DVD Rebuilder (CCE One-Click Beta v0.46) Comments and Suggestions

[nwg]

I just like to say that 0.38 is the best yet.

I recorded a TV series of the TV onto DVD RAM. This was eight half hour programmes. I made a 8GB DVD from it using DVD Author.

Previously I used 037 to make another DVD using Half D1 for the whole project. I couldm't watch it as the picture looked wierd. I believe this might have been the motion blur bug.

I have redone the whole thing with 0.38 and it plays perfect.

[jdobbs]

Originally posted by rui
Yeah :)

For me, it's really a life savior, because i just can't leave the machine turned on over night (it's in a sleep room), and CCE takes it's time to encode.

Sorry for asking this, but anyone that has already used the feature, is this the correct way:

During an CCE encode, i must first press the PAUSE in DVD-RB, only then to cancel the CCE encode, right?

This seems the correct way, because if i press pause in DVD-RB, and wait to CCE to finish the current encode job, DVD-RB will assume that the finished encode actually didn't ended correctly, right?

And if i first press cancel in the CCE encode, it will instantly pass to the next cell encode, and when i press PAUSE in DVD-RB, it will assume that the first one ended sucessfully, when it didn't.

EDIT: Or why i just don't try it, and see :o You are correct. You should always press the STOP/PAUSE button first. Then stop the CCE encode. If you try to stop the encode first, DVD-RB won't know that it wasn't finished and will mark it in the INI as completed. You'll then have an unfinished job that might bomb your REBUILD.

[jdobbs]

Originally posted by nwg
I just like to say that 0.38 is the best yet.

I recorded a TV series of the TV onto DVD RAM. This was eight half hour programmes. I made a 8GB DVD from it using DVD Author.

Previously I used 037 to make another DVD using Half D1 for the whole project. I couldm't watch it as the picture looked wierd. I believe this might have been the motion blur bug.

I have redone the whole thing with 0.38 and it plays perfect. That is so far-out, hip, groovy, and cool. :cool:

[unplugged]

Originally posted by jdobbs
Right now I only turn on zig-zag when it is determined to be progressive. I guess what unplugged is looking for is a "treat it as progressive" flag rather than just removing "interlaced=true". I think i misunderstood when I added the option to the advanced menu.
[BIG EDIT]
OK, yes
I mean give possibility to set just for main movie
- ConvertToYUY2(interlaced=false)
- Progressive-flag true
- zig-zag scan order
All 3 switches should be used with the option you already inserted in DVD-RB (ConvertToYUY2()).
I'm pretty sure that this will exactly fit most DVDs.
Generally other combinations of these options are for really interlaced content.

P.S.: And for really interlaced movies or clips I expect they really need interlaced=true all times, of course.
I think program need some advanced routine (decomb based?) to detect (prior encoding) if a VTS is:
- progressive
- progressive (stored as interlaced)
- interlaced

[jdobbs]

Originally posted by rui
Yeah :)

For me, it's really a life savior, because i just can't leave the machine turned on over night (it's in a sleep room), and CCE takes it's time to encode. Actually I wish I'd thought of this myself...:) I was in the middle of a long CCE encode and just wanted to compile and test a new version. Hmm... hey, wait a minute! (I sez to myself) I have Pause and Resume now! Hot damn.

[jdobbs]

@unplugged

By the way, I like your signature block. Lee Ritenour is one of my favorite artists, as well as Dave Grusin and the entire cast of GRP performers.

[burnout]

nice work so far jdobbs :cool:

just wondering if it would be possible to add support for smooth deinterlacer similar to decomb?

and this is a long shot... but would it be possible to extract the audio during prepare, then it could be downsampled (either by rebuilder, or done manually with besweet) and analysed before encoding (some how calculating a new bitrate to encode the video files at)? i guess this could be done by selecting a new bitrate for it from a drop down menu in advance so that the size could be calculated, or analysing after with new audio, reauthorist style. thought it was worth asking just incase seeing as how many updates have been made ;)

[jdobbs]

Originally posted by burnout
nice work so far jdobbs :cool:

just wondering if it would be possible to add support for smooth deinterlacer similar to decomb?

and this is a long shot... but would it be possible to extract the audio during prepare, then it could be downsampled (either by rebuilder, or done manually with besweet) and analysed before encoding (some how calculating a new bitrate to encode the video files at)? i guess this could be done by selecting a new bitrate for it from a drop down menu in advance so that the size could be calculated, or analysing after with new audio, reauthorist style. thought it was worth asking just incase seeing as how many updates have been made ;) That could be a possible improvement in the long-range. I guess my experience has been that the amount of space you save by downsampling generally isn't significant enough to make noticable improvements in the video quality (except LPCM).

[unplugged]

Originally posted by jdobbs
By the way, I like your signature block. Lee Ritenour is one of my favorite artists, as well as Dave Grusin and the entire cast of GRP performers.
Well then you shouldn't miss the Fourplay smooth jazz band (where there has been Rit. at the guitar), even if I guess you have heard of them.
If not, try to listen album "Elixir", you'll have a clear image of what they all together reach to do.
Fourplay is a *MUST* for whoever like smooth jazz style and melody, FP sound is so dense that our HI-FI seems never enough to express...
Four musicians, four composers, four performers, everyone with own sound personality and united sooo smoootly that I can't beileve still now.
Bob James (piano), Lee Ritenour (now Larry Carlon), Nathan East (bass & vocals), Harvey Mason (drum).

[jdobbs]

@unplugged

I'll find that CD today.

[burnout]

Originally posted by jdobbs
That could be a possible improvement in the long-range. I guess my experience has been that the amount of space you save by downsampling generally isn't significant enough to make noticable improvements in the video quality (except LPCM).

cheers, i find it helps, especially if you get some tracks at 448kbps and downsample to 256kbps, will give you an improvment and not really noticable audio quality loss (if you don't have a huge speaker system for your tv like me)

Originally posted by burnout
just wondering if it would be possible to add support for smooth deinterlacer similar to decomb?

was the answer to this included in the above thing too? just i find decomb.dll leaves white artifacts on high motion scenes and doesn't really look a lot better than the interlaced material.

[jdobbs]

@burnout

I'm looking at a way to make it possible for advanced users to add additional AVS commands at will. It will be one of those "use at your own risk" features...

[burnout]

nice :D looking forward to it

[onesoul]

Originally posted by burnout
nice work so far jdobbs :cool:

just wondering if it would be possible to add support for smooth deinterlacer similar to decomb? There are better alternatives today like kerdeldeint from Donald Graft.
(Smoothdeinterlace should be more useful if you were outputting doublerate frame rate, which could help if you needed to apply temporal/spatial filters like pixiedust or fluxsmooth, etc., then still in avisynth you would separatefields and selectevery(4,0,3)->TFF or selectevery(4,1,2)->BFF and weave to recreate original frame rate. But then again today there are better options like dgbob also from from Donald Graft or even kernelbob(script by Scharfis_brain using kerneldeint.)

[burnout]

cheers, will have a look into that. used smoothdeinterlacer for ages so never though about changing

[Paced]

Just a little unimportant suggestion here, but would it be possible to have the 'overall progress' (in %) of DVD-RB in the title bar (for those of us who minimize it when it's doing its stuff)?

[The_Flash]

Originally posted by unplugged

Fourplay is a *MUST* for whoever like smooth jazz style and melody,

I'm not your typical jazz listener, but boy was I addicted when my roomate played them in college. Good stuff.

[godhead]

@jdobbs
Thanks for making such a great app. I hope you find my donation useful and I'm off to finally go test 0.38

@burnout
Nice avatar! :D

[jdobbs]

Originally posted by godhead
@jdobbs
Thanks for making such a great app. I hope you find my donation useful and I'm off to finally go test 0.38

@burnout
Nice avatar! :D So that was you! (The donations are easy to count) Thanks. :D

[lab-one]

@jdobbs

Not a big "to-do" but did I see you once say that donaters would eventually receive update notifications via email?

[marnum]

Thank you for this great tool jdobbs (also for IFO update!)! It's really making life easier.

I don't know if more experienced users here are going to hit me, or if this has been asked before, but have you already considered using the kvcd notch matrix? *duck*
I'm using it quite often, combined with a lower max bitrate (~7000 kbs gives no blocknoise at all)..

[burnout]

Originally posted by godhead
@burnout
Nice avatar! :D

cheers, right back at ya :cool:

@onesoul thanks for the kerneldeint tip, looks like an improvement on smoothdeinterlacer, and no artifacts like with decomb... (also found out a new NTSC2PAL avs script today, so cheers to any1 who helped with that)

[Axlemar]

I don't know how big of an issue this still is but, Are any of you still getting the 0004 errors about buffer overflow? I redid titles I had backed up before that gave me no problems with the older versions and with the newest version of .38 I get the overflow error. I always use Quenc with no deinterlace and dynamic cell rates in one click mode. The dvd has episodes from a series on it. Thanks for your time and good luck getting the engine stable, I can't wait till the next improvement!

[jdobbs]

Originally posted by Axlemar
I don't know how big of an issue this still is but, Are any of you still getting the 0004 errors about buffer overflow? I redid titles I had backed up before that gave me no problems with the older versions and with the newest version of .38 I get the overflow error. I always use Quenc with no deinterlace and dynamic cell rates in one click mode. The dvd has episodes from a series on it. Thanks for your time and good luck getting the engine stable, I can't wait till the next improvement! The older version were broken and didn't let you know. The newer ones catch this and tell you about it.

[lab-one]

doom9 has posted a DVD-RB guide (http://www.doom9.org/mpg/dvdrb.htm)

[jdobbs]

New Version 0.39 Posted

I've posted a new version of DVD ReBuilder. You can access it through the first page of this thread. Below is a summary of the changes:

- Fixed a bug in which LPCM that was deselected (chosen to be removed) was not being properly removed or calculated for removal.

- Fixed an error related to field based stream processing. This should eliminate many if not all of the "Error code '9'" messages received in the PREPARE phase. A special thanks to Msc_Alex and RB for helping solve this one.

- Added an INI option making it possible to set the minimum and maximum bitrates when doing CCE VBR encoding. They are min_bitrate=nnnn and max_bitrate=nnnn (in Kbps. Example: max_bitrate=9000). Just set them under [Options] in REBUILDER.INI. No error checking is done on the numbers so take care. This should be changed only by advanced users as setting them incorrectly can seriously degrade the quality of the picture. Also note that hanging them can cause oversized or undersized output.

- Implemented new code for handling BFF sources. Instead of converting to TFF as was done in v0.38, this version duplicates the original more closely (if the source cell was bottom field first, the newly created cell will be also.

- Changed logic in batch mode so the status window isn't cleared for each job and all activity can be printed or reviewed upon completion.

[jptheripper]

well it sound like you just did half the work of hat i was asking for.

Instead of putting the min max bitrate in the rebuilder.ini options, could you find a way to make it per titleset, and add avg_bitrate? that way we could set extras to a lower rate.

just realized youwould have to redo the math.. sorry thought you were closer than you are.,

thanx for chruning out new versions anyway

[wmansir]

EDIT: never mind, it was a misunderstanding.

[jptheripper]

Actually it was a complement, and I am a native english speaker with a masters degree, I just typed too fast.

My point, more coherently, was that jdobbs has been doing an amazing amount of work, and in the process of fulfilling someone elses request I though he had inadvertently also made progress in what I had requested.

And I don't see where it came off as demanding at all. Regardless, that comment should have been sent as a PM. Since you posted in the forum I must respond in the forum.

Lets get this back on topic, and if you have any further comments please PM them.

-jp

[wmansir]

I apologize if I misread your comment. The way I read it was you said he 'just' did half the work you asked, as in: You only did half of what I asked.

I made the comment in public because I wanted to discourage that type of attitude.

[m3taPT]

Originally posted by jptheripper
Actually it was a complement, and I am a native english speaker with a masters degree, I just typed too fast.


You probably mean "compl-I-ment". Typed too fast again heh? ;)

[Xitrum]

I have never been able to get DVD-RB to work with CCE V2.67.10 or v2.67.11 or v2.67.23 after numerous attempts. After the first phase "prepare", dvd-rb displayed a popup window saying that the path to CCE was not set, even though I did set it (this is with v0.38 with previous versions, dvd-rb just open up CCE and got stuck there because CCE could not find .ecl files to do works.)

I can get dvd-rb to work with CCE 2.5 or 2.66 (the versions that available to download from doom9).

Did anyone had this problem before? I would appreciated if you can tell me how to fix this problem.

Software I used:
dvd-rb v0.38
CCE v2.67.xxxx with eclecc v1.8b
Avisynth v2.54 with decodefix100.zip file extracted into the plugins directory.

Thanks in advance.

[lab-one]

Did you open ECLCCE and select the CCE executable and also point DVD-RB to ECLCCE in the DVD-RB setup? You can find a more thorough explenation here (http://forum.doom9.org/showthread.php?s=&threadid=74308).

[The_Flash]

I'm sure this is coincidence, but EVERY backup I've done has resulted in a final size of 4.32GB. Any plans to implement an option that would allow users to control how conservative they want the bitrate calculator to be?

[bohemian]

@jdobbs
I have tried backing up my KISS Symphony DVD (PAL - R4) with each version, and always got the runtime error 9 at the start of the Prepare phase.
I have`nt posted any bug reports because many others have already done so, and you were already aware of the problem.
I read a post recently where you mentioned you may have found the cause of the error, well obviously you did because I finaly managed to run the prepare phase successfully with v0.39, congrats :)
I ran a vob through bitrate viewer in case your were interested which is shown below.
Now I can finaly encode this disk and see what sort of quality I can get.



Num. of picture read: 80675
Stream type: MPEG-2 MP@ML CBR
Resolution: 720*576
Aspect ratio: 16:9 Generic
Framerate: 25.00
Nom. bitrate: 6000000 Bit/Sec
VBV buffer size: 112
Constrained param. flag: No
Chroma format: 4:2:0
DCT precision: 10
Pic. structure: Field
Field topfirst: No
DCT type: Field
Quantscale: Nonlinear
Scan type: Alternate
Frame type: Interlaced

[JDay]

@The_Flash

It's documented in rebuilder.txt that you can add "TargetSectors=(X)" to customize the output size. 2236400 is DVD-RB's default, and 2297888 is supposedly the max for dvd-5. Wouldn't recommend setting it to the max. I encoded a movie on the maximum setting, which ended up several megs oversized (had to re-do the whole thing).

@jdobbs

It would be nifty if you could add some options for QuEnc, such as switches for trellis, kvcd, 2-pass, etc. (well, everything except input, output, aspect ratio and bitrate). Or if you could have an option in the .ini such as "QuEncCommandline=(X)". The commandline would the change to: "-i (X) -o (X) -b (X) -aspectratio (X) [custom options from .ini] -auto -close"

Also regarding QuEnc, it seems to output everything as progressive at the moment, so deinterlacing is needed. FieldDeinterlace(blend=false) seems to yield better (on telecined film) or identical (on true interlaced video) results (compared to the default, "blend=true") with everything I've tried it on. Could you make that the default?

[Xitrum]

Did you open ECLCCE and select the CCE executable and also point DVD-RB to ECLCCE in the DVD-RB setup? You can find a more thorough explenation here.

Yes I did all that, as that what I had to do to get CCE v2.50 or 2.66 to work

By the way, since the last post, I read the new DVD-Rebuilder tutorial and found that even though I'm using CCE SP v2.67xxxx, in the "option->CCE options" menu I have to tick the "CCE SP (v2.66+)" for it to work. I mistakenly tick the CCE Basic (v2.67+) before.

[Gunner-GoNad]

Originally posted by Xitrum
I have never been able to get DVD-RB to work with CCE V2.67.10 or v2.67.11 or v2.67.23 after numerous attempts. After the first phase "prepare", dvd-rb displayed a popup window saying that the path to CCE was not set, even though I did set it (this is with v0.38 with previous versions, dvd-rb just open up CCE and got stuck there because CCE could not find .ecl files to do works.)

I can get dvd-rb to work with CCE 2.5 or 2.66 (the versions that available to download from doom9).

Did anyone had this problem before? I would appreciated if you can tell me how to fix this problem.

Software I used:
dvd-rb v0.38
CCE v2.67.xxxx with eclecc v1.8b
Avisynth v2.54 with decodefix100.zip file extracted into the plugins directory.

Thanks in advance.

I had this same problem. You need to point the eclecc to the correct exe. But you need to put eclecc path under (path to CCE SP (NEW)). Then you need to choose CCE SP (v2.66+) as the encoder. That should work.

- Gunner-GoNad

[Fr4nz]

Is it possible to extract the "whole" m2v from the vobs instead of extracting EVERY chapter?

I think this could be useful for CCE, in order to plan better the encoding with a better bitrate distribution over the film.

[SAPSTAR]

[QUOTE]Originally posted by jdobbs
...

The latest version is v0.39, 15 Apr 2004. You can get it here. (http://forum.doom9.org/attachment.php?s=&postid=472904) Read about it here. (http://forum.doom9.org/showthread.php?postid=472904)

....
[QUOTE]

Two little things :

- The link is still pointed to the v0.38 version....

- New error in the Rebuild phase (v0.38) :
Erreur d'execution '75', Erreur dans le chemin d'accés
(English translation : Execution error 75, wrong path)
I checked all needed files are there (m2v, flags, ...).

[jdobbs]

Originally posted by Fr4nz
Is it possible to extract the "whole" m2v from the vobs instead of extracting EVERY chapter?

I think this could be useful for CCE, in order to plan better the encoding with a better bitrate distribution over the film. Certainly its possible. But there are already many other packages that can do that. It defeats many of the features of DVD-RB. For example, neither CCE Basic nor QuEnc have the capability to force I-FRAME insertion at fixed points and therefore you could not set the chapter points accurately for rebuild.

The bit distribution when using "dynamic cell allocation uses the PREPARE scan through the original data as a "pass" of sorts and I think you will find that it yields results equal to an "entire" demux/encode.


ADDED: Corrected two embarrassing typos so I would look smarter than I really am.

[walkistalki]

SAPSTAR

refresh your browser cache. I'm getting version 0.39 fine.

regards

[jdobbs]

Originally posted by SAPSTAR
[QUOTE]Originally posted by jdobbs
...

The latest version is v0.39, 15 Apr 2004. You can get it here. (http://forum.doom9.org/attachment.php?s=&postid=472904) Read about it here. (http://forum.doom9.org/showthread.php?postid=472904)

....
[QUOTE]

Two little things :

- The link is still pointed to the v0.38 version....

- New error in the Rebuild phase (v0.38) :
Erreur d'execution '75', Erreur dans le chemin d'accés
(English translation : Execution error 75, wrong path)
I checked all needed files are there (m2v, flags, ...). :confused: Nothing related to path has changed recently... Does repointing either the source or working path make it go away?

[nwg]

The source/working Path is working for me. I have never had any problems with it.

[Msc_Alex]

Yes great you did it :D

No more runtime 9 for me, glad I could help.
Next BUG :figure this paypal thing out

I get this on my NT station after hitting About
http://img6.imageshack.us/img6/4686/donate.jpg

I will bypass this one
:p

[jdobbs]

Originally posted by Msc_Alex
Yes great you did it :D

No more runtime 9 for me, glad I could help.
Next BUG :figure this paypal thing out

I get this on my NT station after hitting About
http://img6.imageshack.us/img6/4686/donate.jpg

I will bypass this one
:p :eek: Hmm... that's a bug I definitely need to fix (it might explain a lot too). I looked in the forum rules and couldn't find a prohibition on this (unless this would be considered blatent advertising) so here's (https://www.paypal.com/xclick/business=DVD-RB%40comcast.net&item_name=DVD+ReBuilder+Development+Donation&item_number=DVD-RB-001&no_note=1&tax=0&currency_code=USD) where the button is supposed to go.

[Fr4nz]

Originally posted by jdobbs
Certainly its possible. But there are already many other packages that can do that. It defeats many of the featres of DVD-RB. For example, neither CCE Basic not QuEnc have the capability to force I-FRAME insertion at fixed points and therefore you could not set the chapter points accurately for rebuild.

The bit distribution when using "dynamic cell allocation uses the PREPARE scan through the original data as a "pass" of sorts and I think you will find that it yields results equal to an "entire" demux/encode.

Okay tnx! I'll try this dynamic cell allocation feature! Anyway, great work jdobbs! :)

[JDay]

The about screen/donate button worked fine for me (WinXP). Just donated, thanks!

[slafe]

Originally posted by Xitrum
I have never been able to get DVD-RB to work with CCE V2.67.10 or v2.67.11 or v2.67.23 after numerous attempts. After the first phase "prepare", dvd-rb displayed a popup window saying that the path to CCE was not set, even though I did set it (this is with v0.38 with previous versions, dvd-rb just open up CCE and got stuck there because CCE could not find .ecl files to do works.)

I can get dvd-rb to work with CCE 2.5 or 2.66 (the versions that available to download from doom9).

Did anyone had this problem before? I would appreciated if you can tell me how to fix this problem.

Software I used:
dvd-rb v0.38
CCE v2.67.xxxx with eclecc v1.8b
Avisynth v2.54 with decodefix100.zip file extracted into the plugins directory.

Thanks in advance.


I have the same problem..but now it`s works for me....in the dvd-rb- options it`s says cce basic (2.67+)....it`s only if you ran basic cce...if you run sp...2.66 and above then put the flag to cce sp (2.66+) it should work for you......"jdobbs maybe should rename it?"



i have ecl cce 1.81 and cce 2.67.0.0.23 and dvd-rb 0.39...put the eclcce patch to the cce ...and in dvdrb...you take ecl cce..under new cce..

good luck..

[SAPSTAR]

Originally posted by JDay
The about screen/donate button worked fine for me (WinXP). Just donated, thanks!

Same thing just did what I should have done earlier !!! jdobbs I can't imagine how much work it must be !!!