BD Rebuilder Beta - Bug Reports Only

[soneca]

I, too, noticed the RAM problem on my PC.
In the creating m2ts phase everything is fine but in the creating movie only BD structure phase the memory usage causes my system to barely respond as Windows begins swapping. In the first picture you can see the ram usage for the m2ts creation and when it begins to rise that's the creating movie only BD structure(when you look at the process list neither BD-RB nor tsmuxer are shown with that much memory, tsmuxer with 250-300 MB as in the m2ts creation phase).
In the second picture you see the memory usage after BD-RB is finished and more than 3 GB are released.

I also noticed that excessive consumption of memory in "Performing CRF Prediction"...

[Elvis Gratton]

Originally Posted by Sharc

Isn't this the same as I observed in my case? Interlaced sources are encoded progressive ....

Same for me.
Interlaced sources are encoded progressive...
deinterlace=off

Tested on: Paul McCartney - The Space Within Us 2005 US Tour
1080i/29.970fps
BD Rebuilder v0.37.05

[iSeries]

I've had the same problem, tried doing 'Depeche Mode - Tour Of The Universe' blu ray, which is 1080i/29.97. Resultant BD-RB encode was progressive.

[larrsz]

Originally Posted by jdobbs

Try going to the SETUP menu and click/add "und" to your audio selection list (undeterminted).

OK, so I got it to work but not by that method (I cannot see/find a checkbox for UND in the settings area, perhaps you can tell me where to find that).

Instead, I removed all check boxes from the audio section under audio settings. I had been trying to use "do not reencode AC3" thinking it would leave the 384kbps alone but instead I was getting no audio at all. With nothing checked, I am at least getting AC3 but it is being reencoded to 640kbps.

Maybe, if you can tell me where the setting is you are talking about, I can get it to pass the 384k stream without taking it to 640. However, if not, it still works for me.

[larrsz]

Originally Posted by jdobbs

I don't get it. I've done hundreds of interlaced sources... especially 720x480i and they all encode as interlaced.

I haven't had much opportunity to work on BD-RB the last few days. I'll run a 1080i/25fps source through tomorrow. If it encodes incorrectly I will fix it and hopefully get a version out either late tomorrow or Sunday.

@jdobbs

Just an FYI as I saw these comments, the file I asked about yesterday (with the "und" ac3 track) was 1080i. When I ran it through BD-RB, MediaInfo reports the resultant file is progressive. Still reports 29.97fps but "progressive". The file I started with shows 29.97 interlaced.

[Stereodude]

Originally Posted by jdobbs

I don't get it. I've done hundreds of interlaced sources... especially 720x480i and they all encode as interlaced.

I haven't had much opportunity to work on BD-RB the last few days. I'll run a 1080i/25fps source through tomorrow. If it encodes incorrectly I will fix it and hopefully get a version out either late tomorrow or Sunday.

Here's the AVIsynth script for the video:

#Created by BD Rebuilder - v0.37.05 (beta)
DirectshowSource("G:\BDMV\STREAM\00000.m2ts", fps=29.97, framecount=225060, audio=false)
ConvertToYV12().AssumeFPS(30000,1001)

No problems here.


The VID_00000.CHP file is correct (based on 30000/1001 FPS as it should be).


Here are the command lines getting passed to x264 with my 1080i 29.97fps interlaced source:

"C:\BD-RB\tools\x264.exe" "D:\TEMPBD\WORKFILES\VID_00000.AVS" --preset medium --b-pyramid strict --weightp 1 --qpmin=0 --bitrate 18696 --level 4.1 --qpfile "D:\TEMPBD\WORKFILES\VID_00000.CHP" --sar 1:1 --aud --nal-hrd vbr --pic-struct --vbv-bufsize 30000 --keyint 24 --min-keyint 1 --ipratio 1.1 --pbratio 1.1 --vbv-maxrate 35000 --threads auto --slices 4 --thread-input --stats "D:\TEMPBD\WORKFILES\VID_00000.AVS.264.stats" --pass 1 --output NUL

"C:\BD-RB\tools\x264.exe" "D:\TEMPBD\WORKFILES\VID_00000.AVS" --preset medium --b-pyramid strict --weightp 1 --qpmin=0 --bitrate 18696 --level 4.1 --qpfile "D:\TEMPBD\WORKFILES\VID_00000.CHP" --sar 1:1 --aud --nal-hrd vbr --pic-struct --vbv-bufsize 30000 --keyint 24 --min-keyint 1 --ipratio 1.1 --pbratio 1.1 --vbv-maxrate 35000 --threads auto --slices 4 --thread-input --stats "D:\TEMPBD\WORKFILES\VID_00000.AVS.264.stats" --pass 2 --output "D:\TEMPBD\WORKFILES\VID_00000.AVS.264"

This seems to be the problem. It looks like there are a couple of mistakes:
--keyint should be 30
--tff is missing

[rysktkr]

Hi jdobbs,

Here is the inf and log files of the blocky video encode:

-----------------------
[12:10:32] BD Rebuilder v0.37.04 (beta)
- Source: RAGING_BULL
- Input BD size: 34.00 GB
- Approximate total content: [02:09:03.819]
- Target BD size: 22.95 GB
- Windows Version: 5.1 [2600]
- MOVIE-ONLY mode enabled
- Quality: High Quality (Default), ABR
- Audio Settings: AC3=0 DTS=0 HD=1 Kbs=640
[12:10:32] PHASE ONE, Encoding
- [12:10:32] Extracting A/V streams [VID_00000]
- [12:56:07] Reencoding: VID_00000 (1 of 1)
- [12:56:07] Collecting video information
- Source Video: MPEG-4 (AVC), 1920x1080
- Rate/Length: 23.976fps, 185,666 frames
- Bitrate: 19,189 Kbs
- [12:56:07] Reencoding: VID_00000, Pass 1 of 1
- [22:12:56] Video Encode complete
- [22:12:57] Reencoding audio tracks (if req'd)
- [22:12:58] Multiplexing M2TS
[22:25:48]PHASE ONE complete
[22:25:48]PHASE TWO - Rebuild Started
- [22:25:48] Rebuilding BD file Structure
[22:57:26] - Encode and Rebuild complete
[22:57:26]JOB: RAGING_BULL finished.


[Status]
LABEL=RAGING_BULL
VERSION=v0.37.04 (beta)
SOURCE_SIZE=36505282560
SOURCE_VIDEO_SIZE=36505282560
TARGET_SIZE=24641536000
REDUCTION=.675012882299958
RESIZE_1080=0
AUDIO_TO_KEEP=eng;
SUBS_TO_KEEP=all
BACKUP_MODE=1
MOVIEONLY_TYPE=0
USE_LAVF=0
QUICK=0
ENCODE_STEP=0
COMPLETED=1
REBUILD_COMPLETE=1
[00000]
AUDIO=100000000
PGS=111111010000
M2TS_TARGET=24641536000
RATE=19189
NSTART=27000000
NEND=375471840
NSIZE=19052058624
FLINK=0
MLINK=0

[jdobbs]

Originally Posted by Stereodude

Here's the AVIsynth script for the video:No problems here.


The VID_00000.CHP file is correct (based on 30000/1001 FPS as it should be).


Here are the command lines getting passed to x264 with my 1080i 29.97fps interlaced source:This seems to be the problem. It looks like there are a couple of mistakes:
--keyint should be 30
--tff is missing

If it's interlaced, then yes "--tff" should be there. But there's no reason keyint should be 30. It could be any value up to and including 30... and 24 just happens to be is the default.

I've been able to repeat this issue, I'm guessing some change I've made recently must have caused it. I'm working on a fix and should have it out soon.

[jdobbs]

Originally Posted by rysktkr

Hi jdobbs,

Here is the inf and log files of the blocky video encode:

-----------------------
[12:10:32] BD Rebuilder v0.37.04 (beta)
- Source: RAGING_BULL
- Input BD size: 34.00 GB
- Approximate total content: [02:09:03.819]
- Target BD size: 22.95 GB
- Windows Version: 5.1 [2600]
- MOVIE-ONLY mode enabled
- Quality: High Quality (Default), ABR
- Audio Settings: AC3=0 DTS=0 HD=1 Kbs=640
[12:10:32] PHASE ONE, Encoding
- [12:10:32] Extracting A/V streams [VID_00000]
- [12:56:07] Reencoding: VID_00000 (1 of 1)
- [12:56:07] Collecting video information
- Source Video: MPEG-4 (AVC), 1920x1080
- Rate/Length: 23.976fps, 185,666 frames
- Bitrate: 19,189 Kbs
- [12:56:07] Reencoding: VID_00000, Pass 1 of 1
- [22:12:56] Video Encode complete
- [22:12:57] Reencoding audio tracks (if req'd)
- [22:12:58] Multiplexing M2TS
[22:25:48]PHASE ONE complete
[22:25:48]PHASE TWO - Rebuild Started
- [22:25:48] Rebuilding BD file Structure
[22:57:26] - Encode and Rebuild complete
[22:57:26]JOB: RAGING_BULL finished.


[Status]
LABEL=RAGING_BULL
VERSION=v0.37.04 (beta)
SOURCE_SIZE=36505282560
SOURCE_VIDEO_SIZE=36505282560
TARGET_SIZE=24641536000
REDUCTION=.675012882299958
RESIZE_1080=0
AUDIO_TO_KEEP=eng;
SUBS_TO_KEEP=all
BACKUP_MODE=1
MOVIEONLY_TYPE=0
USE_LAVF=0
QUICK=0
ENCODE_STEP=0
COMPLETED=1
REBUILD_COMPLETE=1
[00000]
AUDIO=100000000
PGS=111111010000
M2TS_TARGET=24641536000
RATE=19189
NSTART=27000000
NEND=375471840
NSIZE=19052058624
FLINK=0
MLINK=0

Something is seriously wrong with your configuration. I can't say what it is -- but it is there. There is no way in the world that an encode with a rate of 19Mbs will be blocky unless the input was blocky already.

[Sharc]

Originally Posted by jdobbs

If it's interlaced, then yes "--tff" should be there.

If it's being encoded at 29.97 with 1080 resolution, --tff must be there for blu-ray compliance, even if the the original should (by mistake) have been encoded as progressive.
If the original content is progressive (no temporal shift between odd and even fields) it could be encoded with --fake-interlaced, but I am not shure if BD-RB can automatically decide between --tff and --fake-interlaced options.

[jdobbs]

Originally Posted by Sharc

If it's being encoded at 29.97 with 1080 resolution, --tff must be there for blu-ray compliance, even if the the original should (by mistake) have been encoded as progressive.
If the original content is progressive (no temporal shift between odd and even fields) it could be encoded with --fake-interlaced, but I am not shure if BD-RB can automatically decide between --tff and --fake-interlaced options.

Why are you telling me something that I obviously already know?

The "--fake-interlaced" is used within BD-RB when the source has been deinterlaced, but the format is such that progressive isn't allowed (such as 720x480). That way the output of the deinterlacing can use the efficiency associated with progressive encoding but still be included in a BD (within specifications). BD-RB makes no attempt to try and find progressive content within an interlaced stream (other than IVTC) -- in fact I can't think of any reliable way that it could be done even if I wanted to.

[Stereodude]

Originally Posted by jdobbs

If it's interlaced, then yes "--tff" should be there.

Well, isn't that the issue we're discussing? Specifically that interlaced 29.97i (30000/1001) FPS concerts get encoded as progressive?

It seems there are several of us who have all found the same issue. Every tool I've tried report the video stream as interlaced (eac3to, TSmuxer, & DGdecNV). Visual inspection shows it's interlaced. Even BD-Rebuilder recognizes it's interlaced.

I don't see any ambiguity on whether or not it is interlaced.

Edit: I see you edited the post that you were able to duplicate the issue, so it seems we're all on the same page now.

[jdobbs]

Originally Posted by Stereodude

Well, isn't that the issue we're discussing? Specifically that interlaced 29.97i (30000/1001) FPS concerts get encoded as progressive?

It seems there are several of us who have all found the same issue. Every tool I've tried report the video stream as interlaced (eac3to, TSmuxer, & DGdecNV). Visual inspection shows it's interlaced. Even BD-Rebuilder recognizes it's interlaced.

I don't see any ambiguity on whether or not it is interlaced.

What the hell is it with you people? Can you not read what I wrote?


I KNOW THERE IS AN ISSUE. I'M WORKING IT!!!!! Repeatedly telling me what I already know is only pissing me off.

[Stereodude]

Originally Posted by jdobbs

What the hell is it with you people? Can you not read what I wrote?


I KNOW THERE IS AN ISSUE. I'M WORKING IT!!!!! Repeatedly telling me what I already know is only pissing me off.

Well, you edited your post after I hit the quote button, so my reply got posted before I read your addition.

Sorry...

[jdobbs]

Originally Posted by Stereodude

Well, you edited your post after I hit the quote button, so my reply got posted before I read your addition.

Sorry...

And it was just as correct before I edited it. The edit simply says that I've been able to repeat it.

How about this post from yesterday? (#11501)

Originally Posted by jdobbs

I haven't had much opportunity to work on BD-RB the last few days. I'll run a 1080i/25fps source through tomorrow. If it encodes incorrectly I will fix it and hopefully get a version out either late tomorrow or Sunday.

[jdobbs]

Well, I've found what is causing it... an If/then/else statement that is not acting correctly. The problem is that it appears to be written correctly, but for some reason it isn't hitting the "else" when the "if" is false -- weird... I'll figure it out.

[Edit] Ok. I've fixed it. You'll see it in the next release.

[Sharc]

Originally Posted by jdobbs

Well, I've found what is causing it... an If/then/else statement that is not acting correctly. The problem is that it appears to be written correctly, but for some reason it isn't hitting the "else" when the "if" is false -- weird... I'll figure it out.

[Edit] Ok. I've fixed it. You'll see it in the next release.

Congrats and thanks!
No one wanted to teach you anything. But at times it is helpful to learn if other people experience a similar problem, or if it's only a local issue here. No offense.

[jdobbs]

Does anyone know of some good free software for reporting pulldown and field order in VC-1 interlaced sources? I have a routine that is scanning the picture header and I want to verify that some of the fields are being read correctly (I wrote it a long time ago, and I can't seem to find my reference material).

[Edit] Never mind. I found my reference...

[Elvis Gratton]

Originally Posted by jdobbs

Does anyone know of some good free software for reporting pulldown and field order in VC-1 interlaced sources? I have a routine that is scanning the picture header and I want to verify that some of the fields are being read correctly (I wrote it a long time ago, and I can't seem to find my reference material).

[Edit] Never mind. I found my reference...

May be mediainfo with VC1 in MKV container !!!

Video
ID : 1
Format : VC-1
Format profile : AP@L3
Codec ID : WVC1
Codec ID/Hint : Microsoft
Duration : 1h 55mn
Bit rate : 27.7 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate : 29.970 fps
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Scan order : Top Field First
Compression mode : Lossy
Bits/(Pixel*Frame) : 0.445
Stream size : 22.4 GiB (98%)

[Elvis Gratton]

B.T.W. jdobbs

If you have found other soft to see vc1 information, let me know.

thanks