[lexnasa]

This debate is stupid... I suggest we all leave it to rust. 24 bit is inherently superior to 16 bit at the same sampling rate, but quality of playback depends on the playback medium and decoding hardware and software. A high end 16 bit system may outperform a low end 24 bit system, so what?

 

[bigshot]

I'd agree with you if you change the word "inherently" to "theoretically". Because in playback of music there is no audible difference between the two formats. High end or low end- all of the superiority of 24 bit is beyond the range of human hearing.

See ya
Steve

 

[lexnasa]

Sorry, rubbish. I write and produce music for a living... you're being duped by the scientific studies that pretend to understand human hearing.

I can play you a native 24 bit 44.1khz recording vs a native 16 bit 44.1khz recording of the same sound and there is a difference. By native I mean that no dithering has been used.

To address your human hearing point, what you may have been duped by is the empirical limit of human hearing, on which CDs were based. However, this empirical limit ignores the physicality of playback. It ignores harmonic interraction. So by removing what you can't hear, it also removes the interraction of what you can't hear with what you can hear.

As you're ignoring sampling frequency, and only addressing bitrate, your argument is fundamentally flawed. If you imagine analogue as being a continous flow of data, and digital as a series of samples that are taken and reconstructed, which do you think would be better? And taking more detailed samples should be good? By your argument 4 bit should sound as good as 16 bit... it doesn't! 4 bit or 16 bit or 24 bit at the same sampling frequency are addressing exactly the same range of human hearing.

I can't believe I'm defending myself against someone that thinks 16 bit audio is the endgame... wait, I have to pinch myself!

 

[shinew]

Quote:

Originally Posted by lexnasa /img/forum/go_quote.gif To address your human hearing point, what you may have been duped by is the empirical limit of human hearing, on which CDs were based. However, this empirical limit ignores the physicality of playback. It ignores harmonic interraction. So by removing what you can't hear, it also removes the interraction of what you can't hear with what you can hear.

but isn't it frequncy's job, not bit rate?

 

[nick_charles]

Quote:

Originally Posted by lexnasa /img/forum/go_quote.gif Sorry, rubbish. I write and produce music for a living... you're being duped by the scientific studies that pretend to understand human hearing. I can play you a native 24 bit 44.1khz recording vs a native 16 bit 44.1khz recording of the same sound and there is a difference. By native I mean that no dithering has been used. What you may have been duped by is the empirical limit of human hearing, on which CDs were based. However, this empirical limit ignores the physicality of playback. It ignores harmonic interraction. So by removing what you can't hear, it also removes the interraction of what you can't hear with what you can hear. I can't believe I'm defending myself against someone that thinks 16 bit audio is the endgame... wait, I have to pinch myself!

It really isnt just a matter of theory. There has been a rather well-publicised study examining how well listeners can differentiate between high res and 16/44.1 music. With 60 Audiophile, Engineer and Sound Engineering students as subjects the level of differentiation was 50% i.e complete chance.

If you would like to post some 16/44.1 and 24/44.1 samples taken under otherwise identical conditions i.e recorded simultaneously with identical microphone placement and identical mastering and identical average volume levels I would be happy to audition them under DBT conditions as I guess would several folks here.

And of course you could support your case better by doing so yourself.

There is a plethora of "I can hear a difference" stories out there but very few have been supported by unbiased i.e carefully controlled unsighted listening tests, so far the vast bulk of empirical evidence suggests that we just cannot hear the difference.

Seriously, if you can provide us with some new comparable 16 and 24 bit samples for testing it would be interesting, please not the ones we have all seen from Linn and others, they just are not identical between formats.

And in any case nobody is arguing against 24 bits for recording and mastering, the issue is whether we need 24 bits for playback.

 

[CyberTheo]

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif Seriously, if you can provide us with some new comparable 16 and 24 bit samples for testing it would be interesting, please not the ones we have all seen from Linn and others, they just are not identical between formats.

I've been following this thread with interest. Let me throw a wrench. Why not Linn samples? I feel they are good base for comparison, unless you're proposing another theoretical study here that we've seen beaten to death already. All commercial CDs were recorded in much higher bit and sampling rates than 16/44; they were all down sampled to comply to CD (or whatever) format. Even if someone is successful in proving the theoretical aspect to human hearing audibility between 16/44 and higher rate samples, it's of no value to the end user - who are always getting down sampled digital format from higher rate originals.

 

[gyrodec]

lexnasa, you don't seem to know where the extra 8 bits of 24 bit go. They just tack on the bottom and define detail that is already below any meaingful noise floor (this is for delivery only, for recording the extra bits have obvious advantages in level setting on recording, but we are talking normalized at playback). Unless you think you can hear signal at -90db in your listening room, 16 and 24 bit will be identical.

"As you're ignoring sampling frequency, and only addressing bitrate, your argument is fundamentally flawed. If you imagine analogue as being a continous flow of data, and digital as a series of samples that are taken and reconstructed, which do you think would be better? And taking more detailed samples should be good? By your argument 4 bit should sound as good as 16 bit... it doesn't! 4 bit or 16 bit or 24 bit at the same sampling frequency are addressing exactly the same range of human hearing."

You don't take the range of human hearing, in volume terms, and split it into as many samples as you have. You start at pegging the meters and then each bit defines a sound half as loud. The extra 8 bits define the sounds as quite as nose-hair waving and the self-noise of the microphones. Sampling should make a difference from various secondry isses (anti-alasing filter slopes etc), but bit-depth does nothing but give you more signal-to-noise. And as 16 bit is 96db, and there is only about 70db in even the most dynamic recordings, 16 would seem to cover it.

 

[nick_charles]

Quote:

Originally Posted by CyberTheo /img/forum/go_quote.gif I've been following this thread with interest. Let me throw a wrench. Why not Linn samples? I feel they are good base for comparison, unless you're proposing another theoretical study here that we've seen beaten to death already. All commercial CDs were recorded in much higher bit and sampling rates than 16/44; they were all down sampled to comply to CD (or whatever) format. Even if someone is successful in proving the theoretical aspect to human hearing audibility between 16/44 and higher rate samples, it's of no value to the end user - who are always getting down sampled digital format from higher rate originals.

The problem with the Linn samples is that the high res and red book versions are just not quite the same, they are slightly but definitely different, even within the 20 - 20k range. I have run these through a spectrum analyser and the high res sample is just louder across the whole audible spectrum, about 3db louder on average - I did this at the highest res available where the level was sampled every 5hz.

This subtle difference is enough to render comparisons misleading.

However you can take Linn's high res samples and then downsample them with suitable dither, that would be a valid comparison, if 24 bits downsampled to 16 is audibly different from 24 bits then that would be an interesting finding.

 

[gyrodec]

I'b in ineterested in the results of that test. It shouldn't be audible, but any piece of science is always one experiment away from being found wanting. (Then the real fun starts of trying to work out what really is going on.)

 

[VeipaCray]

The difference between 24 and 16 bit is well.. 8 bit. A third grader could have told you that.

What that equates to in real life...well I'll leave that to the more knowledgeable people who have previously posted on this thread.

 

[nick_charles]

Quote:

Originally Posted by gyrodec /img/forum/go_quote.gif I'b in ineterested in the results of that test. It shouldn't be audible, but any piece of science is always one experiment away from being found wanting. (Then the real fun starts of trying to work out what really is going on.)

Well since I could not tell the difference between 88.2K and 44.1K samples anyway in a blind test I will leave it to younger ears to prove me wrong, or not as the case may be

I may try again when my new super-whizzo 24/96 DAC and headphone amp arrives from China.


For now Linns free samples can be found here

Download our testfiles

NB the 44.1 bit and 88.2 khz samples are not exactly aligned or the same length so some judicious trimming is required.

 

[frankR]

Quote:

Originally Posted by lexnasa /img/forum/go_quote.gif To address your human hearing point, what you may have been duped by is the empirical limit of human hearing, on which CDs were based. However, this empirical limit ignores the physicality of playback. It ignores harmonic interaction. So by removing what you can't hear, it also removes the interaction of what you can't hear with what you can hear.

My advice, you're wasting your time with these guys.

Forty pages ago or so, I posted analysis of 24/96 and 16/44 audio by comparing power frequency spectra which demonstrates exactly what your saving. Namely, 16/44 Redbook audio does not render the audible frequency range as well as 24/96. They continually revert back to a sophomoric half-sampling theorem argument.

Using the Nyquist sampling theorem to claim 16/44 is adequate is a weak argument to begin with. It merely describes the maximum resolved frequency. It is far too inadequate to describe how well sampled a signal is.

I used an imaging analogy to show details throughout the frequency domain are better rendered with greater resolution. I don’t need 1080i HD to watch a baseball game, but it sure is more enjoyable.

It’s also frustrating to hear them say Redbook is all you need when I know SACD sounds unmistakably superior to Redbook.

 

[frankR]

Quote:

Originally Posted by VeipaCray /img/forum/go_quote.gif The difference between 24 and 16 bit is well.. 8 bit.

It's actually 16,711,680.

Ever heard of a logarithm?

That's like saying the difference between a magnitude 6 earth quake and a magnitude 8 is 2.

 

[gyrodec]

frankr - this is the 16/24 bit thread, not the sample rate thread. Nyquist sampling theory has no place here, just db.

Technically, as VC said 8 bit and not 8, and as 8 bit is 16,711,680 (as you stated) he is completely correct. (May be missleading for non-technoical readers, but correct all the same.)

 

[frankR]

Quote:

Originally Posted by gyrodec /img/forum/go_quote.gif frankr - this is the 16/24 bit thread, not the sample rate thread. Nyquist sampling theory has no place here, just db.

We're discussing whether there is a difference between high resolution audio and rebook.

DSD is 1-bit audio.

Quote:

Originally Posted by gyrodec /img/forum/go_quote.gif Technically, as VC said 8 bit and not 8, and as 8 bit is 16,711,680 (as you stated) he is completely correct. (May be missleading for non-technoical readers, but correct all the same.)

No, 8-bit is 256.