Hi folks,
I just wanted to provide an update on the ALAC decoder, in case anyone
was wondering.
With some great help from some other guy at doing some brute force
simplification of part of the code, I've now got to the stage where I
know almost entirely what's going on. And I've almost finished writing
a basic decoder from scratch. (It works for most mono files at the
moment).
For those that are interested, it isn't too dissimilar to FLAC.
Although they use an adaptive FIR algorithm, with the predictor
coefficients changing from sample to sample based on error values /
pervious samples. They encode the prediction error using a highly
modified Rice algorithm.
With this in mind, it should be rather trivial to write an encoder.
Perhaps borrowing bits from FLAC's. The hardest part will be adapting
the prediction stuff, I believe adapting Rice will be pretty easy.
Hopefully I'll feel good enough to make a release by the weeks end.
I've learnt a lot about lossless audio compression these last few weeks
 .
David.
On 17-Feb-05, at 5:23 PM, David Hammerton wrote:
> Hi there,
>
> I'm emailing this to both the ffmpeg-devel list and Electronic
> Frontiers Australia, in the hope that someone will be able to answer
> my query.
>
> A while ago Apple introduced a new audio codec dubbed 'Apple Lossless
> Audio Codec,' often referred to as 'ALAC'. This is a proprietary
> format, Apple is yet to release any specifications about the format.
> Up to now, I'm unaware of anyone reverse engineering the format or
> releasing a non-Apple decoder.
>
> A little while ago, I managed to reverse engineer the relevant parts
> of the Quicktime DLLs (Quicktime being Apple's Media library). I have
> a almost fully working ALAC decoder. I have managed to entirely decode
> in to the original form every single audio file that I've been able to
> throw at it. It works perfectly for all intensive purposes.
>
> However, I'm unsure of my legal position if I were to release it as
> is. Currently it is _very similar_ to Apple's code. Of course I've
> never seen their source code, but it's similar to their binary code.
> My implementation is written in medium-level C.
>
> I originally planned to, after getting to this stage, try and figure
> out the spec and write a mostly-clean-room implementation of the
> codec. However, I'm about to go back to University and am (frankly)
> bored of it! I just want it out there! I've started a cleaner
> implementation, as the next step to figuring out a spec, but yeah,
> just lost interest really.
>
> So - what do I do?
>
> If I release the driver as it is, am I breaching Apple's copyright,
> since it is effectively their code?
>
> I'm also happy to work with another developer to do a proper clean
> room thing, if anyone wants to.
>
> Thanks guys
>
> Regards
>
> David.
>
> PS - you can catch me on IRC, freenode on #lhl, or phone me if you
> like, details in my whois. I'm in Melbourne Australia, so don't
> message me at 3am expecting me to reply! I sleep like a normal person
> :).
>
> --
> David Hammerton
> david <at> craz.net
> http://crazney.net/
> Welfare *NOT* Warfare
>
>
>
> -------------------------------------------------------
> SF email is sponsored by - The IT Product Guide
> Read honest & candid reviews on hundreds of IT Products from real
> users.
> Discover which products truly live up to the hype. Start reading now.
> http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
> _______________________________________________
> Ffmpeg-devel mailing list
> Ffmpeg-devel <at> lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/ffmpeg-devel
>
>
--
David Hammerton
david <at> craz.net
http://crazney.net/
Welfare *NOT* Warfare
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
| 