24bit vs 16bit, the myth exploded! - Page 27

Another point on SNR. If you take the theoretical value on a 20-20khz rms basis, and compare the rms voltage of the cart with a 0db sine vs the total rms thermal noise, you get

20*log(0.8e-3/(0.13e-9*sqrt(30*20000)))/log(10)

= 78 dB

This, it must be said, is 2db better than my OC9 (which for 0.4db/12 ohms yields 76db). On a spot basis that improves to 121db, but CD-DA evaluated on the same basis would yield something like 133-139db depending on how you treat the RMS values.

If you are particularly generous, and allow the use of velocity peaks extending to +15dB, then Clearaudio's numbers more or less equal out. But that's not what they're claiming..... I mean, given that (like I said before) the fact that I am recording 10khz 0db tones at -27dbFS and yet still clipping on some music does imply that records are out there which are punching 110cm/s, but much of that velocity only comes about due to tracking/tracing distortion - not because the cutting head is actually moving that fast to begin with.

Great write-up gregorio! I learned quite a few things there..

I've learned so many things from this thread. Thank you very much, G!!

Now, I am totally convinced that it is a total waste of space to store 24/192kFz materials. What software can I use to down sample these, so there is absolutely no audible differences between this and downsampled 16/44.1k?

G is gone, he seems to believe that we're too retarded...maybe we are, after all

some manufacturer posted some supposedly 16/24bit files w/ a huge SQ difference, but the mods killed it?! http://www.head-fi.org/forums/showthread.php?t=440388

it was called "Real-world comparisons of 24-bit and 16-bit music from Linn Records - free giveaway!"

You must be referring to this.

LinnRecords - The 24bit Comparison - Alyn Cosker - SoundCloud

exactly! can you hear a huge SQ difference? I can't say that I can...I'll try again tomorrow w/ fresh ears

Well anyone can post two different sounding files and call it "16/24 bit real-world comparison".
That's why I asked for the original source.

edit: Will also do some ABX tomorrow, but I will take the 24/96 file as source and convert it myself. I really don't trust these guys.

well, as I wrote in their thread before it was closed...they should indeed let us know how these 2 files were created..

if they come from the same 32float source? same dithering algorithm? if the 16bit was downconverted from the 24bit?

I've got some 7.1 24/48 TrueHD lossless soundtracks that truly lack in 16bit...but it's due to the conversion, integer>integer is lossy as hell.

Yeah, well I did some tests with their 24/96 file. I converted it to 16/44.1 (with fb2k, dithering enabled) and created an audio diff with DiffMaker. The resulting file shows what I expected: no difference.

They said they had two different masters. When I ran some stats on them in Cool Edit Pro they were a bit different the 24 bit had 6db more dynamic range, but both were under 96db anyway. Looking at the spectrograms in Audacity the 24 bit hits close to max (0db) far more frequently than the 16 bit, this would make it appear more dynamic perhaps. These do not look quite the same to me.

I downsampled with CEP and compared the 16 bits and 24 bit-down-sampled versions and then plotted the differences they average out at 0.18db, varying from 0.08db to 0.37db, but that may be my process, even so the differences are quite small though 0.18db might just be audible as an average loduness difference but I doubt it.

Not sure that these samaples prove very much one way or another ?

That was us. Linn Records. We didn't actually say there was a huge SQ difference.

Don't now why the mods killed it ... perhaps I didn't ready the T's and C's closely enough, if anyone could shed any light on that it would be appreciated.

I suppose I should have pointed out that these are the exact files as we sell them on through our website, hence the reason I called it a 'real-world' comparison. We have not doctored the files in any way for effect, in fact quite the opposite...

The 16-bit file is the same as we press to our CDs which are all hybrid HDCDs. Hence it goes through a slightly different mastering process and is encoded for HDCD. We actually try and get it as close to the 24-bit Studio Master as we can, within the limitations of the media.

It is of no benefit to us to dumb down our CD releases in order to highlight the quality 24-bit files, that would just be daft!

The whole idea of the 24-bit Studio Master format is to get as close to the original recording as we can, and have that reproduced for the listener at home. Yes you do need to have a level of investment in your hi-fi and listening environment for it to make the difference - having an in adequate DAC or soundcard can narrow the difference so much that there is little point. We do hope, and believe, that the technology will catch up, and that's part of what Linn as a hi-fi manufacturer aims to do too.

My self and our chief engineer, Philip Hobbs, are devising a more rigorous test from an original analogue tape that should satisfy the audio engineering geeks amongst us. That is a little while away though, as Phil has a very busy recording schedule at the moment.

It should be said though, that we still think CD quality is great. I love my CD collection! And all of our releases are available on hybrid SA/HDCDs.

There are listeners who want the to get as close to the original recording - as the engineer, and musicians intended it - though and have the hi-fi capable of reproducing it, and for them we have 24-bit ... (or of course vinyl ).

A bit harsh! We have been making audiophile recordings since 1983 and take it very seriously. We aren't about to fudge something just for some slim marketing advantage. If you wanted to check me on that, download the tracks direct from the website and then analyse the HDCD, it's just what we have provided through this giveaway.

24-bit isn't for everyone, much like vinyl it takes some time and investment, but for those who are interested the results are beautiful. And when the technology catches up, and storage prices drop, you won't have to buy your music over again!

Jim - Linn Records

ok thanks for the clarifications...but HDCD is not true 24bit, it's actually dithered 20bit AFAIK?

and saying that the only 16/24bit difference makes a very much audible difference in this very thread might require some more informations from your side

Not quite sure what you mean here! The 16-bit file comes from the HDCD master, and the 24-bit has its own session.

Ultimately the only thing that really matters in music is an audible difference surely! I'm not sure if most people care about a visual difference to an audio waveform! :

This has been a very interesting and enjoyable thread, but some of the initial statements on which it was borne are slightly, well, wrong. We going to put together a document setting the case for 24-bit, from an engineering/science end. Just give us a little time - we have a packed recording and release schedule.

The proof of the pudding is always in the listening (and enjoyment) and is always subjective. Hence the 24-bit comparison series we have started.

The objectivity we will get to later,

Jim - Linn Records

Yeah maybe that's a bit harsh, but I didn't want to doubt your competence or skills in any way! /apologizes

The reason, that I was so skeptical, is that there was no information about how these files were "created" and from which source(s). (maybe I've overlooked something?) Also the post looked a bit like advertisement.

As you've said, both files were mastered differently so it's not a genuine test "to hear the difference between 16 and 24bit" (as stated on that soundcloud.com page). To really just compare 16 to 24 bits, the files should have a) the same source, b) same sampling rate and c) the conversion process should be specified.

Nevertheless, I enjoyed listening those samples and I usually don't listen to jazz.

oh ok..so it's not 16bit either as HDCD uses the 2 bottom bits of CDDA infos to encode HDCD 20bit dithered data AFAIK...and also the audio is +6dB louder than when it's properly HDCD decoded.

so this indeed seems to be far from a 16/24bit genuine comparison?! it'd appear to actually be a 14/24 bit comparisons w/ different masterings....so please keep us posted whenever you will be able to provide us w/ genuine 16/24bit files dithered from the same exact 32float source w/ the same exact mastering/dithering algorithms.

that's what annoys the OP of this thread, companies push 24bit....but the mastering is much different, and in that specific case it's even worse because your 16bit file has been HDCD encoded so it's not even true 16bit