I have uploaded some decoded demos -- not tuned perfectly with 100% proper DolbyA levels this time, but is at least an approx example. MP3 does NOT do the decoder justice at all. I don't remember the provenance of the carpenters' demo, but I think that it was HDtracks. This kind of demonstrates that even the premium suppliers can leak DolbyA material!!!
here is the location:
https://spaces.hightail.com/space/z3H68lAgmJ
John
FYI -- and I know that this is slightly off topic, but is indirectly related to the potential 'snake-oil' issue regarding the whole audiophile excess (probably, but just slightly maybe not -- don't really want to hurt ANYONES feelings!!!)
Most importantly -- the source materials for my examples are commonly available, and I found the advantage of using the decoder by accident a year or so ago. That 'accident' of guessing that the 'harsh', excess HF 'digital' sound was that I thought that it JUST MIGHT be DolbyA encoding. I found my guess to be correct perhaps 1/4 to 1/2 of the time. There is one major problem in determining if something is DolbyA encoded or not -- sometimes ones' "ear" and the decoder both can be fooled. Sometimes material can be guessed to be DolbyA encoded and it really isnt. Sometimes the 'harsh sound' problme is really just bad EQ -- but just as often it really is DolbyA encoding. Another frustrating factor -- sometimes DolbyA encoding sounds so bad, that they EQ the DolbyA encoded master -- and therefore the sound is JUST A LITTLE better, but makes it more tricky to properly decode that material -- not only the threshold needs to be determined, but the errant EQ needs to be backed out for optimal results.
The two sets of items on my demo that I believe MIGHT not be DolbyA encoded (but the decoding sounds good) are the Nat King Cole example, and the Dionne Warwick examples. I am more sanquine about the Warwick being DolbyA than the Nat King Cole. It might help me if someone who REALLY knows the material help me determine if the decoded versions that I posted do sound correct!!! Thank you if you can help me with this...
I just uploaded slightly more accurate decoding examples. To me, the most interesting and surprising (even though it still doesn't sound perfect -- I think that it is good as possible, even with a real DolbyA) -- is the Scarborough Fair example. If you listen to the original -- there is lots of LF (yes, LOW frequency) distortion that is cancelled out by the decoder. The LF distortion occurs because of the fast attack/fast decay even at low frequencies (2msec attack/62msec decay.) The 62msec is too fast for a normal compressor at the lowest audio frequencies -- will definitely cause distortion. However, the expander (decoder) fits the curve closely enough to mitigate a lot of the distortion.
BTW -- off direct topic, but talking very technical -- the DolbyA compression/expansion is pure linear WRT the audio signal -- basically a full wave detector with some fancy tricks by the genius Ray Dolby. If the detection would have been RMS, the distortion would have been a lot less (squaring the signal instead of taking the raw signal) has two advantages -- first the resulting gain control signal has less ripple, and the attack/decay time constant would be used with the signal squared -- which slows down the gain control action. So, RMS would 'sound' better without decoding, but wouldn't work as good for hiding gain modulation. Any claims about RMS being better because of 'phase' isn't quite as operative as one might think -- because the bandwidths have been narrowed by the input filters, so the phase advantage is still extistent to an extent, but not as important as if the processing was single band (the whole audio signal.)
Again, the examples are on the site, and will probably disappear within a few days.
Sincerely,
John