24/96 audio

[NDA]

I'm thinking of buying a high res audio player - but there doesn't seem to be much choice in terms of what is available to download in the UK.... Licensing restrictions seem to be the cause.

Any experts on high res lurking?

[Crackie]

These test results may help you make your decision.

http://www.hydrogenaudio.org/forums/index.php?show...

Good on you if you want to read the whole article/thread but if you don't fancy doing that, the summary of several hundred test results concludes that listeners in this test, and many other competently conducted tests, hear no improvement when using anything higher than standard Red Book 16 bit/44.1KHz.

I became very sceptical about hi-res a few years ago after reading various reviews for a product called a Zanden Audio 5000 DAC. Several reviews concluded it was, by a considerable margin, the best digital playback they'd ever heard. It is a 16 bit machine which uses Philips TDA1541 chips from 1985; this suggested that higher bitrates and/or clock speeds are not critical factors at all and that implementation ( PSU quality, Jitter reduction techniques, analogue output stage quality etc etc ) are more important factors determining the overall sound quality a digital player.

Edited by Crackie on Wednesday 22 January 12:41

[theboss]

Linn Records for classical and jazz... all the albums I've bought have been very well performed and recorded too (which counts for far more than bitrate)

[NDA]

Having been a serious audiophile in my younger days, a lot of it comes down to people saying "I think I can hear a difference".

Sony are launching some seriously nice HD Audio Walkmans in February and I'm tempted to buy - which is why I'm looking at 24/96. I use Flac files a fair bit and they do sound richer.... But maybe I only 'think' they do. Hard to say really.

[probedb]

NDA said:

I use Flac files a fair bit and they do sound richer.... But maybe I only 'think' they do. Hard to say really.

It's called the placebo effect. Go look it up. FLAC files are just lossless files, they won't sound "richer" than any other lossless format, or the original CD, or actually any properly encoded lossy formats, contrary to what "audiophiles" like to tell you.

Do some proper ABX tests which are documented on hydrogenaudio and across the internet and then see if you can tell a difference between lossy and lossless files for a start. Also bear in mind 24/96 mixes maybe different than the 16/44.1 mixes rather than just being higher resolution versions so are not directly comparable.

But at the end of the day it's your money and you can spend it on what you like

[conkerman]

I have given up listening to my stereo (Esp now I don't have 2 channel) and am quite happy listening to music at 320kbps. I can't hear a significant difference compared to CD.

[P700DEE]

OMG the luddites have arrived in digital audio! With Vinyl reviving and many still believing that 16-bit audio is a compromised format surely any move to 96/24 should be applauded? Listening to music is a very subjective task; we all have differing opinions and even methods of interpreting what is good/right about the music we listen to. Surely the differences between 16 bit 44KHz and 96/24 are measurable? If so perhaps this points to flaws in the ABX evaluation technique rather than that red book CD is a perfect format?

[JustinP1]

Crackie said:

These test results may help you make your decision.

http://www.hydrogenaudio.org/forums/index.php?show...

Good on you if you want to read the whole article/thread but if you don't fancy doing that, the summary of several hundred test results concludes that listeners in this test, and many other competently conducted tests, hear no improvement when using anything higher than standard Red Book 16 bit/44.1KHz.

The problem is not with hi-res audio, but blind ABX testing as a regimen itself.

Those tests (and by this I mean ABX testing for audio) all sound very robust and scientific, but there is a fundamental flaw in that there is never a baseline taken as to how much of a significant change can be accurately tested by ABX testing in that scenario.

I would suggest that PHers subjected to a properly formatted and normalised ABX test would probably not be able to differentiate between DACs, power amps, cables (of any sort), or, shock horror, CD and vinyl.

So, just because an ABX does not prove positive, it does not prove that the above changes do not change or improve the sound audibly.


I've done orchestral recordings in a variety of formats. 16/44.1 24/44.1 24/96 you name it. I would say that - certainly with my gear - there is an audible difference. I also have masters of professional recordings at 16/44.1 and 24/96 and the difference is very clear.

That said, a £100 24/96 PC sound card may not offer any benefit over a very, very good 16/44.1 DAC, and in fact may produce inferior results. The more data and the faster it needs to be supplied the more errors can be induced.

[Finlandia]

I tried the Astell&Kern AK100 portable player and there is a very notable difference between 16/44 and 24/192 tracks. But you may not hear any difference between 16 and 24 if you are using cheap earphones.

[Crackie]

P700DEE said:

OMG the luddites have arrived in digital audio! With Vinyl reviving and many still believing that 16-bit audio is a compromised format surely any move to 96/24 should be applauded? Listening to music is a very subjective task; we all have differing opinions and even methods of interpreting what is good/right about the music we listen to. Surely the differences between 16 bit 44KHz and 96/24 are measurable? If so perhaps this points to flaws in the ABX evaluation technique rather than that red book CD is a perfect format?

I don't think I'm a Luddite and closed to new ideas; quite the opposite I would hope. I work in the R&D department of an audio manufacturer and research into new electrical, magnetic and mechanical design is at the heart of what we do.

The reason for my post above was to point out that the weight of objective evidence I've seen shows that 16 bit / 44.1Khz is not a limiting factor in digital replay and that there appears to be no audible benefit from higher bit rates or sampling frequencies per se.

The merits of the ABX evaluation methodology used above may or may not be flawed however I've yet to see any test results, other than subjective anecdotal opinion, that shows anyone can hear the difference between 16/44.1 replay and anything higher. If there is any non anecdotal proof that high res improves playback I'd be happy to see it; every day is a school day .

This link might be useful for anyone interested in the subject http://xiph.org/~xiphmont/demo/neil-young.html

Edited by Crackie on Thursday 23 January 07:27

[RedLeicester]

Crackie said:

there appears to be no audible benefit from higher bit rates or sampling frequencies per se.

Quick, you'd better tell the recording industry!

[telecat]

HDTracks should be launched in the UK in March. Thy currently only allow downloads to US ip addresses but you can wander on to the site and take a look at their offerings. Peter Gabriel released "SO" as a 24/48 FLAC and also MUSE released "The 2nd Law" as a 24/96 download available from their Online Shop. I burnt it to DVD-Audio and it does sound very good. There are also a few European Sites selling DSD (SACD files) and Hi Band PCM downloads.

There is Also the "Pure Audio" Blu Ray Music discs. 24/96 and they have just released another batch that includes Tears for Fears "The Hurting".

Hardware wise There are now more than a few DACS available. The latest being the Chord QuteEX. It Accepts most PCM bit streams up to 32/384 and DSD 128 which runs at twice the data rate of DSD itself.

Personally I find that well recorded Music sounds better at higher band widths. However you cannot polish a T*rd.

[sparkyhx]

I haven't heard the high res stuff So my view is somewhere in the middle. You are probably right, the average man in the street cannot tell the difference, but I think people who attuned to 'listening' probably can.

My understanding albeit very ancient understanding was that 16 bit recording artificially cut of the high and low frequencies that were inaudible. The arguement was that music wasn't just about what could be heard, it was about what could be 'experienced', ie. certain frequencies affect the body. and that High res went some way to addressing this.

As we are on a car site I'll use an analogy, mess with suspension on a car and a test/racing driver will be able to tell something is different. Mr joe public would be largely unaware.

Couple that with different levels of amp and speakers etc and its not unreasonable that high res is largely pointless to most people.


I'm just slowly moving over my music to purely digital and am downloading FLAc's for which I have vinyl. And the difference in quality of the FLAC recordings is astonishing , I've no idea why some sound so poor.

[JustinP1]

Crackie said:

The merits of the ABX evaluation methodology used above may or may not be flawed however I've yet to see any test results, other than subjective anecdotal opinion, that shows anyone can hear the difference between 16/44.1 and anything higher. If there is proof that high res is actually better I'd be happy to see it; every day is a school day .

I know this is subjective and anecdotal, but it's as high quality as anecdotal views can be.

Some years ago I had the pleasure of working with some very high end recording engineers and producers in probably the most technically lauded studio in the world at the time. I was writing for my degree at the time, so I did manage through the work day to pick their brains.

At the time, 1999 (IIRC), 2 inch, 24 track magnetic tape was still the staple here, with the alternative being 24/96 through converters. The opinion of the engineers at the time, specifically including someone who you could say has the most 'golden ears' of anyone (he's won the Grammy for best engineer at least 6 times) was that although he still preferred the fidelity of magnetic tape over digital, 24/96 was a big step over 16/44.1 and was pretty damn close enough.

We're not just talking about recording here, because clearly he's listening back to finished mixes.

Of course the sampling rate is not the only factor in improving sound reproduction and it's not clear cut, however, I have seen and heard enough to know there is a difference. These engineers, and indeed the whole industry are not a victim of a huge con.

[Jobbo]

RedLeicester said:

Quick, you'd better tell the recording industry!

There's some benefit to the recording industry - they don't need to worry about getting the levels just right to avoid clipping and noise, so any old monkey will be able to do it

[Crackie]

RedLeicester said:

Crackie said:

there appears to be no audible benefit from higher bit rates or sampling frequencies per se.

Quick, you'd better tell the recording industry!

Hi RedLeicester - Not sure if your post is sarcastic or not but I did put replay in bold for a reason.

[RedLeicester]

More tongue in cheek than sarcastic

It's nice to hear some sense for a change - I work in 24/96 all day every day, with only final masters dithered down to 16/44.1 or 48 (depending on usage obviously), and I'm tired of the whole hifi/audiophile tripe spouted about "HD" audio. Even the "16bit is naff" thing is utter rubbish based upon ropey convertors from the 80s where even the expensive ones would be comprehensively trounced by a boggo Tandy CD player of the 90s or later.

[RedLeicester]

Jobbo said:

RedLeicester said:

Quick, you'd better tell the recording industry!

There's some benefit to the recording industry - they don't need to worry about getting the levels just right to avoid clipping and noise, so any old monkey will be able to do it

[Crackie]

JustinP1 said:

Crackie said:

These test results may help you make your decision.

http://www.hydrogenaudio.org/forums/index.php?show...

Good on you if you want to read the whole article/thread but if you don't fancy doing that, the summary of several hundred test results concludes that listeners in this test, and many other competently conducted tests, hear no improvement when using anything higher than standard Red Book 16 bit/44.1KHz.

The problem is not with hi-res audio, but blind ABX testing as a regimen itself.

Those tests (and by this I mean ABX testing for audio) all sound very robust and scientific, but there is a fundamental flaw in that there is never a baseline taken as to how much of a significant change can be accurately tested by ABX testing in that scenario.

I would suggest that PHers subjected to a properly formatted and normalised ABX test would probably not be able to differentiate between DACs, power amps, cables (of any sort), or, shock horror, CD and vinyl.

So, just because an ABX does not prove positive, it does not prove that the above changes do not change or improve the sound audibly.

Conversely just because an ABX does not prove positive, it does not prove that the above using 16/44.1 instead of 24/192, for example, limits or reduces sound audibly.

Imho the test results from the "16 bit bottleneck" link are much closer to real evidence rather than anecdotal evidence. To summarise, regarding whether various professional recording engineers, students in a university recording program, and dedicated audiophiles could detect a '16 bit bottleneck' ( 16 bit A/D/A loop ) inserted into a high end, hi-res system. " The test results for the detectability of the 16/44.1 loop on SACD/DVD-A playback were the same as chance: 49.82%. There were 554 trials and 276 correct answers."

It would be logical to expect some minor preference for 24/192 bearing in mind the 16 bit bottleneck signal had to endure 2 additional conversions ( 24/192 to 16/44.1 and back again ).

You have far more experience with the recording industry than I do and you have clearly heard improvements when using higher res equipment. As you know there are a huge number of factors involved in sound perception and my point is that the perceived improvements you hear are not necessarily the result of formats with bigger numbers in them. Confirmation bias is a big factor.

The Audio Poll results from the following, admittedly unscientific test, achieved virtually the same 50/50 split/guess/coin toss as the Hydrogen one above.
http://www.trustmeimascientist.com/2012/05/05/resu...

I'm interested to know what type of testing you would advocate as an alternative to ABX ?

Edited by Crackie on Wednesday 22 January 14:11

[JustinP1]

Crackie said:

You have far more experience with the recording industry than I do and you have clearly heard improvements when using higher res equipment. As you know there are a huge number of factors involved in sound perception and my point is that the perceived improvements you hear are not necessarily the result of formats with bigger numbers in them. Confirmation bias is a big factor.

...I'm interested to know what type of testing you would advocate as an alternative to ABX ?

I totally agree. I see it from both worlds as a recording engineer (and owner of a company) who only deals in real world performance and value for money improvement. On the other hand, for fun, I have a home set-up which is audiophile standard. Of course, hi-fi shops have a rather different ethos to my own!

I totally agree with confirmation bias too - that's my major bug-bear with comments on the forum that say I've just installed X and it's night and day from my old Y. In my own work, I've seen (not me of course! ) recording engineers play 'the Emperors new clothes' on a mix with an artist or record label executive who wants more 'sheen' or 'whack' in a mix. The numpty comes back, you tell him you've remixed to give more sheen/whack and they love it and think they are a genius. Of course, they are listening to the ams mix!


I've racked my brains to offer a testing regimen to prove how component/cable X changes a sound. The difficulty is that much of hearing is perception, and the ABX test also is a test of memory, and unless the change can be categorised clearly (like trumpet out of tune) then hearing the difference between two pieces of audio is a fight with the psyche and not your ears!