How much did ADCs in the 80s affect the sound quality of CDs?

[AnalogSteph]

Isn't IRIAA only applied during / just prior to the actual cutting process?

Maybe a mistake on their page then, they show "97 dB (1V)":

Defo a mistake, they're traditionally rated at 97 dB / 1 mW (or 102 dB / 1 V). The Sennheiser website used to be riddled with errors like this (always made me wonder why such a major company couldn't get that straight), but it hadn't been such a major issue in recent times.

Do you know how the album RMS is calculated?

Presumably like it always is?

(n being the total number of samples in the entire album, with or without upsampling, and x being normalized float sample values)

With two different tools giving the same results, I would be reasonably confident that they are correct.

While there could be a discrepancy of 3 dB depending on your definition of 0 dBFS (the raw calculation would give -3 dB for a fullscale sine, hence the "AES +3 dB" option), I don't see where one of 5 dB and change would be coming from.

 

[JIW]

[...]
Presumably like it always is?

(n being the total number of samples in the entire album, with or without upsampling, and x being normalized float sample values)

With two different tools giving the same results, I would be reasonably confident that they are correct.

While there could be a discrepancy of 3 dB depending on your definition of 0 dBFS (the raw calculation would give -3 dB for a fullscale sine, hence the "AES +3 dB" option), I don't see where one of 5 dB and change would be coming from.

I checked my calculation and it is not correct. To use the RMS formula for all samples, I have to average the mean square values not the root of them. If track i of N has level RMS_i and duration T_i while the album has duration T, the album is

RMS = sqrt(sum(T_i*RMS_i^2, i = 1, N)/T).

That way, I get an album RMS of -25.58 dBFS (AES). There is still a discrepancy, but since the track duration is given in seconds, there are still +/- 22050 samples uncertainty.

 

[JIW]

I calculated the uncertainty from track length being in seconds rather than samples using propagation of uncertainty. I get an absolute uncertainty of 2.045*10^(-5) which translates to an uncertainty of 0.0048 dB. This does not even come close to covering the discrepancy.

 

[JIW]

I set my volume control such that 0 dBFS is 7 dBV for my HD600. The highest Peaks should be around 115 dB SPL that way. For speakers in a room that have a sensitivity of 90 dB at 1 W at the main listening position that would require around 300 W peak power.

The DR at the site I linked seems to be the average of the track DRs not album DR.

For a more recent classical recording, Ravel: Daphnis et Chloé - Pavane by Yannick Nézet-Séguin conducting the Rotterdam Philharmonic Orchestra released on BIS in 2015 has a peak of -0.16 dB and RMS of -28.16 dBFS for a crest factor of 31.01 dB (See here).
View attachment 418278

For an even earlier recording, Tricycle by Flim and the BB's from 1983 has a peak of -0.06 dB and a RMS of -25.18 dBFS for a crest factor of 28.13 dB (See here). The 2020 re-release is even more dynamic with peak of -0.57 dB and RMS of -27.51 dB for a crest factor of 29.95 dB (See here). Note that this is not classical but a small jazz ensemble. Recorded using Mitsubishi X-80.
View attachment 418279
View attachment 418281
View attachment 418282
Pictures of full booklet available on Discogs.

I calculated the album RMSs the wrong way. The correct value for the Ravel is -25.91 dBFS RMS and the correct value for the Flim and the BB's is -24.39 dBFS RMS for the original and -26.70 dBFS RMS for the remaster.

I did some more looking around for Mahler albums and found the most dynamic one is the SACD side of Symphony 4 by Iván Fischer conducting the Budapest Festival Orchestra from 2009 on Channel Classics (CCS SA 26109). It has an album RMS of -27.31 dBFS and a peak of -0.01 dBFS for a crest factor of 30.31 dB. See here for track data.

 

[JIW]

I calculated the album RMSs the wrong way. The correct value for the Ravel is -25.91 dBFS RMS and the correct value for the Flim and the BB's is -24.39 dBFS RMS for the original and -26.70 dBFS RMS for the remaster.

I did some more looking around for Mahler albums and found the most dynamic one is the SACD side of Symphony 4 by Iván Fischer conducting the Budapest Festival Orchestra from 2009 on Channel Classics (CCS SA 26109). It has an album RMS of -27.31 dBFS and a peak of -0.01 dBFS for a crest factor of 30.31 dB. See here for track data.
View attachment 421954

I just listened to the 24/192 version on Qobuz. I recorded the last movement using Audio Hijack in 24/192 AIFF. The peak and RMS below are consistent with those the SACD version: -10.44 dB and -33.60 dBFS RMS.

Volume control on DL200 was at 85 in low gain so output is 1.5 dBV at 0 dBFS and thus 109.5 dB SPL peak at maximum digital value for HD600. So a good bit lower than the Respighi where it was 7 dBV at 0 dBFS.

 

[krabapple]

May I suggest all the posts about high crest factor releases be moved to another thread.They arent really on topic here

On topic: my understanding is that the RIAA curve is applied to the cutting head during cutting So it would not be captured on a LP production tape. If it was, then it would have to be turned off every time the tape was used for cutting.

The vinyl production master, in this scenario, captures all the adjustments made *before* the signal reaches the RIAA circuit.