Hi Richard and Matt and all,
There are still lurking incompatibilities with various
versions of saolc out there--I didn't realize that anyone
was actually ready to use the bitstream features (which
are needed for sdif2mp4). Needless to say, I'm pleased
to be wrong!
I'll try to take some time on Monday and make an official
release of the final saolc reference software, which has
differences in the bitstream-handling code from the
last release of saolc (ie, it actually matches the standard).
I believe that the code that Matt and I wrote for sdif2mp4
is compatible with that version.
I haven't been tracking changes between the SDIF libraries
and sdif2mp4, I suggest that a code merge to create a single
code point between me (for the new encoder routines that
match the final release of saolc), Matt (for the new SDIF
libraries), and you (for your own improvements and
simplifications) is in order. Once that is done, other
continued development can proceed in lockstep.
I'm really excited that you're working with sdif2mp4--it's
a really great idea (it was Matt's), and opens up a lot
of interesting possibilities for doing compositional and
analysis/synthesis things in SAOL. Do you have specific
compositional plans yourself, or are you just assembling
the tools for the CDP?
At some point, I'd like to do the same sort of utility for
Jim Beauchamp's ADSYN library (unless that already can write
SDIF?).
Best,
-- Eric
-----Original Message-----
From: Richard Dobson <rwd@cableinet.co.uk>
To: Matt Wright <matt@cnmat.CNMAT.Berkeley.EDU>
Cc: saol-dev@media.mit.edu <saol-dev@media.mit.edu>
Date: Saturday, February 19, 2000 8:28 AM
Subject: Re: updated sdif2mp4 files
>Ah, right; but there is no idication on the SDIF download site that
>sdif2mp4 had been updated.
>
>I have downloaded the new code set, and, after some changes (had to
>remove "malloc_alloc" from the <vector> declarations), got it to build,
>but the generated mp4 file seems to be corrupted - saolc rejects it with
>copious error messages, and crashes. On dumping the orc, most numeric
>values, including the srate, are garbage, as if not written, or back to
>front, or something.
>
>My version generates a mp4 file that is completely accepted by saolc (my
>build), and synthesises to a healthy soundfile. sfront55 (downloaded
>build) reports similar numeric errors, but it looks like it only has a
>problem with floating-point fields - it reads srate and other integer
>fields correctly. I don't have the time to investigate this myself, and
>I am not a bitstream adept anyway!
>
>I see that the bitstream files have had several significant changes from
>the version I am using in saolc and in sdif2mp4. I have not made any
>changes to those files. These changes imply that saolc itself has been
>updated, but there has been no mention of this on the saol lists, and
>there is also no indication in bistream.cpp itself. If this is an
>'official' update which is supposed to replace the one in saolc, it
>really needs to be documented as such in the source file. As I do have a
>working combination of sdfi2mp4 and saolc, I don't want to make any
>changes that may break that; though it would be nice to be abe to run
>sfront on the mp4 files too.
>
>I haven't had an answer to my original question, but I will asssume I
>should not publish my version of sdif2mp4 until advised otherwise.
>
>
>
>Richard Dobson
>
>Matt Wright wrote:
>>
>> We updates sdif2mp4, and the current version should be what you
>> see on the CNMAT SDIF web page. I'll have time to look at your
>> version next week; perhaps a merge is in order.
>>
>> -Matt
>
>--
>Test your DAW with my Soundcard Attrition Page!
>http://wkweb5.cableinet.co.uk/rwd (LU: 17th September 1999)
>CDP: http://www.bath.ac.uk/~masjpf/CDP/CDP.htm (LU: 4th November 1999)
>
This archive was generated by hypermail 2b29 : Mon Jan 28 2002 - 12:03:52 EST