Skip to main content
Topic: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh (Read 1997 times) previous topic - next topic - Topic derived from Vinyl is equivalent t...
0 Members and 1 Guest are viewing this topic.

Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

I guess it's more or less similar to some kind of ADPCM with bit depth somewhere between 6..12 bits, and the usual 44100 sample rate, plus some kind of pre-emphasis. Not like it can be compared to normal PCM, as vinyl also adds nonlinear distortion which is also very dependent on signal characteristics. Anyways, that was a strange question but if the point is to make comparison to CDDA, then it's overwhelmingly obvious that vinyl has far worse fidelity.
I agree.  The biggest typical kind of problem with the digital recordings has not been the actual transmission method (CD, digital download, etc), but has been errors or incomplete final mastering processes.  Sometimes there might be an EQ problem because a the final phases, vinyl does usually require some EQ/processing, and probably more severe is the leaving of the noise reduction on the recording.  DolbyA DOES leak through the process -- how often?   Well how often have you heard that so called 'harsh, digital' sound?

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #1
Well how often have you heard that so called 'harsh, digital' sound?
More often than Dolby A, preemphasis, sourced from a vinyl master, etc. is actually the case, I assure you.

You know how the saying goes, when you're a hammer everything looks like a nail.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #2
how often have you heard that so called 'harsh, digital' sound?
Almost all the time for CDDA, and for many years.
DSD and vinyl rarely have this problem.

So, it's a bit weird that such a superior audio format for so many years has had the worst mixes.

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #3
Well how often have you heard that so called 'harsh, digital' sound?
More often than Dolby A, preemphasis, sourced from a vinyl master, etc. is actually the case, I assure you.

You know how the saying goes, when you're a hammer everything looks like a nail.

I have read quotes from recording engineers doing mastering that they were told not do do the full DolbyA decoding process (just EQ it.)   It DOES happen and it has happened often...  Please refer to my site with demos.  This is a real thing, and it is starting to percolate around in the recording circles.  It does not take a recording engineer to playout a digital master tape.   Simple example (many many more)  Queen Greatest Hits, Hollywood records -- DolbyA.

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #4
how often have you heard that so called 'harsh, digital' sound?
Almost all the time for CDDA, and for many years.
DSD and vinyl rarely have this problem.
There must be something fundamental problematic at your place.
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #5
Well how often have you heard that so called 'harsh, digital' sound?
More often than Dolby A, preemphasis, sourced from a vinyl master, etc. is actually the case, I assure you.

You know how the saying goes, when you're a hammer everything looks like a nail.

I have read quotes from recording engineers doing mastering that they were told not do do the full DolbyA decoding process (just EQ it.)   It DOES happen and it has happened often...  Please refer to my site with demos.  This is a real thing, and it is starting to percolate around in the recording circles.  It does not take a recording engineer to playout a digital master tape.   Simple example (many many more)  Queen Greatest Hits, Hollywood records -- DolbyA.

Not once did I say or even imply that non-decoded Dolby A wasn't a real issue.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

 

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #6
Well how often have you heard that so called 'harsh, digital' sound?
More often than Dolby A, preemphasis, sourced from a vinyl master, etc. is actually the case, I assure you.

You know how the saying goes, when you're a hammer everything looks like a nail.

I have read quotes from recording engineers doing mastering that they were told not do do the full DolbyA decoding process (just EQ it.)   It DOES happen and it has happened often...  Please refer to my site with demos.  This is a real thing, and it is starting to percolate around in the recording circles.  It does not take a recording engineer to playout a digital master tape.   Simple example (many many more)  Queen Greatest Hits, Hollywood records -- DolbyA.

Not once did I say or even imply that non-decoded Dolby A wasn't a real issue.

But -- I am claiming that DolbyA encoding is probably more common than the extreme pre-emphasis that sounds like DolbyA.  Let me explain -- they have to go out-of-their-way to do the pre-emphasis (unless they are taking it off of tape without de-emphasis, which I doubt.)  It is EASY & CHEAPER to leave the DolbyA encoding on the material.

It seems to me that very often the digital masters might have been made by a direct copy, leaving the DolbyA encoding on purpose. Multiple encode/decode cycles is destructive because of the accumulated intermod.  So, doing a DolbyA decode wouldn't be prudent due to the possibility of re-recording onto analog.  So, when 'cheap' companies what their latest compiliation or '90th anniversary edition', then a simple digital copy is 'the way to go' when the bean counters are in control.  Up until now, to ACCURATELY decode the DolbyA for such material, one needs to convert back to analog, copy the material through a DolbyA only in real-time, and then convert back to digital.  That is an expensive process because of the need for a recording engineer.
With my decoder -- the current version only runs in approx real-time, but I plan for newer (pro) versions to be approx 4X real time, and not only that -- my decoder is quite accurate.

So -- DolbyA isn't only real, but is very common.  Nowhere near 100% of the pre-1990 material, but somewhere approximating 25-50%.  That is a very significant issue -- moreso than you might have implied.

And, again, there is NO reason to have excessive pre-emphasis on the recordings to the extent that I have heard, but there is a lot of reason why DolbyA has been left on the recordings.  The excessive EQ is an active event, and 'left over' DolbyA happens without extra work.

John

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #7
You've completely missed the point, but it really doesn't matter.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #8
You've completely missed the point, but it really doesn't matter.
The statement:  "More often than Dolby A, preemphasis, sourced from a vinyl master, etc. is actually the case, I assure you."

Is what I have been disagreeing with.  I don't think that what you said is quite true, and I claim that DolbyA is more common than you implied.  Normal vinyl pre-emphasis before the master producing machine should not be nearly that significant, and the left over DolbyA is quite likely.  AFAIR, the devices used to produce the vinyl are pretty good, even though there might be the need to rolloff the low end and/or limit the HF signal, but the result of that processing shouldn't be anywhere near as intense what I have been hearing on the CDs that I  allege to be DolbyA encoded.  If there has to be 15dB of limiting when creating a vinyl master, someone is pushing the HF way too hard.  15dB of compression or limiting should probably be avoided at any frequency when processing recordings at the last stages.  (That much compression at mixdown might be helpful for certain artistic reasons -- but doing DolbyA levels of compression/limiting  in any kind of phase after being in the artists control is likely destructive.)  Of course, DolbyA is meant to be undone, so the damage shouldn't be permanent in that case.

So, if you were meaning to write something different than you did -- I apologize for misconstruing what you meant.

John

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #9
I was combing preemphasis, etc. with Dolby A.

Where are your stats?  Have you taken a survey or something that clearly indicates Dolby A (+preemphasis, +whatever else you can imagine, no matter how unlikely or obscure) is the reason for the majority of claims that digital sounds harsh?

Right.  I didn't think so.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #10
I was combing preemphasis, etc. with Dolby A.

Where are your stats?  Have you taken a survey or something that clearly indicates Dolby A (+preemphasis, +whatever else you can imagine, no matter how unlikely or obscure) is the reason for the majority of claims that digital sounds harsh?

Right.  I didn't think so.

I guess you like the hiss, compressed highs, limited spatial depth and LF distortion of DA encoded material?   Apparently spending lots of money on a music reproduction system is a waste if you have many of the older recordings.  Maybe some people cannot tell the difference, and that is okay with me (not because I don't care -- because I really do, but the sound isn't broken to some people.)

I have tried decoding practically all of the my music collections that I have on computer (mostly pop, some pop orchestral -- over 100 album equivalents), and the decoder tends to sound bad or really bad on non-encoded recordings (sometimes it is passable, but with no high end, sometimes shfiting L/R, sometimes more distortion, etc.)

So, my tests have been existence proof (no hairbrained theory), and even though my collection (essentially randomly obtained) is about 50% encoded (my recordings are almost 100% pre 1990), I usually suggest DA being 25% of pre-1990 material -- mostly cheap collections that apparently just played out from master tape.   Maybe that stat isn't terribly accurate -- but it is definitely order of magnitude accurate.  Listening to undecoded DolbyA (even if someone does a bit of a treble cut) doesn't sound good to me at all.  I have too many randomly acquired recordings (over the last 20yrs) for anyone to honestly claim that half of my collection is only set of encoded recordings.   There has to be huge percentage of the old recordings left with lousy sounding encoding -- and that is why I used the conservative number of 25%.

I do NOT like the cheap (low-end, Radio Shack-type) sound of compressed highs -- maybe you do -- I suspect that lot of people don't like the sound of DolbyA when applied over the entire recording.   EQ leaves the compression on the recording, it leaves the LF distortion, it leaves the spatial flatness, and cannot fix other problems with encoded material either.  DolbyA is sometimes used for effect by the artists, but that isn't what we are talking about here.  (Also, it is kind of cool when a recording from a cheap collection actually sounds better than a 'premium' or fully branded release, but only after the 'cheap' copy is decoded.)  Even some of the 'normal' album releases are encoded.

It can be more tricky that it really should to decode the material.  It can be problematical, because sometimes the turkeys (bean counters) tell the production people to use EQ (yuck.)  There are a few tricks that I have found to reverse the EQ based on how undecoded DolbyA sounds, and it seems like the first amount of EQ to try to undo is 3dB butterworth at 3k HF shelving boost.   Sometimes adding 3dB at 9kHz might help, but I haven't ended up using the 9k boost.

EVEN IF NOT TOTALLY ACCURATE, the ATTEMPT to DA decode already encoded material usually sounds better than simply 'getting by' with that harsh & incorrect sound.   I have made the mistake of thinking something sounded encoded, but found out otherwise by running the decoder perhaps at most on 5-10 recordings.  I have also varied the threshold (equivalent to setting the tone level) over 1-2dB in some cases, and the decoded version still most-of-the-time sounds better.   The threshold can be based upon the tone, if the tone has been maintained.

If I spent over $200 for an audio system, I'd definitely want better quality than the lousy undecoded sound of DolbyA -- or just as bad, attempt to fix it by EQ -- then ending up with more problems.  It is sad when a $2k or greater system sounds so bad because of an owners choice listening to an undecoded recording. If the owner doesn't have the facilities to run the decoder, then that is regrettable, but is a fact of life.  I do understand it when the decoder CAN NOT be used, but voluntarily not even trying to use it when a system might be so very expensive is a travesty.  It is sad that someone might voluntarily deprive themselves of something so superior  -- esp if the facility to provide that superior quality is essentially free.  Maybe it is a religious thing that I don't understand?  Again, there are people who cannot use the tool -- and I do regret that, and wish I could help.  I have done decodes for people who aren't able to do it, but doing decodes in volume would just take too much time.

DA is often mistaken for bad EQ -- but why would the producers explicitly decide to do bad EQ?  They aren't that stupid to do something EXTRA to make the audio sound bad, but there are failings in the late production processes that allow DolbyA stuff to be distributed -- and it doesn't cost anything of the producers.

If you like compressed HF, more hiss than necessary, distorted lows, flatter spatial image, then enjoy the material without decoding!!!   I think undecoded DolbyA sounds cheap, and I have finally realized that the undecoded DolbyA  was  a major motivating factor to quit the HiFi hobby in the '80's.   CDs often sound harsh -- like DolbyA, but I didn't realize or know that at the time.

John

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #11
Have you taken a survey or something that clearly indicates Dolby A [...] is the reason for the majority of claims that digital sounds harsh?
[extremely long diatribe]
I'll take that as a no.

...then again, perhaps I misunderstood you as saying the common complaint that digital sounds harsh is overwhelmingly due to a lack of Dolby A decoding.  If this is the case (that I misunderstood) then I apologize.

Trust me, I laud your efforts and wish I could have somehow helped get this problem solved back when I worked for Dolby.

I probably have a dozen or so titles that are likely a victim of the very problem you are addressing, FWIW.  Alas, I don't have a 64-bit processor to run your tool and don't see myself getting one any time soon, so I guess I'm out of luck.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #12
CDs often sound harsh...
..subjectively to you.
Not here. Maybe a few here and there, but the vast minority, not "often".
Loudspeaker manufacturer

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #13
CDs often sound harsh...
..subjectively to you.
Not here. Maybe a few here and there, but the vast minority, not "often".
It all depends on the genre.  I find the pop genre from the 1960's through late 1980s to have lots of DolbyA encoded material.  Some people do like the HF compressed sound, and that is okay.  It is easy to get used to that HF boost, and that is why I warn people to listen to the decoded versions first on my demos.   If you get used to listening to the harsh stuff, then it becomes normal.  I DID quit my stereo hobby because of the sound quality, but I have always been sensitive to the quality of recordings.   Now,  have 'old  ears' yet still clearly hear the difference.

Simple example how the decoder tool has helped:  I have some ABBA stuff with a distortion/noise band at between 19.1k and 19.4kHz -- in fact, it seems to appear on commercial recordings also.  That distortion band caused all kinds of problems with my listening -- it messed up lots of things with the qulaity.  I cannot directly hear that frequency, but I am clearly bothered by the 2nd order effects.  I found out that I could take the DA encoded copies of the material that are essentially 'master tapes', notch out that band, and DA decode it, and now the clarity is much better than any other copy that I have (all of the polar, polydor releases, some of the compiliations like the studio series and a few others.)   Apparently, that same kind of distortion had been a problem on the recordings since day one -- and even the so called 'pristine' polar versions are blown away by what I have, and many have that distortion band.  Supertrouper is terrible with a lot of intermod at the beginning on 90% of the versions out there -- but my decoder doesn't produce intermod like that.   Couldn't have my nice copies without the DA decoder.  (And other than the greater compression on the normal releases, my decoded copies have the same tonal balance of even the vinyl ripped copy that I have.)   I don't know where that 19200Hz distortion comes from, but the freq range is suspiciously close to the 9.6k/19.2k/38.4k baud rate -- not sure what is going on.

Almost anyone who has given the demos a chance, and often having harsh recordings themselves, the results have most of the time been positive.  There have been a few who didn't like the lush, clean sound of properly decoded material (very similar to the  sound of vinyl, but without the background rumble, muss and fuss.)  (Didn't like DBXII back in the day, except for vinyl -- really helped with rumble.)   Not all decoding is fully successful, and there is material that is NOT encoded, but almost sounds like it is...  It is a cr*p shoot at times, but I am very happy that I have the tool.  It is definitely not a religious object either way.

Again, I do admit that there is also material out there which has been EQed, and additionally the HF compression sometimes becomes a normal thing.  The EQed versions almost appear to be normal, but it also makes it more difficult to decode.  Geesh, one of my favorite groups is ABBA (pretty dense stuff)-- but also the copies that listen to are fully decoded DolbyA with peak to RMS up to 24dB (typically 20dB), and a crest factor of at least 8.   I don't like some of the recordings like the ABBA studio series with peak to RMS of 14dB and crest factor of 4.   Amazingly, some people like the sound of the compression.  I get shouted down about how good it (13-14dB pk-rms, and CF of 4) sounded!!!

So this is all a matter of taste, and preference.  I have written (and designed) compressor/limiters/expanders for 40yrs, and don't like the sound of compression -- but somtimes it is necessary.  However 14dB peak to RMS and crest factor of 4 is NOT necessary.  (Referring to my copy of the ABBA studio series.)  The DolbyA decoding is similar.

John

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #14
CDs often sound harsh...

It all depends on the genre.
Now we're getting somewhere. "It depends". Also on specifics, like group or album. A far cry from your original claim...

I have some ABBA stuff
Me too. Greatest Hits iirc. Sounds like shit. Unbearably bright and harsh. Try something else. The vast majority of my CDs don't sound like that. At all.

cheers,

AJ
Loudspeaker manufacturer

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #15
Now that we are OT enough so i tried my Super Trooper coming from an early singles CD box.
A prime example for harshness. It still sounds like shit when trying to de-emphasize it and indeed jsdyson, your expander makes it sound pretty ok!
Is troll-adiposity coming from feederism?
With 24bit music you can listen to silence much louder!

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #16
There must be something fundamental problematic at your place.
I don't know.
I have a few CDs that sounds great. It seems a really good recording and mastering job was done. It's mainly recent classical recordings from Harmonia Mundi, BIS, Glossa, Alpha and recently even Naxos.
But the majority of CDs sound harsh, compressed, flat in space and time and very often it is a pity to destroy good music like that.
Not at all so with the DSD and vinyl issues. Of course, recent vinyl often also lack in quality, but I put the same labels responsible for that.

It's not correct to be categoric, but when I look at a large music collection spanning some 70 years, I would enjoy picking let's say 100 really good recordings. Majority would be vinyl, 20 or so would be DSD. Very few would be cd releases.

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #17
But the majority of CDs sound harsh, compressed, flat in space and time
Not in physical reality, only in vinylphile fantasy world. 99% of my CDs sound worlds better than vinyl across all genres.
Something is either terribly amiss with your system/music choices or you just like the sound of vinyl, which is nothing like physical reality. IOW, a vinylphile. Same tired, invalid, old arguments from the echo chamber.
Loudspeaker manufacturer

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #18
But wait, I thought all complaints about harsh digital are due to un-decoded Dolby A, regardless of the delivery format. I guess this must actually only be for CDDA as it appears the non-harsh DSD releases must have instead been sourced by new generation tapes created by acutal engineers in order to decode Dolby A whenever the original source tape had Dolby A encoding.

Must have been a dirty little secret, but now we finally know the true reason!
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #19
The claim here seems to be that for over thirty years now, mastering engineers have often been too dumb/lazy/poorly paid to properly decode Dolby-A encoded analog master tapes when creating digital masters for release.  (Except for DSD, apparently those MEs are different)

Typically tape boxes were marked with their Dolby encoding status.

I have heard/read about a case where a Dolby A-encoded tape lacked calibration tones, and the ME had to reverse-engineer the proper output and input levels from tape deck to decoder.  (e.g. Mobile Fidelity's release of 'The Yes Album').   You can read about that here   (Btw I don't get why jdyson thinks this has to involve another onerous  analog step; could not the output of the decoder could be digitized directly? ) . 

But I have not heard this to be a widespread mistake. 

There also  may be cases too where DolbyA encoded tapes weren't decoded (as it happens another Yes album may have been victim of this, at the *mixing* stage, but that case appears to have been due to lack of communication between various engineers on the album),  but I have not heard it to be a widespread issue either.

 



 

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #20
It also appears that this is the reason given for why people think digital sounds harsh, as opposed to something like, and I'll just come right out and say it, expectation bias/placebo effect/etc. and generally happens regardless of the recording in question, unless it's SACD, a "hi-res" download, played through an apodizing filter, played through a Pono, mastered by Steve Hoffman, etc.
Is 24-bit/192kHz good enough for your lo-fi vinyl, or do you need 32/384?

Re: Vinyl is equivalent to which digital bit-depth and sampling rate?

Reply #21
Not in physical reality, only in vinylphile fantasy world. 99% of my CDs sound worlds better than vinyl across all genres.
Something is either terribly amiss with your system/music choices or you just like the sound of vinyl, which is nothing like physical reality. IOW, a vinylphile. Same tired, invalid, old arguments from the echo chamber.
Maybe you're right. Sound production is quite subjective and maybe you like what I dislike and vice versa.
I actually think dsd and tape sounds better, when available.
In my opinion, CD is the media that destroyed the craftmanship of music production, even if there's nothing wrong with the format.

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #22
Not in physical reality, only in vinylphile fantasy world. 99% of my CDs sound worlds better than vinyl across all genres.
Something is either terribly amiss with your system/music choices or you just like the sound of vinyl, which is nothing like physical reality. IOW, a vinylphile. Same tired, invalid, old arguments from the echo chamber.
Maybe you're right. Sound production is quite subjective and maybe you like what I dislike and vice versa.
I actually think dsd and tape sounds better, when available.
In my opinion, CD is the media that destroyed the craftmanship of music production, even if there's nothing wrong with the format.

Let's for the sake of argument say that it's true that "CD is the media that destroyed the craftmanship of music production", then as you also hint at, it's really the choices being made in the studio by artists, producers and engineers that have destroyed music - not the media itself. And I'm with you to a large extent. Too much music has been overly compressed or made excessively bright - but again, it's not the format. I think the music world might have been better off if digital dynamic range compression had never been invented.
As far as I know blind tests have already been conducted between DSD and PCM and people couldn't hear a difference, despite the fact that DSD seems to be a broken format according to a paper published by Stanley Lipshitz and James Vanderkooy, which just shows us how forgiving our ears really are.
But yes, I'm sure that certain/many CDs actually have had some problems during the digitization proces. J.J. (James Johnston) also spoke about this at one point.
"What is asserted without evidence can be dismissed without evidence"
- Christopher Hitchens
"It is always more difficult to fight against faith than against knowledge"
- Sam Harris

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #23
I actually think dsd and tape sounds better, when available.
In my opinion, CD is the media that destroyed the craftmanship of music production, even if there's nothing wrong with the format.
Digital disorder is a well known subset of Audiophile disorder. My condolences and best of luck with it.

p.s. CD didn't create studio idiocy and/or audiophiles. Its not responsible for anything.
Loudspeaker manufacturer

Re: Undecoded Dolby A: Why People Complain That Digital Sounds Harsh

Reply #24
I am fond of ABBA music, particularly slower ballads, and have to agree that it often sounds harsh. In vague terms, the female vocalists step onto each other, fighting for the same space. The music can't be listened loud. At moderately low level, with rested ears it sounds fine. I suspect this is partly due to reflections from walls in my room, and maybe the production decisions for adapting the sound for small speakers and radio. The CD is unlikely at fault, as the music predates it.

I'm skeptical towards this custom expander processor. I can't run the x64 binaries to experiment with it, and haven't read all of the many posts documenting its development. If there indeed was undecoded Dolby on some tracks during mixing or mastering, then subsequent decisions with levels and EQ are based on this effect being present, and it is now part of the sound. You'd have to make a series of assumptions in attempt to unwind the processing.

How do you set input level without the calibration tones? There is a software Dolby-A en/decoder in U-He Satin tape machine emulator (also included are Dolby-B and dbx), which can opearate at many times realtime with modest CPU requirements. Its effect is very level dependent, and is not subtle. Decoding with too low input level results in aggressive jumpy expansion in "Super Trouper", the quiet intro breaks. Encoding without decoding makes the bass drum disappear and snare turn to puff. I do not hear a lesser degree of this effect in the dynamic original (POCP-2207).

 
SimplePortal 1.0.0 RC1 © 2008-2019