Auto Gordian Knot: current version 2.55

[rmagere]

Ok I understand.

By the way do you still need me to hang on the source vobs or can I get rid of them? (I suspect I can but wouldn't want to just do away with them and you still needing some testing).

EDIT:
|Referring to message below.
|
|Yep deleted them
|
V

[len0x]

It should be OK to delete them now.

[kilivipin]

Just a quick question. I installed the lkatest stable and then updated with the latest beta, 1.73, I believe. I just downloaded 1.75. Can I just run the installer and not uninstall the previous beta?

[len0x]

Originally posted by kilivipin
Can I just run the installer and not uninstall the previous beta?

yes

[kilivipin]

Thank you. You rule! Great program. One of the best really.

[BigDid]

Originally posted by BigDid
... Wondering if using the 1.1 beta and checking the vbv box before doing a normal encode with AGK would do it?
Did [/B]

I am trying the above actually, any links to tools for bitrate statistics?
I would need a tool (easy to use if possible) that reads an avi/xvid file and gives/summarize the bitrates infos for that file: avg, min, max etc..

So far I have found:
- DRFAnalyser (which gives me graphics but no summary)
- OSD of the MPC Player (wich gives me instant input bitrate only, may vary from one play to another)
- FileInfo from VDubMod (gives only frame size for min/avg/max)

Thanks for your infos.

Did

[quake74]

Hi,

First, great piece of software len0x, thanks.
Second, there was a bunch of dvd on amazon.co.uk for 6-7 pounds, and I bougth quite a few of time, time to backup!
I am encoding a sitcom (Coupling). I want to fit 6 episodes per CD, which gives me a target size of 116Mb. Using audio Auto (I guess 128kbs) I get:

[11/10/2004 9:38:21 PM] AutoGK 1.75b
[11/10/2004 9:38:21 PM] OS: WinXP (5.1.2600).2
[11/10/2004 9:38:21 PM] Job started.
[11/10/2004 9:38:21 PM] Input dir: F:\Coupling\VIDEO_TS_2
[11/10/2004 9:38:21 PM] Output file: F:\Coupling\VIDEO_TS_2\Coupling 1x02 Size Matters.avi
[11/10/2004 9:38:21 PM] Audio: English
[11/10/2004 9:38:21 PM] Subtitles: none
[11/10/2004 9:38:21 PM] Codec: XviD
[11/10/2004 9:38:21 PM] Target size: 116Mb
[11/10/2004 9:38:21 PM] Started encoding.
[11/10/2004 9:38:21 PM] Demuxing and indexing.
[11/10/2004 9:39:14 PM] Processing file: F:\Coupling\VIDEO_TS_2\VTS_02_1.VOB
[11/10/2004 9:39:14 PM] Processing file: F:\Coupling\VIDEO_TS_2\VTS_02_2.VOB
[11/10/2004 9:39:14 PM] Source aspect ratio: 16:9
[11/10/2004 9:39:14 PM] Source resolution: 720x576
[11/10/2004 9:39:14 PM] Found PAL source.
[11/10/2004 9:39:14 PM] Analyzing source.
[11/10/2004 9:41:03 PM] Source is considered to be interlaced.
[11/10/2004 9:41:03 PM] Found 43593 frames
[11/10/2004 9:41:03 PM] Encoding audio.
[11/10/2004 9:46:03 PM] Audio size: 28,623,344 bytes (27.30 Mb)
[11/10/2004 9:46:03 PM] Overhead: 1,743,744 bytes (1.66 Mb)
[11/10/2004 9:46:03 PM] Video size: 91,267,728 bytes (87.04 Mb)
[11/10/2004 9:46:03 PM] Running compressibility test.
[11/10/2004 9:49:16 PM] Duration was: 3 minutes 13 seconds
[11/10/2004 9:49:16 PM] Speed was: 11.27 fps.
[11/10/2004 9:49:16 PM] Compressibility percentage is: 31.75
[11/10/2004 9:49:16 PM] Chosen resolution is: 608x352 ( AR: 1.73 )
[11/10/2004 9:49:16 PM] Predicted comptest value is: 34.94
[11/10/2004 9:49:16 PM] Running first pass.

It's still running first pass now. But why didn't it scale down a little? The comptest value is really low! I already encoded the first episode, it decided to use the same resolution and it's too blocky. Am I doing something wrong?

[quake74]

So, the second episode got compressed at 608 width, very very ugly. I tried the third episode (they are all very similar and should have similar compressibility) with a max width of 512 and AutoGK decided for a width of 400 (which is about right at these low bitrates). Here it is:

[11/10/2004 11:13:13 PM] AutoGK 1.75b
[11/10/2004 11:13:13 PM] OS: WinXP (5.1.2600).2
[11/10/2004 11:13:13 PM] Job started.
[11/10/2004 11:13:13 PM] Input dir: F:\Coupling\VIDEO_TS_3
[11/10/2004 11:13:13 PM] Output file: F:\Coupling\VIDEO_TS_3\Coupling 1x03 Sex Death and Nudity.avi
[11/10/2004 11:13:13 PM] Audio: English
[11/10/2004 11:13:13 PM] Subtitles: none
[11/10/2004 11:13:13 PM] Codec: XviD
[11/10/2004 11:13:13 PM] Target size: 116Mb
[11/10/2004 11:13:13 PM] Started encoding.
[11/10/2004 11:13:13 PM] Demuxing and indexing.
[11/10/2004 11:14:07 PM] Processing file: F:\Coupling\VIDEO_TS_3\VTS_02_1.VOB
[11/10/2004 11:14:07 PM] Processing file: F:\Coupling\VIDEO_TS_3\VTS_02_2.VOB
[11/10/2004 11:14:07 PM] Source aspect ratio: 16:9
[11/10/2004 11:14:07 PM] Source resolution: 720x576
[11/10/2004 11:14:07 PM] Found PAL source.
[11/10/2004 11:14:07 PM] Analyzing source.
[11/10/2004 11:16:04 PM] Source is considered to be interlaced.
[11/10/2004 11:16:04 PM] Found 43826 frames
[11/10/2004 11:16:04 PM] Encoding audio.
[11/10/2004 11:21:41 PM] Audio size: 28,797,536 bytes (27.46 Mb)
[11/10/2004 11:21:41 PM] Overhead: 1,753,088 bytes (1.67 Mb)
[11/10/2004 11:21:41 PM] Video size: 91,084,192 bytes (86.86 Mb)
[11/10/2004 11:21:41 PM] Running compressibility test.
[11/10/2004 11:24:35 PM] Duration was: 2 minutes 51 seconds
[11/10/2004 11:24:35 PM] Speed was: 12.79 fps.
[11/10/2004 11:24:37 PM] Compressibility percentage is: 44.74
[11/10/2004 11:24:37 PM] Chosen resolution is: 400x224 ( AR: 1.79 )
[11/10/2004 11:24:37 PM] Predicted comptest value is: 73.63
[11/10/2004 11:24:37 PM] Running first pass.

The first pass is going on, while I am going to sleep. It is gonna compress another 3 episodes and I'll report back.

[BigDid]

@ Sharktooth, thanks for the infos found in the xvid thread
@ LenOx, hope it can help when migrating to xvid 1.1

First of all I was not able to reproduce the stall of my standalone, I suspect some mysterious correlations between the xvid high bitrate, the 2gb file size, the dvd-r media and/or the 192 vbr audio :(

As I had already started I went to the end of the tests to give the results (still seeking a bitrate stats tool :p )
The initial movie was 640 wide (LOTR2) but vobs unavailable at the moment to make a short split so I used LOTR3 which also stalled (same 192 vbr audio, same 2gb size & 624 wide).

I made a 12866 frames/Pal split and increased the size of the encode seeking the stall. I ended with a full size, 95% quality and still no stall :confused:
So I made the tests with a 80Mb size, resolution 672*288, same length (12866 frames), same AGK rev: 1.74, +- same encoded files size :min 81944 max 81968, same bitrate: 1300kbps, same machine: laptop 1.2 Gb celeron.

Find the summary of the results here:

http://img5.exs.cx/img5/9263/Ecran4.png

or here with the snapshots:http://forum.doom9.org/showthread.php?s=&postid=568408#post568408

Any power-user, senior or master feel free to detail the results as I don't feel qualified to do so. If I have the time I will try to post snapshots of the DRFanalyser graph for each input.

Did

/Edit: Still not familiar with picture attachment, if it does'nt show up I will try to re-attach it in this post or a new one/

[Alwayscamp]

i tried again with v 1.74 and got the same results .. before i finished i saw you got 1.75 up. i haven't tried with it although i can if you think you might have changed somthing that would a;ter my results.
here are the log files

divx_agk.log

[11/8/2003 2:11:48 PM] AutoGK 1.74b
[11/8/2003 2:11:48 PM] OS: WinXP (5.1.2600).2
[11/8/2003 2:11:48 PM] Job started.
[11/8/2003 2:11:48 PM] Input dir: C:\Documents and Settings\Administrator\Desktop\titan
[11/8/2003 2:11:48 PM] Output file: C:\Documents and Settings\Administrator\Desktop\titan\Divx.avi
[11/8/2003 2:11:48 PM] Audio: English
[11/8/2003 2:11:48 PM] Subtitles: none
[11/8/2003 2:11:48 PM] Codec: DivX
[11/8/2003 2:11:48 PM] Target size: 700Mb
[11/8/2003 2:11:48 PM] Started encoding.
[11/8/2003 2:11:48 PM] Demuxing and indexing.
[11/8/2003 2:14:20 PM] Processing file: C:\Documents and Settings\Administrator\Desktop\titan\VTS_01_1.VOB
[11/8/2003 2:14:20 PM] Processing file: C:\Documents and Settings\Administrator\Desktop\titan\VTS_01_2.VOB
[11/8/2003 2:14:20 PM] Processing file: C:\Documents and Settings\Administrator\Desktop\titan\VTS_01_3.VOB
[11/8/2003 2:14:20 PM] Source aspect ratio: 4:3
[11/8/2003 2:14:20 PM] Source resolution: 720x480
[11/8/2003 2:14:20 PM] Found NTSC source.
[11/8/2003 2:14:21 PM] Analyzing source.
[11/8/2003 2:29:50 PM] Source has percentage of interlacing in motion areas: 47.26
[11/8/2003 2:29:50 PM] Source has percentage of telecined patterns: 99.80
[11/8/2003 2:29:50 PM] Source has percentage of progressive patterns: 0.20
[11/8/2003 2:29:50 PM] Source has percentage of interlaced patterns: 0.00
[11/8/2003 2:29:50 PM] Source is considered to be FILM.
[11/8/2003 2:29:50 PM] Found 272786 frames
[11/8/2003 2:29:51 PM] Encoding audio.
[11/8/2003 3:03:39 PM] Audio size: 177,198,416 bytes (168.99 Mb)
[11/8/2003 3:03:39 PM] Overhead: 17,597,696 bytes (16.78 Mb)
[11/8/2003 3:03:39 PM] Video size: 539,207,088 bytes (514.23 Mb)
[11/8/2003 3:03:39 PM] Target bitrate is: 379kbps
[11/8/2003 3:03:39 PM] Running compressibility test.
[11/8/2003 3:03:39 PM] Writing the following script to C:\Documents and Settings\Administrator\Desktop\titan\agk_tmp\Divx_comptest.avs
===========================================================
LoadPlugin("C:\PROGRA~1\AutoGK\DGMPGDec\DGDecode.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\autocrop.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\decomb.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\RemoveGrain.dll")

movie = mpeg2source("C:\Documents and Settings\Administrator\Desktop\titan\agk_tmp\Divx.d2v")
cropclip = autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=0,threshold=34,samplestartframe=0)
fixed_aspect = 0.888888888888889
c_width = width(cropclip)
c_height = round(height(cropclip) / fixed_aspect)
input_par = float(c_width)/float(c_height)
input_par = input_par > 1.4 ? input_par : (4.0/3.0)
out_width = 640
out_height = round(float(out_width) / input_par)
hmod = out_height - (floor(out_height / 16 ) * 16)
out_height = (hmod > 4) ? (out_height + (16 - hmod)) : (out_height - hmod)
new_aspect = (float(out_width) / float(out_height)) / fixed_aspect
Telecide(movie,order=1,guide=1,post=2).Decimate(mode=0)
autocrop(mode=0,wmultof=4,hmultof=4,samples=10,aspect=new_aspect,threshold=34,samplestartframe=0)
LanczosResize(out_width,out_height)
RemoveGrain(mode=2)
SelectRangeEvery(280,14)
===========================================================
[11/8/2003 3:16:19 PM] Duration was: 12 minutes 36 seconds
[11/8/2003 3:16:19 PM] Speed was: 18.03 fps.
[11/8/2003 3:16:20 PM] Compressibility percentage is: 17.44
[11/8/2003 3:16:20 PM] Using softer resizer.
[11/8/2003 3:16:20 PM] Chosen resolution is: 320x144 ( AR: 2.22 )
[11/8/2003 3:16:20 PM] Predicted comptest value is: 69.77
[11/8/2003 3:16:20 PM] Running first pass.
[11/8/2003 3:16:20 PM] Writing the following script to C:\Documents and Settings\Administrator\Desktop\titan\agk_tmp\Divx_movie.avs
===========================================================
LoadPlugin("C:\PROGRA~1\AutoGK\DGMPGDec\DGDecode.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\autocrop.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\decomb.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\RemoveGrain.dll")

movie = mpeg2source("C:\Documents and Settings\Administrator\Desktop\titan\agk_tmp\Divx.d2v")
cropclip = autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=0,threshold=34,samplestartframe=0)
fixed_aspect = 0.888888888888889
c_width = width(cropclip)
c_height = round(height(cropclip) / fixed_aspect)
input_par = float(c_width)/float(c_height)
input_par = input_par > 1.4 ? input_par : (4.0/3.0)
out_width = 320
out_height = round(float(out_width) / input_par)
hmod = out_height - (floor(out_height / 16 ) * 16)
out_height = (hmod > 4) ? (out_height + (16 - hmod)) : (out_height - hmod)
new_aspect = (float(out_width) / float(out_height)) / fixed_aspect
Telecide(movie,order=1,guide=1,post=2).Decimate(mode=0)
autocrop(mode=0,wmultof=4,hmultof=4,samples=10,aspect=new_aspect,threshold=34,samplestartframe=0)
BicubicResize(out_width,out_height,0,0.5)
RemoveGrain(mode=2)
===========================================================
[11/8/2003 5:04:27 PM] Duration was: 1 hour, 48 minutes 3 seconds
[11/8/2003 5:04:27 PM] Speed was: 42.07 fps.
[11/8/2003 5:04:27 PM] Running second pass.
[11/8/2003 6:54:19 PM] Duration was: 1 hour, 49 minutes 46 seconds
[11/8/2003 6:54:20 PM] Speed was: 41.42 fps.
[11/8/2003 6:54:20 PM] Job finished.


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

XviD_agk.log

[11/8/2003 6:54:22 PM] AutoGK 1.74b
[11/8/2003 6:54:22 PM] OS: WinXP (5.1.2600).2
[11/8/2003 6:54:22 PM] Job started.
[11/8/2003 6:54:22 PM] Input dir: C:\Documents and Settings\Administrator\Desktop\titan
[11/8/2003 6:54:22 PM] Output file: C:\Documents and Settings\Administrator\Desktop\titan\XviD.avi
[11/8/2003 6:54:22 PM] Audio: English
[11/8/2003 6:54:22 PM] Subtitles: none
[11/8/2003 6:54:22 PM] Codec: XviD
[11/8/2003 6:54:22 PM] Target size: 700Mb
[11/8/2003 6:54:22 PM] Started encoding.
[11/8/2003 6:54:22 PM] Demuxing and indexing.
[11/8/2003 6:56:37 PM] Processing file: C:\Documents and Settings\Administrator\Desktop\titan\VTS_01_1.VOB
[11/8/2003 6:56:37 PM] Processing file: C:\Documents and Settings\Administrator\Desktop\titan\VTS_01_2.VOB
[11/8/2003 6:56:37 PM] Processing file: C:\Documents and Settings\Administrator\Desktop\titan\VTS_01_3.VOB
[11/8/2003 6:56:37 PM] Source aspect ratio: 4:3
[11/8/2003 6:56:37 PM] Source resolution: 720x480
[11/8/2003 6:56:37 PM] Found NTSC source.
[11/8/2003 6:56:38 PM] Analyzing source.
[11/8/2003 7:09:17 PM] Source has percentage of interlacing in motion areas: 47.26
[11/8/2003 7:09:17 PM] Source has percentage of telecined patterns: 99.80
[11/8/2003 7:09:17 PM] Source has percentage of progressive patterns: 0.20
[11/8/2003 7:09:17 PM] Source has percentage of interlaced patterns: 0.00
[11/8/2003 7:09:17 PM] Source is considered to be FILM.
[11/8/2003 7:09:17 PM] Found 272786 frames
[11/8/2003 7:09:17 PM] Encoding audio.
[11/8/2003 7:45:20 PM] Audio size: 177,198,416 bytes (168.99 Mb)
[11/8/2003 7:45:20 PM] Overhead: 10,911,488 bytes (10.41 Mb)
[11/8/2003 7:45:20 PM] Video size: 545,893,296 bytes (520.60 Mb)
[11/8/2003 7:45:20 PM] Running compressibility test.
[11/8/2003 7:45:20 PM] Writing the following script to C:\Documents and Settings\Administrator\Desktop\titan\agk_tmp\XviD_comptest.avs
===========================================================
LoadPlugin("C:\PROGRA~1\AutoGK\DGMPGDec\DGDecode.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\autocrop.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\decomb.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\RemoveGrain.dll")

movie = mpeg2source("C:\Documents and Settings\Administrator\Desktop\titan\agk_tmp\XviD.d2v")
cropclip = autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=0,threshold=34,samplestartframe=0)
fixed_aspect = 0.888888888888889
c_width = width(cropclip)
c_height = round(height(cropclip) / fixed_aspect)
input_par = float(c_width)/float(c_height)
input_par = input_par > 1.4 ? input_par : (4.0/3.0)
out_width = 640
out_height = round(float(out_width) / input_par)
hmod = out_height - (floor(out_height / 16 ) * 16)
out_height = (hmod > 4) ? (out_height + (16 - hmod)) : (out_height - hmod)
new_aspect = (float(out_width) / float(out_height)) / fixed_aspect
Telecide(movie,order=1,guide=1,post=2).Decimate(mode=0)
autocrop(mode=0,wmultof=4,hmultof=4,samples=10,aspect=new_aspect,threshold=34,samplestartframe=0)
LanczosResize(out_width,out_height)
RemoveGrain(mode=2)
SelectRangeEvery(300,15)
===========================================================
[11/8/2003 8:03:10 PM] Duration was: 17 minutes 49 seconds
[11/8/2003 8:03:10 PM] Speed was: 12.76 fps.
[11/8/2003 8:03:11 PM] Compressibility percentage is: 24.52
[11/8/2003 8:03:11 PM] Chosen resolution is: 608x272 ( AR: 2.24 )
[11/8/2003 8:03:11 PM] Predicted comptest value is: 27.32
[11/8/2003 8:03:11 PM] Running first pass.
[11/8/2003 8:03:11 PM] Writing the following script to C:\Documents and Settings\Administrator\Desktop\titan\agk_tmp\XviD_movie.avs
===========================================================
LoadPlugin("C:\PROGRA~1\AutoGK\DGMPGDec\DGDecode.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\autocrop.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\decomb.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\RemoveGrain.dll")

movie = mpeg2source("C:\Documents and Settings\Administrator\Desktop\titan\agk_tmp\XviD.d2v")
cropclip = autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=0,threshold=34,samplestartframe=0)
fixed_aspect = 0.888888888888889
c_width = width(cropclip)
c_height = round(height(cropclip) / fixed_aspect)
input_par = float(c_width)/float(c_height)
input_par = input_par > 1.4 ? input_par : (4.0/3.0)
out_width = 608
out_height = round(float(out_width) / input_par)
hmod = out_height - (floor(out_height / 16 ) * 16)
out_height = (hmod > 4) ? (out_height + (16 - hmod)) : (out_height - hmod)
new_aspect = (float(out_width) / float(out_height)) / fixed_aspect
Telecide(movie,order=1,guide=1,post=2).Decimate(mode=0)
autocrop(mode=0,wmultof=4,hmultof=4,samples=10,aspect=new_aspect,threshold=34,samplestartframe=0)
LanczosResize(out_width,out_height)
RemoveGrain(mode=2)
===========================================================
[11/8/2003 11:01:14 PM] Duration was: 2 hours 58 minutes 2 seconds
[11/8/2003 11:01:15 PM] Speed was: 25.54 fps.
[11/8/2003 11:01:16 PM] Expected quality of first pass size: 27.25
[11/8/2003 11:01:16 PM] Running second pass.
[11/9/2003 2:38:38 AM] Duration was: 3 hours 37 minutes 20 seconds
[11/9/2003 2:38:38 AM] Speed was: 20.92 fps.
[11/9/2003 2:38:38 AM] Job finished.

[quake74]

Well, looking at alwayscamp posts, I bet it's the same bug. I dowloaded divx and will try as soon as this encode finishes. For the rest of my encodes, using maximum width made autogk choose appropriate resolutions
(max width 512 it chose 400 with compressibility 73.63, max width 448 it chose 448 with compressibility 74.51) while, using auto width or min width 400, autogk always chose 608 with low compressibilities (40.21, 31.81, 34.94, 31.80).

[BigDid]

Originally posted by BigDid

Here is a summary of the results:


This one got through, see attachment.

Also the snapshots from DRFAnalyser (hope I have'nt mess up the order :rolleyes: ) here :http://img83.exs.cx/img83/4153/Complet2.th.png (http://img83.exs.cx/my.php?loc=img83&image=Complet2.png)

Did

[Taurus]

Originally posted by BigDid
Hope this one will get through

You can host your pictures at Imageshack (http://www.imageshack.us/)
It's no big hazzle to display them right here at the forum;)

Cheers

Taurus

[len0x]

I found it and fixed, but still checking some other bits for other possible bugs :)

[daberti]

Originally posted by len0x
I found it and fixed, but still checking some other bits for other possible bugs :)

Just one question though: the bug you're talking about affect only 1.7x tree or also 1.60?

[len0x]

@daberti

1.7x mainly. There were some bits wrong in 1.60 as well but they rarely come up (or you don't notice them because you don't check the quality logs after encoding: there might have been large difference between predicted and expected quality).

[quake74]

Thanks len0x, waiting for the new version for some more testing ;)

[BigDid]

Originally posted by Taurus
You can host your pictures at Imageshack (http://www.imageshack.us/) ..
Taurus

Thanks Taurus, I've done that, very good tip.

All the best

Did

[daberti]

Originally posted by len0x
@daberti

1.7x mainly. There were some bits wrong in 1.60 as well but they rarely come up (or you don't notice them because you don't check the quality logs after encoding: there might have been large difference between predicted and expected quality).

@lenOx
I encoded Drugstore Cowboy either with Divx and with Xvid (AutoGk 1.60).
I noticed relevant differences in comp. and res. between two codecs.

Logs follow:

Dvix log:
[10/11/2004 21.07.02] AutoGK 1.60
[10/11/2004 21.07.02] OS: WinXP (5.1.2600).2
[10/11/2004 21.07.02] Job started.
[10/11/2004 21.07.02] Input dir: C:\DVD\Drugstore Cowboy
[10/11/2004 21.07.02] Output file: C:\DVD\Drugstore Cowboy\Drugstore Cowboy Dvix.avi
[10/11/2004 21.07.02] Audio: Italiano
[10/11/2004 21.07.02] Subtitles: none
[10/11/2004 21.07.02] Codec: DivX
[10/11/2004 21.07.02] Target size: 1000Mb
[10/11/2004 21.07.02] Custom audio settings: VBR MP3 with average bitrate: 128Kbps
[10/11/2004 21.07.02] Started encoding.
[10/11/2004 21.07.02] Demuxing and indexing.
[10/11/2004 21.16.02] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_1.VOB
[10/11/2004 21.16.02] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_2.VOB
[10/11/2004 21.16.02] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_3.VOB
[10/11/2004 21.16.02] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_4.VOB
[10/11/2004 21.16.02] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_5.VOB
[10/11/2004 21.16.02] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_6.VOB
[10/11/2004 21.16.02] Source aspect ratio: 16:9
[10/11/2004 21.16.02] Source resolution: 720x576
[10/11/2004 21.16.02] Found PAL source.
[10/11/2004 21.16.02] Analyzing source.
[10/11/2004 21.24.56] Source has percentage of interlacing in motion areas: 0,02
[10/11/2004 21.24.56] Source is considered to be progressive.
[10/11/2004 21.24.57] Found 145903 frames
[10/11/2004 21.24.57] Encoding audio.
[10/11/2004 21.41.36] Audio size: 91,731,128 bytes (87.48 Mb)
[10/11/2004 21.41.36] Overhead: 9,413,760 bytes (8.98 Mb)
[10/11/2004 21.41.36] Video size: 947,431,112 bytes (903.54 Mb)
[10/11/2004 21.41.36] Target bitrate is: 1299kbps
[10/11/2004 21.41.36] Running compressibility test.
[10/11/2004 21.41.36] Writing the following script to C:\DVD\Drugstore Cowboy\agk_tmp\Drugstore Cowboy Dvix_comptest.avs
===========================================================
LoadPlugin("C:\PROGRA~1\AutoGK\DGMPGDec\DGDecode.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\autocrop.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\undot.dll")

movie = mpeg2source("C:\DVD\Drugstore Cowboy\agk_tmp\Drugstore Cowboy Dvix.d2v")
cropclip = autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=0,threshold=34,samplestartframe=0)
fixed_aspect = 1.42222222222222
c_width = width(cropclip)
c_height = round(height(cropclip) / fixed_aspect)
input_par = float(c_width)/float(c_height)
input_par = input_par > 1.4 ? input_par : (4.0/3.0)
out_width = 704
out_height = round(float(out_width) / input_par)
hmod = out_height - (floor(out_height / 16 ) * 16)
out_height = (hmod > 4) ? (out_height + (16 - hmod)) : (out_height - hmod)
new_aspect = (float(out_width) / float(out_height)) / fixed_aspect
autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=new_aspect,threshold=34,samplestartframe=0)
Undot()
LanczosResize(out_width,out_height)
SelectRangeEvery(280,14)
===========================================================
[10/11/2004 21.48.56] Duration was: 7 minutes 18 seconds
[10/11/2004 21.48.56] Speed was: 16,63 fps.
[10/11/2004 21.48.57] Compressibility percentage is: 36,39
[10/11/2004 21.48.57] Using softer resizer.
[10/11/2004 21.48.57] Chosen resolution is: 528x288 ( AR: 1,83 )
[10/11/2004 21.48.57] Predicted comptest value is: 64,69
[10/11/2004 21.48.57] Running first pass.
[10/11/2004 21.48.57] Writing the following script to C:\DVD\Drugstore Cowboy\agk_tmp\Drugstore Cowboy Dvix_movie.avs
===========================================================
LoadPlugin("C:\PROGRA~1\AutoGK\DGMPGDec\DGDecode.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\autocrop.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\undot.dll")

movie = mpeg2source("C:\DVD\Drugstore Cowboy\agk_tmp\Drugstore Cowboy Dvix.d2v")
cropclip = autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=0,threshold=34,samplestartframe=0)
fixed_aspect = 1.42222222222222
c_width = width(cropclip)
c_height = round(height(cropclip) / fixed_aspect)
input_par = float(c_width)/float(c_height)
input_par = input_par > 1.4 ? input_par : (4.0/3.0)
out_width = 528
out_height = round(float(out_width) / input_par)
hmod = out_height - (floor(out_height / 16 ) * 16)
out_height = (hmod > 4) ? (out_height + (16 - hmod)) : (out_height - hmod)
new_aspect = (float(out_width) / float(out_height)) / fixed_aspect
autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=new_aspect,threshold=34,samplestartframe=0)
Undot()
BicubicResize(out_width,out_height,0,0.5)
===========================================================
[10/11/2004 23.10.11] Duration was: 1 hour, 21 minutes 12 seconds
[10/11/2004 23.10.11] Speed was: 29,95 fps.
[10/11/2004 23.10.11] Running second pass.
[11/11/2004 0.26.43] Duration was: 1 hour, 16 minutes 31 seconds
[11/11/2004 0.26.43] Speed was: 31,78 fps.
[11/11/2004 0.26.43] Job finished.
*******************

Xvid log:
[10/11/2004 16.58.40] AutoGK 1.60
[10/11/2004 16.58.40] OS: WinXP (5.1.2600).2
[10/11/2004 16.58.40] Job started.
[10/11/2004 16.58.40] Input dir: C:\DVD\Drugstore Cowboy
[10/11/2004 16.58.40] Output file: C:\DVD\Drugstore Cowboy\Drugstore Cowboy Xvid.avi
[10/11/2004 16.58.40] Audio: Italiano
[10/11/2004 16.58.40] Subtitles: none
[10/11/2004 16.58.40] Codec: XviD
[10/11/2004 16.58.40] Target size: 1000Mb
[10/11/2004 16.58.40] Custom audio settings: VBR MP3 with average bitrate: 128Kbps
[10/11/2004 16.58.40] Started encoding.
[10/11/2004 16.58.40] Demuxing and indexing.
[10/11/2004 17.05.47] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_1.VOB
[10/11/2004 17.05.47] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_2.VOB
[10/11/2004 17.05.47] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_3.VOB
[10/11/2004 17.05.47] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_4.VOB
[10/11/2004 17.05.47] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_5.VOB
[10/11/2004 17.05.47] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_6.VOB
[10/11/2004 17.05.47] Source aspect ratio: 16:9
[10/11/2004 17.05.47] Source resolution: 720x576
[10/11/2004 17.05.47] Found PAL source.
[10/11/2004 17.05.48] Analyzing source.
[10/11/2004 17.12.32] Source has percentage of interlacing in motion areas: 0,02
[10/11/2004 17.12.32] Source is considered to be progressive.
[10/11/2004 17.12.32] Found 145903 frames
[10/11/2004 17.12.32] Encoding audio.
[10/11/2004 17.28.54] Audio size: 91,731,128 bytes (87.48 Mb)
[10/11/2004 17.28.54] Overhead: 5,836,160 bytes (5.57 Mb)
[10/11/2004 17.28.54] Video size: 951,008,712 bytes (906.95 Mb)
[10/11/2004 17.28.54] Running compressibility test.
[10/11/2004 17.28.54] Writing the following script to C:\DVD\Drugstore Cowboy\agk_tmp\Drugstore Cowboy Xvid_comptest.avs
===========================================================
LoadPlugin("C:\PROGRA~1\AutoGK\DGMPGDec\DGDecode.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\autocrop.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\undot.dll")

movie = mpeg2source("C:\DVD\Drugstore Cowboy\agk_tmp\Drugstore Cowboy Xvid.d2v")
cropclip = autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=0,threshold=34,samplestartframe=0)
fixed_aspect = 1.42222222222222
c_width = width(cropclip)
c_height = round(height(cropclip) / fixed_aspect)
input_par = float(c_width)/float(c_height)
input_par = input_par > 1.4 ? input_par : (4.0/3.0)
out_width = 704
out_height = round(float(out_width) / input_par)
hmod = out_height - (floor(out_height / 16 ) * 16)
out_height = (hmod > 4) ? (out_height + (16 - hmod)) : (out_height - hmod)
new_aspect = (float(out_width) / float(out_height)) / fixed_aspect
autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=new_aspect,threshold=34,samplestartframe=0)
Undot()
LanczosResize(out_width,out_height)
SelectRangeEvery(300,15)
===========================================================
[10/11/2004 17.37.32] Duration was: 8 minutes 37 seconds
[10/11/2004 17.37.33] Speed was: 14,09 fps.
[10/11/2004 17.37.33] Compressibility percentage is: 40,74
[10/11/2004 17.37.33] Using softer matrix.
[10/11/2004 17.37.33] Chosen resolution is: 608x336 ( AR: 1,81 )
[10/11/2004 17.37.33] Predicted comptest value is: 59,77
[10/11/2004 17.37.33] Running first pass.
[10/11/2004 17.37.33] Writing the following script to C:\DVD\Drugstore Cowboy\agk_tmp\Drugstore Cowboy Xvid_movie.avs
===========================================================
LoadPlugin("C:\PROGRA~1\AutoGK\DGMPGDec\DGDecode.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\autocrop.dll")
LoadPlugin("C:\PROGRA~1\AutoGK\filters\undot.dll")

movie = mpeg2source("C:\DVD\Drugstore Cowboy\agk_tmp\Drugstore Cowboy Xvid.d2v")
cropclip = autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=0,threshold=34,samplestartframe=0)
fixed_aspect = 1.42222222222222
c_width = width(cropclip)
c_height = round(height(cropclip) / fixed_aspect)
input_par = float(c_width)/float(c_height)
input_par = input_par > 1.4 ? input_par : (4.0/3.0)
out_width = 608
out_height = round(float(out_width) / input_par)
hmod = out_height - (floor(out_height / 16 ) * 16)
out_height = (hmod > 4) ? (out_height + (16 - hmod)) : (out_height - hmod)
new_aspect = (float(out_width) / float(out_height)) / fixed_aspect
autocrop(movie,mode=0,wmultof=4,hmultof=4,samples=10,aspect=new_aspect,threshold=34,samplestartframe=0)
Undot()
LanczosResize(out_width,out_height)
===========================================================
[10/11/2004 19.04.49] Duration was: 1 hour, 27 minutes 14 seconds
[10/11/2004 19.04.49] Speed was: 27,87 fps.
[10/11/2004 19.04.49] Expected quality of first pass size: 57,42
[10/11/2004 19.04.49] Running second pass.
[10/11/2004 20.59.24] Duration was: 1 hour, 54 minutes 34 seconds
[10/11/2004 20.59.24] Speed was: 21,22 fps.
[10/11/2004 20.59.24] Job finished.

**************************
2nd pass in Xvid takes an age compared to Dvix's one.

Is it all going as it should?

BTW: great SW lenOx we'll never thankyou enough

[len0x]

Originally posted by daberti

Is it all going as it should?


I see nothing wrong with the logs.

[daberti]

Originally posted by len0x
I see nothing wrong with the logs.

Thanks lenOx

[Taurus]

@ len0x

...In the meantime

Thank you for your constantly search for bugs.
I really endore your versatile approach to perfection.
Sorry I can't come up with bugs.., the encodes I'm doing are normally
in a higher quality range (above 704:xxx).
I never encountered any problems there so far. Even if I crosschecked with
"normal" GK on Compressibility check, there were almost identical results in both.

So keep up the good work.

(Mpeg4 is not dead) as someone in another thread mentioned. ;)

Cheers

Taurus

P.S.: If you need further help in matrix comparision, let me know.

[daberti]

Originally posted by daberti
Thanks lenOx

@lenOX
I'm encoding always with 1.60 yet giving a max width of 608 (same remaining setting as in above posted log).
Results seem to be quite different from Xvid log posted above even if having BTW the same width.
I.e: the auto-width chooses 608/336 and brings about Compressibility percentage 46,13 while Predicted comptest value is 61.04 (Expected quality of first pass size: 65.01) .
The Max Width = 608 chooses 608/336 and brings about Compressibility percentage 59.79 while Predicted comptest value 66,28 .
First pass is just running....

The delta between the two approaches is too wide and I suspect we're running into the same resizing bug as the 1.7x tree.

Thus I'm thinking about switching to 1.7x (clean 1.60 uninstall required BTW?)

Edit:
1.76 now:
with max width at 624 these are the results
[12/11/2004 16.56.51] Compressibility percentage is: 60,33
[12/11/2004 16.56.51] Using softer matrix.
[12/11/2004 16.56.51] Chosen resolution is: 608x352 ( AR: 1,73 )
[12/11/2004 16.56.51] Predicted comptest value is: 68,95

Edit 2:
12/11/2004 17.23.33] Custom resolution settings: minimum width of 544 pixels
[12/11/2004 17.23.33] Custom audio settings: VBR MP3 with average bitrate: 128Kbps
[12/11/2004 17.23.33] Started encoding.
[12/11/2004 17.23.33] Demuxing and indexing.
[12/11/2004 17.30.14] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_1.VOB
[12/11/2004 17.30.14] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_2.VOB
[12/11/2004 17.30.14] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_3.VOB
[12/11/2004 17.30.14] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_4.VOB
[12/11/2004 17.30.14] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_5.VOB
[12/11/2004 17.30.14] Processing file: C:\DVD\Drugstore Cowboy\VTS_01_PGC_01_6.VOB
[12/11/2004 17.30.14] Source aspect ratio: 16:9
[12/11/2004 17.30.14] Source resolution: 720x576
[12/11/2004 17.30.14] Found PAL source.
[12/11/2004 17.30.14] Analyzing source.
[12/11/2004 17.37.00] Source is considered to be progressive.
[12/11/2004 17.37.01] Found 145903 frames
[12/11/2004 17.37.01] Encoding audio.
[12/11/2004 17.52.28] Audio size: 91,731,128 bytes (87.48 Mb)
[12/11/2004 17.52.28] Overhead: 5,836,160 bytes (5.57 Mb)
[12/11/2004 17.52.28] Video size: 1,076,837,832 bytes (1026.95 Mb)
[12/11/2004 17.52.28] Running compressibility test.
[12/11/2004 18.00.25] Duration was: 7 minutes 56 seconds
[12/11/2004 18.00.26] Speed was: 15,31 fps.
[12/11/2004 18.00.27] Compressibility percentage is: 49,91
[12/11/2004 18.00.27] Using softer matrix.
[12/11/2004 18.00.27] Chosen resolution is: 624x336 ( AR: 1,86 )
[12/11/2004 18.00.27] Predicted comptest value is: 71,66
[12/11/2004 18.00.27] Running first pass.
[12/11/2004 19.28.09] Duration was: 1 hour, 27 minutes 41 seconds
[12/11/2004 19.28.09] Speed was: 27,73 fps.
[12/11/2004 19.28.10] Expected quality of first pass size: 66,70
[12/11/2004 19.28.10] Running second pass.

[Sharktooth]

Originally posted by Taurus
P.S.: If you need further help in matrix comparision, let me know.
A new matrix is coming... so maybe... :D

[len0x]

Originally posted by daberti
The delta between the two approaches is too wide and I suspect we're running into the same resizing bug as the 1.7x tree.


Well, a lot was changed in resizing after 1.60, so I can't say if that's the case, but 1.76 has definitely less problems in that area.
And you logs from 1.76 seem OK to, don't they?

[len0x]

Originally posted by Taurus
Thank you for your constantly search for bugs.
I really endore your versatile approach to perfection.


I'm glad some ppl still can see that :) Its more of a science really...
These days I seem to get more attention as probable violator of GPL and promoter of dead mpeg4 :)

[BigDid]

Originally posted by BigDid

Any power-user, senior or master feel free to detail the results as I don't feel qualified to do so.


Should I rephrase that in
I would appreciate any feedback from anybody with better knowledge than me on the High bitrates, spikes, VBV buffer sujects...?

Because I have still that stall problem on the 2 encodes with 2Gb size (LOTR2 & LOTR3)
Without feedback or directions I am trying to reencode LOTR3 with the different options with no luck at the moment ( near 6 hours each time :mad: )

Thanks in advance

Did

[daberti]

Originally posted by len0x
Well, a lot was changed in resizing after 1.60, so I can't say if that's the case, but 1.76 has definitely less problems in that area.
And you logs from 1.76 seem OK to, don't they?

1.76 works really NICE.
Just a question though: IYO what codec is the best with 1.76?
Xvid or Divx? The only thing I noticed is a quantum leap in speed on Divx side.

[kio]

Hello,

Big congrats for your software. I'm blind and thanks to you for helping me to release my divx movies :)

However, since you released 1.76B version, i try to download it but without success ! :(

I got "module is not actif" or some thing like that... do U know why ?

best regards,
and congratulations for your really nice job ! :d

[len0x]

Originally posted by kio
I got "module is not actif" or some thing like that... do U know why ?

Your browser doesn't have third-party cookies enabled. This hasn't been changed since new AutoGK website was up, so something must have happened to your browser.

[len0x]

Originally posted by daberti
Just a question though: IYO what codec is the best with 1.76?
Xvid or Divx?

Its all up to your taste really. I used to do only DivX (before I even started AutoGK project). But then switched to XviD and happy with it. AutoGK has more control over XviD because of quantizer settings and custom matrixes (that really do make a difference). If you go to the survey section on the website you can see what codec ppl are using with AutoGK.

[kio]

sorry, it doesn't work even i empty my cookies. i tried from another computer, nothing happend :'(

_http://len0x.btothec.com/web3/index.php?name=Downloads&d_op=getit&lid=8

help plz i need the upgrade

[daberti]

@ lenOx
I realized that XVID really saves space at same quality level.
Custom matrices: you are talking about the one shipped with 1.76 or something else?

[daberti]

Originally posted by kio
sorry, it doesn't work even i empty my cookies. i tried from another computer, nothing happend :'(

_http://len0x.btothec.com/web3/index.php?name=Downloads&d_op=getit&lid=8

help plz i need the upgrade

Are you behind a firewall? ;)
Are you sure that under Privacy setting of your browser you allow all cookies from url above to be stored? ;)

[Sharktooth]

@len0x: beta5
8 9 10 12 12 15 16 20
9 10 11 12 15 16 20 21
10 11 13 15 16 20 21 27
11 13 14 17 19 22 26 28
13 14 17 19 22 26 29 38
14 17 19 23 25 30 36 40
18 18 23 25 31 34 43 62
18 24 24 32 34 46 52 80

16 16 17 19 19 20 21 23
16 17 18 19 20 21 23 24
17 18 19 20 21 23 24 30
18 19 20 21 23 24 29 31
19 20 21 23 25 28 32 48
20 21 22 25 27 34 45 53
22 22 25 27 36 42 58 84
22 26 26 38 40 63 72 106

http://ebola.gamersrevolt.it/len0x/eqm_AGKHR_b5.xcm

[len0x]

DTS support is there, plz test it

[Sharktooth]

@Len0x/Taurus: please test beta5. It should keep the same details as V2 but with a bit more compressibility (at least with the sources i tested...) and softer scaling (coefficients follows the scan order) to avoid artifacts and excessive sharpness on CG material.

[therealjoeblow]

I don't pretend to understand the intricacies of IVTC, that's why I use AutoGK ;-)

However, I tried to encode:

http://www.imdb.com/title/tt0077572/

...several times with different versions of AutoGK from 1.25 to 1.5x to 1.60 to 1.7x, and in all cases, Decomb with Telecide/Decimate was chosen by AutoGK to IVTC the movie. This is an older film, with a *very* grainy source, and in all cases (regardless of version or codec used for compression), there were *nasty* decombing artifacts at various points in the movie - best described as 'shimmering or dancing vertical blinds'. (That's why I used so many different versions - I thought maybe one of them would perform the IVTC better since there have been various versions of DGIndex and predecessors, and Decomb used over time; but there was never any noticeable difference).

I assume that DGIndex didn't apply the Force Film algorithm probably because the movie didn't analyze at 100% film, that's why AutoGK took over with Decomb (inner workings obviously not clearly understood). It's been a *major* source of frustration that I couldn't get a decent encode of this after trying repeatedly for several months. So last night on a whim, I used the "Restore/Stop DGIndex and Overwrite .d2v File" trick to enable the "Force Film" algorithm in DGIndex, and the resulting encode is beautiful, none of the artifacts remain, not even in the scenes where it was the worst in the original attempts, where it was almost unwatchable. And there don't appear to be any other problems introduced, encode plays back nice and smooth throughout, no interlacing remains at scene transitions, etc.

So, maybe this is just one of those odd-ball cases where the default analysis algorithm can't work 100% perfectly? I don't know how many situations there may be that fit this pattern, but clearly here the Force Film setting worked way better than Decomb could.

Questions:

1) in cases where such artifacting *doesn't* occur, is there any visual difference between Force Film and Decomb IVTC? IE, does one generally produce a better looking final product than the other on normal material (I've never done any direct comparisons other than this one where the difference was astounding), and

2) would it be possible to get a Hidden Option to enable Force Film in DGIndex to override whatever the analysis decides (or maybe a toggle to use one or the other, so you clearly know which method is going to be used by force)?

Many Thanks,

(PS - donation we spoke about is still coming this week - last 2 weeks have been too hectic to get to the Post Office).

[Taurus]

Originally posted by Sharktooth
@Len0x/Taurus: please test beta5.

Did some short tests on material as discussed before (mostly CG).
I must say this one looks very promising (Beta5). No harshness or artificial sharpening. Smooth as V2 with the benefit of even better
compressibility. Good job.
I will do some more testing on "real life" sources the next days.
Changing resolutions too. Mostly I test on Q=2 720x***.
It will be interesting to see interaction on different resolutions.
Will ring the bell later :D

Oh, thanks len0x for DTS support. Will test it on some Music DVD I own.

Cheers

Taurus

[len0x]

Originally posted by Sharktooth
@Len0x/Taurus: please test beta5.

well, it doesn't look too good from my side :) More block and less details on my my testing material (I'll do more tests anyway), although indeed higher compressibility. I like beta4 much better. May be its true that we need two matrixes for notmal and CG material (i.e. cartoon mode of XviD can be complimented with beta5 and the rest with another matrix).

[len0x]

Originally posted by therealjoeblow
So, maybe this is just one of those odd-ball cases where the default analysis algorithm can't work 100% perfectly? I don't know how many situations there may be that fit this pattern, but clearly here the Force Film setting worked way better than Decomb could.


I'd like to see: a) logs from normal encode, b) d2v file

I do not see any particular reasons for decomb behave badly unless its processing hybrid material. It's supposed to a) produce progressive material and b) remove duplicates. I don't think second step can be at fault, because it would introduce shuttering rather then artifacts. But I'm not sure how decomb actually does first step. may be having out of pattern frames confuse it (when its picks wrong frames during 3:2 pull down). May be manono may have some ideas about that.

[therealjoeblow]

Originally posted by len0x
I'd like to see: a) logs from normal encode, b) d2v file

I do not see any particular reasons for decomb behave badly unless its processing hybrid material. It's supposed to a) produce progressive material and b) remove duplicates. I don't think second step can be at fault, because it would introduce shuttering rather then artifacts. But I'm not sure how decomb actually does first step. may be having out of pattern frames confuse it (when its picks wrong frames during 3:2 pull down). May be manono may have some ideas about that.

SHUTTERING! That's a better term to describe what I saw, rather than 'artifacts'. It looked as if someone was repeatedly pulling the drawstring on the vertical blinds from open to closed position at high frequency.

I'll re-run the encode tonight with default settings and post the 2 logs and 2 d2v's for comparison. I could also clip out a portion of the encoded video of each if that helps understand the effect better?

[rotvel]

Originally posted by len0x
DTS support is there, plz test it

I'm tried to rip a DVD with DTS audio -> 2 CD XviD/mp3 cbr. But I got an


*************************************
EXCEPTION: Audio is not found.
*************************************


I did everything like I use to, ripping with DVD Decrypter etc.

[Taurus]

Originally posted by len0x
DTS support is there, plz test it

Maybe you should mention that ffdshow (audio) also decodes DTS quite well.
It uses libdts and about the same cpu resources as ac3filter.
So something to add for the FAQs and/or the Help files.

Cheers

Taurus

[len0x]

Originally posted by rotvel
I'm tried to rip a DVD with DTS audio -> 2 CD XviD/mp3 cbr. But I got an


please post stream info file

[len0x]

Originally posted by Taurus
Maybe you should mention that ffdshow (audio) also decodes DTS quite well.


see, i didn't know know that as I don't use ffdshow (in fact I don't use latest ac3filter either because its not working with powerdvd...)

[len0x]

Originally posted by therealjoeblow
I'll re-run the encode tonight with default settings and post the 2 logs and 2 d2v's for comparison. I could also clip out a portion of the encoded video of each if that helps understand the effect better?

Well, d2v file says 98.7% film and threshold in autogk to make it FF is 99%, so it just missed the cut :) I can see that only ending of the movie has NTSC source, so if you look at it with FF option on, then you should see pretty heavy shuttering, but if that's credits then it probably won't bother you. I can reduce threshold to say 98.5%, so that your source would get FF automatically (have you seen any other moview like this?).

[rotvel]

Originally posted by len0x
please post stream info file


0x20 - Subtitle - Chinese - [SubPicture 01: Wide/Letterbox] / LBA: 9915 / PTS: 00:00:31.498 / Delay: 31264ms
0x21 - Subtitle - Chinese - [SubPicture 02: Wide/Letterbox] / LBA: 9917 / PTS: 00:00:31.498 / Delay: 31264ms
0x22 - Subtitle - English - [SubPicture 03: Wide/Letterbox] / LBA: 9918 / PTS: 00:00:31.498 / Delay: 31264ms
0x89 - Audio - DTS / 5ch / 48kHz / DRC / Japanese / LBA: 41 / PTS: 00:00:00.233 / Delay: 0ms
0xE0 - Video - MPEG-2 / 720x480 (NTSC) / 16:9 / Letterboxed / LBA: 1 / PTS: 00:00:00.233 / Delay: 0ms

[len0x]

Originally posted by rotvel

0x20 - Subtitle - Chinese - [SubPicture 01: Wide/Letterbox] / LBA: 9915 / PTS: 00:00:31.498 / Delay: 31264ms
0x21 - Subtitle - Chinese - [SubPicture 02: Wide/Letterbox] / LBA: 9917 / PTS: 00:00:31.498 / Delay: 31264ms
0x22 - Subtitle - English - [SubPicture 03: Wide/Letterbox] / LBA: 9918 / PTS: 00:00:31.498 / Delay: 31264ms
0x89 - Audio - DTS / 5ch / 48kHz / DRC / Japanese / LBA: 41 / PTS: 00:00:00.233 / Delay: 0ms
0xE0 - Video - MPEG-2 / 720x480 (NTSC) / 16:9 / Letterboxed / LBA: 1 / PTS: 00:00:00.233 / Delay: 0ms


Hm... does DTS file get demuxed in agk_tmp dir and if yes then what is the name of the file ?

[Taurus]

Originally posted by len0x
DTS support is there, plz test it

DTS -> MP3 vbr = OK
DTS -> MP3 cbr = OK
DTS -> AC3 = NO Dts file just got muxed to video.(Expected)

For all SAP Users who can't decode DTS it would be a nice feature to have the 5.1 or 6.1 DTS Stream converted to AC3.
I know it's more work to do. Did this several times with BeSweet and if I remember correctly from the past it was a mess to setup ;) :D

Cheers

Taurus