CCE One-Click Software Beta

[djan]

Hi,

I ran the firt time an interlaced DVD. I so checked the ConvertToYUY2(interlaced=true) option. Unfortunately, it didn't do well its job. The resulting DVD is always interlaced and not so pretty to look. Someone can explain me why ? Thx. In the other hand, I would like to know if I really need to deinterlace as I intend to look this DVD only on my standalone ?

Thx for any help.

[philos31]

Ahhhh,

AFAIK this is my first posting on this forum...
Read a lot, but never hit the reply button....

This software is worth writing a reply about, I've been using it about 2 weeks now, and after a lot of trouble to get it to run on my very dirty harddisk (the prob was 3 versions of avisync)

Now that it works I can say that the program does what it tell's you.
I am amazed at the quality you get out of such a simple (for the user) program.

I will be posting a bit more here now that my first message is out :)

Just wanted to say thanks, great job!

[Paced]

Originally posted by djan
Hi,

I ran the firt time an interlaced DVD. I so checked the ConvertToYUY2(interlaced=true) option. Unfortunately, it didn't do well its job. The resulting DVD is always interlaced and not so pretty to look. Someone can explain me why ? Thx. In the other hand, I would like to know if I really need to deinterlace as I intend to look this DVD only on my standalone ?

Thx for any help.

Hello, CovertToYUY2(interlaced=true) has no effect on whether or not your output will/won't be interlaced. You won't need to deinterlace if you intend to play this on your stand-alone - you only notice the 'interlacing' on your computer (depending on the program you view the DVD with - use PowerDVD/WinDVD/Ulead DVD Player instead).

[quinn]

Love this program (donated my $15 for it) - have a question and a comment -

question - how does it find dvd2avi? I deleted all copies on my disks but the dvd2avidg to be sure, but it might be nice to be able to control it....

comment - I use 1/2 D1 a lot, but the option seems to be global - it would be nice if a) it didn't do it to 16:9 video, since that doesn't work well with most stand-alones, and doesn't conform to the spec and b) that the conversions (16:9 -> 4:3 and 1/2 D1 (and possibly both)) were on a VTS by VTS basis...

This program rocks!! Thanks a bunch

[jdobbs]

Originally posted by quinn
Love this program (donated my $15 for it) - have a question and a comment -

question - how does it find dvd2avi? I deleted all copies on my disks but the dvd2avidg to be sure, but it might be nice to be able to control it....

comment - I use 1/2 D1 a lot, but the option seems to be global - it would be nice if a) it didn't do it to 16:9 video, since that doesn't work well with most stand-alones, and doesn't conform to the spec and b) that the conversions (16:9 -> 4:3 and 1/2 D1 (and possibly both)) were on a VTS by VTS basis...

This program rocks!! Thanks a bunch Thanks! DVDAVI is never called. DVD-RB does the scanning/breakout function itself. I'll take a look at disallowing 1/2 D1 on 16:9 conversion.

[djan]

Originally posted by Paced
Hello, CovertToYUY2(interlaced=true) has no effect on whether or not your output will/won't be interlaced. You won't need to deinterlace if you intend to play this on your stand-alone - you only notice the 'interlacing' on your computer (depending on the program you view the DVD with - use PowerDVD/WinDVD/Ulead DVD Player instead). Hi, I'm using WinDVD 5 Gold, but it's always interlaced. Is there an option to deinterlace ? I'm a PAL user. And for what is the option interlaced=true so ? Can I backup it without using this option ? Thx.

[StifflerStealth]

@djan:
Try this in the AVS files after the mpeg2source line:

LoadPlugin("x:\Path\to\decomb.dll")
FieldDeinterlace()

I think you still need the interlaced=true, though. I said I _think_.

Well, this must pf helped redfive19, because I haven't seen him in 13 pages. :D

Stiff

[DDogg]

jdobbs, I remembered one of my odd pieces of historical information. You, or somebody in one of these threads, was talking about drag and drop of an ecl into 2.5 throwing that weird "According to the registry, the type of file specified ...blah, blah".

This happens when caps are used for the ecl file suffix. ".ecl" will work, whereas ".ECL" will not. Maybe you already figured that out, but there was too much thread for my eyes to go back through. Just disregard if this is useless information to you.

[jdobbs]

Originally posted by DDogg
jdobbs, I remembered one of my odd pieces of historical information. You, or somebody in one of these threads, was talking about drag and drop of an ecl into 2.5 throwing that weird "According to the registry, the type of file specified ...blah, blah".

This happens when caps are used for the ecl file suffix. ".ecl" will work, whereas ".ECL" will not. Maybe you already figured that out, but there was too much thread for my eyes to go back through. Just disregard if this is useless information to you. If that turns out to be all that is wrong I'm going to slam my head into a wall. Hmm... not that I think about it, in a sense that's all programming really is: Slamming your head against a wall for hours so you can enjoy the feeling when you get to stop slamming.

[DDogg]

If that turns out to be all that is wrong I'm going to slam my head into a wall. Hey, you you would be in good company. This snake has bitten some very smart people over the years because smart people like you would not even consider that such a dumb error could possible exist. That's for people like me to do :)

/add: also, putting any extra information at the top of an 2.5 ecl file can sometimes cause problems (or sometimes not - Voodoo stuff). I note you put "; Created by DVD ReBuilder". It is not causing any problem that I can see, but I specifically remember an instance where it did in another project.

[RB]

Originally posted by DDogg
jdobbs, I remembered one of my odd pieces of historical information. You, or somebody in one of these threads, was talking about drag and drop of an ecl into 2.5 throwing that weird "According to the registry, the type of file specified ...blah, blah".

This happens when caps are used for the ecl file suffix. ".ecl" will work, whereas ".ECL" will not. Maybe you already figured that out, but there was too much thread for my eyes to go back through. Just disregard if this is useless information to you.
DDogg, you're the man! :) I think that's it. Anyway, I already "fixed" it in EclCCE 1.8b so jdobbs, there's really no need to do bad things to your head :)

As for why that is, it's an easy to make programming error. They obviously perform a case-sensitive match against the file extension for files dropped into CCE, e.g. "if (strcmp(szFileExt, "ecl")) { /* Not an ECL file */ }"

[jdobbs]

Originally posted by RB
DDogg, you're the man! :) I think that's it. Anyway, I already "fixed" it in EclCCE 1.8b so jdobbs, there's really no need to do bad things to your head :)

As for why that is, it's an easy to make programming error. They obviously perform a case-sensitive match against the file extension for files dropped into CCE, e.g. "if (strcmp(szFileExt, "ecl")) { /* Not an ECL file */ }" Yep -- just renamed it and it worked fine...

[hypo20]

@jdobbs
Hi. It's weekend, my time. :) Will post some tests. :)

I see you slowed down little bit. That's good. I don't want to hear about you, that you're slamming your head to the wall saying "Rebuilder, Rebuilder, Rebuilder, Rebuilder, ....." somewhere in sanatorium. :)

We all don't want to loose our greatest brain. :D

[casonbang]

@jdobbs

Thanks for adding the check for interleaving! I just tried to backup Princess Mononoke (which I didn't think had anything wanky) and the warning saved me from a few days of scratching my head wondering why it wouldn't work.

Any rough idea of when you'll tackle the interleaving problem? Days, weeks, months...

PS: $10 donation on its way!

[deh707]

this is my first time trying this program

im using DVD-RB 028, with CCE 2.50


i did the one click transcode mode, and it worked.

the only problem was it never took out the audio tracks i told it to remove, making the final VIDEO_TS folder size 4.66GB. (if it removed the other 3 tracks, it would be 4.38GB)




anyone know why it still included the unwanted audio tracks?

[jdobbs]

Originally posted by hypo20
@jdobbs
Hi. It's weekend, my time. :) Will post some tests. :)

I see you slowed down little bit. That's good. I don't want to hear about you, that you're slamming your head to the wall saying "Rebuilder, Rebuilder, Rebuilder, Rebuilder, ....." somewhere in sanatorium. :)

We all don't want to loose our greatest brain. :D I haven't stopped... I've been trying to chase down a couple of fairly elusive bugs. Mainly I'm trying to see what happened between version .26 and .27 that seems to have brought about the "Runtime error '6'"

[John Doe]

Great tool. But I´ve one problem. CCE starts everytime in 2-pass mode, although it ought to start in 3-pass mode. :(
EclCCE is in version 1.8b & CCE SP in 2.67.00.23.

By using the Big3-method everything seams fine, so it have to be a bug of DVD-RB.

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

Sorry, wrong thread

[jdobbs]

Originally posted by John Doe
Great tool. But I´ve one problem. CCE starts everytime in 2-pass mode, although it ought to start in 3-pass mode. :(
EclCCE is in version 1.8b & CCE SP in 2.67.00.23.

By using the Big3-method everything seams fine, so it have to be a bug of DVD-RB.

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

Sorry, wrong thread Use "Search" on this subject in the DVD-RB threads. 2 in CCE = 3 in real life. (3=4, 4=5, etc.)

[Offline]

That "Runtime error '6'" bug is a pain. I'm going back to v0.26
which seemed to work fine. I hope you find a solution.

In the meantime please do not disallow half D1 for 16:9. I know
it is not in the "DVD specs" but it works fine and allows for
superior picture at below 2,000kb rates on those really overstuffed
DVD9's. Perhaps a simple warning instead?

Fantastic program btw!!!

[jdobbs]

Originally posted by Offline
That "Runtime error '6'" bug is a pain. I'm going back to v0.26
which seemed to work fine. I hope you find a solution.

In the meantime please do not disallow half D1 for 16:9. I know
it is not in the "DVD specs" but it works fine and allows for
superior picture at below 2,000kb rates on those really overstuffed
DVD9's. Perhaps a simple warning instead?

Fantastic program btw!!! Actually I just went to the DVD FAQ and checked, and anamorphic 16:9 is acceptable for Half-D1. Here's a quote (I modified the table by adding slashes for formatting):

For anamorphic video, the pixels are fatter. Different pixel aspect ratios (none of them square) are used for each aspect ratio and resolution. 720-pixel and 704-pixel sizes have the same aspect ratio because the first includes overscan. Note that conventional values of 1.0950 and 0.9157 are for height/width (and are tweaked to match scanning rates). The table below uses less-confusing width/height values (y/x * h/w).

Aspect/720x480/720x576/352x480/352x576
4:3/0.909/1.091/1.818/2.182
16:9/1.212/1.455/2.424/2.909

BTW. I already found the solution and will post it today.

[redfive19]

Originally posted by StifflerStealth
@djan:
Try this in the AVS files after the mpeg2source line:

LoadPlugin("x:\Path\to\decomb.dll")
FieldDeinterlace()

I think you still need the interlaced=true, though. I said I _think_.

Well, this must pf helped redfive19, because I haven't seen him in 13 pages. :D

Stiff

@StifflerStealth

You're damned right! I've already informed jdobbs of this bug. After manually adding those two lines to each AVS, my interlaced material is encoded fine. This has worked on two DVD's that did not previously work. I am very suprised that more people haven't seen this as I've found DVDRB to only work (without AVS manipulation) on progressive DVD's. My problem was a bit different than DJAN though, I saw the problems on my standalone, NOT my PC.

Does anyone know an easy way to add lines to multiple AVS's? I've tried ultraedit's macro but you cannot apply it over multiple opened files. Any easier way I am just not seeing?

-redfive

[Roe5685]

on april 2 i tried the lastest rebuilder with cce 2.5.
"process" goes fine.
when you go to "encode" it then processes the d2v files without doing anything so it finishes in seconds rather than hours and of course encode fails.
i examined the ecl file and a few of the many d2v files in the rebuilder d2vavs folder and they seem fine. out of curiosity i started cce and loaded the ecl file and hit encode. over several hours it processed the files. somewhere along the line it hit a checksum error which i too have had on occasion so i did not end up with useable streams to author.

in summary then:
there appears to be some very slight clerical error such that rebuilder does not start cce 2.5 encoding. on the other hand the files made up by rebuilder for cce 2.5 to process seem to be fine.
i hope that the error will be corrected shortly.
thankyou for letting me try this adventurous program.

Rosemary.

[jdobbs]

Originally posted by Roe5685
on april 2 i tried the lastest rebuilder with cce 2.5.
"process" goes fine.
when you go to "encode" it then processes the d2v files without doing anything so it finishes in seconds rather than hours and of course encode fails.
i examined the ecl file and a few of the many d2v files in the rebuilder d2vavs folder and they seem fine. out of curiosity i started cce and loaded the ecl file and hit encode. over several hours it processed the files. somewhere along the line it hit a checksum error which i too have had on occasion so i did not end up with useable streams to author.

in summary then:
there appears to be some very slight clerical error such that rebuilder does not start cce 2.5 encoding. on the other hand the files made up by rebuilder for cce 2.5 to process seem to be fine.
i hope that the error will be corrected shortly.
thankyou for letting me try this adventurous program.

Rosemary. Could you give me a little information on how you are doing the encode?

1. What version of eclCCE are you using with CCE 2.50?
2. After the "PREPARE" -- try opening one of the .AVS files with Windows Media Player and make sure it works.

[smlong426]

The chapter pixelation problems were fixed with .26, but I still see stuttering of either audio or video (or both) on my two hardware players.

Also, is there any way to fix previous Rebuilder images that exhibit the 'no audio' problems on certain players without re-encoding the whole thing all over ?

[jdobbs]

Originally posted by redfive19
@StifflerStealth

You're damned right! I've already informed jdobbs of this bug. After manually adding those two lines to each AVS, my interlaced material is encoded fine. This has worked on two DVD's that did not previously work. I am very suprised that more people haven't seen this as I've found DVDRB to only work (without AVS manipulation) on progressive DVD's. My problem was a bit different than DJAN though, I saw the problems on my standalone, NOT my PC.

Does anyone know an easy way to add lines to multiple AVS's? I've tried ultraedit's macro but you cannot apply it over multiple opened files. Any easier way I am just not seeing?

-redfive You won't have to find another way. This will be in v0.29.

[issa]

I am using DVD-RB 0.28. I got "Run-time Error '6': Overflow"
after I pick the source path. I did try few different DVDs,
and all gave me the same result. Is there a way to fix
this problem?

Also, jdobbs, could you replace "AddAudio()" with "AudioDub(BlankClip())"? That what the "AddAudio()" do.
It is repleacment of "Resample()", since "Resample()"
no longer add empty clip when it used.

[jdobbs]

Originally posted by issa
I am using DVD-RB 0.28. I got "Run-time Error '6': Overflow"
after I pick the source path. I did try few different DVDs,
and all gave me the same result. Is there a way to fix
this problem?

Also, jdobbs, could you replace "AddAudio()" with "AudioDub(BlankClip())"? That what the "AddAudio()" do.
It is repleacment of "Resample()", since "Resample()"
no longer add empty clip when it used. Will this work with all versions of AVISYNTH? If so, I'll remove both options and put this in every AVS. Is there any drawbacks? (speed?)

[StifflerStealth]

Originally posted by redfive19
@StifflerStealth

You're damned right! I've already informed jdobbs of this bug. After manually adding those two lines to each AVS, my interlaced material is encoded fine. This has worked on two DVD's that did not previously work. I am very suprised that more people haven't seen this as I've found DVDRB to only work (without AVS manipulation) on progressive DVD's. My problem was a bit different than DJAN though, I saw the problems on my standalone, NOT my PC.

Does anyone know an easy way to add lines to multiple AVS's? I've tried ultraedit's macro but you cannot apply it over multiple opened files. Any easier way I am just not seeing?

-redfive

I was going to suggest to write a dos batch script, but since it will be in the new version, I guess it's not needed to do that.

Stiff

[jdobbs]

Originally posted by issa
I am using DVD-RB 0.28. I got "Run-time Error '6': Overflow"
after I pick the source path. I did try few different DVDs,
and all gave me the same result. Is there a way to fix
this problem?

Also, jdobbs, could you replace "AddAudio()" with "AudioDub(BlankClip())"? That what the "AddAudio()" do.
It is repleacment of "Resample()", since "Resample()"
no longer add empty clip when it used. Just goes to show you... sometimes you unnecessarily take the hard road. I just ran this and it fixes the CCE bug fine. I'm going to get rid of the options I have now and put this in every VTS.

Thanks for the excellent advice!

[issa]

jdobbs,

In AviSynth 2.0.x, use "Resample()".
In AviSynth 2.5.x, use "AudioDub(BlankClip())".
Since "Resample()" in AviSynth 2.5.x will not produce
empty audio clip, the problem still exsit even you
put "Resample(44100)" in the script.

Also, where can I find the DVD-RB 0.26?
I would like to try out is the problem of my DVD or
the hidden problem in the program.

[StifflerStealth]

Originally posted by jdobbs
Will this work with all versions of AVISYNTH? If so, I'll remove both options and put this in every AVS. Is there any drawbacks? (speed?)

I have an Intel P4 and CCE SP 2.67, so I don't use Addaudio or ResampleAudio at all. I noticed that the encode took _10 hours_ with the AddAudio() option, but the very same project took _6.5 hours_ without the AddAudio option when I tried with a newer version of DVD-RB. So you can have a check thing like you got now, but have that one audio option, so I (and others) can uncheck the audio option. AFAIK, the audio options are there for ppl with CCE 2.50 and an AMD proccessor, right?!?

Stiff

EDIT: Added: "without the AddAudio option"

[jdobbs]

Originally posted by issa
jdobbs,

In AviSynth 2.0.x, use "Resample()".
In AviSynth 2.5.x, use "AudioDub(BlankClip())".
Since "Resample()" in AviSynth 2.5.x will not produce
empty audio clip, the problem still exsit even you
put "Resample(44100)" in the script.

Also, where can I find the DVD-RB 0.26?
I would like to try out is the problem of my DVD or
the hidden problem in the program. Give me a few hours and I'll have 0.29 posted. I advise against the use of 0.26, as there are errors that I've mentioned before. The folks who used it were doing so to help me find the bug that was introduced in 0.27 and I am certainly appreciative for that -- but I've fixed it now, and I hate to have someone build a DVD not knowing that it has some inherent problems.

[issa]

Thanks for give us this great program.

[redfive19]

Originally posted by jdobbs
You won't have to find another way. This will be in v0.29.

@jdobbs
You're the man!!! Thanks!

[issa]

The "AddAudio()" is a sample script to replace "Resample()"
to fix the bug of CCE 2.5 in AviSynth 2.5.x.
You can find the script here (http://www.avisynth.org/index.php?page=AddAudio).
Someone did report CCE 2.6x had memory leak without adding
the empty audio, here (http://forum.doom9.org/showthread.php?threadid=42924&perpage=20&highlight=memory%20leak&pagenumber=3).

[DDogg]

jdobbs, I think some people don't know that AddAudio() depends on having AddAudio.avsi in your plugin dir.

A one size fits all (versions of CCE) solution is to do what dvd2svcd does via an import as part of your script. Please see the note below this for a clarification.

LoadPlugin("PATH:\MPEG2Dec3.dll")
mpeg2source("PATH:\V01B.D2V")
trim(0,14136)
ConvertToYUY2()
Import("PATH:\PITA_CCEWorkaround.avs")
PITA_CCEWorkaround(44100)

where PITA_CCEWorkaround.avs is:

function PITA_CCEWorkaround(clip v1, int rate) {
v2 = Blankclip()
v1 = AudioDub(v1,v2)
return v1
}

Note: Dvd2svcd actually use the name "ResampleAudio" as the function name for some backward compatibility reasons. Although there is actually an internal command which is called ResampleAudio, he gets away with it by expoiting the fact that a function takes precedence over an internal command. I purposely changed the name above so as to not cause confusion. You can name it anything you want. Hopefully, this will not cause you more confusion :)

/Add: hmmm, are you sure you even need this? If I remember it correctly, this was a workaround for loading (or dragging) the avs in CCE, because the audio checkbox was checked. Once you unchecked it and saved the ecl, you did not need the workaround anymore. It occurs to me that since you are creating the ecl with the audio already turned off, and loading via a commandline, you may not even need to be doing this. Crap, I hope I got that correct.

[RB]

Originally posted by redfive19
@StifflerStealth

You're damned right! I've already informed jdobbs of this bug. After manually adding those two lines to each AVS, my interlaced material is encoded fine. This has worked on two DVD's that did not previously work. I am very suprised that more people haven't seen this as I've found DVDRB to only work (without AVS manipulation) on progressive DVD's. My problem was a bit different than DJAN though, I saw the problems on my standalone, NOT my PC.

Hmm, this surprises me. Normally encoding a progressive source as interlaced has almost no negative effect, the Q factors may get a little higher and CCE speed will be a little lower but that's about it. In fact, most DVDs are encoded interlaced although you can clearly see that the source material is progressive. In a software player, this might cause a slight vertical jitter because the player applies a BOB deinterlace when it really doesn't need to, usually fixed by setting deinterlace to "Force Weave" (WinDVD at least).

So how did it look wrong on your TV? I can only imagine that a wrong setting for Upper Field First/Offset Line in CCE could cause problems here.

jdobbs, am I right in that the fix you are mentioning is setting the CCE options (alternate_scan, progressive) according to the source stream? If so, you could also automate the Interlaced=True/False parameter of ConvertToYUY2(). It must be set to interlaced=true if the picture headers in the source don't have the progressive_frame flag set. If there's a mixture in the cell, use interlaced=true too because it has almost no negative impact on progressive sources.

[Roe5685]

your reply got me thinking. thanks.
I inspected the d2vavs folder and saw it was over 2 gigs.
i then used your rebuild third step. it failed the first time due to "overrun" and the second time due to some other error. nevertheless it generated some vob files and i played the first one on my holywood plus decoder.
i had just opened up my cce 2.5 and used that to generate the m2v files. i.e. you open the ecl file in d2vavs and then hit encode.
i have never used eclcce before. when i encode by hand i edit my ecl template and then open cce and hit encode. on rare occasion i have used the method not requiring a template ecl from a guide but then you have to be awake to restart cce after the first pass.
nevertheless i took your hint and opened the latest version of eclcce which i just downloaded in the last 24 hours. i have just used that to start the encoding of your d2v files. it is progressing nicely as we speak but failed on at least two d2v files on checksum. i had erased everything in your d2davs folder so i don't know how this happened. i will author using your third step when it completes despite the checksum errors.
i have no idea how to use eclcce to generate a one click solution.
Any advise?

Thanks again

Rosemary.

[RB]

Originally posted by DDogg
/Add: hmmm, are you sure you even need this? If I remember it correctly, this was a workaround for loading the avs in CCE, because the audio checkbox was checked. Once you unchecked it and saved the ecl, you did not need the workaround anymore. It occurs to me that since you are creating the ecl with the audio already turned off, and loading via a commandline, you may not even need to be doing this. Crap, I hope I got that correct.
Absolutely correct. Ifaud_out=0 is in the ECL, the AddAudio() snippet is not necessary for CCE 2.50. However, funnily enough, it's still advised for CCE 2.66+ because otherwise CCE will continously increase it's memory usage with every VBR pass (not an issue with CCE 2.50!). You can see this in task manager, grabs a new large chunk of memory at the start of every pass and never releases it until you close CCE. This does not happen when there's an audio track in the AVS.

[jdobbs]

Originally posted by DDogg
/Add: hmmm, are you sure you even need this? If I remember it correctly, this was a workaround for loading (or dragging) the avs in CCE, because the audio checkbox was checked. Once you unchecked it and saved the ecl, you did not need the workaround anymore. It occurs to me that since you are creating the ecl with the audio already turned off, and loading via a commandline, you may not even need to be doing this. Crap, I hope I got that correct. [/B] You're right. But since I made the AVS files available to everyone to use, CCE 2.50 will crash whenever someone trys to load it -- so I decided it was best just to add it and kill the confusion.

[jdobbs]

Originally posted by RB
jdobbs, am I right in that the fix you are mentioning is setting the CCE options (alternate_scan, progressive) according to the source stream? If so, you could also automate the Interlaced=True/False parameter of ConvertToYUY2(). It must be set to interlaced=true if the picture headers in the source don't have the progressive_frame flag set. If there's a mixture in the cell, use interlaced=true too because it has almost no negative impact on progressive sources. Great minds think alike. I've done that.

[RB]

Great! :)

[jdobbs]

Originally posted by RB
Hmm, this surprises me. Normally encoding a progressive source as interlaced has almost no negative effect, the Q factors may get a little higher and CCE speed will be a little lower but that's about it. In fact, most DVDs are encoded interlaced although you can clearly see that the source material is progressive. In a software player, this might cause a slight vertical jitter because the player applies a BOB deinterlace when it really doesn't need to, usually fixed by setting deinterlace to "Force Weave" (WinDVD at least).

So how did it look wrong on your TV? I can only imagine that a wrong setting for Upper Field First/Offset Line in CCE could cause problems here.

jdobbs, am I right in that the fix you are mentioning is setting the CCE options (alternate_scan, progressive) according to the source stream? If so, you could also automate the Interlaced=True/False parameter of ConvertToYUY2(). It must be set to interlaced=true if the picture headers in the source don't have the progressive_frame flag set. If there's a mixture in the cell, use interlaced=true too because it has almost no negative impact on progressive sources. I read a post earlier with interest. It said that sometimes a stream is listed as interlaced even when it is progressive??? Are you saying there are times when the picture_structure is set to TF or BF (instead of FRAME) and the the progressive flag is set? If so I have to go back and do some recoding...

Note: Edited to make myself look smarter.

[DDogg]

Originally posted by RB
jdobbs, am I right in that the fix you are mentioning is setting the CCE options (alternate_scan, progressive) according to the source stream? Originally posted by jdobbs
Great minds think alike. I've done that. For clarification - Meaning that progressive streams will have ZigZag and progressive on, and alternate scan off? (I hope)

As for the cce audio workaround, what both of you said makes perfect sense. Leaving it in all the scripts for all versions of cce is the way to go. It certainly hurts nothing. No need to have the settings in dvd-rb anymore, just a universal command or import in every script. That will make it a lot cleaner.

[najt]

What programming language / development enviroment are you using?

[RB]

Originally posted by jdobbs
I read a post earlier with interest. It said that sometimes a stream is listed as interlaced even when it is progressive??? Are you saying there are times when the picture_structure is set to TFF or BFF (instead of FRAME) and the the progressive flag is set? If so I have to go back and do some recoding...
I read that too in relation with BitrateViewer, but it's not true. Many people confuse what BitrateViewer reports and what they see. They run BitrateViewer which for many DVD VOBs will report "interlaced" because that's what the MPEG picture headers say (progressive_frame=off). Then they look at it in a software player or DVD2AVI/whatever and go "hell, but I see no interlacing, so BitrateViewer must be wrong". That's where this "DVDs are often incorrectly flagged interlaced" rumors come from, IMHO. Tools like BV (and I assume your D2V generation code) very accurately report how the stream is encoded, not what it looks like, and this is the information you need, especially for ConvertToYUY2(). Speaking of this, you could also examine the alternate_scan coding extension and set Alternate/ZigZag in CCE accordingly.

And no, I have never seen a stream where progressive_frame was on but picture_structure was TOP_FIELD or BOTTOM_FIELD, it was always FRAME_PICTURE (maybe you made a typo here mentioning TFF/BFF, top_field_first is a separate flag). I actually modified DVD2AVIdg to output a log file when saving the D2V where I record the number of frames, number of progressive frames, number of frame pictures and a lot of other interesting information, I use this for a while now so I have looked at a lot these log files :)

Anyway, what I have recently seen for the first time was all progressive_frames but it was BFF (top_field_first=0). That's how all the cells are in VTS_2, VTS_05 and VTS_06 on Matrix Revolutions Disc 2 (PAL R2), a very weird, complex IFO structure in there BTW. I don't think Upper Field First/Offset Line should be set to 1 in CCE for this, did a quick test encode this way and looked OK.

Another thing I noticed is that it looks like for field pictures (picture_structure TOP_FIELD or BOTTOM_FIELD), the top_field_first flag can't be trusted. I recently had a DVD where some extras where interlaced field pictures and top_field_first was 0, so the D2V was written with all zeroes. But actually it was TFF, checked using AssumeTFF().SeparateFields() AVS script. I think for field pictures, the top_field_first flag is simply not used and whichever field comes first in the stream is the "top field" i.e. the field to decode and display first on an interlaced output device, and presumably this is how Mpeg2Dec3dg outputs it as well.

[jdobbs]

Originally posted by DDogg
For clarification - Meaning that progressive streams will have ZigZag and progressive on, and alternate scan off? (I hope)
Yes. But first I have to clarify whether is is possible to have the progressive flag turned on when a stream's picture_structure is set to TF or BF as opposed to FRAME. I need be sure that by looking at the two parameters I can conclusively decide whether a stream is progressive or interlaced.

Note: Edited to make myself look better

[RB]

Originally posted by DDogg
For clarification - Meaning that progressive streams will have ZigZag and progressive on, and alternate scan off? (I hope)

Yes, that would be a good optimization too. The stream may have progressive_frames on but still use alternate scan rather than ZigZag. I haven't seen that yet and it would make no sense, but you could always set up CCE for ZigZag when source is progressive.

[jdobbs]

And no, I have never seen a stream where progressive_frame was on but picture_structure was TOP_FIELD or BOTTOM_FIELD, it was always FRAME_PICTURE (maybe you made a typo here mentioning TFF/BFF, top_field_first is a separate flag). I was talking about the picture_structure field not the TFF bit -- bad notation on my part.

[issa]

jdobbs,

I still got "Run-time Error '6': Overflow" after I pick
the source path. I did delete the INI and re-configure.
I am using the 0.29a already. Could you help me to
find out what happen to me?