DSDA: a New Format for LP Fans (2 Viewers)

[David Susilo]

Grady,

My point? Everybody here is saying that LP is the better MEDIUM. That's what I want to debunk. I'm not talking about which medium is best for PRERECORDED MAINSTREAM recording. There is nothing mainstream about using LP.

JUST MEDIUM, period.

Tape is still better than vinyl.

 

[Lee Scoggins]

David,

Well that's a matter of opinion not a fact,and 70db is more of max figure[for 45rpm] then a ballpark.

Lewis,

I think you missed my point. What I am saying is that many agree LP sound can be fantastic (clearly an opinion but a very popular one at that) and often is, which makes me think the 70db is a bogus number.

John and Lewis,

Prove to me with references that 70db is an accurate number. I think Vince Maskeeper may have thrown this in on another post, but where does it come from? Is it scientifically backed up at all?

Did the AES study this? on quality turntables? I have heard that 30ips tape has great dynamic range. If this is true, isn't it likely that vinyl can capture all of what's on the 30ips master tape. I have seen it work like this many times in the studio.

David,

I don't know that we narrowed this discussion down to the mainstream market only.

I will research this 70db number, I remain suspicious. My friends at Classic and Cardas will know this one. I'll gladly admit if its the right one - it just does not feel right given the dynamics of good licorice pizzas.

Lee

 

[David Susilo]

Lee,

yes I meant 30ips 1" tape. Afterall, we're talking about hi-end tunrtables too. AFAIK, the topic is about the medium and not specifically mainstream medium as alluded by Grady.

70db for turntable was published by Sony during the early days of CD. I remember the poster clearly, a picture of a giant treble clef and Sony published 65db for cassette, 70db for LP, 80db for Beta Hi-Fi, and 90db (at the time) for CD.

as an aside,

Dynamic range is, well, the range (in db) of the lowest possible to the highest possible loudness that can be handled by a medium with minimal amount of distortion. Therefore, it's not about loudness but about 'how loud can it go from the quietest recordable point'.

 

[John Kotches]

Righto...



Let's start with the published figures from "The Brady Guid to CD-ROM".... (C) 1987, Prentice Hall, by Laura Buddine and Elizabeth Young -- who cite 55-60dB.



The professional audio reference cites "65dB", at

http://www.showorks.com/Library/Glo...g up a few more over the weekend. Regards,

 

[John Sully]

The limit of vinyl is the performance of the turntable itself. The funny thing was that I went to the Rega site and looked for specs on this -- none published. If I remember correctly from the last time I went turntable shopping the best S/N rations I found were between 65 and 70 db down. I doubt that this has changed much in the ensuing time, belt drives provided the best figures then, and my guess is that they still provide the best performance in this respect.

I think that 70db is a generous estimate of the S/N ratio which vinyl can provide.

 

[Philip Hamm]

Lee,
Do you know what "dynamic range" means? It's the difference between the medium's version of silence and the loudest volume the medium can produce without distorting.
It's comically easy to demonstrate the limited dynamic range of vinyl, you don't need any "studies". Just hook the turntable through a phono preamp out, then into some kind of metering device. A Casette Tape deck, MiniDisc Deck, a DAT deck, a computer all work fine. You're not recording here just using the meter to monitor the signal level. Set up record pause and play with the levels until you get the peaks at around 0dB. Now look what the meter reads in the noise between tracks. Absolute value there's your dynamic range. It varies between turntables, and even varies based on the different groove locations on a record. Notice that the noise floor is lower on the inside grooves than the outside ones. After spending some time auditioning some high end (for me) Rega tables at my local dealer I decided that my Denon is fine for me for the forseeable future.
Where the real big variance occurs is from record to record. Some like My Mobile Fidelity Sound Labs Original Master Recording of Pink Floyd's "Dark Side Of The Moon" can easily aaproach 70dB even on my modest system. However, if I put in my mass market MCA LP of The Who's "Who's Next" from the early 80s I'm barely beating 45dB. The surface noise is obnoxious on that record.
I believe that LP is capable of reproducing sounds that are quieter than it's very high noise floor. Someone else should correct me.
Anyway, we've strayed WAY off topic. The topic not vinyl in general (though it has strayed there). The topic is LPs sourced from DSD masters which I still hold is a ridiculous idea (at best).

If LP really is 70db of range, then why does it sound so good?

First of all it's 70dB at it's very best and I've rarely heard vinyl that can approach this number. I think 60 is a more accurate measurelemt of typical vinyl. Though my collection consists mostly of non-"audiophile" mass-market records.
But the reason it sounds so good with such a poor dynamic range is because dynamic range is only one of many factors involved.

 

[Lee Scoggins]

But the reason it sounds so good with such a poor dynamic range is because dynamic range is only one of many factors involved.

Agreed.
Lee

 

[Grady]

For conventional consumer use vinyl is better than tape in the vast majority of circumstances. It really matters little that the master tape is better than the LP cut from it, the real issue is what software playback format is best for consumer use? Unless high quality open reel dubs were made available to consumers it doesn't matter how good they'd be.

All this talk about dynamic range is interesting. I agree that lots of records have limited dynamic range. Putting thirty or more minutes per side can and does lead to compression of dynamics. But, this is a decision made by the engineers that recorded, mastered and cut the record. While in many cases Lp's are compressed, done correctly they have more than enough dynamic range for consumer use. Many DGG classical lp's are badly compressed. The Mahler 2nd Symphony with Kubliek for example. But if you listen to the same piece with Solti and the LSO on London/Decca, you'd be amazed. Done correctly the Lp has enough dynamic range for 95% or more of all music. If the piece is truly massive, like the Mahler 8th, I agree it would fall short. But you'd have to have a pretty massive playback system to notice it.

Grady

 

[Philip Hamm]

Sorry Lee no offense intended, it just seemed from reading your posts that you perhaps misunderstood this extremely simple concept. That confused me since I know you to be very knowledgable. I misread or was mistaken.

However, I want to reiterate that I believe that LP can reproduce sounds that are quiter than the noise floor, thus the real dynamics of the music reproduced in the grooves exceed the "on paper" dynamic range of the format. The "worst case" noise floor on LP records consists of mostly low frequency rumble, perhaps the midrange and high end dynamic range is much better. (kind of like I wouldn't judge SACD's dynamic range by the high frequency noise) I don't know if this is true, but I believe it to be so. If this is the case, this effect could explain the extreme dynamics that LP enthusiasts such as myself observe in the format.

Grammy winner T-Bone Burnett's song 'The Sixties' from his album (not avaibale on CD) "Proof Through The Night" is a great example of dramatic LP dynamics.

 

[Larry B]

Phil:

Thanks for the clarification.

John:

I apologize if I misunderstood your post.

Larry

 

[Frank_S]

There's an interesting article in Stereophile(May issue)latest issue, an interview with Mark Levinson. A lot is covered during the interview and midway through they discuss CD,SACD, and the LP. According to Mark Levinson, many consumers complain about the performance of the CD, including himself, and he finds that many consumers spend a lot of money on hi-end digital gear only to come out still disappointed. He goes on to say that MORE and more people are going back to LP. He also says that he really thinks SACD is the closest thing to analog next to the LP. I won't go any further, iff you're interested read the article in Stereophile.

 

[John Kotches]

Frank,

This is the same Mark Levinson who has stated publically that PCM makes people physically ill.....

Makes one wonder, yes it does.

Regards,

 

[John Kotches]

Lee,

Sorry dude -- you're off the track here:

If you listen to Classic Records albums from Mike Hobson, then you know there must be more than 70db of dynamic range.

No, I don't know that.... Show me.

Regards,

 

[John Kotches]

Lee,
I agree -- David Chesky's SACDs do sound splendid -- he must be using a really good algorithm to transfer from PCM to DSD
Regards,

 

[Lee Scoggins]

This is the same Mark Levinson who has stated publically that PCM makes people physically ill.....

To be fair to Mark Levinson, this is taken out of context. He was discussing the inability of people to sit down and listen to redbook CDs for any length of time.

Lee

 

[Lee Scoggins]

Philip,

However, I want to reiterate that I believe that LP can reproduce sounds that are quiter than the noise floor, thus the real dynamics of the music reproduced in the grooves exceed the "on paper" dynamic range of the format.

Exactly! This is rapidly becoming my conclusion: Dynamic range in dbs does not capture the full beauty of the format.
Lee

 

[John Kotches]

Lee,

Don't start off sentences "I know for a fact", when what you are offering is subjective opinion, that is a long, long ways from being proven.

I've read the glossies: "Look how the pulses mimic the original wave".

I respond with: "Look how closely the PCM samples come to the point in time values in the wave". With 20-bit sampling, we're accurate to within one part in 500,000 (one-bit for +/-)... with 24-bit, we're accurate to within one part in 8 million. That's an awful lot of 0s.

How accurate is DSD to the original voltage of the waveform?

You have to deal with aggregate pulse samples over a period of time to make that judgement.

Here's an interesting experiment that I would love to see performed, yet no one has yet to the best of my knowledge:

Take a sample recording in analog on 30ips tape (the gold standard to many). Send simultaneous outputs to 3 encoders: DSD, PCM @ 24/96K, and PCM @ 24/192K.

You now have a valid comparison.

Synchronize playback, and record the output analog waves from each of the samples, measuring distortion between the original and each of the 3 copies.

Which solution is closest to the original input, and how much does each vary?

That would be an interesting experiment indeed.

Regards,

 

[Lee Scoggins]

I respond with: "Look how closely the PCM samples come to the point in time values in the wave". With 20-bit sampling, we're accurate to within one part in 500,000 (one-bit for +/-)... with 24-bit, we're accurate to within one part in 8 million. That's an awful lot of 0s.

You can't really compare 20 bit sampling in PCM to DSD since it is an entirely different methodology of sampling. With the Delta-Sigma algorithm, sampling rate is even higher, but it is one bit for most of the process. This statement does not really support your argument since the formats are so different.

Also, how do you deal with the "zero-crossing" distortion inherent in PCM that Ed Meitner discovered?

Is this not also a major flaw like you consider SACDs ultrasonic noise?

Seems you can't have it both ways and there will likely never be a perfect format.

The best thing to do is the experiment you suggest with a room full of experienced listeners and engineers (since ear can be trained to listen better) and see what the scorecards are for each format in blind testing. This would be expensive to do, so I rely on what I hear on these formats in the studio and that is why I have drawn my conclusion that right now Super Audio is the better format.

I feel more confident in my assessment now that I have heard "pure DSD path" SACDs which sound spectacular.

Lee

 

[John Kotches]

Lee,

I am merely countering the argument that "more samples" makes DSD better. Sorry you don't follow my logic, bizarre though it may be at times.

My point with the test is to look at the measurable differences -- and quantifying what the results are from each process. Many people think DSD sounds better, so I would think it's worthwhile to investigate what differences it has from PCM. I personally do not have such tools at my disposal to carry out such a test.

In the cases of Michael Bishop's tests, the methods used weren't described as either sighted, single blind or double blind. Michael's words imply that these tests were sighted, so we don't know how much of this is attributable to listener bias.

Regards,

 

[Lee Scoggins]

John,

Has Secrets of HT ever looked into DVDA and SACD differences?

Lee