1970 digital PCM recording, Something, Steve Marcus, 13-bit/32kHz, sounds d@mn good!!!

Steve was an incredibly talented musician. The fact that he stayed so many years with Buddy Rich attests to how highly he was regarded. Saw him up-close and personal many nights when I was chasing the BR Band around PA and NY states in the 70's and early 80's. His Buddy's Buddies tribute CD's are fabulous, too!

 

I thought that Denon would have just padded the 13-bit files with zeroes. Strange that they did not appear to do so. However, there is another explanation.

Denon DID have a 14-bit/47.25kHz system in use. It was coupled with an analog compander that made the SNR and dynamic range equivalent to 15.5 bits, according to the Fine article. Anyway, this could have been output to analog and then resampled back to 16/44.1. In fact, that's probably how it was done rather than using some sort of punitive DSP.

 

Yes I don't think DSP resampling existed at the time. They would have had to play the tape into an ADC to 44.1/16. Explaining why the last 3bits aren't padded with zeroes.

 

Which spectrogram are talking about? Those that were posted are only part of 1 song.

 

I'm sure that's what happened - they played it back to analog and re-digitized it, probably using the original decks.

There wasn't a lot of equipment around in the '80s to convert sample rates I'd imagine, especially oddball rates. And given the linearity issues with early A/D converters, it would probably be best to do your D/A conversion on the same deck, since the channels were probably matched pairs. At least that would be my guess...

 

OK. I thought it was for the entire album. If the spectrogram was from the third track--the one from the bonus track--my comments would still hold.

 

That was supposed to read "primitive DSP."

 

It is the first minute of the first track.

 

Well, the CD says it is a "PCM Recording" so I doubt an analog safety was used. I am going to contact Nippon Columbia for more information and also send an email to Tom Fine. I'll let you know what I find out.

 

Maybe they mistakenly kept the original credits, it wouldn't be the first time that it happens...

 

True!

 

Just received from Japan 3 Denon CDs featuring their early PCM recordings: 2 with CD pre-emphasis in disc TOC (2005 re-issues) & 1 without pre-emphasis (2010 re-issue) which audio likely was pre-amplified by ~+10dB. Wonderful sound quality on all 3 discs:

Smetana Quartet - W.A. Mozart: 2 String Quartets, KV 458 "Hunt" & KV 421 (recorded in Feb-Mar 1982; 1st CD issued in 1982; re-issued in 2005)
A. Nicolet, H.Holliger, C. Jaccottet, et.al.- J.S. Bach / C.P.E. Bach: Trio Sonatas (recorded in Sep 1982; 1st CD issued in 1984; re-issued in 2005)
Heinz Holliger - G.P. Telemann: 12 Fantasies for Oboe Solo (recorded in Nov 1979; 1st CD issued in 1984; Blu-Spec CD re-issued in 2010)

 

I think it's mentioned that the original analogue tape (supposedly recorded in parallel with digital) was used for this release. Check out the following link through Google translate:
http://www.cdjournal.com/main/news/-/50839

 

Early recordings? Ha! I have some Denon's from 1970, 1971, and 1972. Now those are early PCM!

 

Are the CDs featuring these recordings pre-emphasised, by any chance?

 

The ones that I have that are originals probably are. I have one that is a Denon reissue and my guess is that the reissue is not.

 

It doesn't really surprise me. It was recorded and produced with skill and effort. The 13bits is equivalent to analog tape and records and is enough to capture the dynamics of the recording at a reasonable signal to noise ratio.

 

Got a couple of 2010 Denon Blu-Spec CDs featuring recordings made by Nippon Columbia in 1972 (both discs are carrying PCM Recording badge):

Smetana Quartet - Mozart: Two String Quartets, KV 458 & KV 421 (recorded on Apr 24-26, 1972)
Jean-Pierre Rampal - Telemann: Twelve Fantasias for Flute Solo (recorded on Oct 30, 1972)

Looked at the spectrograms of these CDs supposedly sourced from 32 kHz sampled PCM & they did not show any cutoff at ~16 kHz (spectrum of the audio extends to near 22 kHz frequency). There are some peaks showing at ~15.75 kHz though. Not sure, if 32 kHz PCM masters were used for these CD re-issues, seems unlikely (unless some analogue domain futzing was involved)... No pre-emphasis present on these 2 CDs, BTW.

 

That is mental, I've never seen that before - the voices are so bad it's hilarious, never realised George was quite so Irish...

 

The 1972 recordings weren't made with the 32kHz recorder. Rather, a 47.25 kHz recorder was used.

Just read this and all your questions will be answered:

http://www.aes.org/aeshc/pdf/fine_dawn-of-digital.pdf

 

Hmmm. I vaguely recall reading something about some analog trickery they tried using early on to cram more bandwidth into 32kHz sampled PCM.

 

Makes sense, thank you. I should have re-read the article before posting... So, 1972 Denon digital recordings utilized 13/47.25 LPCM.

 

Yes, and sometime later they developed an analog compander circuit that raised bit depth to effectively 15.5 bits. So, in the mid-70's, Denon's 15.5/47.25 gear was achingly close to DAT quality, and Soundstream's 16/50 recorder was already higher.

 

THAT'S what I was thinking of. Thanks!

Didn't the 8-bit PCM used by 8MM video also use an analog compander?

 

Yes, that's the compander circuit they used to increase SNR and dynamic range, giving the effective performance of 15.5 bits.