flac dev - Mar 2013 - [flac 1.3.0pre2 pre-release] Updates to test suite

Jaren Stangret wrote:
> Interestingly, I noticed compression level '0' is omitted, so
I've added it
> in.
Thanks.
> Also, I tested this on a 24bit/96kHz FLAC file and the test failed between
> compression level 0 and compression level 1 -- the file size was greater
> with a compression level of 1, rather than 0.  The test works on
> 'noisy-sine.wav'.  It may be worth trying to run this test on a
variety of
> FLAC files (multiple channels, multiple bps, multiple sample rates, etc).
Compression levels are always highly dependent on the source material. This
particular test was added to test for the most gross kind of regressions
like the one I fixed in fc360735ce4d1aa88a94bfccdd3bea5bdd19a8d6.

When and if we come to update/improve FLAC's core compression code, that
is when we would need a more comprehensive test for compressed file sizes.

At this point, I think what we have is sufficient.

Cheers,
Erik
-- 
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/

Attached is a patch for 'test_grabbag.sh'.  Continuing in the same vein
as
the other patches, this patch consists of minor reworks of
functions/commands as well as heavily commenting what's going on.

More than halfway there!

Thanks,
Jaren

On Tue, Mar 12, 2013 at 9:40 PM, Erik de Castro Lopo <mle+la at
mega-nerd.com>
wrote:
>
> Jaren Stangret wrote:
>
> > Interestingly, I noticed compression level '0' is omitted, so
I've
added it
> > in.
>
> Thanks.
>
> > Also, I tested this on a 24bit/96kHz FLAC file and the test failed
between
> > compression level 0 and compression level 1 -- the file size was
greater
> > with a compression level of 1, rather than 0.  The test works on
> > 'noisy-sine.wav'.  It may be worth trying to run this test on
a variety
of
> > FLAC files (multiple channels, multiple bps, multiple sample rates,
etc).
>
> Compression levels are always highly dependent on the source material.
This
> particular test was added to test for the most gross kind of regressions
> like the one I fixed in fc360735ce4d1aa88a94bfccdd3bea5bdd19a8d6.
>
> When and if we come to update/improve FLAC's core compression code,
that
> is when we would need a more comprehensive test for compressed file sizes.
>
> At this point, I think what we have is sufficient.
>
> Cheers,
> Erik
> --
> ----------------------------------------------------------------------
> Erik de Castro Lopo
> http://www.mega-nerd.com/
> _______________________________________________
> flac-dev mailing list
> flac-dev at xiph.org
> http://lists.xiph.org/mailman/listinfo/flac-dev
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
http://lists.xiph.org/pipermail/flac-dev/attachments/20130313/9c6695e2/attachment-0001.htm
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test_grabbag.sh.patch
Type: application/octet-stream
Size: 10944 bytes
Desc: not available
Url :
http://lists.xiph.org/pipermail/flac-dev/attachments/20130313/9c6695e2/attachment-0001.obj

Jaren Stangret wrote:
> Attached is a patch for 'test_grabbag.sh'.  Continuing in the same
vein as
> the other patches, this patch consists of minor reworks of
> functions/commands as well as heavily commenting what's going on.
> 
> More than halfway there!
Jaren,

Any sign of the patches for the rest of these tests? I have some other
changes to the tests that I'd like to make and don't want my patches
to clash with yours.

Cheers,
Erik
-- 
----------------------------------------------------------------------
Erik de Castro Lopo
http://www.mega-nerd.com/