about evolution (by natural selection)

The "proof" of it is quite clearly there in the genetic information of organic life. Evolution is fact.

Non believers want more...fossils and suchlike...which we also have.

 

[andyjm]

Not sure I understand this but why do 2 recordings, will they be the same? It usual to record in 24 bit now and then reduce to 16bit for cd.

For this (very long) argument's sake, the only sure way would be a simultaneous 16 and 24 bit recording/mastering of a studio performance, and then compare the two and decide which sounds better.

But I don't get Dave's and others' reduction of a high-res file to 16/44 and then deciding it sounds no different! It's this kind of obfuscation I can't stomach.

You are being very illlogical.

If YOU get a 24/96 track and resuce it to 16/44.1yourself then you know it is the same recording. Not hearing any differences in this case will confirm to YOU that they are audibly the same.

Take many people doing the same test and you get a statistically larger result.

ALL recordings are done at the higher bit rate these days. When recording is makes sense as there is room for adjustments to be made.

Just to clarify Fr0g's response:

1. Get a 24/96 track that you think is super

2. Downsample it yourself (many packages will do this) to 16/44.1

3. You now have 2 tracks, the original 24/96 track and a downsample of the same track at 16/44.1

4. Compare the two tracks using an ABX method (I believe foobar allows you to do this)

5. As the two files have exactly the same source, the only difference is sample rate and bit depth. If you can't tell the difference, then this would indicate that there was no extra 'magic' in the 24/96 track that couldn't be captured in 16/44.1.

 

[pauln]

I agree. Although in a way, I don't. The "proof" is in the maths.

And so completely ignore our ears, or the fact we have any...

That's the whole point - as I'm sure you know.

Our perception of what our ears "hear" is influenced by other factors, not just the changes in pressure that our ear drums are sensitive too.

 

[pauln]

Itls similar to the "debate" about evolution (by natural selection)

The "proof" of it is quite clearly there in the genetic information of organic life. Evolution is fact.

Non believers want more...fossils and suchlike...which we also have.

But they still try and discredit the evidence - fossils were created at the same time as the earth (4000 years ago was it?) and were put there by "god" specifically to test man's faith...

Anyway who cares; some people go through life believing in all kinds of odd things. As long as they don't try to con anyone out of their hard earned money, let them get on with it if it makes them happy.

 

[chebby]

But they still try and discredit the evidence - fossils were created at the same time as the earth (4000 years ago was it?) and were put there by "god" specifically to test man's faith...

Strange how they don't try to discredit the quantum physics that makes all their means of 'spreading the word' possible. (TV, cable channels, computers and the internet, mobile phones, PA systems in churches etc.)

Or maybe they do?

That would be funny.

 

[davedotco]

Not sure I understand this but why do 2 recordings, will they be the same? It usual to record in 24 bit now and then reduce to 16bit for cd.

For this (very long) argument's sake, the only sure way would be a simultaneous 16 and 24 bit recording/mastering of a studio performance, and then compare the two and decide which sounds better.

But I don't get Dave's and others' reduction of a high-res file to 16/44 and then deciding it sounds no different! It's this kind of obfuscation I can't stomach.

You are being very illlogical.

If YOU get a 24/96 track and resuce it to 16/44.1yourself then you know it is the same recording. Not hearing any differences in this case will confirm to YOU that they are audibly the same.

Take many people doing the same test and you get a statistically larger result.

ALL recordings are done at the higher bit rate these days. When recording is makes sense as there is room for adjustments to be made.

Just to clarify Fr0g's response:

1. Get a 24/96 track that you think is super

2. Downsample it yourself (many packages will do this) to 16/44.1

3. You now have 2 tracks, the original 24/96 track and a downsample of the same track at 16/44.1

4. Compare the two tracks using an ABX method (I believe foobar allows you to do this)

5. As the two files have exactly the same source, the only difference is sample rate and bit depth. If you can't tell the difference, then this would indicate that there was no extra 'magic' in the 24/96 track that couldn't be captured in 16/44.1.

This is exactly right and as far as I know the only way to ensure that everything is, as it should be, under the control of the experimentor.

Relying on a third part to supply the test material is simply 'not on'.

In reality we are comparing the original hi-res master with the downsampled version, a version that has also undergone another stage if digital manipulation, in this case sample/bit rate conversion. If digital processing is as invasive as some suggest, this should make the differences easier to hear.

Now I have tried this experiment myself, in an informal manner with a third party doing the switching and through a decent playback system. Getting the playback levels identical was quite difficult but once done neither I, nor several colleagues could tell the difference. We even tried the usual 'tricks' changes in noise floor, listening to extended decay of individual notes, none of this helped.

Of course this 'proves' nothing, real experimental proof in a subjective matter of this kind can only be achieved statistically, ie test enough (random) people enough times to give a result that is as close to fact as it is possible to get.

 

[BigH]

But I don't get Dave's and others' reduction of a high-res file to 16/44 and then deciding it sounds no different! It's this kind of obfuscation I can't stomach.

Because that is the normal practice, and we are trying to see if there is a difference between them, ie is it worth paying out more money for 24 bit.

 

[Neuphonix]

Worth a read:

http://www.trustmeimascientist.com/2013/02/04/the-science-of-sample-rates-when-higher-is-better-and-when-it-isnt/

he seems to suggest that any percieved benefits which people may be hearing in higher bit recordings could be due to unintended distortions caused by higher sampling rates. I think? :?

 

[spiny norman]

Oh gawd, must almost be time for sometime to trot out that tired old

link again.

 

[Neuphonix]

Oh gawd, must almost be time for sometime to trot out that tired old

link again.

apologies, just trying to contribute. I take it you don't find the article to be accurate?

 

[spiny norman]

More a question of boredom at these endless circular threads, in which the same old Googled 'evidence' is trotted out.

Can't we just say 'good for you' to those who are happy they can hear the difference between CD quality and 'hi-res' music, and are happy to spend more for that difference, and 'good for you' to those who can't, won't or don't believe there's any difference, and are thus saving their money.

All this self-righteous battling for justice, accusing people of con-tricks, delusion or stupidity, and general cox and box is getting very tiresome. Surely there are more important issues in the world to contemplate while enjoying some good music, whatever the file format?

 

[Neuphonix]

Yeah I agree, although I must say I have found this thread to be a little more interesting than others on the same topic. Slightly more informative and a bit less vitriolic.

i don't feel strongly for or against, however have decided 16 bit is good enough for me, I can't justify the added expense

but am absolutely happy for those who do

 

[cheeseboy]

More a question of boredom at these endless circular threads, in which the same old Googled 'evidence' is trotted out.

Can't we just say 'good for you' to those who are happy they can hear the difference between CD quality and 'hi-res' music, and are happy to spend more for that difference, and 'good for you' to those who can't, won't or don't believe there's any difference, and are thus saving their money.

All this self-righteous battling for justice, accusing people of con-tricks, delusion or stupidity, and general cox and box is getting very tiresome. Surely there are more important issues in the world to contemplate while enjoying some good music, whatever the file format?

but take that away and the only thing left on the internet would be pron and facebook

 

[Anonymous]

[EDITED BY MODS AND USER BANNED - you are Maxflinn and I claim my £5. That makes £4,945 so far]

 

[CnoEvil]

i don't feel strongly for or against, however have decided 16 bit is good enough for me, I can't justify the added expense

but am absolutely happy for those who do

I agree.....but do you hear a difference (even if you think it's not VFM)?

 

[shadders]

Hi,

I have not read all the posts, but from an engineering perspective, the higher number of bits and higher sampling rate will be a benefit.

If we examine the current CD specification, where this is based on 44.1kHz sampling rate, and 16bits per sample, is sufficient for most people.

In the extreme, if we assume that we have infinite number of samples per second, and an infinite number of bits to be assigned per sample, the reconstructed waveform will be an exact replica of the signal at the point of sampling - possibly the mixing desk.

Hence, 24bit at 192kHz (or 96kHz) is between these two extremes - that is the CD standard and the perfect representation.

As long as the engineering process to record and then replay is as accurate as possible - then higher bits and higher sampling rate will be a benefit.

Of course, as others have stated, some Blu-Ray discs may be 96kHz which are just resampled 44.1kHz recordings.

Hifi News regularly prints analysis of recordings and show that this is in fact the case.

So, a higher sample rate and greater number of bits per sample is always a benefit, but whether they have been implemented correctly is another issue.

Regards,

Shadders.

 

[Phileas]

Hi,

I have not read all the posts...

You don't say? :wall:

(Sorry, couldn't resist. )

 

[Anonymous]

More a question of boredom at these endless circular threads, in which the same old Googled 'evidence' is trotted out.

Can't we just say 'good for you' to those who are happy they can hear the difference between CD quality and 'hi-res' music, and are happy to spend more for that difference, and 'good for you' to those who can't, won't or don't believe there's any difference, and are thus saving their money.

All this self-righteous battling for justice, accusing people of con-tricks, delusion or stupidity, and general cox and box is getting very tiresome. Surely there are more important issues in the world to contemplate while enjoying some good music, whatever the file format?

Best post on this thread and relevant other bilge, cringe-inducing, post building threads. Lock the thread.

 

[Neuphonix]

i don't feel strongly for or against, however have decided 16 bit is good enough for me, I can't justify the added expense

but am absolutely happy for those who do

I agree.....but do you hear a difference (even if you think it's not VFM)?

When I first bought the Akurate I purchased quite a few 24bit albums, blindly thinking that they must be superior. They did sound great, no doubt about it, but after reading up a bit I realised that there was more to consider than bit depth.

I can't say that I have ever taken the time to sit down & compare versions, or downsample & blind test as suggested. I simply dont have the time or the inclination to analyse at that level.

One thing i can hear with-out a doubt is albums that are poorly mastered. This seems to me to be a much bigger issue in terms of SQ.

If an album by one of my top ten fav artists came out I could be tempted to buy in 24bit, but otherwise the VFM hurdle is too much for me. If the difference was 10- 15% then i would probably buy the 24bit version. I do understand (in laymans terms at best) the debate about one being better than the other, but you certainly don't seem to hear the point being made that the 24bit version sounds WORSE than the 16 bit, simply that it is no better.

As I said, more than happy to live & let live.

Apologies to Spiny & others for not letting this circular thread die a natural death.

 

[CnoEvil]

When I first bought the Akurate I purchased quite a few 24bit albums, blindly thinking that they must be superior. They did sound great, no doubt about it, but after reading up a bit I realised that there was more to consider than bit depth.

I can't say that I have ever taken the time to sit down & compare versions, or downsample & blind test as suggested. I simply dont have the time or the inclination to analyse at that level.

One thing i can hear with-out a doubt is albums that are poorly mastered. This seems to me to be a much bigger issue in terms of SQ.

If an album by one of my top ten fav artists came out I could be tempted to buy in 24bit, but otherwise the VFM hurdle is too much for me. If the difference was 10- 15% then i would probably buy the 24bit version. I do understand (in laymans terms at best) the debate about one being better than the other, but you certainly don't seem to hear the point being made that the 24bit version sounds WORSE than the 16 bit, simply that it is no better.

As I said, more than happy to live & let live.

Apologies to Spiny & others for not letting this circular thread die a natural death.

The reason I asked, is you have the same system as me, only better.

I now have quite a lot of 24 bit stuff, most of which I got free from Linn, as an offer when I bought my MDS. Some of these tracks I have in 320kbps, 16 bit and 24 bit. In each case I can hear a subtle improvement....mostly in ambient info and subtle details, like the reverb and decay of a piano note, the breathing of artists and the squeak of fingers on double bass strings.

I have no idea if they are mastered differently, but I do believe that 24 bit allows for better mastering.

It is easier to hear on simply recorded stuff.

I also have some 24 bit stuff that is nothing very spectacular.

IMO. ATM. Hi-Res is not worth the premium over well mastered 16 bit.

 

[andyjm]

Hi,

I have not read all the posts, but from an engineering perspective, the higher number of bits and higher sampling rate will be a benefit.

As long as the engineering process to record and then replay is as accurate as possible - then higher bits and higher sampling rate will be a benefit.

So, a higher sample rate and greater number of bits per sample is always a benefit, but whether they have been implemented correctly is another issue.

Regards,

Shadders.

Er no. Not done much engineering then I guess, Shadders.