Log in

View Full Version : AutoMKV 0.95 beta 21 NEED TESTER!


Pages : 1 2 3 4 5 [6] 7 8

Encoder888
25th January 2008, 04:38
buzzqw:

Just wanted to congratulate you on the new aspect ratio improvements. It does work! The half inch 16:9 difference with original DVD versions of the the 1.85 movies that I noticed with 0.93 is gone now and all the movies I tested retain their perfect original aspect ratios.

I have a question about the delaycut changes - What difference does it make whether you use delaycut or you get the delay fixed in the mux, or you use the delayaudio in soundout (not even sure what that is). I mean, if it fixes the audio delay what exactly is the end-result difference? Is there a best option?

buzzqw
25th January 2008, 08:37
@survivant001

is it possible to add audio CUT to get to trim the audio if the track is too long ?

the audio is "cutted" at muxing or is encoded only from frame 0 to XXX ?

@Taddeusz
Thank you! The only issue I ran into is that option won't save in the automkv.ini

already fixed in next update! (and added to queue too)

@valnar

i must admit that i prefer to add fuctions that i know how works.
And the SAR/PAR/DAR are still in a grey area...

anyway on x264 encoding the SAR:
On Non Anamorphic encoding is always --sar 1:1
On Anamorphic encoding is --sar ITU*AR*((cropped_width/total_width)/(cropped_height/total_height))

1) on a 2.35 PAL movie (cropped 72 top, 72 bottom, ITU=53.3333/52, AR=16/9 or 4/3)

53.3333/52*16/9*((720/720)/(432/576))=2.431147578347578

without ITU= 2.37037037037037

@Encoder888

I have a question about the delaycut changes - What difference does it make whether you use delaycut or you get the delay fixed in the mux, or you use the delayaudio in soundout (not even sure what that is). I mean, if it fixes the audio delay what exactly is the end-result difference? Is there a best option?

for your information:
MeGui : use DelayAudio
AutoGK : Fix in Mux
Old AutoMKV: DelayCut
StaxRip: Fix in Mux
RipBot264 (iirc): use DelayAudio

so.. all choice are correct, and no difference should be spotted on end-results

@ALL

i would like to publish the new FINAL 0.95 soon.. so please BUGS are welcome

BHH

mnerec
25th January 2008, 09:16
00:01:25 Dgindex CMD: C:\Tools\AutoMKV 0.95beta\exe\dgindex\dgindex.exe -FO=0 -OM=1 -TN=80 -YR=2 -AIF=[C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\Qu he hecho yo para merecer esto_1984\VTS_01_1.vob] -OF=[C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\movie] -exit -minimize
00:03:23 D2V file size: 717804
00:03:23 Audio 1 :movie t80 2_0ch 192kbps delay 0ms.ac3
00:03:23 Pre Audio Size: 139941888 bytes
00:03:23 Applying DelayAudio of 0 to audio file: movie t80 2_0ch 192kbps delay 0ms.ac3
00:03:23 Encoding C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\movie t80 2_0ch 192kbps delay 0ms.ac3 to NeroAAC with 0.50 quality
00:07:48 Encoded 1' Audio Track: -1 bytes
00:07:48 Unable to encode Audio, try piping with FFmpeg!
00:07:48 NeroAACEnc doesn't accept piping from ffmpeg, switching to LibFaac at 96kbps
00:11:22 Encoded 1' Audio Track: 53906920 bytes
00:11:25 Number of Frames: 145773
00:11:25 Framerate: 25.000000
00:11:25 Movie length in Seconds: 5831
00:11:25 Movie Width/Height: 720/576
00:11:25 DgIndex AR: 16:9
00:14:15 Source is declared tff by a margin of 52/43.
00:14:15 Advanced Deinterlace Routines Log
00:14:15 Analyzed 871 sections of five frames
00:14:15 Found Interlace Sections: 1 0.11%
00:14:15 Found Progressive Sections:870 99.89%
00:14:15 Found Telecined Sections: 0 0.00%
00:14:15 Found Zero Mov: 99.89% 870
00:14:15 Found One Mov: 0.11% 1
00:14:15 Found Two Mov: 0.00% 0
00:14:15 Found Three Mov: 0.00% 0
00:14:15 Found Four Mov: 0.00% 0
00:14:15 Found Five Mov: 0.00% 0
00:14:15 The Movie is declared to progressive
00:14:15 Starting X264
00:14:15 Command Line 1' Pass X264: C:\Tools\AutoMKV 0.95beta\exe\encoder\x264.exe --pass 1 --bitrate 2056 --stats "C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\.stats" --ref 1 --progress --keyint 250 --bframes 3 --qpmin 10 --qpmax 51 --no-psnr --filter 0,0 --subme 2 --direct auto --vbv-maxrate 25000 --me dia --no-ssim --level 4.1 --b-pyramid --analyse none --threads auto --thread-input --sar 1:1 --output NUL "C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\movie.avs"
00:14:15 Command Line 2' Pass X264: C:\Tools\AutoMKV 0.95beta\exe\encoder\x264.exe --pass 2 --bitrate 2056 --stats "C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\.stats" --progress --keyint 250 --bframes 3 --qpmin 10 --qpmax 51 --no-psnr --mixed-refs --trellis 1 --ref 4 --filter 0,0 --direct auto --vbv-maxrate 25000 --me hex --no-ssim --level 4.1 --weightb --b-pyramid --b-rdo --bime --analyse p8x8,b8x8,i4x4 --threads auto --thread-input --sar 1:1 --output "C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\movie.264" "C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\movie.avs"
01:14:02 Finished 1' Pass X264
03:41:59 X264 Final 2 Pass Encoding Stats:
x264 [info]: using SAR=1/1
x264 [info]: using cpu capabilities: MMX MMXEXT SSE SSE2 SSE3 SSSE3 Cache64
x264 [info]: slice I:1057 Avg QP:17.82 size: 41374
x264 [info]: slice P:64797 Avg QP:19.77 size: 15172
x264 [info]: slice B:79919 Avg QP:21.35 size: 5903
x264 [info]: mb I I16..4: 30.7% 0.0% 69.3%
x264 [info]: mb P I16..4: 11.2% 0.0% 6.2% P16..4: 36.4% 27.9% 11.8% 0.0% 0.0% skip: 6.5%
03:41:59 Only Movie Size: 1498654720 bytes
03:41:59 "C:\Tools\AutoMKV 0.95beta\exe\matroska\mkvmerge.exe" -o "C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\Qu he_1984.mkv" --default-duration 0:25000/1000fps --track-name -1:"Qu he_1984" --chapters "C:\Tools\AutoMKV 0.95beta\exe\chapters.txt" "C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\movie.264" --language -1:esp --default-track -1:yes "C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\audio1.mp4" --title "Done with AutoMKV 0.95 beta 17 http://forum.doom9.org/showthread.php?p=854221 "
03:42:00 Final Muxed size: -1 bytes
03:42:00 Encoding finished: 02:27:58 elapsed time
03:42:00 -----------------------------------------------------------------------
03:42:00 Script movie.avs
03:42:00
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\AddGrain.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\autocrop.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\aWarpSharp(Pruned).dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\blockbuster.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\cnr2.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\ColorMatrix.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\Convolution3D.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\corrector.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\DctFilter.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\Decomb.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\Deen.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\degrainmedian.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\DGDecode.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\EEDI2.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\FFMpegSource.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\FFT3DFilter.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\FluxSmooth.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\frfun7.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\hqdn3d.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\LeakKernelDeint.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\MaskTools.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\medianblur.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\MipSmooth.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\MSharpen.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\mt_masktools.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\mvtools.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\NicAudio.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\RemoveGrainS.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\RepairS.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\SangNom.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\SimpleResize.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\SoundOut.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\TBilateral.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\TDeint.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\TIVTC.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\TomsMoComp.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\TTempSmooth.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\UnDot.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\UnFilter.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\VSFilter.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\warpsharp.dll")
03:42:00 #loadpluginstart
03:42:00
03:42:00 #test1
03:42:00
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\BlindDeHalo3_mt2.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\Deblock_QED_MT2.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\FastLineDarken 1.3 MT MOD.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\HybridFuPP.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\LanczosPlus.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\Lem.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\LimitedSharpenFaster.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\MCBob_v03c.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\mvbob.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\QMlib.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\seesaw.avsi")
03:42:00 Import("C:\Tools\AutoMKV 0.95beta\exe\filter\spresso.avsi")
03:42:00 #importstart
03:42:00
03:42:00 #test2
03:42:00
03:42:00 movie = mpeg2source("C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\movie.d2v")
03:42:00 function getOrder(clip c) {
03:42:00 order = GetParity(c) ? 1 : 0
03:42:00 Return order }
03:42:00 #filter1start
03:42:00
03:42:00 #test3
03:42:00
03:42:00 last = movie
03:42:00 crop(0,12,-0,-12)
03:42:00 fixed_aspect = 1.854991
03:42:00 #resizestart
03:42:00
03:42:00 #test4
03:42:00
03:42:00 LanczosResize(800,432)
03:42:00 #filter2start
03:42:00
03:42:00 #test5
03:42:00
03:42:00 RemoveGrain(mode=2)
03:42:00 -----------------------------------------------------------------------
03:42:00 Script mkvmaudio.avs
03:42:00
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\NicAudio.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\SoundOut.dll")
03:42:00 LoadPlugin("C:\Tools\AutoMKV 0.95beta\exe\filter\FFmpegSource.dll")
03:42:00 NicAC3Source("C:\Users\Public\Videos\Movies\_NEEDS SUBS_Almodovar\_que_he\temp\movie t80 2_0ch 192kbps delay 0ms.ac3")
03:42:00 DelayAudio(0.0000)
03:42:00 #Applying STEREO downmixing routines
03:42:00 function stereo(clip a)
03:42:00 {
03:42:00 flr = GetChannel(a, 1, 2)
03:42:00 fcc = GetChannel(a, 3)
03:42:00 lfe = GetChannel(a, 4)
03:42:00 lfc = MixAudio(fcc, lfe, 0.2071, 0.2071)
03:42:00 mix = MergeChannels(lfc, lfc)
03:42:00 lrc = MixAudio(flr, mix, 0.2929, 1.0)
03:42:00 blr = GetChannel(a, 5, 6)
03:42:00 Return MixAudio(lrc, blr, 1.0, 0.2929)
03:42:00 }
03:42:00 #
03:42:00 6==Audiochannels() ? stereo() : last
03:42:00 function addvideo(clip c) {
03:42:00 blankclip(length = Int(100 * AudioLengthF(c) / float(audiorate(c))), fps=100)
03:42:00 Return audiodub(last,c)
03:42:00 }
03:42:00 Normalize()
03:42:00
03:42:00 Soundout(output="cmd",type=1,format=3,autoclose=true,executable="C:\Tools\AutoMKV 0.95beta\exe\besweet\neroaacenc.exe",prefilename=" -ignorelength -q 0.50 -if - -of audio.mp4",postfilename="",nofilename=true,showoutput=false)
03:42:00
03:42:00 - - - - - - - - - - - - FINISHED JOBS - - - - - - - - - - - - - - - - -

Tjek out this line:
00:07:48 NeroAACEnc doesn't accept piping from ffmpeg, switching to LibFaac at 96kbps

Why on earth 96kpbs? I asked for 0.5 quality (~192kbps).

I know 192 to 192kbps transcoding might be a bit silly, but this is what I asked, not some random bitrate.

Why is this?

buzzqw
25th January 2008, 09:31
@mnerec

Why on earth 96kpbs? I asked for 0.5 quality (~192kbps).

i cannot pass 0.35 as "bitrate" to libfaac... :eek:

i will try to "translate to a rational bitrate

BHH

mnerec
25th January 2008, 09:34
Here is the conversion list I use.

Its is from an audio processing program dbPowerAmp and is intended for NeroACC:

5
Quality .05 (estimated bit rate: 16 Kbps)
-q .05
Quality .15 (estimated bit rate: 32 Kbps)
-q .15
Quality .25 (estimated bit rate: 64 Kbps)
-q .25
Quality .35 (estimated bit rate: 100 Kbps)
-q .35
Quality .45 (estimated bit rate: 150 Kbps)
-q .45
Quality .5 (estimated bit rate: 170 Kbps)
-q .5
Quality .55 (estimated bit rate: 200 Kbps)
-q .55
Quality .65 (estimated bit rate: 250 Kbps)
-q .65
Quality .75 (estimated bit rate: 300 Kbps)
-q .75
Quality .85 (estimated bit rate: 350 Kbps)
-q .85
Quality .95 (estimated bit rate: 400 Kbps)
-q .95

buzzqw
25th January 2008, 10:56
@mnerec

i will use MOD16(quality*320) so 0.50=160, 0.35=112

http://www.64k.it/andres/data/a/AutoMKV095beta18.exe

i hope this will be the last beta

BHH

Honeyko
25th January 2008, 11:32
I would consider any behavior which prevents an encoded file from being moved outside of the Temp folder (or which triggers a renaming of the Temp folder) to be a bug, as it will result in destroyed encodes by succeeding queued projects.

buzzqw
25th January 2008, 11:51
I would consider any behavior which prevents an encoded file from being moved outside of the Temp folder (or which triggers a renaming of the Temp folder) to be a bug, as it will result in destroyed encodes by succeeding queued projects.

if file isn't moved it's because mux failed, and if mux fail it's because some stream is missing.

It's better to resolve problem then to walk around

BHH

mnerec
25th January 2008, 12:23
@mnerec
i will use MOD16(quality*320) so 0.50=160, 0.35=112


Hmm, in my tests Nero generates aprox. 170-175 kbps for me at 0.50, not 160kbps.

Doesn't that mean that you'll always generate lower quality than what the user expects/wants when switching encoders?

Maybe you should do a few tests with libfaac to back up your numbers? Personally I don't know how to compress with libfacc, so I can't do it for you.

buzzqw
25th January 2008, 12:39
quality is always variable ...

and since it's a failsave situation .. it's more then enough

BHH

zacoz
25th January 2008, 13:12
I've tested the latest beta and still have the same problem (http://forum.doom9.org/showthread.php?p=1087493#post1087493) as with earlier 0.95 betas. If you use original size (no resizing or croping) with a 16:9 source and output muxing AR set to Automatic then the output MKV is muxed as 5:4 - yuk :(

Is there a flaw in the formula used to calculate the automatic muxing AR when there is no cropping or resize?

Or is it simply an mathematical anomaly such that there needs to be an override in these cases so that Automatic setting simply uses the source AR.

I'll go and run some tests with a DVD source (instead of the below SD Capture) to see if the same issue occurs.

21:39:43 0.95 beta 18 - - - - - - - - - - - - - START JOBS - - - - - - - - - - - - - -
21:39:43 OS Version: Windows XP
21:39:43 Selected Input file: D:\Capture\7 Digital 2008-01-25_213205.m2v
21:39:43 Selected Output dir: D:\Capture\temp\
21:39:43 Selected Final Name: Test
21:39:43 Selected Container type: MKV
21:39:43 Selected Encoder: X264
21:39:43 Selected Size: Unlimited File Size
21:39:43 Selected Profile: Constant_Quality.xml
21:39:43 Selected Resizer: NONE / Original
21:39:43 Selected Filters: NONE
21:39:43 Selected Deinterlacer: AUTO
21:39:43 Selected Field Order: AUTO
21:39:43 Enabled Constant Quality With Unlimited Size (Q 22)
21:39:43 Track Language 1: D:\Capture\7 Digital 2008-01-25_213205.mp2
21:39:43 Audio Codec 1: Nero AAC
21:39:43 Audio Quality 1: 0.35
21:39:43 Audio Channel 1: Stereo
21:39:43 Track Language 2: NONE
21:39:43 Audio Codec 2: Nero AAC
21:39:43 Advanced Audio Settings DUMP Track 1
21:39:43 Nero Option Profiles: AUTO --- Encoding Mode: Quality
21:39:43 CCT Option Profiles: AUTO --- Channel Options: AUTO
21:39:43 Lame Settings: ABR
21:39:43 Audio Normalization: 1
21:39:43 Frequency: AUTO --- Tempo: NONE --- Pitch: NONE --- Other Add:
21:39:43 Selected Unlimited media size
21:39:43 Dgindex CMD: E:\AutoMKV95b\exe\dgindex\dgindex.exe -FO=0 -OM=1 -YR=2 -AIF=[D:\Capture\7 Digital 2008-01-25_213205.m2v] -OF=[D:\Capture\temp\movie] -exit -minimize
21:39:44 Encoding D:\Capture\temp\7 Digital 2008-01-25_213205.mp2 to NeroAAC with 0.35 quality
21:40:00 Encoded 1' Audio Track: 269826 bytes
21:40:03 Number of Frames: 538
21:40:03 Framerate: 25.000000
21:40:03 Movie length in Seconds: 22
21:40:03 Movie Width/Height: 720/576
21:40:03 DgIndex AR: 16:9
21:40:59 Source is declared tff by a margin of 104/0.
21:40:59 Advanced Deinterlace Routines Log
21:40:59 Analyzed 106 sections of five frames
21:40:59 Found Interlace Sections: 106 100.00%
21:40:59 Found Progressive Sections:0 0.00%
21:40:59 Found Telecined Sections: 0 0.00%
21:40:59 Found Zero Mov: 0.00% 0
21:40:59 Found One Mov: 0.00% 0
21:40:59 Found Two Mov: 0.00% 0
21:40:59 Found Three Mov: 0.00% 0
21:40:59 Found Four Mov: 0.00% 0
21:40:59 Found Five Mov: 100.00% 106
21:40:59 The Movie is declared to interlaced
21:40:59 Starting X264
21:40:59 CRF encoding X264: E:\AutoMKV95b\exe\encoder\x264.exe --pass 1 --crf 22 --stats "D:\Capture\temp\.stats" --progress --keyint 250 --bframes 3 --qpmin 10 --qpmax 51 --no-psnr --mixed-refs --trellis 1 --ref 3 --filter -2,-1 --direct auto --vbv-maxrate 25000 --me umh --no-ssim --level 4.1 --weightb --b-pyramid --b-rdo --bime --analyse p8x8,b8x8,i4x4,i8x8,p4x4 --8x8dct --threads auto --thread-input --sar 1:1 --output "D:\Capture\temp\movie.264" "D:\Capture\temp\movie.avs"
21:41:42 X264 Final CRF Encoding Stats:
x264 [info]: using SAR=1/1
x264 [info]: using cpu capabilities: MMX MMXEXT SSE SSE2 SSE3 SSSE3 Cache64
21:41:42 Accepted values are between 697303040 and 741343232 bytes (movie+audio)
21:41:42 Got a file of size: 8632931 bytes + audio size 269826 bytes (8902757)
21:41:42 Only Movie Size: 8632931 bytes
21:41:42 "E:\AutoMKV95b\exe\matroska\mkvmerge.exe" -o "D:\Capture\temp\Test.mkv" --default-duration 0:25000/1000fps --track-name -1:"Test" --aspect-ratio 0:1.250000 "D:\Capture\temp\movie.264" --language -1:und --default-track -1:yes "D:\Capture\temp\7 Digital 2008-01-25_213205.mp4" --title "Done with AutoMKV 0.95 beta 18 http://forum.doom9.org/showthread.php?p=854221 "
21:41:42 Final Muxed size: 8909518 bytes
21:41:42 Encoding finished: 00:00:43 elapsed time
21:41:42 -----------------------------------------------------------------------
21:41:42 Script AutoCrop.log
21:41:42
21:41:42 Crop(2,4,716,572)
21:41:42 -----------------------------------------------------------------------
21:41:42 Script movie.avs
21:41:42
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\AddGrain.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\autocrop.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\aWarpSharp(Pruned).dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\blockbuster.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\cnr2.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\ColorMatrix.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\Convolution3D.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\corrector.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\DctFilter.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\Decomb.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\Deen.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\degrainmedian.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\DGDecode.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\EEDI2.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\FFMpegSource.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\FFT3DFilter.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\FluxSmooth.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\frfun7.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\hqdn3d.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\LeakKernelDeint.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\MaskTools.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\medianblur.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\MipSmooth.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\MSharpen.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\mt_masktools.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\mvtools.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\NicAudio.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\RemoveGrainS.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\RepairS.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\SangNom.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\SimpleResize.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\SoundOut.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\TBilateral.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\TDeint.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\TIVTC.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\TomsMoComp.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\TTempSmooth.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\UnDot.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\UnFilter.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\VSFilter.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\warpsharp.dll")
21:41:42 #loadpluginstart
21:41:42
21:41:42 #test1
21:41:42
21:41:42 Import("E:\AutoMKV95b\exe\filter\BlindDeHalo3_mt2.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\Deblock_QED_MT2.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\FastLineDarken 1.3 MT MOD.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\HybridFuPP.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\LanczosPlus.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\Lem.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\LimitedSharpenFaster.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\MCBob_v03c.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\mvbob.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\QMlib.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\seesaw.avsi")
21:41:42 Import("E:\AutoMKV95b\exe\filter\spresso.avsi")
21:41:42 #importstart
21:41:42
21:41:42 #test2
21:41:42
21:41:42 movie = mpeg2source("D:\Capture\temp\movie.d2v")
21:41:42 function getOrder(clip c) {
21:41:42 order = GetParity(c) ? 1 : 0
21:41:42 Return order }
21:41:42 #filter1start
21:41:42
21:41:42 #test3
21:41:42
21:41:42 movie = LeakKernelDeint(movie,order=1,sharp=true)
21:41:42 last = movie
21:41:42 fixed_aspect = 1.780186
21:41:42 out_width = 720
21:41:42 out_height = 576
21:41:42 #resizestart
21:41:42
21:41:42 #test4
21:41:42
21:41:42 #filter2start
21:41:42
21:41:42 #test5
21:41:42
21:41:42 return last
21:41:42 -----------------------------------------------------------------------
21:41:42 Script mkvmaudio.avs
21:41:42
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\NicAudio.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\SoundOut.dll")
21:41:42 LoadPlugin("E:\AutoMKV95b\exe\filter\FFmpegSource.dll")
21:41:42 NicMPG123Source("D:\Capture\temp\7 Digital 2008-01-25_213205.mp2")
21:41:42 #Applying STEREO downmixing routines
21:41:42 function stereo(clip a)
21:41:42 {
21:41:42 flr = GetChannel(a, 1, 2)
21:41:42 fcc = GetChannel(a, 3)
21:41:42 lfe = GetChannel(a, 4)
21:41:42 lfc = MixAudio(fcc, lfe, 0.2071, 0.2071)
21:41:42 mix = MergeChannels(lfc, lfc)
21:41:42 lrc = MixAudio(flr, mix, 0.2929, 1.0)
21:41:42 blr = GetChannel(a, 5, 6)
21:41:42 Return MixAudio(lrc, blr, 1.0, 0.2929)
21:41:42 }
21:41:42 #
21:41:42 6==Audiochannels() ? stereo() : last
21:41:42 function addvideo(clip c) {
21:41:42 blankclip(length = Int(100 * AudioLengthF(c) / float(audiorate(c))), fps=100)
21:41:42 Return audiodub(last,c)
21:41:42 }
21:41:42 Normalize()
21:41:42
21:41:42 Soundout(output="cmd",type=1,format=3,autoclose=true,executable="E:\AutoMKV95b\exe\besweet\neroaacenc.exe",prefilename=" -ignorelength -q 0.35 -if - -of audio.mp4",postfilename="",nofilename=true,showoutput=false)
21:41:42
21:41:42 - - - - - - - - - - - - FINISHED JOBS - - - - - - - - - - - - - - - - -

Edit: Same result with DVD Source :(

buzzqw
25th January 2008, 13:32
I've tested the latest beta and still have the same problem as with earlier 0.95 betas. If you use original size (no resizing or croping) with a 16:9 source and output muxing AR set to Automatic then the output MKV is muxed as 5:4 - yuk

DUMB question: the AR should be 16/9*576/720 ? or 16/9 ? OR 1 ?

BHH

zacoz
25th January 2008, 15:10
Don't know if the dumb question is directed to me or not (cause I am pretty dumb on this stuff).

Thinking out loud here in layman's terms (well my terms anyway)....

The source DAR of 16/9 is within 720*576 (W*H) pixel video so I'm guessing this makes the Pixel AR 16 / 9 * 576 / 720 = 1.422222

To get the correct Display of 16/9 when playing back the final mkv there would need to be a further horizontal stretching of 1.25 (5/4) applied (1.42222 * 1.25 = 1.777778 = 16/9).

This 1.25 is what is being passed to mkvmerge as the AR when the mkv is muxed.
"E:\AutoMKV95b\exe\matroska\mkvmerge.exe" -o "D:\Capture\temp\Test.mkv" --default-duration 0:25000/1000fps --track-name -1:"Test" --aspect-ratio 0:1.250000 "D:\Capture\temp\movie.264" --language -1:und --default-track -1:yes "D:\Capture\temp\7 Digital 2008-01-25_213205.mp4" --title "Done with AutoMKV 0.95 beta 18 http://forum.doom9.org/showthread.php?p=854221 "


On playback however I'm therefore only getting a DAR of 5/4 (1.25).

Is the encoding in 0.95betas outputing square pixels (PAR of 1.0) or pixels with a PAR of 1.422222?

If the former, then shouldn't mkvmerge be using an AR of 16/9 when muxing rather than 5/4?

If the later, then I'm not really sure myself what should be passed to mkvmerge.

Does this rambling make sense to anyone :stupid:

buzzqw
25th January 2008, 15:19
while we are arguing..

try this http://www.64k.it/andres/data/a/AutoMKV095beta18-zacoz.exe

BHH

zacoz
25th January 2008, 15:31
Actually, just been looking at the encoding line again and wondering if the issue is that the --sar 16:9 is incorrect.

"E:\AutoMKV95b\exe\encoder\x264.exe" --pass 1 --crf 22 --stats "D:\Capture\temp\.stats" --progress --keyint 250 --bframes 3 --qpmin 10 --qpmax 51 --no-psnr --mixed-refs --trellis 1 --ref 3 --filter -2,-1 --direct auto --vbv-maxrate 25000 --me umh --no-ssim --level 4.1 --weightb --b-pyramid --b-rdo --bime --analyse p8x8,b8x8,i4x4,i8x8,p4x4 --8x8dct --threads auto --thread-input --sar 16:9 --output "D:\Capture\temp\movie.264" "D:\Capture\temp\movie.avs"

The SAR for 16:9 in a 720*576 PAL video should be 1.422222 (64:45) shouldn't it? Thus --sar 64:45 should be passed to x264.exe?

Still then trying to work out what the AR passed to mkvmerge should be :confused:

zacoz
25th January 2008, 15:35
Please don't take it as arguing :(

I'm just trying to work out why the auto settings don't work in 0.95 when one doesn't crop or resize, whereas they used to work in 0.93.

I'm by no means right in all my thinking here. I've been trying to apply my own warped logic and my friend Google.

buzzqw
25th January 2008, 15:49
Hi have PM SeeMoreDigital

I hope for a bright light !

BHH

zacoz
25th January 2008, 15:49
while we are arguing..
try this http://www.64k.it/andres/data/a/AutoMKV095beta18-zacoz.exe
BHH

Wow, now that's a wide screen and the characters in my test file seem to have put on quite a bit of weight around the middle (i.e. 2.2222 DAR rather than 1.77778)

Looking at this site: http://aflux.deltaanime.net/Zero1/MP4/x264.html about 2/3rds down the page it talks about SAR and that the SAR for PAL 720 x 576 is 64:45 (1.422222) which seems to match with my fuzzy logic.

Encoder888
25th January 2008, 15:51
buzzqw:

Whenever I try to use Nero AAC for 1.0 (even though I know it's an overkill, but still) and I think for more than 0.75, my audio for some reason does not want to encode. I didn't have that problem with 0.93 and if possible, you may wanna look into that.

Also - that's not a bug, but just a suggestion. Whenever I have to encode a widescreen movie (which for me is every movie), I always have to check the "Anamorphic Encoding" option now to get the correct aspect ratio. Before, it did it automatically. You may want to have AutoMKV automatically determine if the movie is anamorphic and switch to anamorphic encoding, so people don't have to check it every time.

Also, I'm not sure what ITU resize is for... Everything works perfect for me with just Anamorphic checked... Well, I guess that falls into the questions category :))

Otherwise, great build, can't wait for the final release.

zacoz
25th January 2008, 15:57
Hi have PM SeeMoreDigital

I hope for a bright light !

BHH

Yes, SeeMoreDigital is pretty bright so there's a good chance :cool:

Hmmm, that reminds me I think I had a webpage bookmarked somewhere before my latest windows install where SeeMoreDigital explained this stuff. I think my initial reactions (due to my noobness) when trying to read it were :eek: :confused: . I think I'd take a bit more in these days though - will have look tomorrow, getting very late here (1am).

zacoz
25th January 2008, 16:32
Tried along lines of Encoder888's post of ticking Anamorphic, leaving PAR and Muxing AR at Auto, but this didn't work as ended up with SAR in x264.exe cli being 1.77778 and DAR in mkvmerge cli being 1.42222 which gave me a 1.42222 playback instead of 1.77778. Seem to be the wrong way around.

Mucked around a little more and seems that regardless of whether I set the SAR in the x264 to 1:1 or 64:45 etc it seems to come down to what Automkv passes to mkvmerge, so in the case of 720x576 PAL 16:9 source with no resize or cropping it has to be
--aspect-ratio 0:16/9
or
--aspect-ratio 0:1.777778

Here's hoping I haven't annoyed you too much tonight with this buzzqw - not my intent.

Taddeusz
25th January 2008, 16:44
This is the problem that I described above. The way I fix this is to specify the proper SAR value in "Force PAR in avs". Like your 64:45 value. It will not accept a single decimal value, it must be in x:y format. The I have to set "Force Muxing AR" to the correct aspect ratio, in your case 16:9. That seems to fix all AR issues and the final video is correct in both the MKV header and programs that read the bitstream SAR.

buzzqw
25th January 2008, 20:41
@Taddeusz

don't want to warn.. but i have just found that force mux ar don't do nothing... since it is applied on wrong stream...

just launch from dos muxing.bat..
you will see somthing about a missing stream

now , edit the muxing.bat and change --aspect-ratio 1:xxx to --aspect-ratio -1:xxx and watch again the movie

BHH

buzzqw
25th January 2008, 20:55
http://www.64k.it/andres/data/a/AutoMKV095beta18-2.exe

for anyone that want play, only changed force muxing dar and automatic dar

BHH

SeeMoreDigital
25th January 2008, 23:10
Personally I find it easier to calculate "everything" as a fraction, because converting fractions to decimals and then back to fractions "can" often produce conflicting Aspect Ratio Signalling (ARS) values.

MAR = Movie Aspect Ratio (ie: The image element of the movie. After any/all black mattes have been cropped away).

PFS = Pixel Frame Size (ie: The pixel resolution of the encode).

ARS = Aspect Ratio Signalling (ie: PAR, DAR, SAR) value.


The maths bit

If your NTSC movie has an aspect ratio of 1.33:1, then this can be represented as a fraction of 4/3. Meaning: -

MAR PFS ARS
4 480 1920 8
- x --- = ---- lowest dominator - Giving you a ARS value of 8:9 (which decimates to 0.888888)
3 720 2160 9

If your NTSC movie has an aspect ratio of 1.77:1, then this can be represented as a fraction of 16/9. Meaning: -

MAR PFS ARS
16 480 7680 32
-- x --- = ---- lowest dominator -- Giving you a ARS value of 32:27 (which decimates to 1.185185)
9 720 6480 27

If your NTSC movie has an aspect ratio of 1.85:1, then this can be represented as a fraction of 37/20. Meaning: -

MAR PFS ARS
37 480 17760 37
-- x --- = ----- lowest dominator -- Giving you a ARS value of 37:30 (which decimates to 1.233333)
20 720 14400 30

If your NTSC movie has an aspect ratio of 2.35:1, then this can be represented as a fraction of 47/20. Meaning: -

MAR PFS ARS
47 480 22560 47
-- x --- = ----- lowest dominator -- Giving you a ARS value of 47:30 (which decimates to 1.566666)
20 720 14400 30

If your NTSC movie has an aspect ratio of 2.40:1, then this can be represented as a fraction of 12/5. Meaning: -

MAR PFS ARS
12 480 5760 8
-- x --- = ---- lowest dominator - Giving you a ARS value of 8:5 (which decimates to 1.6)
5 720 3600 5

If your PAL movie has an aspect ratio of 1.33:1, then this can be represented as a fraction of 4/3. Meaning: -

MAR PFS ARS
4 576 2304 16
- x --- = ---- lowest dominator -- Giving you a ARS value of 16:15 (which decimates to 1.066666)
3 720 2160 15

If your PAL movie has an aspect ratio of 1.77:1, then this can be represented as a fraction of 16/9. Meaning: -

MAR PFS ARS
16 576 9216 64
-- x --- = ---- lowest dominator -- Giving you a ARS value of 64:45 (which decimates to 1.422222)
9 720 6480 45

If your PAL movie has an aspect ratio of 1.85:1, then this can be represented as a fraction of 37/20. Meaning: -

MAR PFS ARS
37 576 21312 37
-- x --- = ----- lowest dominator -- Giving you a ARS value of 37:25 (which decimates to 1.48)
20 720 14400 25

If your PAL movie has an aspect ratio of 2.35:1, then this can be represented as a fraction of 47/20. Meaning: -

MAR PFS ARS
47 576 22072 47
-- x --- = ----- lowest dominator -- Giving you a ARS value of 47:25 (which decimates to 1.88)
20 720 14400 25

If your PAL movie has an aspect ratio of 2.40:1, then this can be represented as a fraction of 12/5. Meaning: -

MAR PFS ARS
12 576 6712 48
-- x --- = ---- lowest dominator -- Giving you a ARS value of 8:5 (which decimates to 1.92)
5 720 3600 25

While the above information is useful for calculating the ARS values of movies with popular MAR's (ie: 4:3, 16:9, 1.85:1, 2.35:1, 2.40:1 etc, etc). We know from bitter experience that these MAR's are, at best, only approximations!


The Good News

Fortunately for us, AutoMKV is able to work out the movies original aspect ratio as a decimal (see "#test3" in the log). And this decimal value can be used together with the encoded pixel frame size to determine the required ARS value.

Indeed in post #226 I wrote: - Hi Buzz,

I've been looking at the log reports and I'm curious to understand how the following "red highlighted" bit is worked out mathematically?
17:14:52 #test3
17:14:52
17:14:52 last=movie.Crop(30,86,656,412)
17:14:52 fixed_aspect = 2.264410
17:14:52 out_width = 656
17:14:52 out_height = 288
17:14:52 #resizestartHowever it's worked out, the resulting "fixed aspect ratio" appears to be very accurate!

Now, a while ago I conceived a little application that can work out the ARS value for movies displayed at any movie aspect ratio and encoded at any pixel frame size. The application is called the ARS Calculator. Unfortunately, it requires .NET to run, so it can't be used in-conjunction with AutoMKV... Putting that aside for a bit. It would work like this: -

http://i29.tinypic.com/2w2g6ya.png

The resulting ARS value can then be entered into the "Force PAR in AVS" box.


The Not So Good News

I've been tinkering with AutoMKV's many input settings and have found that it's possible to confuse the "fixed_aspect" calculations. For example: -

If I select the following: -

Resize Filters: Simple Resize (Neutral Soft)
Set Width Resolution: Auto Crop Only

.... I obtain the movies correct "fixed_aspect".

However, if I select the following: -

Resize Filters: Simple Resize (Neutral Soft)
Set Width Resolution: Auto Crop Only
Anamorphic Encoding: Checked

Or: -

Resize Filters: Simple Resize (Neutral Soft)
Set Width Resolution: Auto Crop Only
Force ITU Resize: Checked

.... I obtain the movies wrong correct "fixed_aspect".

That's all for now guys.... ;)


Cheers

Honeyko
25th January 2008, 23:25
I would consider any behavior which prevents an encoded file from being moved outside of the Temp folder (or which triggers a renaming of the Temp folder) to be a bug, as it will result in destroyed encodes by succeeding queued projects.if file isn't moved it's because mux failed, and if mux fail it's because some stream is missing. It's better to resolve problem then to walk aroundIf the "problem" involves FFmpeg kicking because the user chose a non-MOD16 crop/resize, or because his 45mb audio puked, that's no reason to lose a perfectly good 1gb video file when the next queued project empties the Temp folder. -- I suppose I could keep "Don't Mux" checked permanently, but that would be annoying.

Suggestion: Any error resulting in no mux when the user desires a mux (because he does not have "Don't Mux" checked), would flag "Don't Mux" on a one-time basis in order to preserve the work done so far.

buzzqw
25th January 2008, 23:26
However, if I select the following: -

Resize Filters: Simple Resize (Neutral Soft)
Set Width Resolution: Auto Crop Only
Anamorphic Encoding: Checked

i must admit i have based my calculation on megui project (thinking that was correct)

so.. on case of "Anamorphic Encoding: Checked" the fixed_aspect value on script how must be computed ?


When "Anamorphic Encoding: NOT Checked" i use
fixed_aspect=itu*AR*(cropped_width/total_width)/(cropped_height/total_height))

When "Anamorphic Encoding: IS Checked" i use
fixed_aspect=(cropped_width/cropped_height)
so no computing ITU or AR. if movie isn't cropped cropped_width=total _width and cropped_height=total_height

Wrong ? :confused:

Resize Filters: Simple Resize (Neutral Soft)
Set Width Resolution: Auto Crop Only
Force ITU Resize: Checked

.... I obtain the movies wrong correct "fixed_aspect".

isn't ITU=53.33333/52 (for PAL) and 1.02536086 for NTSC ?

so is above math correct ?... :confused:

:thanks:!

BHH

GenjyoSanzo
26th January 2008, 00:40
I have tested beta's 5-8 & 15-17; here is what I have found:

Direct Show Source:
Works in beta's 5-8; doesn't work with beta 15 or higher.
I don't have any beta's between 8-15 to test.
It doesn't matter if I use a vob and select to use DSS for video, or if I just use a mkv, same results:
AutoMKV will just sit there with about 30%-40% cpu usage on my dual core.


14:12:14 0.95 beta 17 - - - - - - - - - - - - - START JOBS - - - - - - - - - - - - - -
14:12:14 OS Version: Windows XP
14:12:14 Selected Input file: D:\test.VOB
14:12:14 Selected Output dir: D:\temp\
14:12:14 Selected Final Name: test
14:12:14 Selected Container type: MKV
14:12:14 Selected Encoder: X264
14:12:14 Selected Size: Unlimited File Size
14:12:14 Selected Profile: Constant_Quality.xml
14:12:14 Selected Resizer: SimpleResize (neutral soft) / 704
14:12:14 Selected Filters: NONE
14:12:14 Selected Deinterlacer: AUTO
14:12:14 Selected Field Order: AUTO
14:12:14 Enabled Constant Quality With Unlimited Size (Q 17)
14:12:14 Using Matrix: M4G_High_Detail_V3.1.cfg
14:12:14 Track Language 1: NONE
14:12:14 Audio Codec 1: Nero AAC
14:12:14 Track Language 2: NONE
14:12:14 Audio Codec 2: Nero AAC
14:12:14 Enabled Use DirectShowSource for AUDIO and VIDEO
14:12:14



Autocropping + QMTrim:
All tested beta's will work with "light_all_around_denoise",
they all fail with "anime_denoise", "aggressive_all_around_denoise", & "aggressive_all_around_slow_denoise"
I have not tested any of the other scripts.
After x264 fails it will then fail back onto ffmpeg.


Evaluate: System exception - Access Violation (D:\Temp\movie.avs line 89)


The last line in the following is line 89:


#test3

movie = tfm(movie,d2v="D:\temp\movie.d2v").tdecimate()
last=movie.Crop(6,0,712,480)
fixed_aspect = 1.318485
out_width = 712
out_height = 544
#resizestart

#test4

#filter2start

#test5

C0 = Spline36Resize(out_width,out_height).deen("c3d",1,1,2,2)
C1 = BicubicResize(out_width,out_height,0,0.75).deen("c3d",1,3,3,2)
C2 = BicubicResize(out_width,out_height,0,0.50).deen("c3d",1,3,4,2)
C3 = BicubicResize(out_width,out_height,0.333,0.333).deen("c3d",1,4,4,2)
C4 = BicubicResize(out_width,out_height,0.333,0.1775).deen("c3d",1,4,5,1)
C5 = BicubicResize(out_width,out_height,0.38,0).deen("c3d",1,5,5,1)
C6 = BicubicResize(out_width,out_height,0.5402,0).deen("c3d",1,5,6,1).DCTFilter(1,1,1,1,1,1,1,.5)


Dirty work around is to manually set width resolution to something MOD16 in GUI before encoding, or manually edit out_width in movie.avs to something MOD16.


Autocropping + Threshold:
The default threshold of 34 is good enough for most movie's + shows, but it is not good enough for everything.
In previous versions up to version 0.93, I could set "Pause before encoding" and manually change the threshold value to something as high as 90 in the movie.avs file for the material that needed a significantly higher threshold to remove the black bars on the sides, mainly on the right. While this work around was annoying, at least it was there. Now with 0.95 and your new internal use of autocrop this is not possible. Also in versions before 0.95, with AutoCrop Only, if you had a source material that needed multiple values of cropping; as in one scene needed a value of x, but another scene would need a value of y, I could get Autocrop to work correctly if I changed the sample rate & threshold rate depending on the source before letting x264.exe do it's thing.

Suggestion:
In the "Advanced Settings" tab, have a drop down box to be able to change autocrop threshold & sample rate. This would be great for queued jobs since I would not have to manually change each show's job.


Thank you for making a great program.

survivant001
26th January 2008, 03:23
@survivant001



the audio is "cutted" at muxing or is encoded only from frame 0 to XXX ?


exactly.. 0->totalframecount

buzzqw
26th January 2008, 09:33
@GenjyoSanzo

First of all Welcome to the Forum!

and thanks for your suggestion and bugs

Direct Show Source:

fixed

http://www.64k.it/andres/data/a/AutoMKV095beta19.exe

Autocropping + QMTrim:
deen want a compulsory mod16 width... and only deen
as you see the height is mod16 (i force it to mod16).. but te width on case of autocrop only is mod4 ...

i will try to solve

Autocropping + Threshold:
i added the possibility to use only Manual Crop without specifing the width/height

for difficult autocrop i suggest to use the manual way

@survivant001

exactly.. 0->totalframecount
... i don't know .. in a TFM/Decimate situation..

but. i'll try

BHH

zacoz
26th January 2008, 10:17
Done some comparisons (source as noted below) selecting specific width (e.g. 640), AutoCrop Only, and Original, all with and without Anamorphic selected. AutoMKV seems to handle then all fine (visually) except for when Original width is selected.

When Original is selected, this is enforcing the same width AND height. That's fine.
This being the case if the original detected Display Aspect Ratio is greater than the Original Width / Original Height then the video is anamorphic (as mentioned by Encoder888 (http://forum.doom9.org/showthread.php?p=1091885#post1091885) AutoMKV should treat is as Anamorphic in such cases even if advanced option hasn't been ticked)
As the original width and height are both being retained, then the Muxing AR used in mkvmerge should be the detected Display Aspect Ratio of the original file.
Working backwards then, the SAR should be DAR / Width * Height.

Strangely, if I use an anamorphic PAL 720*576 16:9 source video that has no black bars and process using Autocrop Only and Anamorphic ticked then AutoMKV calculates a fixed_aspect = 1.251748 and uses a SAR of 1.4321678 and Muxing AR of 1.780186 and I get a PAL 716*576 16:9 output video. Perfect :)
17:34:43 CRF encoding X264: E:\AutoMKV95b\exe\encoder\x264.exe --pass 1 --crf 22 --stats "D:\Capture\temp\.stats" --progress --keyint 250 --bframes 3 --qpmin 10 --qpmax 51 --no-psnr --mixed-refs --trellis 1 --ref 3 --filter -2,-1 --direct auto --vbv-maxrate 25000 --me umh --no-ssim --level 4.1 --weightb --b-pyramid --b-rdo --bime --analyse p8x8,b8x8,i4x4,i8x8,p4x4 --8x8dct --threads auto --thread-input --sar 14321678:10000000 --output
...
"D:\Capture\temp\movie.264" "D:\Capture\temp\movie.avs"
17:35:23 "E:\AutoMKV95b\exe\matroska\mkvmerge.exe" -o "D:\Capture\temp\7 Digital 2008-01-25_213205.mkv" --default-duration 0:25000/1000fps --track-name -1:"7 Digital 2008-01-25_213205" --aspect-ratio 0:1.780186 "D:\Capture\temp\movie.264" --title "Done with AutoMKV 0.95 beta 18 http://forum.doom9.org/showthread.php?p=854221 "
...
17:35:23 last=movie.Crop(2,4,716,572)
17:35:23 fixed_aspect = 1.251748
17:35:23 out_width = 716
17:35:23 out_height = 576

If I process the same video using Original (width & height) and Anamorphic ticked then AutoMKV calculates a fixed_aspect = 1.251748 (same as above) but uses a SAR of 1.7777778 and Muxing AR of 1.422222 and I get a PAL 720*576 5:4 output video. Very Bad :mad:

18:26:40 CRF encoding X264: E:\AutoMKV95b\exe\encoder\x264.exe --pass 1 --crf 22 --stats "D:\Capture\temp\.stats" --progress --keyint 250 --bframes 3 --qpmin 10 --qpmax 51 --no-psnr --mixed-refs --trellis 1 --ref 3 --filter -2,-1 --direct auto --vbv-maxrate 25000 --me umh --no-ssim --level 4.1 --weightb --b-pyramid --b-rdo --bime --analyse p8x8,b8x8,i4x4,i8x8,p4x4 --8x8dct --threads auto --thread-input --sar 17777778:10000000 --output "D:\Capture\temp\movie.264" "D:\Capture\temp\movie.avs"
...
18:27:27 "E:\AutoMKV95b\exe\matroska\mkvmerge.exe" -o "D:\Capture\temp\7 Digital 2008-01-25_213205.mkv" --default-duration 0:25000/1000fps --track-name -1:"7 Digital 2008-01-25_213205" --aspect-ratio 0:1.422222 "D:\Capture\temp\movie.264" --title "Done with AutoMKV 0.95 beta 18 http://forum.doom9.org/showthread.php?p=854221 "
...
18:27:27 last = movie
18:27:27 fixed_aspect = 1.251748
18:27:27 out_width = 720
18:27:27 out_height = 576


So when using Original it's coming up with wildly different SAR & Muxing AR values for a video that is essentially the same (4 pixels width difference).

So it's not the fixed_aspect calculation that seems to be an issue but the calculation of the SAR & Muxing AR values when Original is chosen.

Is there a variable used in the calculation of SAR & Muxing AR that is not set right when Original is selected?

Getting closer to source of the problem I think (hope).

buzzqw
26th January 2008, 15:02
yes...i hope we are getting very close

http://www.64k.it/andres/data/a/AutoMKV095beta20.exe

the --sar should be fixed and the muxing part.. i have many doubt.. but i hope i have understud varius reading

BHH

valnar
26th January 2008, 15:54
Something is wrong with beta20 and automatic width resolution. This may have happened for a few betas - I don't know. The resulting video is very choppy in MPC and looks terrible. Perhaps all this perfectionism in math is hurting compatibility? (I'm not an expert in AR, but unless anamorphic is ticked, I like the idea of keeping PAR 1:1)

Here are the avinaptic results between beta10 and beta20 with an NTSC 2.35 MAR movie.

Beta 10 - plays okay

[ About file ]

Name: Leon_b10_automatic.mkv
Date: 26/01/2008 09:20:42
Size: 8,050,553 bytes (7.678 MB)

[ Generic infos ]

Play duration: 00:01:14 (74.144 s)
Container type: matroska
Creation time: 26/01/2008 14:20:41 UTC
Number of streams: 2
Type of stream nr. 1: video (V_MPEG4/ISO/AVC) {Leon_automatic}
Type of stream nr. 2: audio (A_AC3)
Audio streams: 1
Muxing Application: libebml v0.7.7 + libmatroska v0.8.1
Writing Application: mkvmerge v2.1.0 ('Another Place To Fall') built on Aug 19 2007 13:40:07

[ Relevant data ]

Resolution: 704 x 304
Width: multiple of 32
Height: multiple of 16

[ Video track ]

Codec ID: V_MPEG4/ISO/AVC
Resolution: 704 x 304
Frame aspect ratio: 44:19 = 2.315789
Pixel aspect ratio: 1:1 = 1
Display aspect ratio: 44:19 = 2.315789
Framerate: 23.976024 fps

[ Audio track ]

Codec ID: A_AC3
Channels (container): 2
Sample rate: 48000 Hz
Mode: dolby

[ About H.264 encoding ]

User data: x264
User data: core 57 svn-714
User data: H.264/MPEG-4 AVC codec
User data: Copyleft 2005
User data: http://www.videolan.org/x264.html
User data: cabac=1
User data: ref=3
User data: deblock=1:-2:-1
User data: analyse=0x3:0x113
User data: me=umh
User data: subme=6
User data: brdo=1
User data: mixed_ref=1
User data: me_range=12
User data: chroma_me=1
User data: trellis=1
User data: 8x8dct=1
User data: cqm=0
User data: deadzone=21,11
User data: chroma_qp_offset=0
User data: threads=1
User data: nr=0
User data: decimate=0
User data: mbaff=0
User data: bframes=3
User data: b_pyramid=0
User data: b_adapt=1
User data: b_bias=0
User data: direct=3
User data: wpredb=1
User data: bime=1
User data: keyint=250
User data: keyint_min=25
User data: scenecut=40
User data: rc=crf
User data: crf=20.0
User data: rceq='blurCplx^(1-qComp)'
User data: qcomp=0.60
User data: qpmin=10
User data: qpmax=51
User data: qpstep=4
User data: ip_ratio=1.40
User data: pb_ratio=1.30
SPS id: 0
Profile: High@L3.1
Num ref frames: 4
Chroma format idc: YUV 4:2:0
PPS id: 0 (SPS: 0)
Entropy coding type: CABAC
Weighted prediction: No
Weighted bipred idc: B slices - implicit weighted prediction
8x8dct: Yes

[ Profile compliancy ]

Profile to check: Home Theater NTSC
Resolution: Ok
Framerate: 23.976024 <> 30
Warning: If you need a more complete report, then click on "DRF analysis"

This report was created by AVInaptic (18-11-2007) on 26 gen 2008, h 09:48:32


Beta 20 - plays choppy

[ About file ]

Name: Leon_b20_automatic.mkv
Date: 26/01/2008 09:33:35
Size: 5,946,998 bytes (5.671 MB)

[ Generic infos ]

Play duration: 00:01:14 (74.144 s)
Container type: matroska
Creation time: 26/01/2008 14:33:35 UTC
Number of streams: 2
Type of stream nr. 1: video (V_MPEG4/ISO/AVC) {Leon_b20_automatic}
Type of stream nr. 2: audio (A_AC3)
Audio streams: 1
Muxing Application: libebml v0.7.7 + libmatroska v0.8.1
Writing Application: mkvmerge v2.1.0 ('Another Place To Fall') built on Aug 19 2007 13:40:07

[ Relevant data ]

Resolution: 640 x 368
Width: multiple of 32
Height: multiple of 16

[ Video track ]

Codec ID: V_MPEG4/ISO/AVC
Resolution: 640 x 368
Frame aspect ratio: 40:23 = 1.739130
Pixel aspect ratio: 368:427 = 0.861826
Display aspect ratio: 640:427 = 1.498829
Framerate: 29.970000 fps

[ Audio track ]

Codec ID: A_AC3
Channels (container): 2
Sample rate: 48000 Hz
Mode: dolby

[ About H.264 encoding ]

SPS id: 0
Profile: Main@L5.1
Num ref frames: 2
Chroma format idc: YUV 4:2:0
PPS id: 0 (SPS: 0)
Entropy coding type: CAVLC
Weighted prediction: No
Weighted bipred idc: No
8x8dct: No

[ Profile compliancy ]

Profile to check: Home Theater NTSC
Resolution: Ok
Framerate: Ok
Warning: If you need a more complete report, then click on "DRF analysis"

This report was created by AVInaptic (18-11-2007) on 26 gen 2008, h 09:48:25


On an unrelated note, I specified Profile: High@L3.1 in both, but only beta 10 did that. Beta 20 was totally messed up, and the avinaptic report didn't even show my x264 parameters.

-Robert

buzzqw
26th January 2008, 16:20
just tryed with beta20... don't get any issue :confused:

i tested automatic width and ANAMORPHIC NOT CHECKED

and got this from avinaptic


[ About file ]

Name: 1201364332__aaa.mkv
Date: 26/01/2008 16:18:52
Size: 1,056,064 bytes (1.007 MB)

[ Generic infos ]

Play duration: 00:00:13 (12.6 s)
Container type: matroska
Creation time: 26/01/2008 15:18:52 UTC
Number of streams: 2
Type of stream nr. 1: video (V_MPEG4/ISO/AVC) {_aaa}
Type of stream nr. 2: audio (A_AAC (AAC Low complexity))
Audio streams: 1
Muxing Application: libebml v0.7.7 + libmatroska v0.8.1
Writing Application: mkvmerge v2.1.0 ('Another Place To Fall') built on Aug 19 2007 13:40:07

[ Relevant data ]

Resolution: 688 x 288
Width: multiple of 16
Height: multiple of 32

[ Video track ]

Codec ID: V_MPEG4/ISO/AVC
Resolution: 688 x 288
Frame aspect ratio: 43:18 = 2.388888
Pixel aspect ratio: 1:1 = 1
Display aspect ratio: 43:18 = 2.388888
Framerate: 25 fps


last=movie.Crop(0,72,720,432)
fixed_aspect = 2.370267
out_width = 688
out_height = 288
#resizestart

may i suggest to close automkv and retry ?

BHH

zacoz
26th January 2008, 16:22
Test with beta 20

Source: ..................... 720 * 576 with DAR 16:9
Set Width Resolution: ... Original
Anamorphic Encoding: ... Selected
Force PAR in avs: ......... Auto
Force Muxing DAR: ........ Automatic

fixed_aspect = 1.251748
SAR ............. 1.4222222 Good!
Muxing DAR ... 1.422222 Not Good (should be 1.777778)

Resulting Output Video: ...... 720 * 576 with DAR 5:4 Not Good (should be 16:9)

00:20:19 CRF encoding X264: E:\AutoMKV95b\exe\encoder\x264.exe --pass 1 --crf 22 --stats "D:\Capture\temp\.stats" --progress --keyint 250 --bframes 3 --qpmin 10 --qpmax 51 --no-psnr --mixed-refs --trellis 1 --ref 3 --filter -2,-1 --direct auto --vbv-maxrate 25000 --me umh --no-ssim --level 4.1 --weightb --b-pyramid --b-rdo --bime --analyse p8x8,b8x8,i4x4,i8x8,p4x4 --8x8dct --threads auto --thread-input --sar 14222222:10000000 --output "D:\Capture\temp\movie.264" "D:\Capture\temp\movie.avs"
...
00:21:00 "E:\AutoMKV95b\exe\matroska\mkvmerge.exe" -o "D:\Capture\temp\beta20 1.mkv" --default-duration 0:25000/1000fps --track-name -1:"beta20 1" --aspect-ratio -1:1.422222 "D:\Capture\temp\movie.264" --title "Done with AutoMKV 0.95 beta 20 http://forum.doom9.org/showthread.php?p=854221 "
...
00:21:01 #test3
00:21:01
00:21:01 movie = LeakKernelDeint(movie,order=1,sharp=true)
00:21:01 last = movie
00:21:01 fixed_aspect = 1.251748
00:21:01 out_width = 720
00:21:01 out_height = 576


----------------------------------------------------------------------------------------------------------------------------

Now I thought what if I test using Manual crop values of 0 - this should be exactly the same as using Original. But it's not. :confused: Using manual crop of 0 both the SAR and the Muxing DAR are correct.

Source: .............................. 720 * 576 with DAR 16:9
Set Width Resolution: ............ 720
Manual Crop: ....................... Selected
Top/Bottom/Left/Right CROP: .. 0
Anamorphic Encoding: ............ Selected
Force PAR in avs: ................. Auto
Force Muxing DAR: ................ Automatic

fixed_aspect = 1.251748
SAR .............. 1.4222222 Good!
Muxing DAR .... 1.777778 Good!

Resulting Output Video: ......... 720 * 576 with DAR 16:9 Good!


00:28:56 CRF encoding X264: E:\AutoMKV95b\exe\encoder\x264.exe --pass 1 --crf 22 --stats "D:\Capture\temp\.stats" --progress --keyint 250 --bframes 3 --qpmin 10 --qpmax 51 --no-psnr --mixed-refs --trellis 1 --ref 3 --filter -2,-1 --direct auto --vbv-maxrate 25000 --me umh --no-ssim --level 4.1 --weightb --b-pyramid --b-rdo --bime --analyse p8x8,b8x8,i4x4,i8x8,p4x4 --8x8dct --threads auto --thread-input --sar 14222222:10000000 --output "D:\Capture\temp\movie.264" "D:\Capture\temp\movie.avs"
...
00:29:42 "E:\AutoMKV95b\exe\matroska\mkvmerge.exe" -o "D:\Capture\temp\beta20 autocrop an.mkv" --default-duration 0:25000/1000fps --track-name -1:"beta20 autocrop an" --aspect-ratio -1:1.777778 "D:\Capture\temp\movie.264" --title "Done with AutoMKV 0.95 beta 20 http://forum.doom9.org/showthread.php?p=854221 "
...
00:29:42 #test3
00:29:42
00:29:42 movie = LeakKernelDeint(movie,order=1,sharp=true)
00:29:42 last = movie
00:29:42 crop(0,0,-0,-0)
00:29:42 fixed_aspect = 1.250000
00:29:42 out_width = 720
00:29:42 out_height = 576


----------------------------------------------------------------------------------------------------------------------------

Just to be sure what was happening by just setting 720 width and 0 crop I also ran a test using Automatic width and explicitly setting the width and height.

Source: ................................ 720 * 576 with DAR 16:9
Set Width Resolution: ............. Automatic (greyed out after selecting Use Manual Crop/Resize though)
Use Manual Crop/Resize: ......... Selected
Manual Crop: ........................ Selected
Top/Bottom/Left/Right CROP: .. 0
Specify Width/Height: ............ 720 x 576
Anamorphic Encoding: ........... Selected
Force PAR in avs: ................. Auto
Force Muxing DAR: ............... Automatic

fixed_aspect = 1.251748
SAR .............. 1.4222222 Good!
Muxing DAR .... 1.777778 Good!

Resulting Output Video: ...... 720 * 576 with DAR 16:9 Good!


00:59:49 CRF encoding X264: E:\AutoMKV95b\exe\encoder\x264.exe --pass 1 --crf 22 --stats "D:\Capture\temp\.stats" --progress --keyint 250 --bframes 3 --qpmin 10 --qpmax 51 --no-psnr --mixed-refs --trellis 1 --ref 3 --filter -2,-1 --direct auto --vbv-maxrate 25000 --me umh --no-ssim --level 4.1 --weightb --b-pyramid --b-rdo --bime --analyse p8x8,b8x8,i4x4,i8x8,p4x4 --8x8dct --threads auto --thread-input --sar 14222222:10000000 --output "D:\Capture\temp\movie.264" "D:\Capture\temp\movie.avs"
...
01:00:35 "E:\AutoMKV95b\exe\matroska\mkvmerge.exe" -o "D:\Capture\temp\beta20 720 576 man crop0 an.mkv" --default-duration 0:25000/1000fps --track-name -1:"beta20 720 576 man crop0 an" --aspect-ratio -1:1.777778 "D:\Capture\temp\movie.264" --title "Done with AutoMKV 0.95 beta 20 http://forum.doom9.org/showthread.php?p=854221 "
...
01:00:36 #test3
01:00:36
01:00:36 movie = LeakKernelDeint(movie,order=1,sharp=true)
01:00:36 last = movie
01:00:36 crop(0,0,-0,-0)
01:00:36 fixed_aspect = 1.250000
01:00:36 #resizestart
01:00:36
01:00:36 #test4
01:00:36
01:00:36 SimpleResize(720,576)


----------------------------------------------------------------------------------------------------------------------------

So why should asking for an "Original" output width and height (which are 720 * 576) give a different (incorrect) result than asking for a 720 * 576 output (which is the original width and height)?

buzzqw
26th January 2008, 16:39
@zacoz

... i think you found the key

i checked the code and found an old hugly hack from early beta

maybe this is the last one... (figer crossed)

http://www.64k.it/andres/data/a/AutoMKV095beta21.exe

:thanks:

BHH

zacoz
26th January 2008, 16:51
You're the Man ! :thanks:

"Original" width selection is now producing correct 16:9 output.

Sorry it took me so many posts to get around to the root of the issue....it's been quite a learning experience.

I'll go run some tests on other width selections just to be thorough.

buzzqw
26th January 2008, 16:57
Thanks Zacoz!

many general usage test must be done, but if these issues (about anamorphic/original) are resolved.. well it's a goood work!

thanks again zacoz and valnar!

BHH

valnar
26th January 2008, 17:14
just tryed with beta20... don't get any issue :confused:

i tested automatic width and ANAMORPHIC NOT CHECKED


@buzzhq

OK, I tried automatic with the 2_Pass_Balanced preset and it worked fine, but i never use that.
[ Video track ]

Codec ID: V_MPEG4/ISO/AVC
Resolution: 704 x 304
Frame aspect ratio: 44:19 = 2.315789
Pixel aspect ratio: 1:1 = 1
Display aspect ratio: 44:19 = 2.315789
Framerate: 23.976024 fps

When I used my own custom preset, it does not work. It used to..
[ Video track ]

Codec ID: V_MPEG4/ISO/AVC
Resolution: 640 x 368
Frame aspect ratio: 40:23 = 1.739130
Pixel aspect ratio: 368:427 = 0.861826
Display aspect ratio: 640:427 = 1.498829
Framerate: 29.970000 fps

Here is the XML preset I use for DVD's (CQ and H@L3.1)
<?xml version="1.0"?>
<VideoProfile xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns:xsd="http://www.w3.org/2001/XMLSchema">
<Name>_CQ_Robert_Q2_eq(crf)</Name>
<Settings xsi:type="x264Settings">
<EncodingMode>9</EncodingMode>
<BitrateQuantizer>18</BitrateQuantizer>
<KeyframeInterval>250</KeyframeInterval>
<NbBframes>3</NbBframes>
<MinQuantizer>10</MinQuantizer>
<MaxQuantizer>51</MaxQuantizer>
<Turbo>false</Turbo>
<V4MV>false</V4MV>
<QPel>false</QPel>
<Trellis>1</Trellis>
<CreditsQuantizer>40</CreditsQuantizer>
<Logfile>.stats</Logfile>
<CustomEncoderOptions />
<FourCC>1</FourCC>
<Zones />
<MaxNumberOfPasses>3</MaxNumberOfPasses>
<NbThreads>1</NbThreads>
<NoDCTDecimate>true</NoDCTDecimate>
<PSNRCalculation>false</PSNRCalculation>
<noFastPSkip>false</noFastPSkip>
<NoiseReduction>0</NoiseReduction>
<MixedRefs>true</MixedRefs>
<X264Trellis>1</X264Trellis>
<NbRefFrames>3</NbRefFrames>
<AlphaDeblock>-2</AlphaDeblock>
<BetaDeblock>-1</BetaDeblock>
<SubPelRefinement>6</SubPelRefinement>
<MaxQuantDelta>4</MaxQuantDelta>
<TempQuantBlur>0</TempQuantBlur>
<BframePredictionMode>3</BframePredictionMode>
<VBVBufferSize>-1</VBVBufferSize>
<VBVMaxBitrate>14000</VBVMaxBitrate>
<METype>2</METype>
<MERange>12</MERange>
<MinGOPSize>25</MinGOPSize>
<Profile>2</Profile>
<Level>3.1</Level>
<IPFactor>1.4</IPFactor>
<PBFactor>1.3</PBFactor>
<ChromaQPOffset>0</ChromaQPOffset>
<VBVInitialBuffer>0.9</VBVInitialBuffer>
<BitrateVariance>1.0</BitrateVariance>
<QuantCompression>0.6</QuantCompression>
<TempComplexityBlur>20</TempComplexityBlur>
<TempQuanBlurCC>0.5</TempQuanBlurCC>
<SCDSensitivity>40</SCDSensitivity>
<BframeBias>0</BframeBias>
<Deblock>true</Deblock>
<Cabac>true</Cabac>
<WeightedBPrediction>true</WeightedBPrediction>
<AdaptiveBFrames>true</AdaptiveBFrames>
<BFramePyramid>false</BFramePyramid>
<BRDO>true</BRDO>
<biME>true</biME>
<ChromaME>true</ChromaME>
<P8x8mv>true</P8x8mv>
<B8x8mv>true</B8x8mv>
<I4x4mv>true</I4x4mv>
<I8x8mv>true</I8x8mv>
<P4x4mv>false</P4x4mv>
<AdaptiveDCT>true</AdaptiveDCT>
<Width>false</Width>
<Sar>false</Sar>
<aq-strength>false</aq-strength>
<aq-sensitivity>false</aq-sensitivity>
<deadzone-inter>false</deadzone-inter>
<deadzone-intra>false</deadzone-intra>
<fpel-cmp>false</fpel-cmp>
<me-prepass>false</me-prepass>
<SSIMCalculation>true</SSIMCalculation>
<Interlaced>false</Interlaced>
<Lossless>false</Lossless>
<QuantizerMatrix />
<QuantizerMatrixType>0</QuantizerMatrixType>
</Settings>
</VideoProfile>

-Robert

Edit: this is beta 20. I didn't try beta 21 yet.

zacoz
26th January 2008, 17:16
BTW, have you had opportunity to think on Encoder888's suggestion to automatically set Anamorphic flag where Source is Anamorphic and "Original" width is selected?

As it presently stands if you use an Anamorphic Source and select "Original" but don't go to Advanced to select Anamorphic you'll end up with a 5:4 encode. I can't think of any reason why a user would want this to occur. It also has potential to trip up quite a few new users who don't venture near the Advanced tab (as is normally suggested to them here in the forum until they know what they're doing).

An approach I might suggest is to not try to test for this during setup rather adjust it as part of the processing:

1. User clicks Start Encoding
2. Analysis phase established DAR of source material
3. Test if "Original" has been selected for output width go to step 4 (otherwise step 7)
4. If source DAR > source width / source height then go to step 5 (otherwise step 7)
5. Enter line into log indicating "Anamorphic source identified and Original width selected therefore Anamorphic encoding activated"
6. Silently (i.e. don't change GUI) set Anamorphic Encoding to true
7. Continue processing as normal

buzzqw
26th January 2008, 17:27
@valnar

quick dumb question.. why using a crf profile AND automatic width ?

@zacoz

As it presently stands if you use an Anamorphic Source and select "Original" but don't go to Advanced to select Anamorphic you'll end up with a 5:4 encode. I can't think of any reason why a user would want this to occur

old 4/3 tv ?

1. User clicks Start Encoding
2. Analysis phase established DAR of source material
3. Test if "Original" has been selected for output width go to step 4 (otherwise step 7)
4. If source DAR > source width / source height then go to step 5 (otherwise step 7)
5. Enter line into log indicating "Anamorphic source identified and Original width selected therefore Anamorphic encoding activated"
6. Silently (i.e. don't change GUI) set Anamorphic Encoding to true
7. Continue processing as normal

... don't know.. i mean i thing it's good.. but... don't know..

should be "erroneus" full 720*576(or 480) at 5:4 ?

"must" be every 16/9 full resolution encoded at anamorphic ?

don't know.. sorry .. i will look around

BHH

valnar
26th January 2008, 17:43
BTW, have you had opportunity to think on Encoder888's suggestion to automatically set Anamorphic flag where Source is Anamorphic and "Original" width is selected?

As it presently stands if you use an Anamorphic Source and select "Original" but don't go to Advanced to select Anamorphic you'll end up with a 5:4 encode. I can't think of any reason why a user would want this to occur.
I agree.


@buzzhq
quick dumb question.. why using a crf profile AND automatic width ?

You're right. No reason except that I'm just testing. But I think I found the problem. When I select my custom profile, it resets the width to original and resize filter to NONE (why?). When I change it back to Automatic, I did not also change the resize away from NONE. It also seems to force ffmpeg when I choose this combination. My fault... but it would nice if AutoMKV prevented such stupidity.

-Robert

Honeyko
26th January 2008, 23:55
BTW, have you had opportunity to think on Encoder888's suggestion to automatically set Anamorphic flag where Source is Anamorphic and "Original" width is selected?

As it presently stands if you use an Anamorphic Source and select "Original" but don't go to Advanced to select Anamorphic you'll end up with a 5:4 encode.

Well, we don't the 5:4 encodes, but it shouldn't be automatically assumed that the user is looking for an anamorphic 720x576 PAR file if he selects "Original" width. He might be looking for smaller 720x400 file-size, particularly if he doesn't have "Don't Care" chosen for file-size.

(Something I'd like to see one of these days is a "checklist" that would pop up when you Start Encoding or Add to Queue which details everything that will happen to the file in terms of PAR and DAR. The user can then immediately see if anything is screwed up, or if he doesn't like way one axis is rounding to MOD16.)

Honeyko
27th January 2008, 04:03
b21 BUG:

Problems remain with the Start buttons.

1. "Start Encoding" a project
2. "Add to Queue" button grays out; "Start Encoding" remains clickable
3. click "Start Encoding" again

....and a second project begins, using the same Temp folder, while the first is still in progress, causing file loss and eventually a crash of AutoMKV (and the encoder may continue running, hidden, until terminated with the Task Manager).

Suggestion: ALL encodes are shunted through a Queue list (i.e., Start Encoding will just set up a single queued item, then start it. If there are already items in the Queue list, a pop up will inform the user).

==//==

BUG: ...the Add to Queue button appears to be permanently grayed...at least I haven't seen it enabled yet it b21.

Honeyko
27th January 2008, 04:33
Suggestion: Accept (and encode properly without "rounding up" or having the encoder halt or crash) odd-numbered crops values if they come in dimensional pairs -- top and bottom are both odd, for example.

(I'm not certain how this would be done....adding a "fake" padding line to either side prior to cropping...?)

zacoz
27th January 2008, 08:47
Well, we don't the 5:4 encodes, but it shouldn't be automatically assumed that the user is looking for an anamorphic 720x576 PAR file if he selects "Original" width. He might be looking for smaller 720x400 file-size, particularly if he doesn't have "Don't Care" chosen for file-size.

If you select "Original" in the width dropdown though you are actually enforcing original width AND HEIGHT. If you want a 720x400 pixel size then you must select either width "720" or width "Autocrop" (and not have ticked Anamorphic of course).

Below is a post I wrote last night but couldn't post due to server database errors. Take from it what you will in light of other posts since:


old 4/3 tv ?

But 5:4 isn't right for an old 4:3 TV anyway is it? As it stands you would end up with your image all squashed horizontally, instead of having a correctly proportioned image which the TV would display either with black bars top and bottom or by chopping off the left and right sides. The 4:3 TV would be doing one of these to the original source too.

"must" be every 16/9 full resolution encoded at anamorphic ?

My understanding is that a video is anamorphic if the ratio of the physical display width to it's physical display height is greater than the ratio of it's pixel width to it's pixel height. I.e. for it to display the picture in the correct proportions it has to be stretched horizontally (or squashed vertically).
Of course if you resize it so that the pixel width/height ratio equals the display width/height then it's not anamorphic.
So this issue is very specifically linked only to where the source is anamorphic and there is no change being made to width and height (original selected for width).

don't know.. sorry .. i will look around
BHH
I'm hoping some others here can provide some input to the discussion too as I'm looking at all this on the basis of what goes in and what comes out and whether it displays correctly by default. I'll be the first to admit I don't understand the intricacies and may still be overlooking a genuine reason for not forcing anamorphic in these cases.

Honeyko
27th January 2008, 10:33
"must" be every 16/9 full resolution encoded at anamorphic?My understanding is that a video is anamorphic if the ratio of the physical display width to it's physical display height is greater than the ratio of it's pixel width to it's pixel height. I.e. for it to display the picture in the correct proportions it has to be stretched horizontally (or squashed vertically).

A non-16x9 example is an NTSC DVD of a cartoon series encoded at 720x480, and which displays at 720x540 for proper 4x3. (In this case the vertical is squashed).

buzzqw
27th January 2008, 11:42
@Honeyko

Problems remain with the Start buttons.

Fixed

Suggestion: Accept (and encode properly without "rounding up" or having the encoder halt or crash) odd-numbered crops values if they come in dimensional pairs -- top and bottom are both odd, for example.

i added a nice (i hope) coloring system on manual crop, and now added a colered check on manual width and height values

i hope it's enough, since i don't want to change manual values users inserted

@ALL

about all forcing and not forcing anamorphic hype around, i will leave as is.
I will just add on statusbar a suggestion, on case of no resize/no crop, of checking anamorphic encoding.

OK?

also.. some test must be done even on mp4 container... ;)

BHH

SeeMoreDigital
27th January 2008, 12:10
@ALL

about all forcing and not forcing anamorphic hype around, i will leave as is.
I will just add on statusbar a suggestion, on case of no resize/no crop, of checking anamorphic encoding.

BHH Hi Buzz,

I'll take a more detailed look at this over the next few days and come up with some suggestions ;)

buzzqw
27th January 2008, 12:42
@SeeMoreDigital

:thanks:

my knowledge isn't enough, i have best hope on you!

and.. btw use the latest beta (now 21)

thanks again SeeMoreDigital!

BHH