Tech Insider					   Technology and Trends


			   USENET Archives

From root@ihack.net Mon, 1 Nov 1999 10:09:26 -0500 (EST)
Date: Mon, 1 Nov 1999 10:09:26 -0500 (EST)
From: Charles M. Hannum root@ihack.net
Subject: [Livid-dev] Observed decoding problems

I have observed the following difficulties so far:

1) Demuxing of multiple camera angles doesn't work.  If I select video
   stream 0, I get all the angles, in alternating blocks.  If I select
   another video stream, I don't get anything.  I haven't yet
   determined how to actually differentiate the data streams.  [Tested
   with `Sarah McLachlan: Mirrorball'.]

2) Also associated with the above, I get a `repeating' effect on the
   audio in the same chapters.  (It repeats each chunk 4 times.)

3) There seem to be some (possibly buggy) AC-3 streams that cause the
   decoder to barf with lots of `Invalid mantissa' and `Invalid
   exponent' messages.  [I'm forgetting now which disc I noticed this
   on.  Could possibly be related to getting out of sync due to
   problem #2.]

4) I've observed at least one disc to have a pair of
   Sequence_end_code/Sequence_start_code buried in the middle of the
   video stream.  Currently the Sequence_end_code causes the video
   decoder to just stop (even though the audio will keep playing).

5) At the beginning of `Strange Days', there's a very noticable
   `warbling' effect during the 20th Century Fox trailer.  I tried
   disabling the normalization hack in output_play() just to be sure,
   and I can vouch that it wasn't that code at fault.  (Note: This
   happens on both the English and French 2.0/ProLogic tracks, but not
   the English 5.1 track.)  I'll try this on a standalone player later
   today.

From shadow@dementia.org Tue, 2 Nov 1999 21:44:09 -0500 (EST)
Date: Tue, 2 Nov 1999 21:44:09 -0500 (EST)
From: Derrick J Brashear shadow@dementia.org
Subject: [Livid-dev] Observed decoding problems

On Mon, 1 Nov 1999, Charles M. Hannum wrote:

> 3) There seem to be some (possibly buggy) AC-3 streams that cause the
>    decoder to barf with lots of `Invalid mantissa' and `Invalid
>    exponent' messages.  [I'm forgetting now which disc I noticed this
>    on.  Could possibly be related to getting out of sync due to
>    problem #2.]

I noticed it on X Files; Decoding all or 0xbx instead of just 0xbc fixed
it. I should see if they were all just 0xbc or 0xbb

-D

From root@ihack.net 03 Nov 1999 14:03:43 -0500
Date: 03 Nov 1999 14:03:43 -0500
From: Charles M. Hannum root@ihack.net
Subject: [Livid-dev] Observed decoding problems

Derrick J Brashear <shadow@dementia.org> writes:

> On Mon, 1 Nov 1999, Charles M. Hannum wrote:
> 
> > 3) There seem to be some (possibly buggy) AC-3 streams that cause the
> > decoder to barf with lots of `Invalid mantissa' and `Invalid
> > exponent' messages. [I'm forgetting now which disc I noticed this
> > on. Could possibly be related to getting out of sync due to
> > problem #2.]
> 
> I noticed it on X Files; Decoding all or 0xbx instead of just 0xbc fixed
> it. I should see if they were all just 0xbc or 0xbb

I looked at the stream IDs, and there are no `strange' ones in either
disc (`Sarah McLachlan: Mirrorball', and `Who Framed Roger Rabbit?').

			   USENET Archives


The materials and information included in this website may only be used
for purposes such as criticism, review, private study, scholarship, or 
research.


Electronic mail:			      WorldWideWeb:
   tech-insider@outlook.com		         http://tech-insider.org/