dvdrb 1.0rc5.1 and trim issues

[Carpo]

lately i have been having issues where the trim value will stop the encoder dead in its tracks (just wont encode the segmant) if i change the first trim vaule by 4 (random thought up number for the test) the encoder happily goes on its way.

I have only had this once b4 - but am now also getting it on Road Warriors dvds :(

trim(42004,46335) is what it was, trim(42000,46335) is what i changed it too

its a pal dvd - if you need any more info just ask - not sure what you want and have to leave for work :( god damn 7 am starts

[jdobbs]

That sounds a lot like a problem in the source that is confusing DGDECODE.DLL in some way... I think by changing the start point of the trim() statement you have just bypassed the troublesome frame.

[jptheripper]

but he is going earlier, encluding all frames (troublesome) plus 4 extra, no? sounds like tis having trouble starting on a particular frame

[jdobbs]

Hmmm... you're right. That makes it even odder.

@Carpo

Could you show the contents of the previous AVS?

[Carpo]

its done it twice no :( once overnight and while i have been at work

i will post post prvious avs files as asked

#------------------
# AVS File Created by DVD Rebuilder
# VOBID:01, CELLID:09
#------------------
LoadPlugin("C:\Program Files\DVD-RB PRO\DGDecode.dll")
mpeg2source("E:\dvd-work\D2VAVS\V01.D2V")
trim(32784,42003)
Undot()
ConvertToYUY2(interlaced=true)


#------------------
# AVS File Created by DVD Rebuilder
# VOBID:01, CELLID:22
#------------------
LoadPlugin("C:\Program Files\DVD-RB PRO\DGDecode.dll")
mpeg2source("E:\dvd-work\D2VAVS\V01.D2V")
trim(94832,97531)
Undot()
ConvertToYUY2(interlaced=true)

dunno if its an issue with my setup but have un installed and re-installed everything 3 times now just to be sure using the dvd-rb installer

[Carpo]

no one have any idear then ? :(

[jptheripper]

have patience, its only been 3 hours

[Carpo]

ye sorry - just that its an annoying thing :/

[jptheripper]

since you are will to try, try changing the previous avs trim down 1 and same with offedning avs

i.e make them 42002 and 42003

just a thought

[Carpo]

i'll try it out overnight again and see what happens - although have been looking at the big 3 guide (feeling adventerious) :P


have done as you suggested - will have to see what happens

[SpazzHH]

Out of curiosity, have you tried it with a different encoder? How about without Undot, just to eliminate it as a possible cause?

[writersblock29]

@Carpo

I can vouch for the R1 version of this disk... but the only thing I changed beyond Rebuilder's default settings was adding the Undot filter and certain CCE settings. No problems whatsoever.

Are you having problems with all the disks (Mad Max, Road Warrior, MM Beyond Thunderdome) or just one? If all are bugging you, maybe doing a fresh install would clear things up.

[Carpo]

all encoders are doing it - i have tried with undot on and off makes no diff - if i change the trim it works, it stuck out seg 23 over night even tho i did change the trim on that (guess not enuff).

i will try some other discs after i get back from work today and see if it happens on them

[Carpo]

That sounds a lot like a problem in the source that is confusing DGDECODE.DLL in some way... I think by changing the start point of the trim() statement you have just bypassed the troublesome frame.

would this be a poss reason to update the dll ?

[Trahald]

you may want to see if this is something that should be brought to neuron2's attention. if you can cut a peice around that spot you around 30 mbs or so and still can duplicate the issue with it.. send it to him and he will look at it. (its possible its a avisynth issue) ... you would want to cut a few gops before the error and see if the trim command starting from the offset will still crash.. meaning if you cut from frame say 41900 to 43000 ... then use trim( 104, 200) against your sample which will start at the same frame and see if the crash still occurs.

[jdobbs]

I wasn't saying there was a bug in DGDECODE. I was saying the source many have an illegal construct that is confusing DGDECODE. Just speculating, but it may be possible that the GOP that is the first in that segment is a part of an open GOP set. While that might cause problems for DGDECODE.DLL (not sure) with the trim statement -- the real problem would be poor authoring.

Remember -- the fault for problems can also be the software/person that was responsible for the original disc... I've seen lots of examples of that.

[Trahald]

yeah.. exactly what i was wondering.. . its very unlikely a dgdecode bug but possibly something that could be worked around. within some reason he is generally up for applying workarounds for problem sources.
if he could reproduce it with a small enough segment, dg is usually willing to take at least take a look.

[Carpo]

you may want to see if this is something that should be brought to neuron2's attention. if you can cut a peice around that spot you around 30 mbs or so and still can duplicate the issue with it.. send it to him and he will look at it. (its possible its a avisynth issue) ... you would want to cut a few gops before the error and see if the trim command starting from the offset will still crash.. meaning if you cut from frame say 41900 to 43000 ... then use trim( 104, 200) against your sample which will start at the same frame and see if the crash still occurs.

which tools would i need for that ?

[Carpo]

the real problem would be poor authoring.

Remember -- the fault for problems can also be the software/person that was responsible for the original disc... I've seen lots of examples of that.

well is a disc from wwe ;)

[Trahald]

there are cutting utils in the vob section of doom9.org's downloads (make sure you click the full downloads link )

[Carpo]

on a bit of further investigation it looks like the 2 segmants in question are at the end of the vobs - would this have any relovance to the situation?

[jdobbs]

@Carpo

It might.

@Everyone

Is anyone else experiencing this?

[Carpo]

they prob arent - i bet its an issue with my system (not long re-installed i might add) - although i did do one disk im having issues with using big 3 and it went ok - so im wondering if i should let dvd rb prepare the disc and use an editor to remove the trim values from the avs files

I have just got bottom series 3 (comedy series here in uk) which is also pal so might give that a go and see what happens if i get same issue i will re-install the system just to see - although that prob wont happen till i get back of holiday in first week of september

[Trahald]

i tried to duplicate it with at a spot there was a open gop starting the next vob segment with various trim values but could not replicate the issue.

[Carpo]

well have re-installed the other pc (on linux box atm) and only things i have put in is xp - xp updates - avisynth - dvdrb and cce and its sticking at the same point seg 10 (15.6% in) - so im at a loss

yet big three guide i followed did it ok - so im wondering if i should just remove the trim values - it will have to wait now anyway as i have to get other pc back up and running and then get ready for my hols :)

[Trahald]

well... hopefully can cut the peices of the vob before you go so it can be looked at. other than that.. have fun ;)

[Carpo]

i'll try that once i have everything back in and working

right think i have cut out the right spot ifoedit roughly says were it is and i hope its the right space (trying to work it out by time instead of frames).

would anyone know of a tool that will cut vobs by frame? i have chopperxp

[jptheripper]

womble mpeg2vcr does very good frame accurate cutting

[Carpo]

little update - i have lost one of the hdds that was in my machine :(, it was the one i used for backing up so have lost the source vobs for the wwe disc i was having probs with - also the other disc i was testing.

But - after testing a disc i bought during the week and backed up using the same setup i have always used all went through ok - so i am thinking it was the drive to blame - even though all checks on that disc said it was fine :/

Still all is working ok again now :) thanks for all the suggestions and help

[Pasqui]

@jdobbs
I experience the same issue as Carpo with DVD-RB Pro 1.0 RC5.1 and Lost In Translation movie.
the following script:
#------------------
# AVS File Created by DVD Rebuilder
# VOBID:03, CELLID:10
#------------------
LoadPlugin("F:\DVD\DVD-RB\DGDecode.dll")
mpeg2source("H:\LOST_IN_TRANSLATION\D2VAVS\V01.D2V")
trim(83598,91689)
ConvertToYV12(interlaced=true)
both hangs HC_GUI encoder, MPC 6.4.8.4 and VirtualDub 1.6.10.
Changing the trim command from trim(83598,91689) to trim(83597,91689) allows me to open the file in applications.
I also used DGIndex 1.4.1 to create a .d2v file from the vob input and created a new avs file:
LoadPlugin("F:\DVD\DGIndex\DGDecode.dll")
mpeg2source("H:\LOST_IN_TRANSLATION\VTS_01_1.d2v")
trim(83598,91689)
ConvertToYV12(interlaced=true)
This time I can open the avs file in any application. Latest's neuron2 release seem more robust so maybe you should think about upgrading DGDecode.dll version you use.
BTW, it's the first time I face such an issue. So far DVD-RB worked great !

Regards,

[jptheripper]

as a side note, dg1.4.1 went final, so maybe it is upgrade time

[Carpo]

im wondering if it would be an idear for dvd-rb to use lastest gd dll it may solve some issues (if any) and would be able to take advantage of the new functions that have been added to it (there are too many to list - and thats what readme/changelogs are for) :)

[jdobbs]

It may. But is is a lot more likely that it will cause more... and I don't want to start chasing errors this close to final release of v1.00. I'll wait until it has been out for a while. The version included with DVD-RB does everything that is required... upgrading only for the sake of upgrading is always a bad idea.

[Carpo]

how about giving it thought for inclusion in 1.10 - as you say after it has been out for a while and is known to be more error solving then causing ? :)

[psychoxxx]

I am also having problems with WWE Road Warriors DVD PAL. I use DVDRB 0.95 free with CCE 2.67 retail. CCE just came to a halt after a few miniutes and stayed there even overnight.

This is the rebuilder.ini:
[Options]
Mode=1
OneClick=1
ReduceOpt=0
NoWarn=1
AdditionalOutput=1
LogFile=1
QuEncHQ=1
EncoderMinimized=0
RemoveDTS=1
HC_Quality=1
HC_Matrix=1
CCE=1
SkinVersion=3
HalfD1=00
Convert_16_9=00
DisableInterlace=00
Completed=13
HalfExtras=0
AVSFilter01=
Reminder=10
AudioDub=1
[Paths]
QuEnc=C:\Program Files\DVD-RB\Encoders\QuEnc\QuEnc.exe
ReJig=C:\Program Files\DVD-RB\Encoders\ReJig\ReJig.exe
HC=C:\Program Files\DVD-RB\Encoders\HC Encoder\HCbatch.EXE
ProCoder=C:\Program Files\DVD-RB\Encoders\EclPro\EclPro.exe
DECOMB=C:\Program Files\AviSynth 2.5\plugins\Decomb521.dll
MPEG2DEC=C:\Program Files\DVD-RB\DGDecode.dll
Working=D:\WORKFOLDER\
Output=N:\OUTPUT\
CCENew=C:\Program Files\Cinema Craft Encoder SP v2.67.00.27\cctsp.exe
Source=Q:\WWERWDVD2\
[CCEOptions]
VBR_bias=25
Quality_prec=24
eclPasses=4
OPV=1
[Audio]
Selected=11
[Subpictures]
Selected=

This is the rebuilder.log:
- Versions:
-- CCE SP Version: 2.67.0.27
-- ReJig Version: 0.3.0.0
-- QuEnc Version: 0.6.1.0
-- HC Encoder Version: 0.15.0.0
-- DECODER Version: UNKNOWN
-- AVISYNTH Version: 2.5.5.0

[18:01:42] One Click encoding activated...
-----------------
[18:01:42] Phase I, PREPARATION started.
- "One Pass VBR (w/analysis)" mode is enabled.
- VTS_01: 3,639,641 sectors.
-- Scanning and writing .D2V file
-- Processed 257,865 frames.
-- Building .AVS and .ECL files
- Reduction Level for DVD-5: 57.6%
- Overall Bitrate : 3,199Kbs
- Space for Video : 4,028,244KB
- Analyzing VTS_01 for optimal Q factor.
-- TargetSize (sectors):2,044,334
-- Sampling 2580 of 257865 frames.
-- Predicted size (sectors) at Q=32: 1,727,096
- Q Value selected: 32
- HIGH/LOW/AVERAGE Cell Bitrates: 3,589/3,197/3,199 Kbs
[19:50:49] Phase I, PREPARATION completed in 109 minutes.
Aborted.

Any suggestion appreciated. Would upgrading to Pro version solve the problem? Thanks a lot.

[jdobbs]

If you are overclocking, set it back to the rated value.

[Rippraff]

@jdobbs

Jerry, I'm wondering why there isn't a second Q factor prediction? With Q=32 it's more than 18 % below the target size.

By the way IMHO this has nothing to do with trim issues. ;)

Cu Rippraff

[psychoxxx]

If you are overclocking, set it back to the rated value.

I am not overclocking. It's Pentium 4 3.4Ghz, 1GB RAM, WinXP SP2. Is it a PAL problem? Thanks.

[jdobbs]

@jdobbs

Jerry, I'm wondering why there isn't a second Q factor prediction? With Q=32 it's more than 18 % below the target size.

By the way IMHO this has nothing to do with trim issues. ;)

Cu RippraffYou're right. That makes no sense at all. It looks like CCE hung attempting to do the prediction -- and he killed the job.

@psychoxxx

There is a bug in the CCE Trial Version. Try changing the "Run Encoders Miimized" setting in the Setup screen.