sox-users@lists.sourceforge.net unofficial mirror
 help / color / mirror / code / Atom feed
From: John Dyson via Sox-users <sox-users@lists.sourceforge.net>
To: "sox-users@lists.sourceforge.net" <sox-users@lists.sourceforge.net>
Cc: John Dyson <johndyson10@yahoo.com>
Subject: I have developed a bona-fide DolbyA decoder -- wanna know if there might be interest in a SOX plugin.
Date: Mon, 30 Apr 2018 02:56:01 +0000 (UTC)	[thread overview]
Message-ID: <1693181382.2280435.1525056961897@mail.yahoo.com> (raw)
In-Reply-To: 1693181382.2280435.1525056961897.ref@mail.yahoo.com


[-- Attachment #1.1: Type: text/plain, Size: 4301 bytes --]

Gang:
Before reading anything else -- the intent here is to make a SOX plugin version of the DolbyA decoder available.
The DolbyA decoder sounds almost EXACTLY the same as a real DolbyA unit when doing decoding -- been working with a recording expert and his 'golden ears' person (along with my DSP/EE/software expertise.)  Currently the program is .wav in and .wav out, fp/signed int, 44.1k -> 192k.  It does the processing at the input/output data rate.  The processing is pretty darned CPU heavy, but can run 3X faster than realtime at 96k on the single core of a Haswell at 3GHz.The complexity is MUCH greater than one might initially surmise, but also is not a behmoth.

Here is the motivation:  lots of material is stealth DolbyA encoded -- refer to the examples on the site below.  The source materials are 'standard' CDs, some stuff from HDtracks and other places.  Much of the 'digital sheen' that is so common nowadays comes from lack of proper DolbyA decoding.  Also, OFF TOPIC, we are considering this for pro applications and will be doing a VST3 plugin (at least we are expecting to start the porting effort soon.) 

In the examples, the files with -raw in the name are direct rips from the source.  The files with -DA in the name are either just DA decoded or have some additional processing as an example of 'finalization'.  Please refer to the HISS and HF sheen in many of the examples, and the rather 'flat' sounding of the results.  After removing the untoward DolbyA compressed sound, then a better quality (for listening) compression/EQ scheme can be applied.  The DolbyA compression is meant to be quick so as to avoid modulation type noise, and has a primitive/odd intermodulation mitigate scheme (which this decoder has a greatly advanced version.)  Trying to 'DolbyA' decode material that hasn't been encoded results in various failure syndromes most of the time -- the only way that one can partially get by with decoding material that is not DolbyA encoded is when a certain kind of fast sidechain compression is used, and there is really little reason to use the kind of compression that DolbyA uses except for noise reduction applications.

There will be an upcoming pro version, but I am wanting to gift this to the community.  Here are the problems:  I don't want to give away the source code -- there is only one other DA decoder that I know of (our tests show that it doesn't work quite as well), and this development required some tricky DSP work and reverse engineering work FAR beyond anything that I was initially planning.  For example, I had to unfold a feedback compressor scheme (that implements an expander) and then make sure that the timeconstants and gain curves were identical.  The conversion is not as intiutive as one might expect -- also I want to be able to fund a future DolbySR decoder effort.  DolbySR makes DolbyA look trivial.  Might give it away some day, but not yet.
I believe that once people (I mean high end hobbyists and pros) find out how much an improvement that the decoder can make, that it will be very desirable.
I don't really know the required SOX licensing, but I could probably make the source of a weaker version available -- there is a LOT of intermod mitigation and filters that avoid aliasing due to the nonlinear operations/etc -- but could possibly remove some of the professional level quality, but really don't want to do so.  I'll probably eventually release the source in its entirety.
So, after reviewing all of the info, do you think 1) is this worthwhile to the userbase? 2) Maybe an partial object distribution might be possible, with a shim of some kind (temporarily) untiil I can get the funding for the SR effort by other means.  I do intend EVENTUALLY ot free the software, it actually has some educational stuff in it.
Please refer to my examples before a quick judge about 'isn't any DolbyA material available.'  1yr ago (even several months ago), I didn't believe it either.

What do you think?
John













 Hightail Spaces

  
|  
|   
|   
|   |    |

   |

  |
|  
|    |  
Hightail Spaces
 Hightail Spaces lets you send unlimited files, of any size, to anyone. Collaborate with your team and evolve you...  |   |

  |

  |

 

[-- Attachment #1.2: Type: text/html, Size: 12356 bytes --]

[-- Attachment #2: Type: text/plain, Size: 202 bytes --]

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot

[-- Attachment #3: Type: text/plain, Size: 158 bytes --]

_______________________________________________
Sox-users mailing list
Sox-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sox-users

           reply	other threads:[~2018-05-02 16:03 UTC|newest]

Thread overview: expand[flat|nested]  mbox.gz  Atom feed
 [parent not found: <1693181382.2280435.1525056961897.ref@mail.yahoo.com>]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-list from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://lists.sourceforge.net/lists/listinfo/sox-users

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1693181382.2280435.1525056961897@mail.yahoo.com \
    --to=sox-users@lists.sourceforge.net \
    --cc=johndyson10@yahoo.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
Code repositories for project(s) associated with this public inbox

	https://80x24.org/mirrors/sox.git

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).