DVD Rebuilder (CCE One-Click Beta) Bug Reports

[Pedro Gouveia]

Originally posted by jdobbs
I just tested it on mine and it works correctly. If you look in the resulting .ECL file you should see a parameter that says vbr_pass=n -- n will equal the number you entered minus 1 (vaf pass plus n more). Looking at the file .INI you sent above the value should be vbr_pass=3.

I'm back. If we stop the encoding process, then make a change in the number of passes, the REBUILDER.ECL doesn't get updated.

It only remembers the number of passes defined when the prepare process started.

[StifflerStealth]

@redfive19: Glad you found the picure quality your looking for. I agree with you on the fact that DVD-RB is an amazing piece of software. Hmmm, wait, scratch that! It's a _tool_, not some _piece_ of software. :-D

Stiff

[jdobbs]

Originally posted by Pedro Gouveia
I'm back. If we stop the encoding process, then make a change in the number of passes, the REBUILDER.ECL doesn't get updated.

It only remembers the number of passes defined when the prepare process started. I see. I create the .ECL file as a part of the PREPARE. So you're correct, changing the value only affects anything that has not yet been prepared.

[jdobbs]

Originally posted by StifflerStealth
@redfive19: Glad you found the picure quality your looking for. I agree with you on the fact that DVD-RB is an amazing piece of software. Hmmm, wait, scratch that! It's a _tool_, not some _piece_ of software. :-D

Stiff I think that unless I get a couple of these bugs out it's going to be called a "piece of" something else completely.

[StifflerStealth]

Originally posted by jdobbs
I think that unless I get a couple of these bugs out it's going to be called a "piece of" something else completely.

Are you kidding or joking? It's hard to tell when ppl are when using text. Anyways, it's great with or without bugs.

Stiff

[MeridiusUK]

how do you know if the film is interlaced or no interlaced ????

[redfive19]

@Meridius

I tell with the app Bitrate Viewer. It comes up with all the info. I am sure there are other ways to tell. I know DoItFast4U has some command line app that scans the m2v file and determines the field order.

@jdobbs
bro, you're being way to hard on yourself (that is, if you are serious) I hope you've not taken anything I've said as an insult! I already told you, once this app gets worked out, I'd be willing to pay for it and I am sure others would too (although I don't know if you have plans to turn it to shareware). I think that's a huge compliment. Also, let me reiterate (sp?), this app worked on a DVD that according to DIF4U did not need vobID demuxing but must have anyway. You're app was the only way I was able to back it up with CCE. And it's an expensive set of dvd's I bought (Total Training for Adobe After Effects 6 - 13 DVD set! $450). Just with that, it's worth it's price. That's a huge investment and your app let me ensure that I don't have to purchase it again. Just my two cents!

-redfive

[robw]

just want to report that I did a batch process of 4 DVDs. I also turned on WinOFF so when the process was finished (20 hours later, whew!) my machine shut down.

These were very full DVDs of episodic material. I set up the ISO ripped files with Daemon's Tools virtual disk. After DVD-RB (3 passes since the original material was ~7.5 GB) finished, I took the output and ran it through DVDStripper to remove a small nag screen (the usual) and also to let it recreate the navigational stuff. The final DVD size was within 40-50 MB of full (i.e. around 4434 to 4444 MB). Video quality looks great.

[P3gasus]

Originally posted by redfive19
@Meridius

I tell with the app Bitrate Viewer.



Be careful! Bitrate viewer tells you the information written in the vob file witch is not always true. In Europe there are many dvds that declare themselves as interlaced (and so bitrate viewer tells you so) but they are not. Some examples are The lord of the rings the fellowship of the ring (Extended) and Pirates of the carribean.
Just in case you didn't know.
Ciao

[daxab]

I just tested it on mine and it works correctly. If you look in the resulting .ECL file you should see a parameter that says vbr_pass=n -- n will equal the number you entered minus 1 (vaf pass plus n more). Looking at the file .INI you sent above the value should be vbr_pass=3.Is this correct for 2.6x?

If I set the VBR passes option in the GUI to n, the .ECL file contains vbr_pass=(n-1), and CCE actually does n-1 passes.

So to get 3 passes, I have to set the number of passes to 4 in the GUI.

[psdos]

can someone say me what is VBR_bias adjust?

[robw]

Ideally you would set the vbr_bias according to the bitrate that is being encoded. As an illustration, although I have not confirmed these numbers, is that for a bitrate of 4000 kbit/s you might use a vbr_bias of 40, for a bitrate of 3000 kbit/s you might use a vbr_bias of 30, etc. Trilight has compiled a nice page outlining how the bias can be chosen.

http://www.trilight.com/dvdguides/flowchartguide/BiasAnalysis.htm

[wmansir]

Originally posted by psdos
can someone say me what is VBR_bias adjust?

VBR_bias governs the amount of variation CCE gives the bitrate between different parts of the movie. A high number means there will be little fluctuation in the bitrate, so action scenes may not be able to get all the bits they need to look good. A lower number means CCE will give more bits to scenes that require them, possibly at the expense of slower scenes.

[MeridiusUK]

so when you select the interlaced and non interlaced option how do you know then what you are picking.

as i have R2 dvds alot so will it make any diffrence to the film if this options is wroung

and if so why can dvd rebuilder not auto scan the film and then select it on its own ????

[P3gasus]

Originally posted by MeridiusUK
so when you select the interlaced and non interlaced option how do you know then what you are picking.

The best way to know if a mpeg2 file is interlaced or progressive is to open it into virtualdubmod, move the bar in a scene - change frame and look at it. If you don't know how to recognize an interlaced video frame this (http://mp3dom.altervista.org/guide/interprog/download/interprog.rar) can be helpful for you.

as i have R2 dvds alot so will it make any diffrence to the film if this options is wroung

The difference is really visible when you try to deinterlace a progressive file. In this case you will see visible encoding errors (strangely coloured blocks). If you encode a progressive video with interlaced settings but without deinterlacing it, you will not notice big differences.

and if so why can dvd rebuilder not auto scan the film and then select it on its own ????

I don't know

Bye

[robw]

This is a very good page on interlacing.

When is it harmful to deinterlace? When is it better to leave a picture interlaced?

http://www.100fps.com/video_resolution_vs_fluidity.htm

[P3gasus]

Originally posted by robw
This is a very good page on interlacing.

When is it harmful to deinterlace? When is it better to leave a picture interlaced?

http://www.100fps.com/video_resolution_vs_fluidity.htm

Very interesting link. :thanks: robw.:D
Bye

[jdobbs]

Originally posted by MeridiusUK
how do you know if the film is interlaced or no interlaced ???? It's flagged one way or the other within the MPEG stream

[jdobbs]

Originally posted by wmansir
VBR_bias governs the amount of variation CCE gives the bitrate between different parts of the movie. A high number means there will be little fluctuation in the bitrate, so action scenes may not be able to get all the bits they need to look good. A lower number means CCE will give more bits to scenes that require them, possibly at the expense of slower scenes. You're right here. Here is a quote from the CCE manual:

"Breaking into bit allocation strategy The encoder allocates bits based on the original evaluation standard, so that all images have the same visual quality. Changing the value of the Bias part breaks into this evaluation standard. 0 to 100 can be set here. The initial value is 30. As the value becomes smaller, more bits are allocated to complicated scenes, and at value 0, the bitrate fluctuation is largest. As this value becomes larger, more bits are allocated to simple scenes, and at value 100, streams closer to CBR are output."

I like to set the value between 10 and 15. Then more bits can be allocated to the scenes that are most likely to get blocky (high action).

[jdobbs]

Originally posted by StifflerStealth
Are you kidding or joking? It's hard to tell when ppl are when using text. Anyways, it's great with or without bugs.

Stiff I'm never serious about anything... :D

[P3gasus]

Originally posted by jdobbs
It's flagged one way or the other within the MPEG stream

Again what is written in the mpeg stream is not always true. At least for R2 pal DVDs! Please Jdobbs read my post in this thread at page 16 where I quoted a single line of a Meridius message about bitrate viewer. I think it can be helpful for your development. If you just read the mpeg stream information and set avisynth avs and cce conseuently, here in Italy (but I think also in other countries) we will have to change both avs and ecl by hand.
Bye

[wakebrder]

Wakebrder sees..............

*Jdobbs hard @ work*

*wakebrder hands Jdobbs a beer*

Donation is on the horizon.

Thanks

[Observer]

Most (95%) PAL movies are progressive (but often tagged as interlaced).

An easy way to manually check if a movie is interlaced or progressive is to open a VOB or M2V file in DVD2avi. Step through some part of the movie where there is motion. An interlaced movie will show horizontal lines around the moving object. You can not trust tools such as Bitrate Viewer when checking if a movie is interlaced.

For more information, please check the CCE faq in the doom9 forums.

[jdobbs]

Originally posted by daxab
Is this correct for 2.6x?

If I set the VBR passes option in the GUI to n, the .ECL file contains vbr_pass=(n-1), and CCE actually does n-1 passes.

So to get 3 passes, I have to set the number of passes to 4 in the GUI. Yes it is correct for all versions of CCE except Basic (it will only do two passes). The source of the confusion is that CCE doesn't count the .VAF creation as pass 1. But it is.

Try this: Create a small file containing just a 1000 frames. Then set the parameters in CCE to 3 passes. Now run it and watch what happens. You will see a .VAF run of 1000 frames followed by 3 additional runs of 1000 each -- or 4 passes. If you are using v2.50 you will see a single .VAF run followed by a pass that appears to be 3000 frames (again, 3 additional passes).

The bottom line is: the VAF pass is a pass. It takes the time of a pass and serves the purpose of a pass -- so I'm calling it a pass. Setting the ECL value to 3 takes four times as long as single pass VBR (setting Q). The last thing I want is for people who don't understand the realities of VBR to waste hours doing passes that don't really provide significant improvement. So my program is setting the value to the number that TRULY represents what is happening.

But for those that are adamant about it -- you can always set the value in DVD-RB to one higher than reality.

[Forko]

first of all thanks for the hard work!

the problem i am having is after it scans the vts's in the beginning it then trys to load the .ecl file into cce. cce then gets a avifileopen error. i noticed that when i try to load the ecl by hand i get the same error, but it i cange the file extension to lowercase it then loads it fine.

BTW i am running winxp pro and cce 2.50

thanks

[jdobbs]

Originally posted by Forko
first of all thanks for the hard work!

the problem i am having is after it scans the vts's in the beginning it then trys to load the .ecl file into cce. cce then gets a avifileopen error. i noticed that when i try to load the ecl by hand i get the same error, but it i cange the file extension to lowercase it then loads it fine.

BTW i am running winxp pro and cce 2.50

thanks Yeah, contrary to what some say there is still a bug associated with CCE 2.50 and audio. Under the avs options turn the ResampleAudio(44100) flag on, redo the Prepare, and it will likely work.

[Forko]

im sorry, where do i change that?

[jdobbs]

All,

Attached is v0.25 -- I didn't get everything in that I wanted, but I figured I'd post this version as it has been a couple days since I've posted. The changes are outlined below:

- Fixed the output sizing problem that was introduced in v0.22 -- DVD-RB should now fully size CCE output to take advantage of an entire DVD-R(W) disc.

- Added code to SETUP dialog that enables you to add the path to the MPEG2DEC.DLL file. I started to see errors related to load failures over-and-over -- if you set this and check the corresponding flag, the LoadPlugin() command is added as the first line to every .AVS file.

- Set a new default .VOB filesize (to match some other applications)

- Added code that disables entering either of the second two processes (ENCODE or REBUILD) unless the previous necessary processes (PREPARE or ENCODE) have been completed. I found that some folks were trying to skip steps and as a result were getting strange errors.

I'm still working on the pixelation/chapter stuttering that some are experiencing -- hard to do for me cause I'm having problems reproducing it. I'm still working and will post again soon...

ATTACHMENT REMOVED AFTER 2293 DOWNLOADS.

A newer attachment (v0.26) is available.

Click here to download (http://forum.doom9.org/attachment.php?s=&postid=466021)

[Forko]

duh, i see...

it is already checked off though

thanks

[acido]

Originally posted by jdobbs
All,

Attached is v0.25 -- I didn't get everything in that I wanted, but I figured I'd post this version as it has been a couple days since I've posted. [ ...CUT ]

Ciao jdobbs,
it's unclear to me if this version fixes the navpacks (chapters skipping) problem fixed by postprocessing with ifoedit.
Thanks, ciao

[Forko]

tested it with the new version and the menu item is checked off and it still wont work.

[JvD]

Do any of these fixes solve the runtime error 9?

[robot1]

I was looking at the working directory.
Does DV-RB create .avs and .ecl even in ReJig mode?

[jdobbs]

Originally posted by robot1
I was looking at the working directory.
Does DV-RB create .avs and .ecl even in ReJig mode? Yes. I had to do the pass anyway, so I thought I'd create them either way. I personally like to use them for reference when I'm doing advanced features. I could turn them off they are annoying...

[jdobbs]

Originally posted by acido
Ciao jdobbs,
it's unclear to me if this version fixes the navpacks (chapters skipping) problem fixed by postprocessing with ifoedit.
Thanks, ciao That has turned into quite the project. It's the main reason no release has come out in 2 days.

[robw]

jdobbs
Thanks for all your hard work. Every two days for an update is unreasonable. Take a break!

[robot1]

Originally posted by JvD
Do any of these fixes solve the runtime error 9?

I've got for the first time the runtime error 9:
the source is an .iso mounted with daemon tools.
I had other backups ripping in filemode with no problem.

Do you rip in iso, or in file mode?

[KungFuCow]

I think there has been some talk about problems with Daemon Tools. You might want to try Alcohol 120% and see if that helps.

[robot1]

I normally rip in filemode, so, no problem. As Jdobbs can't repeat the runtime error, I was guessing.

[jdobbs]

Originally posted by robot1
I normally rip in filemode, so, no problem. As Jdobbs can't repeat the runtime error, I was guessing. I always run in filemode. I guess I need to mount an image in Daemon tools and see what causes this. Just as an aside -- what is the advantage of mounting using Daemon tools instead of just copying using file mode?

[StifflerStealth]

Originally posted by jdobbs
I always run in filemode. I guess I need to mount an image in Daemon tools and see what causes this. Just as an aside -- what is the advantage of mounting using Daemon tools instead of just copying using file mode?

Mainly ppl coming over from the "Big 3" method use the source as an iso to ensure that no changes are ever made to the source. Using the "Big 3" method can generate a lot of files, and it can be easy to overwrite an orginal file.

Stiff

[jdobbs]

Originally posted by acido
Ciao jdobbs,
it's unclear to me if this version fixes the navpacks (chapters skipping) problem fixed by postprocessing with ifoedit.
Thanks, ciao I need help with this problem. Since I can't repeat it I have to rely on what the folks experiencing it can tell me. The big issue now is that I can run it through IFOEdit -- and compare the values to the original or I can run it through DVD Shrink nad compare it to the original -- but even IFOEdit and DVD-Shrinks output files don't match... so who (if any) are at fault (especially on the "first audio packet" offsets -- which seem to be the only differences I can find. Could you please tell me:

1. Does it happen on NTSC, PAL, or both?
2. If you run it through IFOEdit VOB Extras:
2a. Does is fix it if you don't check the "Correct Original IFOs"? If so, what happens if it isn't checked?
2b. Are any of the "VOB EXPERT OPTIONS" checked? What happens if you don't check them?

Your help would be appreciated.

[flaystus]

edit: oops moved suggestion to correct thread.

[KungFuCow]

Just finished Lost In Translation again. Sizing issues look to have been resolved.

Im running it through IFOedit now and about to burn it. Ill report back shortly.

[KungFuCow]

Off and burning. This DVD was corrected using ONLY the Correct Nav Packs option. All advanced options were unchecked as was Correct Original IFOs.

[jdobbs]

Originally posted by KungFuCow
Off and burning. This DVD was corrected using ONLY the Correct Nav Packs option. All advanced options were unchecked as was Correct Original IFOs. It was skipping/pixelating before you ran it through IFOEdit and then worked fine after? This is important!

[KungFuCow]

With above option ONLY checked, stuttering on chapter changing is still there. Ill run it through next with Correct Original Files enabled with no advanced options and reburn

[jdobbs]

Originally posted by KungFuCow
With above option ONLY checked, stuttering on chapter changing is still there. Ill run it through next with Correct Original Files enabled with no advanced options and reburn Thanks, I appreciate it. Can you try it with just "Correct VOB-Unit (NAVPACK.." and with "Adjust VOB-Unit pointer" and "Adjust Audio/Subp" checked in the VOB EXPERT area? I'm trying to see if the IFO has anything to do with the problem at all...

[KungFuCow]

I actually had it all ran through IFOEdit with Correct Original IFOs and NAV Packs only checked so I went ahead and burned it. Chapter issues and other wierdness still there. Im going to run it through now with the settings you wanted and reburn.

[jdobbs]

Originally posted by KungFuCow
I actually had it all ran through IFOEdit with Correct Original IFOs and NAV Packs only checked so I went ahead and burned it. Chapter issues and other wierdness still there. Im going to run it through now with the settings you wanted and reburn. Thanks. If that fixes the problems, I might have this fixed tomorrow.