Jump to content
IGNORED

Lies about vinyl vs digital


Recommended Posts

3 hours ago, semente said:

 

The thing is you are not appreciating because you are not comprehending what @John Dyson is doing, and worst of all you can't even identify the problem in the first place which is a very serious issue indeed.

 

John is creating what you call an "original, 'untainted' one, the first off the rank - it has the guts, the verve, the sparkle, the depth of captured music". Have you ever heard of mastering?

John is correcting inadequatelly mastered recordings.

 

 

Please provide examples of:

 

1) "inadequately mastered" recordings:

 

 

 

2) What you consider "adequately mastered" material:

Link to comment
12 minutes ago, The_K-Man said:

 

 

Please provide examples of:

 

1) "inadequately mastered" recordings:

 

 

 

2) What you consider "adequately mastered" material:

 

598212536_ThisisNOTRemastering!04Faceboo

 

Here's an example, same track, different mastering (top has DR = 10, bottom has DR = 6):

 

GJnN9wV.jpg

 

@John Dyson raised a relevant question regarding the existence of recordings made with Dolby A that were not de-emphasised for CD.

 

I listen mainly to classical so I really can't complain much.

"Science draws the wave, poetry fills it with water" Teixeira de Pascoaes

 

HQPlayer Desktop / Mac mini → Intona 7054 → RME ADI-2 DAC FS (DSD256)

Link to comment
11 hours ago, John Dyson said:

Thanks for the suggestion :-).  I have to document the monstrosity first.  I have written a lot of complex things, fthis DA decoder is A LOT more complex than anything else that I have done myself.   The DA decoder is the single most complex piece of software that I have written, and when it comes to DSP, it is probably the most complex piece of software that I have ever worked on.

 

Today, I just finally started documenting it piece by piece.  It is fun to revisit, but just as a small example:  a real challenge how to describe a complex time variant filter behaviors and how they all interact to do 'whatever is needed'.  I have procrastinated on the documentation for the last 2yrs, the basic concepts haven't changed much, but the techniques have advanced.   There are probably 1000 different filters in the decoder, many Hilbert transform filters, lots of FIR filters, even some IIR filters.  There are nonlinear attack/release filters, just everything that one might think of.  A lot of the filter types can have their time constants modulated -- very important capability in the DHNRDS DA.

 

* Note -- I hint at potentially useful technologies in this note.  Anyone implementing a compressor, limiter or other kind of gain control device  (e.g. expander, etc), and interested in any of the advanced techniques, I am willing to supply information on the techniques.  Some of the techniques can almost do the previously 'impossible'.

 

My guess is that the most generally useful techniques developed are the anti-IMD and anti-MD portions of the code, but the algorithms for DolbyA attack/release calculations are technically 'interesting.  The attack/release algorithms are what is missing in the Sony DolbyA patent.  In the Sony patent, they just said 'fast attack/slow release' -- like yea :-).  Mmmmm....   Creating the attack/release algorithms was more complicated than figuring out a way to emulate the feedback compressors.  A straight audio feedback compressor (like a FET feedback compressor) is not possible to directly implement in DSP.   I wouldn't be surprised if Dolby would have been interested in the DHNRDS DA DolbyA SW technology.  They told us that they were no longer interested and wouldn't consider us to be competition...  If they were interested,  the strangely detailed/complex attack/release algorithms and parameters would be super interesting to them.  The idea of attack/release code might seem simple -- just emulating some diodes and some resistors/capacitors.  If the emulation was that simple, it probably would already have been done!!!   The effective time constants wobble all over the place based upon signal level, state of each section of the attack/release emulation, and even the DolbyA gain.  Each of these has to be implemented for each band of each channel.  That is a lot of math operations!!!

 

The 'fancy' things are the anti-MD and anti-IMD, that is, these are 'extra credit' sections, not really needed to emulate a DolbyA,but makes the DHNRDS DA so very nice :-).

 

The anti-MD is great for getting rid of the effects of the fast gain control creating modulation products in the audio -- the basis of the DolbyA fog.   I developed an incremental method of doing the traditional 'gain * signal' calculation used everywhere in gain control -- but my version doesn't create the modulation products in the same way as normal methods, essentially hiding them, but still remaining in the signal.  (The modulation products are mathematically necessary, but BAD modulation products are not mathematically necessary.  There are patents that do a part of the solution, by my method is *really* *really* good at hiding the MD.)

 

 

The anti-IMD is probably more useful for DolbyA than other compressors/expanders, but might still be helpful.  The DolbyA attack/release calculation creates beat products, which then interfere with the signal by modulating the gain.  Makes ugly sound.  The anti-IMD massages the measured signal so that it no longer has normal 'audio frequencies' in it, but still has the same statstics as a normal audio signal.  Without the 'audio frequencies' in the level measurement, then there is no opportunity for the DolbyA attack/release calculation to create those ugly beat products.  The creation of this 'fake audio' signal is nonlinear and time variant.  Fast attacks are still fast, but 'ripple' as in a level detection ripple doesn't happen.  Also, fast ripple doesn't happen.  The signal going to the attack/release code is really, really clean.

 

One more thing that can be patented also is the DHNRDS DA structure itself.  It is very ad-hoc but also very simple.  it avoids infringing on the Sony patent, and isn't even recognizably the same.  It might be fun to document it -- but no one ever will write a DolbyA decoding emulator again.  I just might write a simple DA decoder for others to use in plugins.  Wont' sound nearly as good as the DHNRDS DA, but about the same as a true DolbyA...  The simple decoder would be good for helping others, but I doubt that anyone is crazy enough to replicate the DHNRDS DA...  It does things that I'd suspect that most real experts in the field would say are impossible.

 

I was thinking about patents, but instead I think that if I do an AES article for each major innovative subsystem, that might be more beneficial than a patent to me.  Some of the concepts could be very helpful for very specific types of equipment...  Mostly compressors/limiters-- stuff like that.   Little people like me don't make money on special purpose patents, but a bit of name recognition could be helpful for future endeavors.

 

But, thanks again for the thought -- I keep trying to figure out how to do the documentation, but only take a few hours :-).

 

John

 

 

 

John,

 

I'm not a patent lawyer but you do need to talk to one.  Your documentation for the process is not the amount of detail you would ever include with a patent.  With a patent you want enough information to protect your invention, but not too much that it can be easily duplicated and then modified to be something else.  

 

Stephen

 

I

Link to comment
1 hour ago, mansr said:

Didn't Dolby already patent the thing back when they invented it?

 

My concern is: Where, in the process of transferring A-encoded masters to CD back then, did the breakdown - neglecting to decode A-NR, or decoding unencoded master by mistake - occur?

Link to comment
1 hour ago, SJK said:

John,

 

I'm not a patent lawyer but you do need to talk to one.  Your documentation for the process is not the amount of detail you would ever include with a patent.  With a patent you want enough information to protect your invention, but not too much that it can be easily duplicated and then modified to be something else.  

 

Stephen

 

I

Oh -- I know about patents, and in fact on a few.  We had patent lawyers for the leg work in the past.  I doubt that I will patent anything, but will go the route of disclosing the technology publically (maybe even GPL -- yech!!!) therefore making it prior art.  IP issues DEFINITELY must/will be better considered, but the very good news is that the actual algorithms and everything except one-support- library is owned entirely by me, and not even the DHNRDS project!!!  (The DHNRDS project has perpetual use and redistribution of each version that is added to it.)

 

The algorithms are so complex that I doubt anyone could infer the details enough to replicate the work based on anything that I have previously written.  Even my private communications are obtuse enough not to have comprehensible details.  *Even* if I disclosed the source code today, it would require someone much brighter than me to understand the methods -- it is written as a bunch of rather non-trivial/well-considered C++ classes that make it easy for me as the author with a certain 'mindset', but without understanding perhaps 20-50 interesting C++ classes, it wouldn't be easy at all to work through.  * Major frustration for me is to document that 'mindset' for the C++ classes so that other people can understand and *utilize* the concepts.  When I claim that the software is non-trivial, I am not understating it...

 

You are motivating me to start working *seriously* on the IP issues, and I might be forced to patent the methods, but would really like to avoid patents.   It is truly mind-boggling for me to explain, but the algorithms are developed about as far as they can be -- I have ferretted out every little niche where the current direction cant really be pushed much further.

 

BUT -- the IP issues are important to make sure that the software continues to exist.  My goal is that of LEGACY, not MONEY.  Somehow, in the past when I have done 'good works', money has followed without trying very hard.  (I used to have the 'Biden' thing happen to me all of the time, that is, can we pay you just to say that we have paid you?)   That hasn't happened in over 20yrs now though :-(.  It might not happen again, but might create some actual opportunities.

 

Thanks for the sentiments -- the IP issues have been in the back of my mind, and I have done patent searches on gathering ideas for my implementation, but could find nothing useful.  The term 'not-applicable' comes to mind.  INTERESTING non-applicable stuff for sure, but not applicable.

 

You might be right -- I might be forced to patent, but I pray not.

 

John

 

 

 

Link to comment
1 hour ago, mansr said:

Didn't Dolby already patent the thing back when they invented it?

Methods and designs can be patented.  I doubt that the software looks ANYTHING like the HW or even the architecture.  The architecture is foreign even to the Sony patent.

I must be careful about 1) cannot represent that Dolby endorses or accepts my project or product, 2)  Cannot formally represent that the DHNRDS DA is a DolbyA unit, but instead that it can decode DolbyA encoded material.   I probably encroach into item (2) a little too far, but NEVER have advocated the purchase of the software, and immediately pass any 'business' stuff off somewhere else.   I talk about the DHNRDS DA more as a research project instead of commercial.

 

If I could get more support as a 'free' software project, I'd probably go wholly in that direction, but I have gotten support from those with commercial intentions, and respect that relationship (as friendship, comrades and a kind business involvement) 100%.

 

John

Link to comment

Just an afterthought about 'documenting' the DA decoder (part of the DHNRDS project.)  There is also the feralA filters.

 

I'll be much more disciplined about documentation when I implement the built-in corrective EQ filters, and will write a public note about the EQ, and the various schemes used.  All of the schemes are similar (except in a few cases), and there even appears to be a commonly used EQ.  That commonly used EQ appears to be chosen more often than the others.

 

-------------

 

Side-note...   As a 'research project' and not really for decoding per-se, I am doing an ABBA decode based upon DA decoding effects instead of the 'ABBA' sound.  I am getting VERY INTERESTING results.   Some of the albums come out more traditional sounding, others are not amenable to the more normal EQ.

 

So far, I really like the sound of SuperTrouper when compared with my 'compatible' decodes or even the original versions.  The sound *is* different, but can tell where they did some things to get a certain 'sound-effect'.

 

I like ABBA (the group) a lot.  It isn't the individual members, but instead it is a certain style and a certain 'camp' that they portray.  Somehow, they are likeable and human.   Their music is 'nice', but always sounded bad to me.  This is the 2nd time that I am trying a more careful reconstruction of their recordings.  ABBA is the only group where I might be caught 'remastering'.

 

John

 

Link to comment
10 hours ago, The_K-Man said:

 

YES.  The issues you are hearing have zero to do with the format.   Digital audio has gotten beyond all that.


well I don’t know about you but my new phono amp has elevated my vinyl enjoyment beyond what I get from digital. 
 

it’s a personal opinion. No need to get defensive. 

No electron left behind.

Link to comment
27 minutes ago, AudioDoctor said:


well I don’t know about you but my new phono amp has elevated my vinyl enjoyment beyond what I get from digital. 
 

it’s a personal opinion. No need to get defensive. 

 

Just be aware though: digital does not 'change' or mess with the audio, unlike your brand new phono pre-amp.

Link to comment
14 hours ago, semente said:

 

Frank, I am sorry to inform you that your standards are just too low. I get the benefits of tweaking, I've experienced first hand how some modifications can provide significant improvement to equipment performance, but you really should be a (big) bit more demanding. We are reproducing recorded vocals and instruments and detail is not all that music is about: audio is not music in black and white, it has colours and the system must get them right too.

And the same is valid for the recording process, of course.

 

I am very demanding ... which is why I dismiss, immediately, nearly all high end rigs I come across - they are getting so much wrong, and making  a mess of a recording I know well - to me, at that moment, they are no better than an AM radio - a very old VW, in Porsche clothing ...

 

14 hours ago, semente said:

 

The thing is you are not appreciating because you are not comprehending what @John Dyson is doing, and worst of all you can't even identify the problem in the first place which is a very serious issue indeed.

 

The "problem" is that the mastering is of a type that makes added distortion from the playback chain very, very obvious ... you see, I use those styles of recordings that John is processing, in their original form, to assess my progress in optimising  - they "make it difficult" for the system, and tell me how 'robust' the playback integrity really is.

 

Otherwise, it's like having a Mercedes which you can't drive over certain, everyday roads - they have "nasty bumps", and the suspension bangs and crashes trying to navigate them - the engineering is not up to scratch.

 

14 hours ago, semente said:

 

John is creating what you call an "original, 'untainted' one, the first off the rank - it has the guts, the verve, the sparkle, the depth of captured music". Have you ever heard of mastering?

John is correcting inadequatelly mastered recordings.

 

Why are there all these remastered Yes albums? Because people didn't like how the original one, with "the guts, the verve, the sparkle, the depth of captured music" sounded on their setups - so, there was a market for the "baby food" versions. "Inadequate" is a subjective decision, not an objective one.

Link to comment
Just now, mansr said:

The only times I've bought another version of something I already owned has been getting older used CDs to escape the loudness-compressed remasters.

 

Perhaps you aren't their target customer.

 

I, on the other hand, have at least a dozen different versions of Kind of Blue.

Sometimes it's like someone took a knife, baby
Edgy and dull and cut a six inch valley
Through the middle of my skull

Link to comment
41 minutes ago, AudioDoctor said:


well, that’s roughly half of the Lampizator Pacific so maybe I’ll have to check it out. Haha. 

Interesting point and back on topic. And that being with FPGA where the result of the hardware is predicate upon the software, is there any difference with digital reproduction design other than the imagination of the programmer?

Link to comment
2 hours ago, mansr said:

The only times I've bought another version of something I already owned have been getting older used CDs to escape the loudness-compressed remasters.

 

 I am fortunate in having started early with CD and having quite a few albums from before the Loudness Wars

 

How a Digital Audio file sounds, or a Digital Video file looks, is governed to a large extent by the Power Supply area. All that Identical Checksums gives is the possibility of REGENERATING the file to close to that of the original file.

PROFILE UPDATED 13-11-2020

Link to comment
19 minutes ago, sandyk said:

 

 I am fortunate in having started early with CD and having quite a few albums from before the Loudness Wars

I was never impressed with CDs, at least not in the early days.  By that I’m guessing you mean the early 80’s?
 

If memory serves correctly I was running with a Denon DP75/DK404/DA401/Dynavector Karat Ruby, Dynavector DH something head amp, Revox B77, Revox B780 etc.

 

And then, small silver shiny things. OK, I can do that.  Revox B225. Got it covered. 
 

But wait.  That Denon turntable designed for good looks but dead, flat sound made for a radio station that couldn’t or wouldn’t ever see any feedback sounded better. 
 

CDs were a convenience, and a great one. You couldn’t play LPs in your car or with your friends on the beach. They’ve come a long ways, no question. And, if nothing else, we can play Beethoven’s Ninth uninterrupted, that being what set the maximum play time of the CD by Sony. 
 

 

 

Link to comment
53 minutes ago, SJK said:

Interesting point and back on topic. And that being with FPGA where the result of the hardware is predicate upon the software, is there any difference with digital reproduction design other than the imagination of the programmer?


I think the answer is that it depends on the goal of the designer. You do make a good point though. 

No electron left behind.

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...