Jump to content
IGNORED

'FeralA' decoder -- free-to-use


Recommended Posts

28 minutes ago, semente said:

Hi @John Dyson,

 

Sorry if this has been asked before but would it be possible for you to list all the steps that you take when you are performing "de-emphasis/decoding" of feral A albums editions/masterings?

 

My second question may become irrelevant after you reply to the first but I will ask anyway: are you performing only "decoding" or are you on top of that trying to "remaster" the tracks in order to perceptually improve their sound?

 

It's been some time since I last followed your work for mutiple reasons, mostly because I don't listen to most of the music that you've chosen to "decode", but there are albums which I still sometimes listen which I believe could benefit from the "treatment".

Also because all my computers are Mac and I cannot use the programme that you've kindly made available to the community.

 

One more question that's just come up to my mind, possibly a dumb one.

Wikipedia describes Dolby A steps as follows:

The input signal is split into frequency bands by four filters with 12 dB per octave slopes, with cutoff frequencies (3 dB down points) as follows: low-pass at 80 Hz; band-pass from 80 Hz to 3 kHz; a high-pass from 3 kHz; and another high-pass at 9 kHz. The compander circuit has a threshold of −40 dB, with a ratio of 2:1 for a compression/expansion of 10 dB.

Are feral A albums the result of recording masters without Dolby A pre-coding which went through Dolby A de-decoding process during mastering, thus explaining the exaggerated treble and limited bottom end?

 

master tape with no Dolby A pre-coding mastered with Dolby A de-coding CD

 

Thanks,

Ricardo

 

 

Sure -- will do it, but until you mentioned it, I didn't think about 1st level tech docs as an immediate priority, but now I do.  I am happy that you mentioned this.

 

I'll try for Sunday, but it might be 3-4 days.   It will be in a rough, but decorated text form -- I have an almost medical inability to draw or structure graphical images.  All of the applicable/useful information will be in the document.  As you can imagine, I have to refer to the program itself, looking for every subtle FA detail, and plan to do so this weekend.   Since the decoder really is very close to completion, I think  the technical information must be memorialized.  I am really glad that you reminded me.

 

 

 

 

 

 

Link to comment
2 hours ago, KSTR said:
  • When you revisit the project, make a sane first things first prioritisation. From my point of view, wrt to the code itself, the top topic is the house EQ. This EQ spoils everything, really, I mean it. Neither you nor anybody else will ever be able to do any meaningful comparisons as long as the main and very dominant effect is that EQ, spanning some 12(!!!) dB (and even if it were only +-1dB that still is too much). Comparing/judging the low-level dynamics requires same large-signal frequency response to +-0.1dB and +-0.1dB of level matching, no way around it. We know this EQ is not helping, you know it is not helping. It is completely superfluous. If some post-EQ is deemed necessary to polish up the result (which may or may not be the case), this can always be done in a extra pass, with different means.

 

Klaus, not sure if you've tried it, but this is what I do to de-EQ John's files. I simply use DeltaWave to match the RAW and Decoded file, and use non-linear level EQ correction only (uncheck phase). The result is to undo the large-level EQ in processed files. You can then play or export the corrected files. Here's an example:

 

Uncorrected EQ (RAW vs. DEC):

image.thumb.png.0ead7018dcc0efe045dc380e4790ad87.png

 

After DeltaWave frequency correction:

image.thumb.png.e9d49ccee8afc3be9cfb49a40d493f09.png

 

But, I'm thinking this will make the decoded files sound a lot more like their RAW versions. At least it does to me when listening to these. 

Link to comment
1 hour ago, KSTR said:

Yes, this is getting out of control.

 

@John Dyson, If I were in your position this would be my ToDo list (my 2ct.):

  • zip the source tree and important toolchain settings etc and any relevant docs and stash it away in a safe place (real or in a private net cloud) where someone, say your best friend or family, has access to. Just in case something fatal happens.
  • Make a public note so people know and then take a real break from the project. Something like two to four weeks, minimum. Don't read/write forums etc, shut down all non-essential computers etc etc. Try to enjoy your time concentrating on other beautiful things of life.
  • When you revisit the project, make a sane first things first prioritisation. From my point of view, wrt to the code itself, the top topic is the house EQ. This EQ spoils everything, really, I mean it. Neither you nor anybody else will ever be able to do any meaningful comparisons as long as the main and very dominant effect is that EQ, spanning some 12(!!!) dB (and even if it were only +-1dB that still is too much). Comparing/judging the low-level dynamics requires same large-signal frequency response to +-0.1dB and +-0.1dB of level matching, no way around it. We know this EQ is not helping, you know it is not helping. It is completely superfluous. If some post-EQ is deemed necessary to polish up the result (which may or may not be the case), this can always be done in a extra pass, with different means.

Have you listened to the latest demos?   There isn't much difference that I'd call "EQ".   There never was any intentional EQ -- I will be doing a diagram of what is going on.  If you ask, I can add a few mods to show you the results with JUST the inter layer EQ -- I think that you'll find the results 'challenging' to listen to.

 

This response is long, but claims of triviality are made a solution that cannot be trivial.  First thing that might help those who might want to assist -- simply listen to the output of a DolbyA.   That would be an interesting experience to get started.   It isn' timpossible to listen to, but it is WRONG for listening, also there are lot more odd things than is immediately audible.  The solution is multi-dimensional -- not just 'frequency response'.  If it was just 'frequency respose', the project would have been completed in minutes.

 

If the EQ isn't done exactly as I write, the levels/gains/signals won't track in the correct levels for the correct gain control to get results that don't gate, sound like there are amplifier dead-zone problems...   I have tested proven non-FA material, and the result there is a really ugly deadzone beahvior at probably 5kHz on up,  it does have the same emotional effect of 'screeching fingernails on the chalkboard), and really does sound like distortion, but of course it is just dynamics processing emulating a deadzone beahvior.

 

First thing. after you make sure that you have the correct parameters BETWEEN the layers to keep your speaker cones intact,  after that,  you'll need to do is to add on output:

three 1 pole at 9kHz -7+-dB (the 7dB is actually 9kHz/21kHz, still an 'even' number because of the calculation.)   After thatt, the respose will be a little 'wrong' (very wrong), then add some LF boost at 75Hz, about 2 1 pole, and a 2pole at 75Hz to top it off.

 

 

Once you do that, then the sound might start getting plausible.   There are various other things like pre/de-emphasis that push the dynamics processing in the right direction.

 

The in-between EQ MUST be exactly1-3kHz EQ (3 1pole HF shelf -3dB at1kHZ, also 3 1pole HF shelf +3dB at 3kHz) -- must be 6 poles but this is also simplified (just makes it easier to read) -- and 1pole 500Hz/-3dB and 1pole75Hz -3dB.   These in between EQ must be exactly the above, or the more accurate version that I can describe offline, but will document this weekend.  If you don't do this,,the associated freq range (e.g. 1kHz to 3kHz or <approx 1kHz will blast your speakers.)   The numbers that I give will keep the gains in the correct place forproper tracking and keeping your speaker cones from being popped.)

 

Basically, there is NO disableable eq beyond the minimum.   There is some 'secret sauce' stuff, but that mitigates some lingering distortion, but doesn't change the EQ.  (Similar to the pre/de-emph needed to keep the dynamics straight -- the secret sauce EQ mods change how the EQ is done with careful phase compensation, hiding some of the artifacts from dynamics processing, but does essentially the same thing as the simple EQ.)

 

If you don't do the above, you get a real mess.   Note that there is, in-total 8 ABSOLUTELY NECESSARY 1poles between layers.   Also, there is absolutely necessary gain (even dB, of course) between each layer.   If the gain isn't used, then the planned 10dB calibraton diffs will drift from layer to layer, also the output levels will be very different than the input level (by on the order of -12 to -15dB because of accumulated gain error.)


I forget the input EQ -- but I do know that there are at least 2 1 poles for pre-emphasis,  3-4 1 poles for de-emphasis (needed beacause of original engineeers design choice and behavior of DolbyA on average.)  The output EQ was the three 1 pole (ABSOLUTELY NECESSARY) and the final LF EQ, whic ends up being several 1pole, one 2pole.  Without the output EQ, it will both shreak and blow your speaker cones (maybe overvoltage electrostatics and shock the user) :-).

 

After all of this, every filter rollof parameter is EVEN dB or directly related to a frequency ratio.   Likewise, absolutely every filter frequency is an even number out of a standard:

on the low end:  1250, 1000, 750, 500, 250, 150 (not used), 125(not used), 100(not used), 75, 50, 37.5, 18.75 and 9.375Hz, which DOES have some audible effects because it

is single pole.   ALL dB numbers are:  (plus or minus) 1.5,  3.0, 6.0 and might be a single 9dB in there somewhere.  Most filter shelf gains are a composite of multiple 1.5 or 3.0dBs.

 

If any frequencies are varied, or gains changed, the results will be VERY bad.  The only serious trouble, sans pre/de-emphasis, has been the signal between 1kHz and <20Hz.  That is a wide freq range and has profound effect on the sound.  The HF problems would never had problems if I could hear -- since we have so many 'golden ears' out there, a specific quantification of the kind of HF problem would have helped.  However, I kept getting wierd nonsense, even when true, didn't normally address the technical problems that needed solution.

 

Some people might claim that the pre/de-emphaiss was not needed, but they are only looking at the static frequency response.   There are a lot more things in dynamics processing than 'static frequency response'.   The dynamics processing world is a LOT more complicated than that.  Next, I'll be told that the DolbyA attack/release calculations can be done with a simple R/C timeconstant.


If someone sees this project as being 'simple', I would appreciate backing it up with a working copy of the simple version.  Then, I might have something to learn.

 

 

 

 

 

Link to comment
22 minutes ago, pkane2001 said:

 

Klaus, not sure if you've tried it, but this is what I do to de-EQ John's files. I simply use DeltaWave to match the RAW and Decoded file, and use non-linear level EQ correction only (uncheck phase). The result is to undo the large-level EQ in processed files. You can then play or export the corrected files. Here's an example:

 

Uncorrected EQ (RAW vs. DEC):

image.thumb.png.0ead7018dcc0efe045dc380e4790ad87.png

 

After DeltaWave frequency correction:

image.thumb.png.e9d49ccee8afc3be9cfb49a40d493f09.png

 

But, I'm thinking this will make the decoded files sound a lot more like their RAW versions. At least it does to me when listening to these. 

 

Before reading this -- if you can resolve the needed diffs into 1st order EQ (the only kind of EQ normally used), please give me the information.  I'll immediately implement it, and we can hear how it sounds!!!  Of course, we also need to know which EQ to do BEFORE the decoding and AFTER the decoding.   Remmeber, all I did was 1st order EQ above 1kHz.  Since we are working in 1st order EQ space, it should be easy for your program to resolve it into 1st order EQ.  If not, then there is something more complicated going on.  Again, 2nd order EQ is used in one architectural place,  with a very specific and defendable purpose.

 

When doing comparisons, it isn't fair to do comparisons against material that is already multi-band compressed.  Just like when I was working on DolbyA,

you don't compare the decoder output and input for listenability, you compare through the entire encode/decode cycle.  Once we have an original copy,

the FA copy, then we can discuss frequency reponse errors on the calculated decoder ouput.   I don't have a good 'original' and FA copy combination

to begin with.  If I had a good example of those two items, from almost any recording, then the decoder can be made very accurate.  As is now,

comparing FA with the Decoder output isn't a lot different than comparing DolbyA input and output or DBX input and output.   I do agree that

the output of the FA decoder should sound SIMILAR to the FA RAW.   However, the FA RAW should also sound similar to the original

recording, but the encoding process is probably not flat also.


We need to compare pre-encoded copy with decoded copy, I really don't care about the FA copy for comparison about technical accuracy.   Once we have pre-encoded, FA generated from pre-encoded, the decoder source, and myself to fix it -- within hours, the decoded copy and the input copy will be essentially exactly the same.

 

Since we have no reference specs, schematics, reference recordings, test recordings, etc, we are stuck using reverse engineering, listening for response balances, and hearing 'tells'.  'Tells' really do work, but require alot of experience to hear and interpret.  This is lot like hearing dynamics distortions -- most people will assert that it doesn't sound correct.  Instead, I will be able to describe the problem.

 

I don't think that the static (I mean static) frequency response makes any sense on a mult-band dynamics processor with a lot of necessary EQ around it.

There are some situations where the frequency response should be flat, but claims that there must be a flat response has no supporting facts.

 

There are most likely instances where the decoder should be flat, and maybe it really should be,but unlikely after listening how FA signal sounds.   The

HF (esp over 9kHz) sounds smushed.

 

 

Link to comment
1 hour ago, pkane2001 said:

Klaus, not sure if you've tried it, but this is what I do to de-EQ John's files. I simply use DeltaWave to match the RAW and Decoded file, and use non-linear level EQ correction only (uncheck phase). The result is to undo the large-level EQ in processed files. You can then play or export the corrected files.

Hi Paul,

I thought about it but haven't tried. I've used my own de-embedding, which, as you might know, suffers in that it applies the correcting IR to the original, rather than than applying 1/IR to the decoding. But, it compensates 100% all mag and phase errors to full precision (good enough to do a "simple load" only in DW and get the best null possible).

The overall EQ is still present but now it's the same for both files. And then I would apply a simple min-phase correction, a curve fit EQ filter parameter set obtained from REW and transformed to an IR in RePhase. For quick checks, I only apply the latter... which leaves 1dB of wiggle room in the difference and more importantly, does not correct those linear phase level jumps at 3kHz an 9kHz.

 

With level EQ only applied by DW, that would mean applying a linear-phase correction(?), which isn't fully correct. Most of John's EQ is minimum phase, except for those mentioned step changes. Probably not a big deal as the overall curvature is low-Q, reducing chances of nasty pre-ringing.

Link to comment

I don't know if you know about the necessary 80Hz 2nd pole EQ needed on the output.  I know that they must be 3/+3dB each.   However, the Q value has been in question.  Usually, in well designed high tech analog audio, you see numbers like 0.50, 0.8409, 1.0, 1.19.  I did skip 0.7071, because at least, in my experience, Butterworth filters were never the correct answer.  On DolbyA, he used 1.17 for  the 80Hz filter, probably was a round-off error from 1.19.  However, on Anne Murrays' recordings I never got the indication of 'proper balance' on the LF until this morning, whenl I made a rather odd change.   Instead of the normal choices that I had tried like 0.8409, 1.0 and I even went as high as 1.19. I didn't get that 'Bingo' sound until 1.414.  Of course, it must be very close to 1.414, and 1.5 doesn't work.  Very interesting that they used a Q value so very high.  (Q=1.414 is fairly high for high fidelity audio -- unlike RF where you might see Q=200 on legacy IF stages.)

I get surprised almost every day with things like this in the design.   I really do try to follow 'the rules', in fact, do so very religious.  The reverse engineering does stay sane if one does follow 'the rules', but it seems like there are always exceptions.

 

Link to comment
59 minutes ago, John Dyson said:

Have you listened to the latest demos?   There isn't much difference that I'd call "EQ".   There never was any intentional EQ

Not yet, John, but will do (and analyzed them, or the new binary directly).

At the moment there IS difference that is plain EQ, as explained, anything more than +-0.1dB is counterproductive for the task. It doesn't matter if the EQ was introduced intentionally or by accident.

Link to comment
17 minutes ago, KSTR said:

Hi Paul,

I thought about it but haven't tried. I've used my own de-embedding, which, as you might know, suffers in that it applies the correcting IR to the original, rather than than applying 1/IR to the decoding. But, it compensates 100% all mag and phase errors to full precision (good enough to do a "simple load" only in DW and get the best null possible).

The overall EQ is still present but now it's the same for both files. And then I would apply a simple min-phase correction, a curve fit EQ filter parameter set obtained from REW and transformed to an IR in RePhase. For quick checks, I only apply the latter... which leaves 1dB of wiggle room in the difference and more importantly, does not correct those linear phase level jumps at 3kHz an 9kHz.

 

With level EQ only applied by, that would mean applying a linear-phase correction(?), which isn't fully correct. Most of John's EQ is minimum phase, except for those mentioned step changes. Probably not a big deal as the overall curvature is low-Q, reducing chances of nasty pre-ringing.

The original design was analog, you really do want to use 1 pole EQ, or it will not be complementary.  You can sometimes get by with linear phase, and I have a linear phase religion when doing my own designs from scratch.  When doing certain analog applications where the phase could be important, unless very careful, use the normal analog style emulation (minimum phase.)   On the DolbyA front end, I got by with linear phase for only one reason -- the Q was 0.420 and 0.470.   For the 80Hz Q=1.17 I simply used an IIR filter.  An IIR filter wouldn't work for the 3kHz and 9kHz because they were not a close enough match because of the bilinear transform.    This is one reason why I am going to write my totally general purpose S-Z bilinear converter, so I can more easily use carefully crafted higher order EQ without a lot of scribbling on paper.  Also, all of the DA EQ is calculated at startup time to adapt to sample rates.

* I have thought about running the band splitting at a higher sample rate, which is a reasonable alternative.

 

ADD-ON:   the reason why close to exact emulation is important in the specific DolbyA situation is that the phase between the 0-80Hz and 80Hz to 3kHz bands is critical.  The compression/expansion is so fast that waveshape coherency is very very important.  Any time there is dyanmics processing involved, especially with the eccentric DolbyA involved, you gotta be somewhat aware of phase.   If your FIR emulation also includes the phae characteristics (esp below 1kHz), then there is chance of hope.

 

For those who are interested, effort would be better spent hunting for truely pure material and the FA equivalent.   Once we have that, then the solution is quick and easy.   I had looked around quite a bit, maybe you can do better.


We have an engine that can do whatever is needed -- we simply do not have accurate information for the specs.   Getting the specs (or input output examples) is MUCH MUCH more helpful than measuring static frequency response -- REALLY!!!

 

 

 

 

 

Link to comment
5 minutes ago, KSTR said:

Not yet, John, but will do (and analyzed them, or the new binary directly).

At the moment there IS difference that is plain EQ, as explained, anything more than +-0.1dB is counterproductive for the task. It doesn't matter if the EQ was introduced intentionally or by accident.

There is no +-0.1dB unless you compare with the original, pre--encoded materials.

It might be fun to measure the DolbyA decoder also input/output instead of through the system.  It will compare similarly at a limited range of levels, but what has been learned?   Steady state? 

 

The above is rule #1 in the Audio Processing 101 class.

 

Link to comment

We do not have general access to pre-encoded masters (other than the occasional old vinyl edition). We only have CD's where DolbyA units (sometimes even modifed ones) have been intentionally mis-used artistically to "improve" the final sound. Most certainly there was additional EQ and sum compressors etc applied before the master for a CD was send out. All of this is a great unknown.

 

And therefore undoing the specific part of low-level dynamic compression of the DolbyA is obviously close to impossible with a one-size-fits-all approach. Exactly as you say, way too many unknowns. For example, you cannot know the pre-/de-emphasis required to undo a potential final mastering EQ and hence the levels for DolbyA de-embedding will be off anyway even if the general level match were close otherwise.

 

FWIW, I happened to find the Simon&Garfunkel LP in my wife's collection and sorry, no way your decoded version (from two weeks back or whatever) is any close tonally to that vinyl.

 

Therefore, best leave the post EQ alone. That can be applied seperately if needed (which again is mostly personal preference).

Link to comment
32 minutes ago, KSTR said:

Hi Paul,

I thought about it but haven't tried. I've used my own de-embedding, which, as you might know, suffers in that it applies the correcting IR to the original, rather than than applying 1/IR to the decoding. But, it compensates 100% all mag and phase errors to full precision (good enough to do a "simple load" only in DW and get the best null possible).

The overall EQ is still present but now it's the same for both files. And then I would apply a simple min-phase correction, a curve fit EQ filter parameter set obtained from REW and transformed to an IR in RePhase. For quick checks, I only apply the latter... which leaves 1dB of wiggle room in the difference and more importantly, does not correct those linear phase level jumps at 3kHz an 9kHz.

 

With level EQ only applied by DW, that would mean applying a linear-phase correction(?), which isn't fully correct. Most of John's EQ is minimum phase, except for those mentioned step changes. Probably not a big deal as the overall curvature is low-Q, reducing chances of nasty pre-ringing.

 

Yes, linear phase correction in DW if phase is not selected, so you're right,  if minimum phase filters were used this will not undo them fully. Engaging phase correction will, though, but I think this will just negate all the main effects of the decoder :)

 

I did listen to a couple of tracks, RAW, Decoded, and de-EQed by DW. I think the latest decoder does a better job with frequency balance, although to me, there's an overemphasis on upper-mid frequencies. Vocals sound brighter than they should, a bit more sibilance. Not something unexpected if you look at the frequency bump from 2 to 7kHz, right where the ear is most sensitive. This gets fixed with the de-EQ, but, like I said, I can hardly tell the difference between the RAW and Decoded with de-EQ.

Link to comment
12 minutes ago, KSTR said:

FWIW, I happened to find the Simon&Garfunkel LP in my wife's collection and sorry, no way your decoded version (from two weeks back or whatever) is any close tonally to that vinyl.

 

Which S&G album are you referring to?

"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
17 minutes ago, KSTR said:

We do not have general access to pre-encoded masters (other than the occasional old vinyl edition). We only have CD's where DolbyA units (sometimes even modifed ones) have been intentionally mis-used artistically to "improve" the final sound. Most certainly there was additional EQ and sum compressors etc applied before the master for a CD was send out. All of this is a great unknown.

 

And therefore undoing the specific part of low-level dynamic compression of the DolbyA is obviously close to impossible with a one-size-fits-all approach. Exactly as you say, way too many unknowns. For example, you cannot know the pre-/de-emphasis required to undo a potential final mastering EQ and hence the levels for DolbyA de-embedding will be off anyway even if the general level match were close otherwise.

 

FWIW, I happened to find the Simon&Garfunkel LP in my wife's collection and sorry, no way your decoded version (from two weeks back or whatever) is any close tonally to that vinyl.

 

Therefore, best leave the post EQ alone. That can be applied seperately if needed (which again is mostly personal preference).

Between two weeks ago and now, might as well be comparing with the single band toy.

 

Looking at frequency response is cool...   It will be interesting to compare before and after, once we get raw non FA material.

In fact, if you do have the S&G album without FA, and I have FA versions -- if you could, lets make arrangements for some longish snippets of the original, non-FA material, and I can complete the decoder to be totally accurate.

 

This would give the before and after that I have been looking for.

 

 

 

 

Link to comment
1 hour ago, John Dyson said:

 

Before reading this -- if you can resolve the needed diffs into 1st order EQ (the only kind of EQ normally used), please give me the information.  I'll immediately implement it, and we can hear how it sounds!!!  Of course, we also need to know which EQ to do BEFORE the decoding and AFTER the decoding.   Remmeber, all I did was 1st order EQ above 1kHz.  Since we are working in 1st order EQ space, it should be easy for your program to resolve it into 1st order EQ.  If not, then there is something more complicated going on.  Again, 2nd order EQ is used in one architectural place,  with a very specific and defendable purpose.

 

When doing comparisons, it isn't fair to do comparisons against material that is already multi-band compressed.  Just like when I was working on DolbyA,

you don't compare the decoder output and input for listenability, you compare through the entire encode/decode cycle.  Once we have an original copy,

the FA copy, then we can discuss frequency reponse errors on the calculated decoder ouput.   I don't have a good 'original' and FA copy combination

to begin with.  If I had a good example of those two items, from almost any recording, then the decoder can be made very accurate.  As is now,

comparing FA with the Decoder output isn't a lot different than comparing DolbyA input and output or DBX input and output.   I do agree that

the output of the FA decoder should sound SIMILAR to the FA RAW.   However, the FA RAW should also sound similar to the original

recording, but the encoding process is probably not flat also.


We need to compare pre-encoded copy with decoded copy, I really don't care about the FA copy for comparison about technical accuracy.   Once we have pre-encoded, FA generated from pre-encoded, the decoder source, and myself to fix it -- within hours, the decoded copy and the input copy will be essentially exactly the same.

 

Since we have no reference specs, schematics, reference recordings, test recordings, etc, we are stuck using reverse engineering, listening for response balances, and hearing 'tells'.  'Tells' really do work, but require alot of experience to hear and interpret.  This is lot like hearing dynamics distortions -- most people will assert that it doesn't sound correct.  Instead, I will be able to describe the problem.

 

I don't think that the static (I mean static) frequency response makes any sense on a mult-band dynamics processor with a lot of necessary EQ around it.

There are some situations where the frequency response should be flat, but claims that there must be a flat response has no supporting facts.

 

There are most likely instances where the decoder should be flat, and maybe it really should be,but unlikely after listening how FA signal sounds.   The

HF (esp over 9kHz) sounds smushed.

 

 

 

John, you keep saying static EQ doesn't make sense with dynamic processing, which is true. Nevertheless, the decoder produces a consistent static frequency change across all the processed tracks. This can't be the effect of dynamic processing, since that would change based on the signal and wouldn't be the same across two different recordings. For whatever reason, there's a static frequency curve that appears to be applied to all decoded content. That's what Klaus has been reporting, and that's what I've been saying to you for a while, even in our previous private conversations. Although the curve has changed over time and gotten better, it's still very distracting.

 

Now, if you tell me that DolbyA requires such a static EQ, I could understand it, but I'd like to see this documented somewhere -- the static correction that's applied now seems to unbalance overall frequency response.

Link to comment
11 minutes ago, John Dyson said:

In fact, if you do have the S&G album without FA

 

Which album are you guys talking about?

"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
10 minutes ago, pkane2001 said:

 

Yes, linear phase correction in DW if phase is not selected, so you're right,  if minimum phase filters were used this will not undo them fully. Engaging phase correction will, though, but I think this will just negate all the main effects of the decoder :)

 

I did listen to a couple of tracks, RAW, Decoded, and de-EQed by DW. I think the latest decoder does a better job with frequency balance, although to me, there's an overemphasis on upper-mid frequencies. Vocals sound brighter than they should, a bit more sibilance. Not something unexpected if you look at the frequency bump from 2 to 7kHz, right where the ear is most sensitive. This gets fixed with the de-EQ, but, like I said, I can hardly tell the difference between the RAW and Decoded with de-EQ.

Yea -- imagine the true distortion created  below 1kHz if they don't match.   It really DOES distort if the phase is screwed up...  It isn't just phase issues that happen, it is true distortion that happens in the encode/decode cycle.

 

Any EQ other than simple 1st order is JUST WRONG.  (That is, except in the <100Hz region, when used, the 2nd order Q values must reasonably match.)   I was warned about this, by a relatively well known recording pro, a long time ago.  Trying to fix something with a dynamics processor by simply adding EQ is a cr*pshoot.  Might help, but probably not.   Even if fixed at one frequency, it will drift all over the place.


The most effective activity is to find a few non-FA references and their FA equivalents.  I can usually tell within about 20seconds of play if something is FA or not.  Then,  I can determine where the EQ needs to go, it isn't likely just in one place, even if needed.  I have no references, no specs.   With references, without exaggerating -- the next day, the decoder WILL be as  accurate as possible.   Fixed curves are 'interesting', but probably wont' help make the decoder any more accurate.  Maybe they might help a little, and that is one reason why I am interested.

 

Because of all of the crazy dynamics processing, the location of portions of the EQ must be chosen in one of three places.   Part of the needed EQ might be pre-emphasis, part of it de-emphasis.  It is unlikely that there would be any changes between, because ANY changes will mess things up TERMINALLY.

 

Link to comment
18 minutes ago, pkane2001 said:

 

John, you keep saying static EQ doesn't make sense with dynamic processing, which is true. Nevertheless, the decoder produces a consistent static frequency change across all the processed tracks. This can't be the effect of dynamic processing, since that would change based on the signal and wouldn't be the same across two different recordings. For whatever reason, there's a static frequency curve that appears to be applied to all decoded content. That's what Klaus has been reporting, and that's what I've been saying to you for a while, even in our previous private conversations. Although the curve has changed over time and gotten better, it's still very distracting.

 

Now, if you tell me that DolbyA requires such a static EQ, I could understand it, but I'd like to see this documented somewhere -- the static correction that's applied now seems to unbalance overall frequency response.

If we are talking about that kind of error, lets start with the latest decoder.  Anything else really is a waste.

In fact, I just changed the Q of a filter, because the 'C' release was probably wrong.

 

I know that there were probably fixed errors, but a lot more needs to be 'fixed' when making those changes.  I can internally make drastic static changes,

and still show the same frequency response as being measured here.  It will then sound different, with the same static response.

 

Lets start with a decoder that actually sounds more similar first -- that would be the 'C' release, or I can make a 'D' release with the Q correction in 10minutes.   I DO want help, but lets start from something that is as good as it can be NOW, before the changes.

 

 

I am trying not to be cranky, but these really are complicated matters -- and I have been working with zero references.  If we can find a reference, most likely I can make the decoder work well within a few dB.   Adding EQ statically on an older version that doesn't sound at all like the RAW doesn't seem all that productive.   After I get the decoder accurate within a few dB, THEN lets do our science project.   The reference material will help use more than anything that we could do right now.

 

Sometimes I get a 'feel' there is something wrong and get frustrated -- I do NOT intend to dissuade, but please ask me for what is needed!!!  We can make this thing work well very quickly and simpler science projects if we have the correct original reference information.

 

 

 

 

Link to comment
5 minutes ago, John Dyson said:

If we are talking about that kind of error, lets start with the latest decoder.  Anything else really is a waste.

In fact, I just changed the Q of a filter, because the 'C' release was probably wrong.

 

 

 

I would be interested in providing feedback but I would rather do it using a track I know, perhaps something by Supertramp or Vega's Solitude Standing?

I was never happy with the decoded Suzanne Vega but it seems that your software has come a long way since.

"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
13 minutes ago, semente said:

 

I would be interested in providing feedback but I would rather do it using a track I know, perhaps something by Supertramp or Vega's Solitude Standing?

I was never happy with the decoded Suzanne Vega but it seems that your software has come a long way since.

I'd suggest some of the more decoder challenging recordings -- maybe 'Crime: School, Crime: dreamer, Crisis: Easy Does it?

 

School because of the kind of bass.

Dreamer because of the vocal mix

Easy Does it because of the sustained low level material, might be gating?

 

For Vega, I'll look around on Solutude standing, see if I have good (instead of garbge) copy.  Send at least one cut of that?

 

Do these sound like good choices?

 

(Because of my time allocatioins today, I'll make a 'D' release,publish it on the site, then do the decodes.)

 

I'll make them available privately.  (People involved -- tell me if you are interested....  once available, I'll make them available immediately.)

 

 

Link to comment

I should have them ready by +2Hrs.  Will announce after making public/private versions.

 

Also, I will leave the lowest level of phase descrambling enabled.   It really IS needed because the DA decoder is very clean/accurate no slop, and the phase scrambling on the recordings is terrible.  There might be minor freq response wobbles because of it, but they WILL be wobbles that are small enough to be unnoticeable.   The phase scrambling is horrendous.  The decoder can REALLY clean them up, but with the frequency response police active now -- humor intended, I'll have to follow more rules than usual :-).

 

I just double checked everything, making sure that I didn't do anything stupid in the source, and will start the process!!!

 

PS:  I do have a REALLY GOOD copy of Solitude Standing...  Didn't even realize that.

 

 

Link to comment

About those who help.

I am adding a contributors list to the decoder that will be displayed on a new 'license' switch.  Where else can I put it?  Let me  know.

 

No matter if we are 'best friends' or not --I am happy to add credits lines for those who substantively contribute, or really try.   I tried to start doing it

in the docs, but the goal will be to add it there also.  When the time comes up, we should talk privately about this.   I will not start the effort until maybe a week or so from now.

As long as the decoder is good, I'll be happy to do it.  If the decoder is not deemed good by the person who helps, I wont' add the byline if they don't want.


If most people think that the decoder is good, Alex will be in the list also.

 

All of this assumes that people feel that the decoder is working.  I don't want to put someone elses name on something that might be embarassing to them.

 

 

Link to comment

Snippets for Crime of The Century -- next will Be Crisis, then Solitude Standing.  Sorry that they must be snippets:

https://www.dropbox.com/sh/iazyjhbyksjs54e/AAChBL9dLKkY75-ksQLKvoe4a?dl=0

 

I need input on this about the bass.   When I reviewed how the bass should be, everything pointed to this.  It seems extreme, but also it has the indications

that it really was recorded that way.  Most other recordings are not this 'interesting' in this regard.

Anne Murray's stuff sounds REALLY GOOD with this current setting, and I am strongly bass adverse.  Somehow this setting sounds correct?!??!

 

Am I wrong?  Should the decoder be pulled back a little?   This has NOTHING to do with EQ gain levels, it is all about the Q value.

 

Link to comment

Snippets for Crisis, What Crisis?

Again, tell me about the bass -- I don't know...

 

https://www.dropbox.com/sh/lgccx2jdrj1dgtf/AACCQwLxlAyEZPE_1AHX8QWOa?dl=0

 

The Supertramp examples are kind of strange.  Is this what they are supposed to sound like?

This is a bit of a quandry for me.

They are definitely expanded, but the strong midrange and lows definitely change the character that I am used to hearing.  Have I been hearing wrong.

 

Everything else seems to work right with similar settings (The calibration level was a little odd, but would be normal for super high quality recordings.)

Solitude standing is coming soon, but it has the crazy bass again.

 

Link to comment
28 minutes ago, John Dyson said:

 

John, despite your hints about your perception of "bass", you might want to know that in the Crime album there plainly is no bass in the first 50 seconds. It all starts right after that (by heart, and knowing the tracks. A Supertramp feature, I suppose. Only just over half way on Crime of The Century (track) itself, there's supposed to be bass. But there is still virtually nothing; what is there is heavily underwhelmed.

 

You should never take Crime of the Century as a test case because the digital album is flowed for output level to begin with (and way too dynamic because of that - lacking compression if you will).

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

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...