IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Has anyone tested SACD vs. CD of same album, e.g. for dynamic range?, [moved from General Music Discussion]
Mikkel
post Mar 4 2013, 17:03
Post #1





Group: Members
Posts: 8
Joined: 4-March 13
Member No.: 107014



Hi everyone,

Just registered because I have a pressing question, which I think only can be answered honestly here at hydrogenaudio. I hope you would be so kind as to help me find an answer.

I have been searching for answer on the question i raise in the topic description: has anyone tested the difference of dynamic range between the two versions of the same album?

I read the double-blind test results from THIS thread. But as far as I can see it concerns 44/16 vs 96/24 and not the difference between the same album but in different distributed formats.

From different places I often notice a comment that says high resolution versions often are mastered differently (and better) than the corresponding CD version. These comments are never substantiated with any kind of measure.

I could of course buy a downloadable album and compare it (yet I haven't). Searching for either SACD or 96 at http://www.dr.loudness-war.info/ does seem to indicate that SACDs (when it comes to dynamic range) does not necessarily translate to better dynamic range when mastering the album. Heuristically, neither does my own collection of CD-only albums to have worse DR-rating than the SACD albums (I listen only to classical music, which may or may not have a bias towards better DR).

Anyway, this is of course no proof of anything. So therefore my question whether any of you know of places that have actually compared different versions of the same album?


Best regards,
Mikkel

This post has been edited by Mikkel: Mar 4 2013, 17:03
Go to the top of the page
+Quote Post
knutinh
post Mar 5 2013, 12:49
Post #2





Group: Members
Posts: 569
Joined: 1-November 06
Member No.: 37047



I believe there are ample anecdotal evidence of "hirez" albums that are indistinguishable (perceptually and/or numerically) from their red-book counterparts.

There are also lots of cases where one red-book release is significantly different from another red-book release of the same album.

So what you would probably want is perhaps to compare a number of albums that are released as both CD and SACD with some kind of DR measure, possibly filtering out remix/remasters that obviously target a totally different experience.

It sounds like a lot of work to me, only to end up with some empirical data that will not matter to the hirez "believers" anyways ("I hear/feel what I hear/feel, no matter what your numbers show").

-k
Go to the top of the page
+Quote Post
AndyH-ha
post Mar 5 2013, 21:35
Post #3





Group: Members
Posts: 2207
Joined: 31-August 05
Member No.: 24222



Even if SACD has a greater potential dynamic range than CD (???), it seems unlikely to matter. A CD that approaches using 96 dB must be a rare creature. It will probably have parts that are inaudible in almost all listening environments. That does not mean that some people might not enjoy it if they have a way to actually hear all of it, but that is probably so small a part of the market that it would not be produced. It has the potential to annoy a much larger number of people because, for them, the recording would just disappear in various parts, under any circumstances that are reasonable for them.

Of course, if the question is just about actual practice,
are some pop albums put out with 8dB of dynamic range on SACD instead of the 6 dB on CD?
the potential isn't relevant.
Go to the top of the page
+Quote Post
Dynamic
post Mar 6 2013, 03:48
Post #4





Group: Members
Posts: 803
Joined: 17-September 06
Member No.: 35307



I haven't tested myself, but an old thread mentioned that Dark Side of the Moon by Pink Floyd had the hybrid SACD/Redbook disc.

The SACD was 5.1 with a large dynamic range. The Redbook CD layer on the same disc was stereo with a lot more dynamic range compression, according to that thread (Dark Side Of The Disc was the thread title)
Go to the top of the page
+Quote Post
saratoga
post Mar 6 2013, 07:10
Post #5





Group: Members
Posts: 4903
Joined: 2-September 02
Member No.: 3264



QUOTE (Mikkel @ Mar 4 2013, 11:03) *
Searching for either SACD or 96 at http://www.dr.loudness-war.info/ does seem to indicate that SACDs (when it comes to dynamic range) does not necessarily translate to better dynamic range when mastering the album. Heuristically, neither does my own collection of CD-only albums to have worse DR-rating than the SACD albums (I listen only to classical music, which may or may not have a bias towards better DR).


Just looking at some of those masters, there is often a difference between CD and DVD-(V/A) versions of tracks.

See e.g.

http://www.dr.loudness-war.info/details.php?id=23919
http://www.dr.loudness-war.info/details.php?id=7144

or

http://www.dr.loudness-war.info/details.php?id=24291
http://www.dr.loudness-war.info/details.php?id=32273

However I suspect that if you listen to classical this won't be as common, as mastering of classical tracks has not been impacted by the loudness wars very much.
Go to the top of the page
+Quote Post
AliceWonder
post Mar 6 2013, 11:33
Post #6





Group: Members
Posts: 119
Joined: 13-July 12
From: California
Member No.: 101393



I will have to look for it but I read a paper with a double blind study using "audiophile" stereo equipment.

What they did is took a bunch of hi-res audio that allegedly showcased the superior quality of high def audio.
They downsampled those tracks to CD quality (16-bit 44.1kHz) and did ABX with a variety of different users, from self described audiophiles to regular people. No one was able to tell the difference in ABX between the un-molested high def audio and the standard CD quality version.

That suggests to me that high def audio is probably a fraud.

My understanding is that due to the math of digital signal processing, the sampling rate only needs to be twice the highest frequency rate we can hear, plus a little head room. 2 x 20kHz = 40kHz so 44.1kHz does that. Additional sampling only provides useful information for reproducing frequencies we can't hear.

My understanding is that the dynamic range of most music could sufficiently be represented with 12-bits per sample, and that all music with 16-bits per sample. In order to benefit from higher bit depth, the song would have to have extreme range AND the volume would have to be turned up high enough that it would cause permanent hearing damage.

But that test I read pretty much sealed it for me. The audio equipment used was very high end, higher end than I could ever justify spending, and people could not tell the difference (could not match X with A or B more accurately than guessing).
Go to the top of the page
+Quote Post
Mikkel
post Mar 6 2013, 12:17
Post #7





Group: Members
Posts: 8
Joined: 4-March 13
Member No.: 107014



@ AliceWonder:
Thanks for your post. I think the paper is referenced in one of the links in my original post.

While we have quite some evidence that we cannot discriminate 16/44 from higher resolution recordings my intention with this post was slightly different.

What I am interested in is if anyone has/had done test comparisons of the same album but from different distributed media; e.g. if Mahler's 5th symphony was distributed on both CD and some high resolution media.

Surfing different fora one runs into the idea that SACDs are better mastered than their CD counterparts. I never see any references to any tests, though... perhaps because mastering quality is a highly subjective topic, I don't know.

Anyway, at least one can compare the dynamic range, which in many cases is a good quality indicator. And comparissons of dynamic range between CDs and their higher resolution counterparts is what I'm searching for.

It would be really interesting to read.


Best regards,
Mikkel
Go to the top of the page
+Quote Post
Mikkel
post Mar 6 2013, 12:26
Post #8





Group: Members
Posts: 8
Joined: 4-March 13
Member No.: 107014



QUOTE (saratoga @ Mar 6 2013, 07:10) *
QUOTE (Mikkel @ Mar 4 2013, 11:03) *
Searching for either SACD or 96 at http://www.dr.loudness-war.info/ does seem to indicate that SACDs (when it comes to dynamic range) does not necessarily translate to better dynamic range when mastering the album. Heuristically, neither does my own collection of CD-only albums to have worse DR-rating than the SACD albums (I listen only to classical music, which may or may not have a bias towards better DR).


Just looking at some of those masters, there is often a difference between CD and DVD-(V/A) versions of tracks.

See e.g.

http://www.dr.loudness-war.info/details.php?id=23919
http://www.dr.loudness-war.info/details.php?id=7144

or

http://www.dr.loudness-war.info/details.php?id=24291
http://www.dr.loudness-war.info/details.php?id=32273

However I suspect that if you listen to classical this won't be as common, as mastering of classical tracks has not been impacted by the loudness wars very much.


Hi saratoga,

Thank you for your response smile.gif . Hopefully, though, if I begin to go through the database (which I probably won't due to too much work already), the differences will be bigger than the examples you list. EDIT: In the examples the difference is 1 ?dbfs?, and in the other the SACD version seems worse when considering individual levels (although the mean is still 10 for both).

Or well, hopefully the difference won't be bigger, because then I can continue to buy the CD or 16/44-version knowing that spending money on the high res format won't matter (much) except to my wallet!


Best regards,
Mikkel

This post has been edited by Mikkel: Mar 6 2013, 12:47
Go to the top of the page
+Quote Post
Kohlrabi
post Mar 6 2013, 14:47
Post #9





Group: Super Moderator
Posts: 1016
Joined: 12-March 05
From: Kiel, Germany
Member No.: 20561



The point is that technologically Redbook CD is sufficient, any improved audio quality on SACDs is due to better mastering. So, high-res formats (SACD/DVDA) don't offer any intrinsic improvement, it's just that sometimes you might get a better master than on CD.

Of course, multichannel audio is a big advantage, if you're into that.

This post has been edited by Kohlrabi: Mar 6 2013, 16:55


--------------------
PRaT is the new jitter.
Go to the top of the page
+Quote Post
Mikkel
post Mar 6 2013, 17:07
Post #10





Group: Members
Posts: 8
Joined: 4-March 13
Member No.: 107014



QUOTE (Kohlrabi @ Mar 6 2013, 14:47) *
The point is that technologically Redbook CD is sufficient, any improved audio quality on SACDs is due to better mastering. So, high-res formats (SACD/DVDA) don't offer any intrinsic improvement, it's just that sometimes you might get a better master than on CD.

Of course, multichannel audio is a big advantage, if you're into that.


Exactly, but the question whether they indeed are better mastered remains unanswered.
Yes, multichannel audio is a huge advantage, so that would make a case for multichannel high res audio (although they could as well have saved us the space and made 16/44 multichannel media).


Best regards,
Mikkel
Go to the top of the page
+Quote Post
greynol
post Mar 6 2013, 17:31
Post #11





Group: Super Moderator
Posts: 10000
Joined: 1-April 04
From: San Francisco
Member No.: 13167



In some cases heavily compressed audio mat actually be troublesome for SACD.
http://www.hydrogenaudio.org/forums/index....st&p=516353


--------------------
Concern trolls: not a myth.
Go to the top of the page
+Quote Post
knutinh
post Mar 6 2013, 22:32
Post #12





Group: Members
Posts: 569
Joined: 1-November 06
Member No.: 37047



QUOTE (Mikkel @ Mar 6 2013, 17:07) *
Exactly, but the question whether they indeed are better mastered remains unanswered.

"better" is a term that means double-blind listening panels instead of simple scripts running on a 1000 albums on a hard-drive.

-k
Go to the top of the page
+Quote Post
saratoga
post Mar 6 2013, 22:38
Post #13





Group: Members
Posts: 4903
Joined: 2-September 02
Member No.: 3264



QUOTE (knutinh @ Mar 6 2013, 16:32) *
QUOTE (Mikkel @ Mar 6 2013, 17:07) *
Exactly, but the question whether they indeed are better mastered remains unanswered.

"better" is a term that means double-blind listening panels instead of simple scripts running on a 1000 albums on a hard-drive.

-k


I think in this context, its clear that he meant "better able to use the dynamic range of the format".
Go to the top of the page
+Quote Post
db1989
post Mar 6 2013, 22:45
Post #14





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



I hope not, seeing as he was replying to a post that quite rightly stated that technical potential is not limiting in either case and therefore is not genuinely relevant.
Go to the top of the page
+Quote Post
Mikkel
post Mar 7 2013, 08:39
Post #15





Group: Members
Posts: 8
Joined: 4-March 13
Member No.: 107014



QUOTE (db1989 @ Mar 6 2013, 22:45) *
I hope not, seeing as he was replying to a post that quite rightly stated that technical potential is not limiting in either case and therefore is not genuinely relevant.


To quote myself:
Surfing different fora one runs into the idea that SACDs are better mastered than their CD counterparts. I never see any references to any tests, though... perhaps because mastering quality is a highly subjective topic, I don't know.

Anyway, at least one can compare the dynamic range, which in many cases is a good quality indicator. And comparissons of dynamic range between CDs and their higher resolution counterparts is what I'm searching for.

So:
1) Yes, blind tests to evaluate subjectively perceived differences in mastering quality between the media would be required but that is not what I am looking for.
2) I limited my question to the use of dynamic range as a quality-indicator. That is: whether SACDs are mastered with better dynamic range than their CD-counterpart; *not* whether one can hear differences between the two masters. This is a relevant question if one has to choose between buying high-res vs. "normal"-res albums (if dynamic range is taken as the only quality indicator, which it clearly isn't. But it is one of the more easy things to test).
3) The dynamic range-testing does not require any blind tests, except if I was interest in perceived quality differences between albums with numerically better or worse dynamic range, or if interested in the issue in point 1 above.

I hope it is clear, what I was asking for. Thanks to everyone for their contributions.


Best regards,
Mikkel

This post has been edited by Mikkel: Mar 7 2013, 08:39
Go to the top of the page
+Quote Post
Mikkel
post Mar 7 2013, 08:43
Post #16





Group: Members
Posts: 8
Joined: 4-March 13
Member No.: 107014



QUOTE (greynol @ Mar 6 2013, 17:31) *
In some cases heavily compressed audio mat actually be troublesome for SACD.
http://www.hydrogenaudio.org/forums/index....st&p=516353


Thank you very much for the link :-). I'll do a read when I'm back from work.

Best regards,
Mikkel
Go to the top of the page
+Quote Post
knutinh
post Mar 7 2013, 14:03
Post #17





Group: Members
Posts: 569
Joined: 1-November 06
Member No.: 37047



QUOTE (Mikkel @ Mar 7 2013, 08:39) *
2) I limited my question to the use of dynamic range as a quality-indicator. That is: whether SACDs are mastered with better dynamic range than their CD-counterpart; *not* whether one can hear differences between the two masters. This is a relevant question if one has to choose between buying high-res vs. "normal"-res albums (if dynamic range is taken as the only quality indicator, which it clearly isn't. But it is one of the more easy things to test).
3) The dynamic range-testing does not require any blind tests, except if I was interest in perceived quality differences between albums with numerically better or worse dynamic range, or if interested in the issue in point 1 above.

What is the algorithm for estimating the dynamic range of a track?

-k
Go to the top of the page
+Quote Post
Mikkel
post Mar 7 2013, 14:31
Post #18





Group: Members
Posts: 8
Joined: 4-March 13
Member No.: 107014



QUOTE (knutinh @ Mar 7 2013, 14:03) *
QUOTE (Mikkel @ Mar 7 2013, 08:39) *
2) I limited my question to the use of dynamic range as a quality-indicator. That is: whether SACDs are mastered with better dynamic range than their CD-counterpart; *not* whether one can hear differences between the two masters. This is a relevant question if one has to choose between buying high-res vs. "normal"-res albums (if dynamic range is taken as the only quality indicator, which it clearly isn't. But it is one of the more easy things to test).
3) The dynamic range-testing does not require any blind tests, except if I was interest in perceived quality differences between albums with numerically better or worse dynamic range, or if interested in the issue in point 1 above.

What is the algorithm for estimating the dynamic range of a track?

-k


I would not know. But perhaps the author of this site (and plug-in) could help you: http://www.jokhan.demon.nl/DynamicRange/
That is the plug-in I have used to evaluate my own collection and comparing with the dynamic range database linked to in an earlier post. As far as I understand it the plug-in compares the RMS level with the peak-level of each track of an album and outputs the mean across all tracks as the dynamic range score. The difference between mean album RMS and peak-level is the dynamic range score.


Best regards,
Mikkel
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Mar 7 2013, 14:50
Post #19





Group: Members
Posts: 3688
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (knutinh @ Mar 7 2013, 08:03) *
QUOTE (Mikkel @ Mar 7 2013, 08:39) *
2) I limited my question to the use of dynamic range as a quality-indicator. That is: whether SACDs are mastered with better dynamic range than their CD-counterpart; *not* whether one can hear differences between the two masters. This is a relevant question if one has to choose between buying high-res vs. "normal"-res albums (if dynamic range is taken as the only quality indicator, which it clearly isn't. But it is one of the more easy things to test).

3) The dynamic range-testing does not require any blind tests, except if I was interest in perceived quality differences between albums with numerically better or worse dynamic range, or if interested in the issue in point 1 above.

What is the algorithm for estimating the dynamic range of a track?


Measurement techniques are oftenbased on definitions.

From Wikipedia:

"Dynamic range, abbreviated DR or DNR,[1] is the ratio between the largest and smallest possible values of a changeable quantity, such as in signals like sound and light. It is measured as a ratio, or as a base-10 (decibel) or base-2 (doublings, bits or stops) logarithmic value."

In audio "possible values" can describe an irrational ratio since at least in digital media, values of zero are possible. We obviously have to throw out irrational data.

If human perception is involved, then the issues of the duration of the extreme values and the possibility of weighting the vastly varying audibility of different frequencies seems to be part of the equation. This is an old issue, and the use of "A weighting" is a classic approach.

In specifying DACs and ADCs the common technique is to process a -60 dB 1 KHz signal and use its weighted or unweighted THD+N + 60 dB as the spec.
Go to the top of the page
+Quote Post
_if
post Mar 8 2013, 23:25
Post #20





Group: Members
Posts: 55
Joined: 18-December 10
From: Ohio
Member No.: 86607



I think you'd have to take it on a case-by-case basis whether the high-res is different or not, but I'd always have my doubts about it until proven otherwise. Very often they're likely to be the exact same mastering. I have one example where the high-res was royally screwed up, though it wasn't an SACD so it didn't have a standard CD layer joined with it: the Queen A Night At The Opera DVD-audio has a nice surround sound mix that was done well, but I was listening to the original stereo mix and thought it sounded like it was clipping. When I ripped it, I found I was right. Severe clipping and lots of it. 24 bits. 96kHz. No limiter. Likely any CD edition of it on the market would sound better.

Bill Frisell's Richter 858 was released on SACD and converted for sale on HDTracks, and that actually does seem to me to sound better than the CD, but not because of dynamic range. It sounds clearer, like the mastering for the CD was kinda grainier in some way (to be clear, I'm not saying that it sounded worse as a result of simply being a CD). But the CD had two releases, one a standard CD and one later as a hybrid SACD, and I don't remember which one I had that I did the comparison to the HDTracks version with and I don't have the CD anymore to check. Maybe the CD layer on the SACD would sound the same, I don't know for sure. Maybe someone here has it and could give a sample so we could find out. I know I've seen someone here who is a Bill Frisell fan.

This post has been edited by _if: Mar 8 2013, 23:28
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 21st August 2014 - 07:32