Early digital recordings on vinyl vs. CD?

Although I would strongly disagree with Doug that digital recording is not inherently flawed, I do agree with him on his digital vinyl selections. Al also has some great ones, and he is spot on in his discussion of why the Telarc recordings were better. Many musicians still consider those some of the best digital recordings ever made. I would certainly say that early digital sounds better on vinyl. The performances are always the top priority for me, and really should be for any music lover, and there are a great many of them. Two from the early digital years that immediately come to mind not mentioned so far are Claudio Abbado's Mahler cycle (DG, not known for their sound quality, in fact sometimes the reverse) and also Charles Dutoit's many excellent recordings with the Montreal Symphony.

Often, the extreme macro dynamics are the most distinctive feature about most early digital recordings on vinyl such as several from Telarc that I recall compared to others, including same recordings on CD, where the format puts a limit on dynamics to some extent.

Some of the early Telarc recordings on vinyl are very good test records to determine the tracking ability of a vinyl rig. IF they do not mis-track totally, break-up or audibly distort during the loud passages, your vinyl rig is in good shape as well as your amps ability to drive your speakers in terms of dynamic headroom.

Learsfool, I agree there are inherent flaws in digital recording technology, as there are in every technology. Sorry if I gave a different impression. The last thing I'd want to be is a digital apologist, lol!

Of course there are flaws in analog recording technology too, albeit different ones that are less objectionable to many ears. Still, if I play an analog recording for a young person, the first thing they invariably notice is the tape hiss. Most of us grew up hearing that and we listen through it without even thinking, but if you've never heard it before it really grabs your attention (especially if you have a youngster's HF sensitivity).

Anyhow, we all seem to agree that early digital recordings of classical on vinyl are highly listenable. The flaws noted by Almarg in a few Telarc LPs were flaws of microphone selection and placement, not anything inherent in digital or analog recording.

The early digital classical LPs have the advantage that they were probably mastered from the native mode of the digital recording, and definitely done through a pro-quality DAC. With CDs, regardless of the original sampling rate and word length, the mastering is downconverted to 16/44.1 and most of the market plays it back through an inexpensive DAC or built-in chipset. With LPs you're usually playing back a high quality analog made from a full-res master.

I do remember a digitally recorded classical record I bought for my brother that I thought was a bit brash, but the ones that I own (including Josh Bell, The Planets, some classical guitar) all sound good--low noise floor, good dynamics and clarity, reasonably rich and full sounding. I do like a full analog signal chain better, but I won't avoid an LP at $2-4 because it's digital, and I end up enjoying many of them very much.

05-22-12: Johnnyb53
The early digital classical LPs have the advantage that they were probably mastered from the native mode of the digital recording, and definitely done through a pro-quality DAC. With CDs, regardless of the original sampling rate and word length, the mastering is downconverted to 16/44.1 and most of the market plays it back through an inexpensive DAC or built-in chipset. With LPs you're usually playing back a high quality analog made from a full-res master.

Johnny makes a good point. The Soundstream digital recorder which Telarc used in those days was 50 kHz/16 bits, as Phasecorrect mentioned earlier, but the difference between 50 kHz and 44.1 kHz sampling is more significant than it may seem from a numerical standpoint (even putting aside differences in implementation quality).

44.1 kHz sampling means that the anti-aliasing filter that precedes A/D conversion has to cut off all frequencies above 22.05 kHz, which only allows about a 10% margin relative to the highest frequency (20 kHz) that is intended to be captured accurately. 50 kHz sampling, on the other hand, allows an anti-aliasing cutoff frequency of 25 kHz, which increases that 10% figure to 25%. That represents a considerable relaxation of the sharpness of the filter rolloff that is necessary, which (everything else being equal) can be expected to significantly reduce the side-effects the filter may have at audible frequencies.

Also, I should add to the comments in my earlier post that (as Mapman alluded to) another major reason for the sonic quality of the early Telarc's was the fact that they applied far less dynamic range compression to their recordings than was typical of releases from the major labels, if indeed they applied any at all. Which means that you'll play them with the volume control set to a higher position than it would be set to for most other recordings. Which will make that bass drum even louder :-)

Regards,
-- Al