Post by johndyson on Nov 4, 2019 23:01:09 GMT 10
I created a separate thread because this doesn't apply only to the DHNRDS, but the general state of the old recordings originally created/mixed down in the 1960s through early 1990's...
instead of
Recording -> DolbyA decode -> distribute...
So, what the remastering does is this:
distribution copy -> inverse EQ -> DolbyA decode -> 'better sound results'
First, the original 'Harsh digital sound' complaints didn't come from only one cause, as there were at least three reasons (not in the order of impact):
1) Early state of 'digital audio' technology.
2) Unexpected signal detail that was partially covered up by complex dynamics in vinyl production
3) Different final/production mastering for digital media vs. analog media.
I believe that items 1 and 2 have been well understood all along, and the relative magnitudes of items 1 and 2 have been discussed for 30+yrs.
Item 3 has been underestimated or ignored in the public/consumer arena. This is a real problem, and has varying magnitude depending on the exact difference.
I dont' know enough about the final mastering (I am not speaking of mixing the material, I am writing here about the two track stereo being prepared for distribution),
to know all of the steps, but there are some deviations between the handling of digital vs. analog material.
---
This discussion is not solely intended to shill for the DHNRDS -- that is NOT the goal here, but maybe to explain why the 'love' for vinyl and other analog media had
been sparked and even sustained for so long. There are definitely recordings which cannot sound as good as they should, some are simply NOT available in their
natural form. Instead, some material is still mostly available in the 'harsh digital' form unless on vinyl or properly mastered tape.
For an example of the 'natural' Carpenters sound -- refer to the link below -- it must be removed in a few days, but it is here for illustrative purposes. TRY to find
a natural sounding Carpenters distribution... You might have problems doing so -- unless you have vinyl or a VERY VERY special digital copy. Even the
Carpenters 'singles' on HDtracks is NOT DolbyA decoded!!! It does NOT have the natural sound. (I haven't been able to 100% accurately decode the
Carpenters until recently, only possible once I found the correct inverse equalization...)
---
I have listened to a LOT of CDs made from older recordings, both CDs mastered/produced in the early days of digital, and CDs/digital produced more recently, and
it is extremely clear to me that the differences can be enumerated as below:
1) No DolbyA decoding, substituted by EQ
2) No DolbyA decoding, not substituted by EQ
3) Additional manipulation, incl compression.
In recent years, becoming much more prominent starting in the middle 1990s, and even more egregious in the 2000's and beyond, dynamic range has been deemphasized as a desirable trait, and loudness seems more important to the distributors. This dynamic range matter is well understood, and even some of the motivations for decreased dynamic range are understood -- it is the other items that I am writing about here. Actually, this missing step of DolbyA decoding, and the oassociated decrease in high frequency dynamic range MIGHT have been a contributing factor to the current 'loudness wars' and the consumer toleration for decreased dynamic range!
---
Here is the scenario that has happened REPEATEDLY in the past:
DolbyA Recording -> EQ -> distribution to consumer
Instead, the following should have happened:
DolbyA Recording -> DolbyA decode -> distribution to consumer
One my ask: Why did this happen? I have no firsthand knowledge, but just existance proof. It has taken some time to collect the information, but there might be several causes:
1) Material already in digital form, archived by LOC procedures that make NR decoding before archiving an optional step.
2) Insufficient metadata/documentation and/or missing calibration tones making the decoding effort inconvenient.
3) Time/cost of realtime limited for HW DolbyA decoding, much slower & inconvenient than copying digial files directly.
When looking at these items above, and the fact that distribution moves a commodity around, ti is not an artistic endeavor from the standpoint of the business people, I'd
suspect that there is sometimes an explicit financial decision to do EQ to attempt to hide the DolbyA encoding instead of the actual decoding operation itself. EQ
can be done digitally fast, while decoding actually takes 1:1 time on recording vs. time to decode.
Mentioning the details of the problem again: when the mastering was done by the distributor, instead of DolbyA decoding, they did an EQ operation like this:
Recording -> EQ -> distributeinstead of
Recording -> DolbyA decode -> distribute...
So, what the remastering does is this:
distribution copy -> inverse EQ -> DolbyA decode -> 'better sound results'
The problem is that the original EQ isn't documented, and also the ongoing problem was that I didn't realize how precise the EQ needed to be...
----
What is my goal here -- mostly to do what I can do to resurrect the old recordings. These recordings are important to history and will NOT be heard correctly unless
correctly and accurately processed....
John
