Sony pulling out of SA-CD? (1 Viewer)

[Michael St. Clair]

Lee,

Too bad Sadie is processing everything as DSD-Wide ("PCM Narrow") instead of 1-bit DSD. Looks like everybody else is too.

I'm starting to think that traditional 1-bit DSD is unworkable for mixing or any substantial processing. And the 'DSD-Wide' moniker, to me, seems like an attempt to avoid using the dreaded "P" word.

Then again, it was originally touted as an archival format, not a production solution.


Well, the 5.1 mix of DSotM sounds warm to me. And it came from a 24/96 PCM source. So either 24/96 also can capture that 'warmth', or the later conversion to DSD is adding something that wasn't there.

 

[Lee Scoggins]

You can capture some warmth with 24/96 when done right, but 24/96 is also less accurate than DSD and this prevents some of the natural tone and timbre of the instruments from getting through. I've worked on too many jazz albums to know the shortcomings of 24/96, particularly on saxophone or percussion where transients are involved.

 

[Michael St. Clair]

Actually, what is really silly is the workstation companies liking to claim that they 'maintain the DSD signal' when they are really converting it to a flavor of PCM.

 

[Jeff Ulmer]

I can capture warmth with anything when done right. If you are losing "warmth" in the recording process, you should look at your front end equipment before looking at the storage. Your mics, preamps and monitoring environment are far more likely to be adding or subtracting warmth than any digital format. What are the specs of your signal chain? What mics and preamps are you using? What A/D?

If DSD is adding warmth, I'll pass.

I'll also agree with Michael again about the marketing hooey surrounding all this "pure DSD" nonsense. Any time you are converting out of native DSD you are adding distortion to the signal, even if only on the very lowest bits. I'm not going to claim to be able to hear it, but it is there nonetheless.

 

[Lee Scoggins]

Even if they was a PCM intermediary like the 352k used on the earlier machines: Who cares if you can't hear it?

I have spoken to several workstation technical support folks. They feel they are not changing the datastream since everything is done at 2.8mhz.

It seems to me Mike that you just can't find any legitimate grounds to attack Super Audio so you focus on one very small point and you get that wrong as well.

Let me ask you this...if Super Audio is so bad why does it dust most PCM recordings? Why do the vast majority of top engineers laud its sonics?

Why do Michael Bishop and Ying Tan and Mark Levinson feel that DSD is closer than 24/96 to the real event?

Why aren't audiophile labels sticking with the less expensive PCM-based mastering?

People have decided there is sonic value here and are changing even when it represents an added expense.

 

[Lee Scoggins]

As I stated, there is no sonic addition. The warmth comes from the instruments in the form of realistic timbre. This analog-like sound is just accuracy but it sounds to some like the effect of an amp with a good clean midrange. This is all good as we are hearing more of the actual music event.

 

[Jeff Ulmer]

Which is a function of mic placement and the recording chain. I'm still interested in what mics you are using to capture all this high frequency content, and the preamps you are using. 99% of the world class mics out there have significant rolloff above 20kHz, so I'm wondering where all these transients I've been missing are coming from, or does DSD just add them by magic?

 

[Michael St. Clair]

Sometimes you have to know what to put on the label; i.e. marketing. Why did so many audiophiles buy green pens, shakti stones, and tice clocks? I'm not necessarily putting DSD in the same category, but the realities of all markets is that what sells is not necessarily superior or as-advertised.

 

[Lee Scoggins]

I tire of arguing with St. Clair, so I won't respond to his negative posts. There is too much good happening in hirez to debate whether PCM can include 2.8Mhz sampling.



Jeff, if you are an engineer you should know that roll-off above 20k is a completely different issue than transients. Transients have to do with the attack-sustain-decay of a note. This happens in micro-second speed which is why faster DSD sampling which addresses the time element in finer gradations does a better job. Think of a sine wave as a simple example. DSD is "riding the wave" by examining changes in the wave up or down. The X axis is time which captures where on the curve one is. In PCM, the height of the wave is captured, but the sampling (along the X axis) is less frequent.

By the way, I use AKG 414s, C24 modified tube mics, and Neumann's latest U mics which are superb.

 

[Michael St. Clair]

Lee, between the two of us, there is only one person who has repeatedly (or ever) called one of the formats inferior to the other. And it's not me.

Think about it.

No aspect of any technology is beyond debate. Nothing negative about that.

 

[Lee Scoggins]

I have only stated a preference for Super Audio based on my personal experience in both critical listening and record engineering.

Every time over the past two weeks there has been some good news reported on a title, you look at the glass as half full and complain about such things as "day and date" release (neither of which format has been strong in), whether a PCM master was used (which did not stop Gaucho from sounding superb), and such technically archane as whether the intermediate stage in a workstation is preserving the 64fs data stream.

Why don't we just agree to disagree and focus on the more positive things happening?

We have a nice stream of good things to enjoy like:

1. Great titles - The Dylan is superb but the really amazing albums I bought recently include Good Bye Yellow Brick Road and Tommy.

2. Continued growth in the format - some 200 titles for this year have already been announced by just a handful of labels including 42 from Telarc. More audiophile labels sign up.

3. More superb hardware: New versions of SACD players are getting even more sonics out of the format like the Krell, the Bel Canto, and the new Sony 9000 which is amazing to my ears.

4. Format extensions like car audio systems are on the horizon. Philips and Harman Kardon are working right now on several. If it's anything like the Levinson system in my Lexus (owned by HK) then we are going to be happy as consumers.

5. Hints at more big name albums. James Guthrie has mentioned several times that he would expect more Pink Floyd albums to come out given the DSOTM success. I am willing to bet there is a lot of lobbying for the Beatles records going on. Sony has been fairly successful at getting a number of artists catalogs released...maybe Springsteen or others will be up soon.

It's an open and perhaps early question as to whether the mainstream acceptance will occur but I definitely see the glass as half full, maybe two thirds.

 

[Michael St. Clair]

For what it's worth, I thought we once had an agreement to not respond to each others posts. It's silly that it comes to that, but I gave it a shot. I went for months without replying to one of your posts, but you wouldn't do the same. Eventually I gave up.

 

[Jeff Ulmer]

Yes and no. If the recording device is incapable of accurately reproducing time line events beyond a certain frequency, then any transient which happens to occur in a shorter span of time is going to be affected. In the case of a microphone, that means the diaphram must be able to respond to a change in amplitude in a certain amount of time. If it is unable to move that fast, the transient is not captured accurately. This is true of sampling rates as well, which is why a higher sampling rate will work better on transients, however there is a limit to how short of a duration a transient can be due to the signal chain before the sampling.

At 96k - four times the limit of human hearing - the transient response up to 20k is limited only by the input chain. If you are hearing differences between 96K, 192K and DSD, I would suggest that those differences are due to anything but the sampling frequency.

Nice mics, by the way (I use all but the C24 myself), but prefer small diaphram condensors over the larger diaphrams on transient critical instruments. This isn't intended as a criticism, but I'm a little surprised you aren't using any of the Earthworks products, which are at least relatively flat up to 40kHz, and not anywhere near as colored as the AKGs or Neumanns.

 

[Lee Scoggins]

We would politely disagree since we have kept a stable and consistent chain in all of our format experiments and adjusted for any volume differences on playback. We also hear our results on both Stax and Grado RS-1 and RS-2 headphones which are ultra-transparent.

Dr. Mark Waldrep at AIX has also spoken on the improvement you get from moving from 96k to 192k on the DVDA side.

Neil Young seems to like 176k as well and his upcoming DVDAs will feature this faster sampling rate. I can't wait to hear them.

 

[Jeff Ulmer]

Well, I would also look at the performance of your preamps and most especially the A/D and D/A, which I would suggest have a far greater impact on the sound than the sampling rate - at least at 96k and above.

While I'm not going to dispute your preference for the sound of the Neumanns, they are not neutral. Both the U87 and the USM69 have an exagerated high end, and the U89 is scooped in the 5k range, both of which work quite well for a variety of material.

 

[Lee Scoggins]

As I said, we kept everything in the recording chain the same during the comparison. We used state of the art Prism ADCs and for DSD the Meitner converters.

When you work with a very familiar recording chain across multiple sessions detecting changes is actually not that hard. You drop in new gear and note what you find. In the case of listening to mic feeds, it's real easy since you have just heard the live event you are trying to capture.

 

[Robert James Clark]

Micheal,


Keep fighting the good fight...


The Master would approve....