High-resolution audio: clarity or confusion?

[Overdose]

The information is not audible and therefore irrelevant. In the broadest terms.

There is a very simple experiment that virtually anyone can do.

Find a good quality hi-res download that you like.

Down sample to Cd standard, 16 bit, 44.1 Khz.

After carefully level matching the two files, listen and compare, but do so blind.

Report what you hear.

Dave you state things with such finallity.

Why would you compabe a down sampled Hi-res audio file to a CD? You would just be making it the same as the CD surely? That sounds stupid to me.

A DTS HD-MA sound track clearly sounds better and more dynamic than a DTS track becuause there is more information there and it's not been compressed. Simple as that.

Blind tested people would say that DTS track would be better because they can hear more more easily due to the compression or dynamic limitation. That doesn't make it better or even as good as uncompressed audio.

Of course a movie mix or album will never be true to life completely because its a fake copy of real life sound. Unless you are listening t one instrument then you could leave it's full dynamic range and nothing would be there to cover up the small subtle sounds.

You are miss-understanding the test, probably my fault, if I am not being clear.

Comparing a CD, or a rip of a CD, to a hi-res download is indeed pointless, we have no idea of the provenance of the material in either case.

In this case the test is to use the best hi-res file that you have and downsample it to 16/44.1 as a copy.

You then compare the original hi-res file to the new 16/44.1 copy in a level matched blind test.

I believe that this can be done in Audacity, though you may need some extra software (plug in) to do so.

Certainly, Audacity can downsample, but I'm not sure you could AB test with it. I believe that Foobar has this facility though.

 

[davedotco]

The information is not audible and therefore irrelevant. In the broadest terms.

There is a very simple experiment that virtually anyone can do.

Find a good quality hi-res download that you like.

Down sample to Cd standard, 16 bit, 44.1 Khz.

After carefully level matching the two files, listen and compare, but do so blind.

Report what you hear.

Dave you state things with such finallity.

Why would you compabe a down sampled Hi-res audio file to a CD? You would just be making it the same as the CD surely? That sounds stupid to me.

A DTS HD-MA sound track clearly sounds better and more dynamic than a DTS track becuause there is more information there and it's not been compressed. Simple as that.

Blind tested people would say that DTS track would be better because they can hear more more easily due to the compression or dynamic limitation. That doesn't make it better or even as good as uncompressed audio.

Of course a movie mix or album will never be true to life completely because its a fake copy of real life sound. Unless you are listening t one instrument then you could leave it's full dynamic range and nothing would be there to cover up the small subtle sounds.

You are miss-understanding the test, probably my fault, if I am not being clear.

Comparing a CD, or a rip of a CD, to a hi-res download is indeed pointless, we have no idea of the provenance of the material in either case.

In this case the test is to use the best hi-res file that you have and downsample it to 16/44.1 as a copy.

You then compare the original hi-res file to the new 16/44.1 copy in a level matched blind test.

I believe that this can be done in Audacity, though you may need some extra software (plug in) to do so.

Certainly, Audacity can downsample, but I'm not sure you could AB test with it. I believe that Foobar has this facility though.

Thank you.

I have only ever done this the old fashioned way, in hardware.

 

[CJSF]

...occasionally I find a digital offering that has been well recorded according to natures analogue hearing mechanism called ears . . .

Arguably, much of the way that human hearing works, the stimulation of hairs in the cochlea at different frequencies and the subsequent firing of nerve cells has more in common with spectrum analysis and digital systems than it has in common with analogue recording.

Oh yes . . . Then I am an 'android' . . . ?

 

[steve_1979]

...occasionally I find a digital offering that has been well recorded according to natures analogue hearing mechanism called ears . . .

Arguably, much of the way that human hearing works, the stimulation of hairs in the cochlea at different frequencies and the subsequent firing of nerve cells has more in common with spectrum analysis and digital systems than it has in common with analogue recording.

Oh yes . . . Then I am an 'android' . . . ?

Inside the Cochlea there are little hairs that vibrate at different frequencies. Each set of hairs only vibrate at a specific frequency. While there are a lots of these hair that vibrate lots of different frequencies there is still only a finite number of them. This means that there is only a finite number of different frequencies that can be heard. I think this is what andyjm was referring to when he said the ear is 'digital'.

Watch this video and take particular notice of the bit from 4:58 - 6:02

http://www.youtube.com/watch?v=PeTriGTENoc

 

[Vladimir]

It's all a matter of cabling, really.

 

[andyjm]

...occasionally I find a digital offering that has been well recorded according to natures analogue hearing mechanism called ears . . .

Arguably, much of the way that human hearing works, the stimulation of hairs in the cochlea at different frequencies and the subsequent firing of nerve cells has more in common with spectrum analysis and digital systems than it has in common with analogue recording.

Oh yes . . . Then I am an 'android' . . . ?

Inside the Cochlea there are little hairs that vibrate at different frequencies. Each set of hairs only vibrate at a specific frequency. While there are a lots of these hair that vibrate lots of different frequencies there is still only a finite number of them. This means that there is only a finite number of different frequencies that can be heard. I think this is what andyjm was referring to when he said the ear is 'digital'.

Watch this video and take particular notice of the bit from 4:58 - 6:02

http://www.youtube.com/watch?v=PeTriGTENoc

Steve,

Thank you for the link, that is an excellent description of how the ear works. You are right in your comments about cochlea hairs and frequency discrimination, but my comment was also trying to capture the nerve cells firing - which is a digital process, not analogue (not quite a correct analogy but very similar). While CJSF may have a magic analogue ear, the rest of us have to make do with something that looks a lot like a spectrum analyser and a form of A2D converter.

 

[matt49]

As these posts about the physical nature of the ear have shown, the question whether an information system is fundamentally analog or digital throws up some surprising and counter-intuitive answers.

If you drill down far enough into the physical nature of an information system, you’ll always reach a point at which things are both quantized (lumpy) and linear (continuous). This is because, according to quantum mechanics, all matter exhibits both wave and particle behaviour.

But doesn’t common sense tell us that the behaviour of a vinyl LP is linear, because the material of the grooves has a smooth, continuous shape? Well yes, but as usual common sense is wrong. The lumpy nature of the polymers that make up vinyl combined with the elliptical form of a stylus means that there’s only a certain number of possible interactions between the stylus and the groove surface. And if you zoom in far enough, those interactions can be adequately expressed in quantized (i.e. digital) form.

Now the degree of resolution you need to express vinyl in quantized form isn’t actually that great. In fact it’s rather less than the available resolution of the red book CD specification. In other words, CD is actually more analog than vinyl.

There’s an easy way to show that digital is able to represent vinyl with complete accuracy. If you record the output of your phono stage on a good enough digital sound card and then play the resulting wav file, then assuming your electronics are transparent it’ll sound identical to the original phono output.

And as Dave will confirm, the better your vinyl replay system, i.e. the greater its resolution, the more like digital audio it’ll sound.

The reason vinyl sounds “smoother” than digital is that it adds noise and distortion that we find pleasing. By contrast, digital lacks this distortion and can sound unpleasantly “dry”. As T. S. Eliot put it in the Four Quartets, “humankind cannot bear very much reality”.

If you think this is nonsense and that no discriminating audiophile could believe such stuff, do have a read of Jim Lesurf’s excellent Information and Measurement, especially chapter 12 (“Analog or digital?”).

:read:

Matt

 

[tino]

One can argue over the technical merits of high resolution audio, but ultimately I am not going to pay any extra for it given that there is already a huge amount of new and used CD red book quality music out there at very agreeable prices. I'd rather put my money there for now ... I find the sound quality more than acceptable on my hifi system/s.

 

[Jota180]

..and amplitude and frequency are the only parameters for sound reproduction?

Example, let's say a (theoretical) perfect recording of a piano being played was created. A single note played would have a variety of different freqencies that, when converging at exact (and differing points in time) govern the overal timbre of that note. A digital representation of that is only samples in time of the original. 16 bit (and more crucially) 44.1khz is not sufficient for rebuilding the original waveforms in their complexity without massive corrective measures in the digital domain to recoup some of the losses.

There are a lot of things in physics and engineering where a smart man with a good general education can grasp the fundamentals. Digital sampling is not one of those things. Much of it is counter-intuitive, as is illustrated by your comments.

44.1KHz sample rate is enough to adequately sample all frequencies up to 20KHz without loss. As your ears can't respond to anything above 20KHz, there can be no component or timbre in the note that is lost to your hearing.

This ^^

CD isn't missing any of the original audio you can hear, only frequencies you can't hear. Unless you're a dog or a bat.

The biggest impact on CD sound quality is the recording and mastering process where record companies appear to be taking part in a loudness war to the detriment of the dynamics of the original.

 

[Jota180]

Here's my take:

http://www.whathifi.com/blog/high-resolution-audio-clarity-or-confusion

What do you think?

I think as a well known magazine on the topic of hifi you have a duty to present the facts to your readers. Fair enough report on what the big companies are saying but you cannot just leave it at that otherwise you run the risk of being viewed as the PR wing of those big companies.

Now I understand that this topic especially is very technical and beyond the understanding of your average punter. This is about some serious maths and physics, so why not invite some experts in field who have no connection to the big hifi companies and record companies to give their unbiased opinions?

I think I've suggested this before but never got any response, but I seriously think the first thing in your editors head upon reading the press release about Hi Rez should have been, 'lets talk to some experts about it and get their views.'

If I was editor of your magazine I'd have been on the phone to some of the UK's top universities looking for experts in the field.

As the most popular magazine on the topic the onus really is on you to supply people, and they're your customers as much as they're customers of Naim, Roksan, Sony etc, with all the facts and arguments on this issue.

You could maybe have a thread where people suggest questions.

 

[ID.]

Here's my take:

http://www.whathifi.com/blog/high-resolution-audio-clarity-or-confusion

What do you think?

Flagged as offensive

Blatant trolling and EDITED stirring. Go on, start a thread on cables next :roll:

 

[cheeseboy]

It's pretty blooming obvious really that Hi-Res is better than standard Redbook 16 bit / 44.1kHz. Only a fool could argue otherwise. If you think about it vinyl cut entirely from analogue sources without any digital futzing in the chain is "infinite" resolution. No one can tell me they can't tell a huge difference between vinyl and CD. Vinyl is better because it isn't digits and numbers limited by file size. Hi-Res audio sounds much better because there is greater resolution. I can see a time not too far away when Hi-Res will match vinyl as far as the human ear can detect a difference anyway. It's already getting there.

wow, thank you for that informative well researched with referenced post there...... :rofl: :help:

 

[CnoEvil]

By contrast, digital lacks this distortion and can sound unpleasantly “dry”.

I have yet to go to a Classical concert that sounded "unpleasantly dry", so "if" a digital recording is like this, it has certainly lost something along the way! :shifty:

 

[hifikrazy]

Why is everything in hifi somehow a big conspiracy to con consumers of their money for no sonic gain, with the prescribed remedy always the boring old double blind test? If you don't want to spend your money, that's totally fine by me, but there's no need to always create a bunch of excuses to justify your inaction.

Was that a rhetorical question?

I too would try to bury my head in the sand, rather than face the fact that I'd been sold a line and spent small fortune where I needn't have done so.

I have absolutely no post purchase consumer regrets, so you guys are doing nobody any favours with your unwelcome advice. In fact, the only ones who get any benefit from your wingeing and moaning are the others who share your opinion, so that you can provide one another support and validation for being tightwads.

 

[TrevC]

It's pretty blooming obvious really that Hi-Res is better than standard Redbook 16 bit / 44.1kHz. Only a fool could argue otherwise. If you think about it vinyl cut entirely from analogue sources without any digital futzing in the chain is "infinite" resolution. No one can tell me they can't tell a huge difference between vinyl and CD. Vinyl is better because it isn't digits and numbers limited by file size. Hi-Res audio sounds much better because there is greater resolution. I can see a time not too far away when Hi-Res will match vinyl as far as the human ear can detect a difference anyway. It's already getting there.

wow, thank you for that informative well researched with referenced post there...... :rofl: :help:

Once hi res reaches the dizzy heights of vinyl what will we have to strive for? The awesome sound of 78rpm shellac?

 

[Overdose]

Why is everything in hifi somehow a big conspiracy to con consumers of their money for no sonic gain, with the prescribed remedy always the boring old double blind test? If you don't want to spend your money, that's totally fine by me, but there's no need to always create a bunch of excuses to justify your inaction.

Was that a rhetorical question?

I too would try to bury my head in the sand, rather than face the fact that I'd been sold a line and spent small fortune where I needn't have done so.

I have absolutely no post purchase consumer regrets, so you guys are doing nobody any favours with your unwelcome advice. In fact, the only ones who get any benefit from your wingeing and moaning are the others who share your opinion, so that you can provide one another support and validation for being tightwads.

Whatever, as long as you can take comfort from your ignorance. It really is no skin off my nose.

 

[davedotco]

By contrast, digital lacks this distortion and can sound unpleasantly “dry”.

I have yet to go to a Classical concert that sounded "unpleasantly dry", so "if" a digital recording is like this, it has certainly lost something along the way! :shifty:

Nice try Cno.......

But using the same words, "unpleasantly dry" to describe two different effects doesn't work either.

I doubt that you hear much of a soundstage at a classical concert either, in this case the recording has 'gained' something along the way.

With very few exceptions, classical recordings are just as artificial a construct as pop and rock albums, if such a recording sounds "unpleasantly dry", that is a production issue.

A live orchestra in concert does not sound "dry", neither does it sound 'warm' or 'involving', it just is. Asking your home system to produce a sound that is, as close as possible, to your perception of what the event should sound like is fine, in fact I consider it an excellent measure of a systems performance.

It is not however the only measure, reproducing what is on the recording, as accurately as possible, is also a valid measurement for someone of a different mindset.

Using both methods, sometimes with some thought, often seemingly at random, seems to be my prefered method, maybe a mood thing?

Or maybe my natural contraryness.......

 

[andyjm]

It's pretty blooming obvious really that Hi-Res is better than standard Redbook 16 bit / 44.1kHz. Only a fool could argue otherwise. If you think about it vinyl cut entirely from analogue sources without any digital futzing in the chain is "infinite" resolution. No one can tell me they can't tell a huge difference between vinyl and CD. Vinyl is better because it isn't digits and numbers limited by file size. Hi-Res audio sounds much better because there is greater resolution. I can see a time not too far away when Hi-Res will match vinyl as far as the human ear can detect a difference anyway. It's already getting there.

Steve,

This is a frequently misunderstood point. Any analogue system has 'resolution', this is set by the noise floor. By definition, the system can't resolve anything smaller than its noise level - otherwise how would you know if a change in the signal was really signal or noise? Resolution in an analaogue system is is usually taken to be the ratio of maximum signal to noise floor or SNR. In a digital system this is comparable to the ratio of the smallest step to the maximum value.

For CDs this is 96dB (a mathematical derivation from 16bits of resolution)

For LPs it varies, depending on the quality of the LP, the cutter, playback and so on. It is somewhere between 50dB and 70dB

I understand that many people prefer vinyl, but this is not because it has better resolution. Redbook CDs already have resolution that is greater than is detectable by a human ear in a domestic setting (and at least 30dB better 'resolution' than LPs)

 

[Kamikaze Bitter]

CD succeeded because it was something people wanted. Vinyl was a very inconvenient medium, fragile and difficult to store. Also vinyl sound quality was appalling. Remember most people listened to vinyl on a Bush radiogram, not a LP12. So CD was something people wanted, and when the discs and the players reached an affordable price point it soared away.

Similarly compressed formats, mobile music players and streaming services deliver something that people want.

But who wants HD music beyond a very few technophiles, audiophiles and Neil Young? And even these can't agree that it makes an audible difference. I just can't see where the demand will come from. At the moment there are distinct drawbacks in terms of file size and bandwidth. Of course in time storage capacity and bandwidth will expand. But I just can't see HD audio as anything other than a dead end – like SACD. For me HD audio is solution without a problem.

 

[matt49]

By contrast, digital lacks this distortion and can sound unpleasantly “dry”.

I have yet to go to a Classical concert that sounded "unpleasantly dry", so "if" a digital recording is like this, it has certainly lost something along the way! :shifty:

Live classical concerts don't sound dry because of the acoustic effects of the hall being full of people. Take those people out of the hall, and you'll have a different and dryer sound.

I used the word "can": i.e. CD sound can sound dry in some circumstances, and those circumstances tend to be artificial production environments.

If you've heard a "dry mix" from a recording studio, you'll know what I mean. It lacks the reverb that we find so pleasing. By accident as it were, dry mixes sound better on vinyl, especially on cheaper systems, because TTs have their own inbuilt reverb: it's a fault of analog playback that's accidentally beneficial in some circumstances.

Matt

 

[matt49]

Steve,

This is a frequently misunderstood point. Any analogue system has 'resolution', this is set by the noise floor. By definition, the system can't resolve anything smaller than its noise level - otherwise how would you know if a change in the signal was really signal or noise? Resolution in an analaogue system is is usually taken to be the ratio of maximum signal to noise floor or SNR. In a digital system this is comparable to the ratio of the smallest step to the maximum value.

For CDs this is 96dB (a mathematical derivation from 16bits of resolution)

For LPs it varies, depending on the quality of the LP, the cutter, playback and so on. It is somewhere between 50dB and 70dB

I understand that many people prefer vinyl, but this is not because it has better resolution. Redbook CDs already have resolution that is greater than is detectable by a human ear in a domestic setting (and at least 30dB better 'resolution' than LPs)

Yes, this is another aspect of the point I made about resolution last night.

Matt

 

[pauln]

Deleted. What's the point. Deaf ears and all that.

 

[BigH]

It's pretty blooming obvious really that Hi-Res is better than standard Redbook 16 bit / 44.1kHz. Only a fool could argue otherwise. If you think about it vinyl cut entirely from analogue sources without any digital futzing in the chain is "infinite" resolution. No one can tell me they can't tell a huge difference between vinyl and CD. Vinyl is better because it isn't digits and numbers limited by file size. Hi-Res audio sounds much better because there is greater resolution. I can see a time not too far away when Hi-Res will match vinyl as far as the human ear can detect a difference anyway. It's already getting there.

Infinite resolution, sounds like you know nothing about vinyl.

 

[Alec]

Why is everything in hifi somehow a big conspiracy to con consumers of their money for no sonic gain, with the prescribed remedy always the boring old double blind test? If you don't want to spend your money, that's totally fine by me, but there's no need to always create a bunch of excuses to justify your inaction.

Was that a rhetorical question?

I too would try to bury my head in the sand, rather than face the fact that I'd been sold a line and spent small fortune where I needn't have done so.

I have absolutely no post purchase consumer regrets, so you guys are doing nobody any favours with your unwelcome advice. In fact, the only ones who get any benefit from your wingeing and moaning are the others who share your opinion, so that you can provide one another support and validation for being tightwads.

"Nonsense"!

 

[cheeseboy]

I have absolutely no post purchase consumer regrets, so you guys are doing nobody any favours with your unwelcome advice. In fact, the only ones who get any benefit from your wingeing and moaning are the others who share your opinion, so that you can provide one another support and validation for being tightwads.

so you don't think anybody should be able to call bullcrap on manufacturers if they are talking out of their bumhole and we should all just blindly believe what they tell us?