24-192 a scam?

Your best bet if you want to compare, is take 192 KHz, 24 bit material, then convert your self (with a reputable program) down to 44KHz 16 bit, then do your comparing. This way you know you are comparing the same effective overall loudness, as often mastering for different media can compress the dynamic range. If you have an ABX program you can try to locate which of the two (192 or converted 44KHz) you prefer, or you might not be able to hear any difference between the two.

Whether one different issue of Carlos Kleiber's Vienna Philharmonic recording of Beethoven's Fifth Symphony would be genuinely significant, would depend on it it being a transcendental revelation of a performance.

His father Erich made a much finer performance - recorded on Decca - decades earlier, and it is not only musically much finer, but recorded in a much finer musical balance. DG did not learn enough from Decca how to record an orchestra. But the performance is the thing ... Priceless.

Why this chase for the best version of something that is far from the most significant in performance terms?

Makes me wonder if sound is the main aim, rather than what the musicians do with it ...

ATB from George

Interesting that mrspoon should mention resampling ;  the caveat below is in the paper by Stuart and Craven that Simon pointed me to (page 12) :  

If a recording has been archived at 192 kHz and it is
required to produce an 88.2-kHz version, a suitable
procedure would be firstly to convert the sample rate to
176.4 kHz by conventional means, using severe filtering
to suppress aliases, and then to convert to 88.2 kHz by
the methods described here. This second conversion can
be expected to provide substantial suppression of ringing
and other artefacts near 88.2 kHz caused by the first
sample rate converter, so one may hope that the audibly
deleterious effects of conventional resampling will
largely be avoided.

Which makes me wonder if the differences sometimes heard between 16/44.1 versions and their higher resolution originals are due to artefacts in the resampling... or am I missing something here ?

24/192 isn't a scam, but it is a bit 2012.  DSD and DXD are where it's at now....

Being serious, my experience tells me that hi res media work best on modern recordings which (I presume) have been recorded and mastered for the intended format from the outset.  So the Kleiber (which I have and sounds great on 24/96) is unlikely to sound much better on 24/192.  Conversely, the recent Schumann symphonies from Linn are for me, the very best advert for 24/192.  But then again, native DSD recordings, for example, Ivan Fischer Bruckner and Mahler symphonies on Channel Classics, show the superiority in turn of DSD over PCM.

You can't research what sounds better online because the only person who can answer that question is you. However, you can inform your decision to investigate further or not, and to try to match versions so that you are comparing like with like. Google is your friend. Also there are forums, a good example being Computer Audiophile where this information can often be got. If it can be got. The labels can be obtuse on occasion. Like many things, what you get out of the research will hopefully be proportional to what you put in. 

If you just want to satisfy yourself that a difference can be heard it might be best to stick with known quantities where want goes into the HIRes is known to be the same as a particular mastering which is still available in 16/44. Some music is available as both 96 and 192 so obtain examples of each. You can buy single tracks in many instances. The easiest material to test may not be your favourite music but there is enough stuff out there to allow you to form an opinion based on what you actually hear, as opposed to what other people tell you you should be hearing.

And once that's all at cruising altitude, we're back to providence, best mix, remasters to avoid, the fact you can't polish a turd at any resolution and all the usual stuff, which is thrown into the HiRes debate as a caution but is not specifically applicable because providence and quality have been paramount since the beginning of recorded music and are no more or less important at 24bit.

I wish you some fun. It's important that fun is had.

I believe a better  comparison is between 24/192 and 16/48 rather than 16/44.1, as then the aggressive filtration isn't required.  I believe that there will be no significant difference between 44.1kHz and 48kHz sample rates (and if there is a discernible difference it will favour the 48kHz version).

For mastering of the CD, I would hope that the conversion is done as described via 176.4 and that complex dithering algorithms are used so that the noise shaping will redistribute the artefacts to higher frequencies, mainly above the audio spectrum band.

George

I agree and disagree!

I agree the performance is paramount.  I qualify that by saying that a better sounding medium can be revelatory.  The difference between, at the extreme, say 24/192 and cassette can make you listen totally anew to a recording you thought you knew

I disagree about the Kleibers' 5ths.  I've always preferred Carlos's but you have made me go back tonight and listen again to Erich, admittedly only in 1644.  I'm now going to listen to Carlos's 1644

Huge, I'm writing this without referring to my books on this and having enjoyed a few beers  so forgive me if not entirely spot on.... , but dithering is one thing and really addresses aliasing. Oversampling usiing zero value samples has the wonderful ability of reducing the effective timing noise (jitter) in the original pass band and pushing the oversampling noise well above the passband so as to more easily filter and reconstruct. The  random noise energy  from the recording / ADC jitter is constant and by oversampling on the DAC process you are distributing that timing noise across a wider spectrum, so you have effectively reduced the noise in the original pass band when reconstructing the signal. I think that is really neat, and reasonably straightforward to folow mathematically if you are that way inclined.

Dear JSH,

The fault is that Carlos makes the entering three quavers a set of triplets from a down down beat [the second of the 4-4 bar], and it is not.

It is an upbeat quaver onto the second beat plus two more in four time  in the second beat onto the third strong beat as a start. Repeated [rhythmically] as a held fermata. It is incredibly easy to get right actually with a silent strong first downbeat in time. The only way to get it right rhythmically. 

Before the performance even starts Carlos is sunk as a basic failure to launch one of the more or lessTtinanic symphonies with basic rhythm failure. He hit the iceberg less than two seconds in ...

Apart from that the actual sonic is grey and typical of DG at the time. One could get over this if the performance was more than a pale imitaion of his father's effort.

And the Concertgebeow was a much finer band in the early fifties than the VPO when Carlos got there ...

Just opinion of course ...

ATB from George

Not a great transfer on youtube, but the Decca CD is splendid, and you soon will hear that the son had only a half respect for Beethoven's text, and the father really did grasp the Titanic element, and his Dutch orchestra responded as only such an orchestra could. Not Berlin. Not Vienna, but even more acute to the sensibility. 

https://www.youtube.com/watch?v=1fnq4u1YL2g

Simon, you are correct and don't need your books:

Dithering addresses the aliasing that occurs when down-sampling by non-integer ratios.

Doing the non-integer ratio down-sampling at higher sampling rate allows the use of oversampling and noise shaping to be applied; this ensures that you don't re-quantise the original quantisation noise (which would otherwise add, at least some degree, due to the non-integer ratio).

At least that's the theory I found when I last looked into it!

And this is how Beethoven Five really goes. 

https://www.youtube.com/watch?v=NgrxLR_8PEE

1955. London, Klemperer.

ATB from George

George

Naughty; we're going off topic!

I'm not sure about the triplet; I think the rhythmic drive is there and that's what matters.  But listening again, for the first time for a while, one or two of the dramatic pauses are a little posed

Carlos is more operatic in approach; dramatic, flourished etc etc.  I heard him conduct the 7th with the LSO in about 1979 at the RFH (the LSO were awful and he never conducted them again!) and we came away exhausted by the sheer intensity.  Almost too intense and listening to this again re-inforces that.  Erich/ Carlos perhaps a bit like Furtwangler/ Toscanini?

The 1955 Klemperer is amazing ( from memory).  The 1962(?) one which is part of the set never quite achieved that; the 7th even more so

The Carlos sound is grey as you say; this 16/44 doesn't bite me.  Which is where I came in; would the 24/192?

I had no idea Brits called eighth notes "quavers."  I learn a lot here!

Hello,

i did not read the whole post. So i maybe write something that some allready wrote.

You dont listen to your ears you do it with your brain. You cant propperly hear a consistant tone up to 20khz. It depents on your age and a few more Parameters........

However, Music does not exist with long consisant tones, this would be a pain in the a...Music are very short tones. A human brain is capable to listen to a tone with a length of 5µs, which is equivialent to 100khz.

So yes , we can hear and difference the differnt Audio Resolution files. Buit i have to admit your ears/brain needs Training

Mike

Not really!

The mastering of the CD issues I quoted are in hi-res [actually 24/96], and the CDs are obviously at Redbook standard. They are far finer than the original LP issues.

So as you can hear even on youtube [MP3] the proginal recordings are the point, not the sampling rate, and of course the competence of the remastering ...

But in a way the real point of this whole forum - let alone this thread - is great music in great performances ...

So very much ON topic really ....

ATB from George

I've realised that there's an interesting contradiction here.

Several people have claimed that the capability of Redbook exceeds the capability of the human ear-brain combination in every way.

Other people claim that vinyl is better than CD.

...interesting...

If it were the same people making both points it would also be amazing!

ATB from George

I've just finished listening to the Pono 24/192 release of Neil's own Harvest and I have to say it sounds amazing! Somehow deeply satisfying. In fact I only intended to listen to a few tracks but couldn't help letting the album roll out to the end. The technical detail of the hi-res debate is largely over my head, but I realise there are some potentially scientifically convincing arguments for why it shouldn't make any difference. What I just listened to however, seemed to have a quality that I don't often if at all associate with 16/44. Perhaps there is another explanation, but I'm certainly one step closer to being a believer and this definitely does make me more interested in buying more hi-res material. Good on Neil Young for what I perceive as a genuine (and possibly quite effective, time will tell) effort to bring musical enjoyment to people in a deeper and more meaningful way.

I'm a little surprised there hasn't been more discussion about Pono here. Sure the catalogue may be a bit thin at this early stage, but as far as I can tell the store seems simple, honest and legitimate. There doesn't seem to be a huge number of hi-res releases on there at present but hopefully that will change.  

There are no scientific arguments as to why one resolution might sound different to another. Just people quoting technical information and basing an opinion on it and/or on anecdotal observation. Until a mufti centre double blind controlled trial in some thousands of subjects has been published which shows that an effect can or cannot be heard, no progress towards an explanation will be made. Until then we are stuck with doing what pleases us and enjoying ourselves. Shocking. Shouldn't be allowed. I know some of us would rather be told what we are hearing but you can't please everyone.

I agree with most of what Harry says......but it doesn't help with the question of whether it is worthwhile buying 24/192 issues when one has an earlier 16/44 or whether 24/192 is worth the extra money for new purchases

My take on the fascinating discussion above is that 24bit is worth getting where it is the format for an original recording, and for re-masterings of older recordings where the extra headroom helps the re-mastering engineer.  But the increase in bit rate from 96 to 192 is a very, very marginal gain and so probably generally not worth it, with pot luck playing a part.

Question:

Which would you rather listen to? A great musical performance on a 78 recording, or dull as ditch-water music in a technically perfect recording?

If you are happy to listen to eat music on a 78 rpm recording then you probably do not need to worry about high resolution compared to the CD standard.

If the music is so dull that it sounds dull on a dull old CD, then it will not sound less dull on a high resolution release! But it may sound more hifi-isn. I'll explain that one if you fail to grasp the point.

I would also advise that you do some investigating with someone who has some high resolution music recordings and the same recordings in CD releases. If you find yourself enjoying the hi-res versions better then go for it. There are no guarantees, one way or the other!

Only a thought! ATB from George

I saw today on a What HiFi post that Tidal are planning a hirez streaming service. They expect the price to drop for the current service and for hirez to sit above that. Will be interesting to see how that develops and whether people sign up or not

Sorry I have to take issue with that.

There are hypotheses that indicate that possibility.  To state that there are no arguments, is to assert that all these hypotheses have been dis-proven and that is not so.

The possibility exists.

It's true that, as yet, the hypothesis that differences can be heard remains unproven.

It's also true that, as yet, the hypothesis that no differences can be heard remains unproven.

Unproven means that we must release the prisoner in the dock!

ATB from George

There is some more discussion about that here: 

https://forums.naimaudio.com/to...-mqa-hires-streaming

One thing to note is that in order to get the best quality stream you will need either a software or hardware MQA playback device.