videolan
/
ffmpeg
mirror of https://git.videolan.org/git/ffmpeg.git
1
Fork 0
Code Releases Activity
ffmpeg/tests/ffmpeg.regression.ref

144 lines
5.9 KiB
Plaintext
Raw Normal View History

added first version of regression tests - simply type 'make test' to test every codec and some mux/demuxes - added simple synthetic audio and video bitstreams generators so that no big streams needs to be used Originally committed as revision 532 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-05-19 01:18:22 +02:00
ffmpeg regression test
change qscale -> lambda for the motion estimation finetune bit/distortion weighting factor used in motion estimation, the old coeffs where finetuned relative to incorrect mv_penalty tables which where then fixed later but the coeffs where not this _may_ fix the long standing blocking artifacts, but may also introduce mudding artefacts theoretically, so please tell us if u stumble across any so we can either fix them or export this variable so the user can change it Originally committed as revision 3189 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-06-03 21:29:47 +02:00
511897938cd4a9dd5d0ab38b4658be18 *./data/a-mpeg1.mpg
734274 ./data/a-mpeg1.mpg
eff964f940a35427fa9e8d35fb4f5699 *./data/out.yuv
1000000l (using uninitalized variable for initalizing bits per MV table) Originally committed as revision 2900 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-03-16 17:57:27 +01:00
stddev: 7.66 PSNR:30.43 bytes:7602176
change qscale -> lambda for the motion estimation finetune bit/distortion weighting factor used in motion estimation, the old coeffs where finetuned relative to incorrect mv_penalty tables which where then fixed later but the coeffs where not this _may_ fix the long standing blocking artifacts, but may also introduce mudding artefacts theoretically, so please tell us if u stumble across any so we can either fix them or export this variable so the user can change it Originally committed as revision 3189 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-06-03 21:29:47 +02:00
5e1e8136cf2565f7cfbcfc7eda8d8f2b *./data/a-mpeg2.mpg
750226 ./data/a-mpeg2.mpg
27ddfd3e44537661eb14b96129a055ef *./data/out.yuv
stddev: 7.69 PSNR:30.40 bytes:7602176
004245d17ae9322ea3c6dc4b0f0554d6 *./data/a-mpeg2.mpg
746381 ./data/a-mpeg2.mpg
c7b9a71245912671889849ed560b67d3 *./data/out.yuv
stddev: 7.69 PSNR:30.40 bytes:7602176
2a77ebec455e17c4ab897044c9b3040a *./data/a-mpeg2i.mpg
754366 ./data/a-mpeg2i.mpg
ddf889858529d09c8fe31c701deacd16 *./data/out.yuv
stddev: 7.71 PSNR:30.37 bytes:7602176
10l (fixes delayed frames during encoding) Originally committed as revision 3279 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-07-01 01:41:29 +02:00
d3d2fd4dd186d8c9de7666398d5af13a *./data/a-mpeg2thread.mpg
816506 ./data/a-mpeg2thread.mpg
9db3aca554bb8ee337a109a3b8629575 *./data/out.yuv
stddev: 7.64 PSNR:30.45 bytes:7602176
mv/mb reuse regression test Originally committed as revision 3686 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-11-16 15:02:41 +01:00
052bcdb13286f4f7137a50c4db032104 *./data/a-mpeg2reuse.mpg
2115703 ./data/a-mpeg2reuse.mpg
526e6e6818296a1b63d91fc165e0bda8 *./data/out.yuv
stddev: 7.68 PSNR:30.41 bytes:7602176
change qscale -> lambda for the motion estimation finetune bit/distortion weighting factor used in motion estimation, the old coeffs where finetuned relative to incorrect mv_penalty tables which where then fixed later but the coeffs where not this _may_ fix the long standing blocking artifacts, but may also introduce mudding artefacts theoretically, so please tell us if u stumble across any so we can either fix them or export this variable so the user can change it Originally committed as revision 3189 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-06-03 21:29:47 +02:00
e7ff3a8780ebdc6687352646c8fbe876 *./data/a-msmpeg4v2.avi
642666 ./data/a-msmpeg4v2.avi
7646e5cc93452b5fb380c8c4c5f535f4 *./data/out.yuv
stddev: 8.02 PSNR:30.03 bytes:7602176
acd696f8967a996b3fbcfac94dd0d889 *./data/a-msmpeg4.avi
644964 ./data/a-msmpeg4.avi
cdc5d7403a16dc9450a2dbacb349dc70 *./data/out.yuv
stddev: 8.01 PSNR:30.04 bytes:7602176
82ca511f87695a1ee24c1681fed75bfa *./data/a-wmv1.avi
648074 ./data/a-wmv1.avi
09253222ab4eb95628c931a86006a2b1 *./data/out.yuv
stddev: 8.02 PSNR:30.04 bytes:7602176
a5bd577163968edab00058f2c8d5efab *./data/a-wmv2.avi
682132 ./data/a-wmv2.avi
09253222ab4eb95628c931a86006a2b1 *./data/out.yuv
stddev: 8.02 PSNR:30.04 bytes:7602176
H261 fixing and cleaning: -corrected wrong value in mv data -set correct mb_type after adjusting index -don't use H263 loop filter when the loop filter flag is set but when using the H261 encoder -use the same unquantizer as H263 (which is optimized btw) -removed unused members in H261Context patch by (Maarten Daniels <maarten.daniels >at< luc >dot< ac >dot< be>) regression test checksum update by me Originally committed as revision 3669 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-11-12 02:21:34 +01:00
bcf34887269746ac7973f88cde336609 *./data/a-h261.avi
735098 ./data/a-h261.avi
0a6e6dd4f09df9fe77ff29581c1a39c3 *./data/out.yuv
stddev: 9.14 PSNR:28.90 bytes:7602176
change qscale -> lambda for the motion estimation finetune bit/distortion weighting factor used in motion estimation, the old coeffs where finetuned relative to incorrect mv_penalty tables which where then fixed later but the coeffs where not this _may_ fix the long standing blocking artifacts, but may also introduce mudding artefacts theoretically, so please tell us if u stumble across any so we can either fix them or export this variable so the user can change it Originally committed as revision 3189 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-06-03 21:29:47 +02:00
fa556e599181bf9328a811a1ce9aa022 *./data/a-h263.avi
682226 ./data/a-h263.avi
f2b7fcff9de17f5aecfeb1090fe1963b *./data/out.yuv
stddev: 8.08 PSNR:29.97 bytes:7602176
fb27fbe102cae4d8b7c47955db87b11d *./data/a-h263p.avi
2414342 ./data/a-h263p.avi
28fd12ac0b168252d81df6f6e60a5d17 *./data/out.yuv
stddev: 2.07 PSNR:41.76 bytes:7602176
9153a152c1937cfb453b0388a1621a05 *./data/a-odivx.mp4
554440 ./data/a-odivx.mp4
e400c9175dd1811cdeee7be5555d33f1 *./data/out.yuv
stddev: 7.99 PSNR:30.06 bytes:7602176
split ffhuffyuv into 2 codecs: "huffyuv" is compatible with the official version "ffvhuff" contains our improvements Originally committed as revision 3720 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-11-28 19:29:38 +01:00
b415b4d46c0c693d973dd45b1898de52 *./data/a-huffyuv.avi
7932708 ./data/a-huffyuv.avi
wmv2 & huffyuv regression test Originally committed as revision 1650 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-03-08 12:50:26 +01:00
799d3db687f6cdd7a837ec156efc171f *./data/out.yuv
fixpoint log() for tiny_psnr so it can output actual PSNR Originally committed as revision 2640 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-12-31 03:32:54 +01:00
stddev: 0.00 PSNR:99.99 bytes:7602176
Always write VOL control parameters, needed by old xvid patch by (Loren Merritt <lorenm at u dot washington dot edu>) Originally committed as revision 3432 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-09-03 04:32:39 +02:00
788092f090eb1b11b78369f83e430fba *./data/a-mpeg4-rc.avi
820964 ./data/a-mpeg4-rc.avi
10l (fixes delayed frames during encoding) Originally committed as revision 3279 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-07-01 01:41:29 +02:00
8458b4ddaca643a3ec1264c1ee2f4986 *./data/out.yuv
stddev: 10.41 PSNR:27.77 bytes:7602176
7aab4613f634d23b6c66eec7270baaff *./data/a-mpeg4-adv.avi
change qscale -> lambda for the motion estimation finetune bit/distortion weighting factor used in motion estimation, the old coeffs where finetuned relative to incorrect mv_penalty tables which where then fixed later but the coeffs where not this _may_ fix the long standing blocking artifacts, but may also introduce mudding artefacts theoretically, so please tell us if u stumble across any so we can either fix them or export this variable so the user can change it Originally committed as revision 3189 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-06-03 21:29:47 +02:00
600240 ./data/a-mpeg4-adv.avi
0b7935cee2070f3b838478cea6dc4b3c *./data/out.yuv
stddev: 10.25 PSNR:27.90 bytes:7602176
Always write VOL control parameters, needed by old xvid patch by (Loren Merritt <lorenm at u dot washington dot edu>) Originally committed as revision 3432 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-09-03 04:32:39 +02:00
0000cc8c66f88d0e25ecdd63e85e457a *./data/a-mpeg4-thread.avi
10l (fixes delayed frames during encoding) Originally committed as revision 3279 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-07-01 01:41:29 +02:00
756550 ./data/a-mpeg4-thread.avi
7e4e2cf2659aa158f5ffed2be45a9c3d *./data/out.yuv
stddev: 12.31 PSNR:26.31 bytes:7602176
Always write VOL control parameters, needed by old xvid patch by (Loren Merritt <lorenm at u dot washington dot edu>) Originally committed as revision 3432 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-09-03 04:32:39 +02:00
545245d3169814f03dad599cac4b9ada *./data/a-mpeg4-Q.avi
899514 ./data/a-mpeg4-Q.avi
10l (fixes delayed frames during encoding) Originally committed as revision 3279 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-07-01 01:41:29 +02:00
cdb77818e108eab89c9ac4bcaf350e19 *./data/out.yuv
stddev: 5.69 PSNR:33.00 bytes:7602176
f28aea7181f680687ea9522f0066d18a *./data/a-error-mpeg4-adv.avi
change qscale -> lambda for the motion estimation finetune bit/distortion weighting factor used in motion estimation, the old coeffs where finetuned relative to incorrect mv_penalty tables which where then fixed later but the coeffs where not this _may_ fix the long standing blocking artifacts, but may also introduce mudding artefacts theoretically, so please tell us if u stumble across any so we can either fix them or export this variable so the user can change it Originally committed as revision 3189 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-06-03 21:29:47 +02:00
750788 ./data/a-error-mpeg4-adv.avi
83c96db61822c4ee2b9d46f1055264c5 *./data/out.yuv
stddev: 17.22 PSNR:23.40 bytes:7602176
10l (fixes delayed frames during encoding) Originally committed as revision 3279 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-07-01 01:41:29 +02:00
cc818f4e46e0ea2f93d61e0c7f83ebed *./data/a-mpeg4-nr.avi
change qscale -> lambda for the motion estimation finetune bit/distortion weighting factor used in motion estimation, the old coeffs where finetuned relative to incorrect mv_penalty tables which where then fixed later but the coeffs where not this _may_ fix the long standing blocking artifacts, but may also introduce mudding artefacts theoretically, so please tell us if u stumble across any so we can either fix them or export this variable so the user can change it Originally committed as revision 3189 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-06-03 21:29:47 +02:00
689772 ./data/a-mpeg4-nr.avi
79f4e8263e40822c0cbf35725f375b24 *./data/out.yuv
stddev: 7.03 PSNR:31.18 bytes:7602176
10l (fixes delayed frames during encoding) Originally committed as revision 3279 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-07-01 01:41:29 +02:00
6aebee12c0ff95756e11d32f305d7e11 *./data/a-mpeg1b.mpg
1037657 ./data/a-mpeg1b.mpg
09a77de112e144c562bc1398b111b61b *./data/out.yuv
stddev: 6.35 PSNR:32.06 bytes:7602176
jpeg style yuv fixes Originally committed as revision 3852 to svn://svn.ffmpeg.org/ffmpeg/trunk
2005-01-20 01:30:20 +01:00
8e69394fdcb19995dfe5cb8c562d9441 *./data/a-mjpeg.avi
1566544 ./data/a-mjpeg.avi
18c3a76f984e717dd886d21fa04355f6 *./data/out.yuv
stddev: 7.93 PSNR:30.13 bytes:7602176
a534c6fc190682842efc0defda2cd442 *./data/a-ljpeg.avi
6263462 ./data/a-ljpeg.avi
lossless jpeg regression test Originally committed as revision 2036 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-07-12 18:26:08 +02:00
799d3db687f6cdd7a837ec156efc171f *./data/out.yuv
fixpoint log() for tiny_psnr so it can output actual PSNR Originally committed as revision 2640 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-12-31 03:32:54 +01:00
stddev: 0.00 PSNR:99.99 bytes:7602176
rm encoding fix Originally committed as revision 3736 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-12-06 22:38:05 +01:00
49c87af74432890cadb28b93da1c653d *./data/a-rv10.rm
change qscale -> lambda for the motion estimation finetune bit/distortion weighting factor used in motion estimation, the old coeffs where finetuned relative to incorrect mv_penalty tables which where then fixed later but the coeffs where not this _may_ fix the long standing blocking artifacts, but may also introduce mudding artefacts theoretically, so please tell us if u stumble across any so we can either fix them or export this variable so the user can change it Originally committed as revision 3189 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-06-03 21:29:47 +02:00
677483 ./data/a-rv10.rm
f2b7fcff9de17f5aecfeb1090fe1963b *./data/out.yuv
stddev: 8.08 PSNR:29.97 bytes:7602176
rv20 regression test Originally committed as revision 3740 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-12-07 14:48:25 +01:00
a00fbb82c030fe41febe1d03eacb338e *./data/a-rv20.rm
652897 ./data/a-rv20.rm
203e9790cb25b1d308fd602275afbb0e *./data/out.yuv
stddev: 8.27 PSNR:29.76 bytes:7602176
asv1 regression test Originally committed as revision 1912 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-05-28 00:43:34 +02:00
89ad858355f9b0a92d024c5ba6e369dd *./data/a-asv1.avi
add size of encoded stream to the regression tests Originally committed as revision 2918 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-03-23 04:05:51 +01:00
2292416 ./data/a-asv1.avi
asv1 regression test Originally committed as revision 1912 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-05-28 00:43:34 +02:00
b3e1696d73071c8f7f0ff4e10c1677ca *./data/out.yuv
fixpoint log() for tiny_psnr so it can output actual PSNR Originally committed as revision 2640 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-12-31 03:32:54 +01:00
stddev: 19.19 PSNR:22.46 bytes:7602176
asv2 regression test Originally committed as revision 2188 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-09-01 15:05:21 +02:00
3909db39c3d2e9a72a952254e8685ff7 *./data/a-asv2.avi
add size of encoded stream to the regression tests Originally committed as revision 2918 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-03-23 04:05:51 +01:00
2221012 ./data/a-asv2.avi
asv2 regression test Originally committed as revision 2188 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-09-01 15:05:21 +02:00
551b263bfb2acb9ea6dd2000f184494f *./data/out.yuv
fixpoint log() for tiny_psnr so it can output actual PSNR Originally committed as revision 2640 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-12-31 03:32:54 +01:00
stddev: 17.98 PSNR:23.02 bytes:7602176
change qscale -> lambda for the motion estimation finetune bit/distortion weighting factor used in motion estimation, the old coeffs where finetuned relative to incorrect mv_penalty tables which where then fixed later but the coeffs where not this _may_ fix the long standing blocking artifacts, but may also introduce mudding artefacts theoretically, so please tell us if u stumble across any so we can either fix them or export this variable so the user can change it Originally committed as revision 3189 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-06-03 21:29:47 +02:00
de42cb5a15bfa8e70cf60edda0b32355 *./data/a-flv.flv
657991 ./data/a-flv.flv
4d9c04e7017dc68da7891048c46c1c37 *./data/out.yuv
stddev: 8.08 PSNR:29.97 bytes:7602176
switch ffv1 to the put/get_symbol() from snow Originally committed as revision 3659 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-10-31 13:22:00 +01:00
255e21b18af133a5196551aac321b5eb *./data/a-ffv1.avi
switch ffv1 from CABAC to the range coder Originally committed as revision 3658 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-10-31 11:08:50 +01:00
2653642 ./data/a-ffv1.avi
ffv1 regression test Originally committed as revision 2477 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-11-03 22:59:21 +01:00
799d3db687f6cdd7a837ec156efc171f *./data/out.yuv
fixpoint log() for tiny_psnr so it can output actual PSNR Originally committed as revision 2640 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-12-31 03:32:54 +01:00
stddev: 0.00 PSNR:99.99 bytes:7602176
try to select w/h from tables Originally committed as revision 3443 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-09-08 19:57:25 +02:00
b92d6d901574ea17e43e96ea97637078 *./data/a-svq1.mov
1383999 ./data/a-svq1.mov
svq1 mv array size fix Originally committed as revision 3361 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-07-28 12:13:45 +02:00
ccc201054669e94717022bb4f2aea4ce *./data/out.yuv
stddev: 10.99 PSNR:27.30 bytes:7602176
using only the fast int dct for regression tests Originally committed as revision 884 to svn://svn.ffmpeg.org/ffmpeg/trunk
2002-08-30 21:49:12 +02:00
21f8ff9f1daacd9133683bb4ea0f50a4 *./data/a-mp2.mp2
10l (fixes delayed frames during encoding) Originally committed as revision 3279 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-07-01 01:41:29 +02:00
95712 ./data/a-mp2.mp2
dithering for the mpeg audio decoder Originally committed as revision 3903 to svn://svn.ffmpeg.org/ffmpeg/trunk
2005-01-30 15:04:56 +01:00
6956f2c5185c4edbae648c614992a14b *./data/out.wav
stddev:117.02 PSNR: 6.76 bytes:1054720
AC3 encoding patch ba (Ross Martin <ffmpeg at ross dot interwrx dot com>) Originally committed as revision 2129 to svn://svn.ffmpeg.org/ffmpeg/trunk
2003-08-20 09:57:00 +02:00
d056da679e6d6682812fffb28a7f0db6 *./data/a-ac3.rm
regression checksum changes due to new resampling code Originally committed as revision 3232 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-06-18 04:59:30 +02:00
97983 ./data/a-ac3.rm
100l (dont reset the sample position at the block end) Originally committed as revision 3618 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-10-21 22:01:32 +02:00
2d081e0e1f2e9bd4514e9ac8ec41884c *./data/a-g726.wav
24268 ./data/a-g726.wav
a719ab6d47d8d601520edb13bf6136b4 *./data/out.wav
quick and dirty PSNR check for audio Originally committed as revision 3902 to svn://svn.ffmpeg.org/ffmpeg/trunk
2005-01-30 14:26:11 +01:00
stddev:134.28 PSNR: 5.56 bytes:96256
ima & ms adpcm regression test Originally committed as revision 2766 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-02-08 03:35:37 +01:00
bdc512b78e2ccb1d815aec08794650eb *./data/a-adpcm_ima.wav
10l (fixes delayed frames during encoding) Originally committed as revision 3279 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-07-01 01:41:29 +02:00
266288 ./data/a-adpcm_ima.wav
write shorter headers if possible Originally committed as revision 3494 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-09-22 22:01:56 +02:00
1316b03da3bdc8df9785c7995e3c9b2a *./data/out.wav
quick and dirty PSNR check for audio Originally committed as revision 3902 to svn://svn.ffmpeg.org/ffmpeg/trunk
2005-01-30 14:26:11 +01:00
stddev:129.55 PSNR: 5.87 bytes:1054720
write shorter headers if possible Originally committed as revision 3494 to svn://svn.ffmpeg.org/ffmpeg/trunk
2004-09-22 22:01:56 +02:00
d2eee867856d2bdb6d08e936d4ceec0c *./data/a-adpcm_ms.wav
267308 ./data/a-adpcm_ms.wav
91a84bb4f319a3a0bf0c0441b3d3a529 *./data/out.wav
quick and dirty PSNR check for audio Originally committed as revision 3902 to svn://svn.ffmpeg.org/ffmpeg/trunk
2005-01-30 14:26:11 +01:00
stddev: 73.79 PSNR:10.76 bytes:1054720