|
Post by ROWUK on Oct 31, 2019 16:06:39 GMT 10
John, will you also be working on DBX decoding too?
|
|
sandyk (RIP Alex, 1939 - 2021)
Global Moderator
Posts: 226
About Me: Retired ex Principal Telecommunications Technical Officer with 43 years at Telstra (Australia)
I am also a Moderator in Hi Fi Critic Forum
Electronics hobbyist for >65 years with DIY projects including Loudspeakers, Stereo FM tuner, S/W Regen Receiver, Superhet AM ,
Synchrodyne PLL AM tuner (Phase Lock Loop),Stereo Tape Deck, Amplifiers including I.C. types, Class A, Class AB 100W/Ch. (ETI5000) 240W/Ch. Mosfet (AEM6000) ,several DACs , numerous PSUs including VERY low noise (<4uV) types etc.for myself and friends
Audio Industry Affiliation: NIL
|
Post by sandyk (RIP Alex, 1939 - 2021) on Oct 31, 2019 16:45:27 GMT 10
Good news -- I have to announce this somewhere because it IS a major breakthrough in the DHNRDS DA decoder.
This is a long background story that no-one wants to hear, but the bottom line, I ended up looking again at the DHNRDS DA decoder for distortion issues. After chasing down some rabbit holes and revisiting the design goals, I had realized that my original goals were 'bug for bug', but better than DolbyA HW. I believe that goal was wrong. I changed the 'goal' temporarily -- I believe it is now a permanent change. NOW -- the goal is 'Better than DolbyA HW... period.'. I stripped out some of the emulation code that supports some of the DolbyA HW bugs. For example, one of the bugs is that the signal attacks have a limited slew rate because the effective drive voltage is limited when running in expander mode. I stripped out the 'drive voltage' limitation in the code (of course, an emulation), and lo-and-behold, the garble that happens on complex/mixed large signals that happens with DolbyA units and also the previous DHNRDS versions is now pretty much GONE.
The previous DHNRDS DA version was pretty good at getting rid of the DolbyA HW decoding fog at low levels, but that doesn't happen nearly as often as the louder mixed chorus and mixed instruments. With this modification, the DHNRDS DA can now track the signal more accurately than previously and more accurately than the DolbyAHW -- thereby reducing the high level garble.
My initial evaluation of the improvement produces results almost as important as the initial versions that were often better better than DolbyA HW. There is a very significant improvement in sound quality.
I will check into a few matters regarding forum message distribution/etc, but if it is small enough, I'll provide a repository for examples in the next few days.
If this was not a major improvement -- I would NOT have bothered posting this message. This upcoming change will make a much more significant improvement in decoding results over previous DHNRDS versiona AND DolbyA versions. I wouldn't be surprised that a lot of material will be heard much more similarly to the original input to the DolbyA/Tape recorder complex.
Some very difficult material, including ABBA (the recordings that I use for my ACID tests), are sounding very errie.... Stuff like 'SuperTrouper' actually sounds almost natural?!?! Bread is really, really, really nice sounding, and the Carpenters recordings are coming out better than I have EVER heard.
Examples will be forthcoming -- next few days. Anyone who has REALLY EVIL DolbyA material that they'd like to resurrect -- I can supply a decoder, or even do a decoding operation gratis (I am very good about keeping private/unpublished/NDA material private, and happily destroy temporary copies.) The DHNRDS can make DolbyA material sound less foggy, have better vocal chorus (less rough mixing), lots of other improvements.
John
|
|
sandyk (RIP Alex, 1939 - 2021)
Global Moderator
Posts: 226
About Me: Retired ex Principal Telecommunications Technical Officer with 43 years at Telstra (Australia)
I am also a Moderator in Hi Fi Critic Forum
Electronics hobbyist for >65 years with DIY projects including Loudspeakers, Stereo FM tuner, S/W Regen Receiver, Superhet AM ,
Synchrodyne PLL AM tuner (Phase Lock Loop),Stereo Tape Deck, Amplifiers including I.C. types, Class A, Class AB 100W/Ch. (ETI5000) 240W/Ch. Mosfet (AEM6000) ,several DACs , numerous PSUs including VERY low noise (<4uV) types etc.for myself and friends
Audio Industry Affiliation: NIL
|
Post by sandyk (RIP Alex, 1939 - 2021) on Oct 31, 2019 17:02:33 GMT 10
Good news -- I have to announce this somewhere because it IS a major breakthrough in the DHNRDS DA decoder.
This is a long background story that no-one wants to hear, but the bottom line, I ended up looking again at the DHNRDS DA decoder for distortion issues. After chasing down some rabbit holes and revisiting the design goals, I had realized that my original goals were 'bug for bug', but better than DolbyA HW. I believe that goal was wrong. I changed the 'goal' temporarily -- I believe it is now a permanent change. NOW -- the goal is 'Better than DolbyA HW... period.'. I stripped out some of the emulation code that supports some of the DolbyA HW bugs. For example, one of the bugs is that the signal attacks have a limited slew rate because the effective drive voltage is limited when running in expander mode. I stripped out the 'drive voltage' limitation in the code (of course, an emulation), and lo-and-behold, the garble that happens on complex/mixed large signals that happens with DolbyA units and also the previous DHNRDS versions is now pretty much GONE.
The previous DHNRDS DA version was pretty good at getting rid of the DolbyA HW decoding fog at low levels, but that doesn't happen nearly as often as the louder mixed chorus and mixed instruments. With this modification, the DHNRDS DA can now track the signal more accurately than previously and more accurately than the DolbyAHW -- thereby reducing the high level garble.
My initial evaluation of the improvement produces results almost as important as the initial versions that were often better better than DolbyA HW. There is a very significant improvement in sound quality.
I will check into a few matters regarding forum message distribution/etc, but if it is small enough, I'll provide a repository for examples in the next few days.
If this was not a major improvement -- I would NOT have bothered posting this message. This upcoming change will make a much more significant improvement in decoding results over previous DHNRDS versiona AND DolbyA versions. I wouldn't be surprised that a lot of material will be heard much more similarly to the original input to the DolbyA/Tape recorder complex.
Some very difficult material, including ABBA (the recordings that I use for my ACID tests), are sounding very errie.... Stuff like 'SuperTrouper' actually sounds almost natural?!?! Bread is really, really, really nice sounding, and the Carpenters recordings are coming out better than I have EVER heard.
Examples will be forthcoming -- next few days. Anyone who has REALLY EVIL DolbyA material that they'd like to resurrect -- I can supply a decoder, or even do a decoding operation gratis (I am very good about keeping private/unpublished/NDA material private, and happily destroy temporary copies.) The DHNRDS can make DolbyA material sound less foggy, have better vocal chorus (less rough mixing), lots of other improvements.
John
I can vouch for what John is saying here, as I have been privileged to have heard many of John's corrections, and the numerous small bugs he has ironed out on the way due to poor documentation (if any) of what the Recording Engineers did to attempt to correct many of the Dolby A related problems. In fact, I believe that in many cases the improvements are so substantial that the Record companies would do well to release these recordings as they were meant to be heard, not just preserved for future generations. One example was with "Peggy Lee -Fever" which is so pure sounding and clean, that I couldn't help but to turn the volume way up !
The Record companies should be ashamed of the poor quality that they provided to consumers, as this track alone will even leave most High Definition releases (not those from Barry) way behind, and bear in mind too, how long ago this recording was made.
Alex
|
|