Sound Quality/Music Reviews and Comments on SACD/DVD-A - Page 11

Any audible difference would likely be caused by differences in mastering and not by the delivery format.


The point with reissues is that you can remaster the material and fix problems (or create new ones) with earlier issues. Sometimes the reissue will sound worse than earlier issues and sometimes it will sound better.
It's all about the quality of the mastering and it makes little to no difference (as long as it's a mono or stereo recording) if the material is released on DVD-A, SACD or CD, regardless of when or how it was originally recorded.

Here's some more info:
http://people.xiph.org/~xiphmont/demo/neil-young.html

^^^^^^ This may stir the pot.

Very interesting/educational article. Fully concur with mastering having a much greater impact over a particular and acceptable delivery container (flac, cd, dvd-a, sacd).

I have to admit, I'm in the "It just sounds better" crowd...comparing same mastering (not title), found on many hybrid SACDs (i.e. MFSL). Not subtantiated by a correctly performed ABX test...just my biased perception. My brain may be overriding what my ears hear, but the article won't keep me from buying or enjoying my "well mastered" Hi-Res stuff, even in 2CH.

The combo between my ears, my brain, my system, hi-res discs/flacs in my room works for my enjoyment ...I do have some 16/44 redbooks that sound much better than some SACD/DVD-A too.

This was why I bought the 96/24 and 192/24 versions of the same track: to see if I could hear any difference. If not, I would buy the less-expensive 96/24 version, and could use standard DVD-Video disks, which are easier to master and work in any DVD player.

It was with some chagrin that I found the difference between those two versions to be clear.

This was not the first track I tried this experiment with - the first was with a contemporary digital recording of the classic jazz standard "Black Coffee" by singer Louise Rogers, accompanied only by an acoustic guitarist. The difference there was even more pronouced than with the Bill Evans track: the 96/24 sounded like a really great recording, while the 192/24 sounded like hearing the musicians playing live.

The Evans test was a "stress test": would there even be anything on an old analog recording to enable me to hear the difference?

Physicists may opine all they like, but the difference was there - to the point that when the 96/24 version started to play, my fiancee - not an audiophile - asked me to please let her listen to "the good version" again instead.

PS No, I didn't do a computer-driven ABX test - but that would put me at the mercy of my computer's sound card. The whole point of making disks is to get the advantage of my Denon's electronics. I did the next-best thing: I alternated versions on a disk a few times to see if I heard a difference when jumping back and forth with the remote.

If there was no audible difference, that would have been clear. But there was.

The problem is that you don't know if the mastering is identical. Even a slight difference in level is enough to make a comparison problematic.


And there's nothing wrong with that because you're more likely to find well mastered recordings among the "Hi-Res" releases since they're usually produced with audiophiles in mind.
The point was that the key to good sound is in the mastering and not in the format.

With respect to the tracks I used for comparison, they came from the same source - HDTracks - and, at least with respect to the Bill Evans track, the description was such that the two tracks would appear to have done the same way, just run at different sample rates. No clear reason why that wouldn't apply to the Louise Robbins, which was a new recording - and the level seemed the same on each pair of tracks.

I downloaded the tracks in question and found (from the spectrum plot) that, as far is I can tell, there is no real difference between the versions except for some faint distortion at ~50kHz in the 192kHz file. There is no significant content (except for what I suspect to be the tape BIAS signal at ~29kHz) above 20kHz in any of the versions:

192kHz


96kHz



I don't doubt that you experienced a difference, the question is why you experienced it.
There are many possible explanations, like: placebo (you need to do a proper ABX-test to rule it out), problems in your playback equipment (the Denon player may not handle all sample rates properly), problems caused by the software used to burn the DVD or maybe there really are differences in the mastering.

You can't draw any conclusions in this specific case yet.

Those two plots don't look the same to me. The 96k plot looks fatter vertically almost across the board, while the 192k plot looks more sharply-defined.

That vertical spread could well correlate with the "impulse current" injected into the signal by sharp Nyquist filtering.

A few months ago, I found an excellent five page article published 15 years ago, in September 1997 - attached - theorizing why 96/24 sounds better than CD's 44/16 even though the Nyquist rule says that a 44khz sampling rate should pass frequencies up to 22kHz.

It turns out that the sharp roll-off used to remove audio above 22kHz (or even 24kHz) has the nasty side-effect of injecting high-frequency ringing energy into the signal (it has to go somewhere) which has a wide temporal spread of about a millisecond.

His theory is that has the effect of interfering with and thus "temporally smearing" the high frequencies whose arrival times we use for the imaging of sound sources. (It also may make the signal more tiring to listen to - my additional inference.) At 96/24 the ringing energy's temporal spread is about 100 microseconds - only a tenth as wide, and at 192/24 it's half of that: 50 microseconds.

I have a theory that some of the problems with lower bit rates could be avoided somewhat by sampling at a high rate like 96 or 192 and then downsampling. This would side-step the ringing current issue at lower bit rates by doing the Nyquist filtering up high and then - in the digital domain - downsampling. On this theory, the 96k Evans track would have sounded better by starting with the 192k sample.

I mentioned this idea to a friend who is knowledgable about digital audio who replied that recording studios standardized on 96/24 quite a while ago - so I may be reinventing the wheel here, but this may account for why modern CDs sound better than first generation CDs that were recorded using a sample rate of 44/16 in the first place.

PS - You didn't mention whether you heard any difference between the two versions.

 

Why high bit-rate sampling sounds better.pdf 288.904296875k . file

Err... no it's because the scale is slightly different. The 96kS/s plot goes to ~47kHz and the 192kS/s plot goes to ~95kHz (and the scale is, of course, logarithmic).
It has nothing to do with filtering.


That depends on the type and quality of the filter and its implementation. With properly designed hardware/software there is no audible ringing when running at 44.1 or 48kS/s.

There are, of course, poorly designed/implemented filters on the market and they can cause audible problems (including those that are much worse than ringing).
In that case they may work slightly better at higher sample rates, but there is no reason to use them in the first place.
The fact that some designers are incompetent isn't much of an argument for higher sample rates.

There is also no evidence that ringing is a major reason (though it may be in some isolated cases) behind perceived audible differences between sample rates.


Almost every converter on the market for the past 20+ years has used delta/sigma (or similar) modulation and operate at MHz-speeds. There are no modern A/D-converters that sample the signal at 96 or 192kS/s.
Running at MHz-speeds means you can use an extremely gentle analog filter and do the critical filtering (and decimation to the desired output sample rate) in the digital domain.

You could say that a superior version of the solution you propose has been standard practice for more than two decades.

And here's a good introduction to digital filters:
http://www.users.qwest.net/~volt42/c...ng/Filters.pdf


The DAC i have connected to my computer at the moment doesn't support 192kHz, but I'll try to listen to them the next time I bring home an audio interface that supports it.

I thought that differential Megaherz sampling was a DSD/SACD specialty. This is the first time I've heard it said that all modern sampling is done that way. Are you saying that all PCM starts out as DSD?

There's just one thing that confuses me [striking best Colombo pose]: if the DAC connected to your computer doesn't support 192kHz, what was feeding that signal to the spectrum analyzer?

DSD is actually based on the first generation of delta/sigma A/D-converters. They used 1 bit and ran at a speed of 64x (resulting in 2.8224MHz of you use a base rate of 44.1kHz), which is why DSD is 1bit/2.8224MHz.
It is basically the modulator signal (the signal before the digital decimation filter) of that type of converter.

The original purpose of DSD was as an archival format that could be converted to any of the standard PCM-formats. It was never intended as a playback format and that caused quite a few problem when Sony later turned it into one (SACD).
Basing a format on the technology available at the time makes it hard to keep it in line with future developments.

1bit/64x delta/sigma was abandoned by the industry for higher rates and eventually more bits (the first multi-bit delta/sigma converter was released in 1999) as well.
Today the converters run at speeds of at least 5-6MHz and use at least 3 bit quantization.

So PCM doesn't start out as DSD (unless you use one of those old 1bit/64x A/D-converters), but DSD (which is actually a form a PCM) is based on an early implementation of the technology used in most modern converters.


The signal never left my computer so no DAC was involved at all. I used an audio editor called iZotope and it supports 192kHz (I think it supports sample rates up to 384kHz).

If I want to use my PC to sample an old analog tape to disk, I go through the line input of its Sigma Tel audio circuitry on the motherboard - with its "HD" driver, which I've read on the net is a 192/24 driver - or I could use my laptop's plug-in Echo Indigo io PC card, which is rated, at Echo's site, at 96/24.

Audacity can be set to record at up to 192/24.

Are all of those really sampling at megahertz speed and then downshifting to kilohertz speed, or are you only talking about professional recording studio equipment?

Bill,

Thanks for the review and getting back on topic...Cannonball's - Somethin' Else is excellent as well (with Miles Davis). You loose the Bill Evan's connection, but it's a classic and AQ is excellent as well.

+1. I have been reading this thread and why nobody had mentioned this fantastic ELP release. Perhaps because they are a band that people love to hate?

Call me an ELP fan from way back and a unabashed one at that, but this particular DVD-A occupies a special place in my collection and it's one I will spin up to show friends MCH audio, together with Hotel California and the bootleg (?) DSOTM that Parson's allegedly mixed.

This is also one of my favorite DVD-A's as well.
The sound quality is nothing short of exceptional.
I'm sure this is one of the reasons why this disc is highly sought after and commands rather steep prices.
I have the SACD as well but I'm pretty sure I will be going back to the DVD-A when it comes to listening time.

Another DVD-A I like for SQ anyway is Britney Spears In the Zone.

Actually I don't know what goes through the minds of record executives when they choose discs for DVD-A or SACD since from the surface one would not expect the Britney Spears fans (when that album was released) to be audiofiles.

I bought it sight (ear) unseen as a blind (deaf) buy and was quite impressed with the audio mix swirling around the speakers. I had only heard a few Britney Spear's songs before including her most famous but idly curious what a modern pop singer would sound like in lossless audio 5.1 surround.

It must be a collector's item now since I see it's available on Amazon for $178!

I cannot help but notice a distinct lack of reviews and/or comments of music specifically written for a multi-channel format such as SACD.

May I direct your attention to:



For those of you unfamiliar with Berlioz' Grande Messe des Morts aka Requiem, it was written for a very large orchestra, huge choir, 16 tympani, and 4 separate brass choirs distributed around the hall at the four points of the compass. If any work (this one was written in 1837) by any composer was designed to take advantage of our modern technology, this is it.

Hector Berlioz: Requiem

Radio-Sinfonieorchester Stuttgart des SWR
SWR Vokalensemble Stuttgart
MDR Rundfunkchor Leipzig
Toby Spence, Tenor
Roger Norrington, conductor

hänssler Classic 93.131 (2006)

Any of you who may be familiar with the first well known recording of this work, that by Charles Munch and the Boston Symphony Orchestra, originally released on RCA Soria and last released as a Living Stereo 3-channel SACD, will notice that Sir Roger has taken a somewhat different approach to tempos and phrasing. Norrington has long prided himself on being historically accurate in both his conducting and in the instrumentation of his orchestras. I doubt that he has changed course for this work. Which is certainly not all bad.

Berlioz himself stated that if only one work of his were to survive, a desert island piece as it were, this would be his choice. And for good reason. It is a choral work first and foremost. But the colorful, audacious writing for the orchestra is almost Mahlerian in his use of the instruments. And when the trumpets call in the Tuba Mirum you'd better be strapped in as the brass sound their demonic calls from everywhere in the room. Likewise the Lachrymosa has the brass resounding once again.

Most of the work is relatively low key, as one would expect from a Requiem. But it's Berlioz genius that turns even the quietest moments into spectacular visions. The piccolo - trombone duets, the fading, cascading Amens at the end.

This a work to be cherished. The 5.1 SACD goes way beyond what I ever expected. I hope you get to experience it.

Needless to say the sound quality is IMHO flawless throughout.

I am far from a big fan of Norrington's work which, typically, seems to be all wrong although for the right reasons. However, I think it is an excellent recording of the work.
Let me also suggest a great recording of it:

Hmmm....Brittany or Berlioz? Berlioz or Brittany? Which one of these two discs should I order this week? Guys, some help please.....

I don't have that problem as I own those and other versions of the Berlioz.

Tough question. I have friends that wouldn't last through 5 minutes of the Requiem. Yet they continue to buy music they won't ever listen to in 3 or 4 years. I've been enjoying the Requiem for 45 years now. It's that good. Hits a spot in my cerebral cortex - goosebump central.

My take on ...

Era: The Very Best Of Era

This will be a tough SACD to find but SACD.net offers a few links to foreign distributors. I was delightfully surprised to listen to the multichannel surround version of this 2004 recording. The CD is generally available but I really think this album flexes its muscle in multichannel surround.

It is a "fusion" album of sorts. There's plenty here including male and female chorus, female lead singer, full orchestration, techno beat, rock rhythm, classical, soundtrack style, and deep bass. All speakers are used to their full extent and your room is enveloped in sound.

The disc was produced exceptionally well. Check SACD.net for the details. The sonics are superb and it is evenly mixed.

Marcus

This thread seems to have gone into hibernation. I think it's time to wake it up!

I had to recalibrate my 5.1 speaker levels and distance setting a few nights ago - which I've taken to doing "by ear" as I don't like the results metering gives me - after which I put the DVD-A of Gaucho, Steely Dan's first digitally-recorded album, into my recently-acquired Oppo BDP-93, and played it at a healthy volume.

Wow!

To paraphrase the final cut in that album: "Phil's playroom is a bunker full of sound!"

(But I'm not a 3D man!)

Hi there, friends.

As my own contribution to the "reawakening" of this thread, I just wanted to chime in with a little bit of info for the old-school prog enthusiast: the Steven Wilson's 5.1 mix of King Crimson's Lark's Tongues in Aspic will be officially released on October 22nd in a CD/DVD-A package. There's even a deluxe edition with 15 discs (!) which includes the album in CD, DVD-A and BD (with DTS-MA and LPCM 24/96 audio tracks), as well as 12 more CDs with live concerts from the era. The package is limited to 7.000 copies.

It being my favorite KC album, I'm absolutely looking forward to it, and given Wilson's mixing resume, I'm sure this is bound to be another astounding surround reissue. I'll be happy to report my impressions when I receive my copy. Too bad I could only dust the wallet long enough to get the "simple" CD/DVD-A edition.

By the way, although it is not SACD nor DVD-A, I absolutely recommend Wilson's own live BD Get All You Deserve, which should at least be cursorily checked out by any audiophile or music enthusiast, not only prog-heads, as it is a masterpiece of great music, great sound and great visuals.

Take care!

Abel

Thanks for your welcome, Bill.

That would be great. Although not so many yet, there are some very interesting music BDs around, from bands and artists such as Pink Floyd, Rush and Steven Wilson, as well as some classical music. There are also some great concerts BDs worthy of our attention. It would also be interesting to make comparisons between SACD releases and their counterparts in BD, such as PF's Dark Side of the Moon and Wish You Were Here, which are available in both formats.

It is indeed very nice to have a place to read and learn so much about this hobby of ours.

Abel