BD Rebuilder Beta - Bug Reports Only

[jdobbs]

@cartman0208
# aType=Output Audio Type, 0=AC3 (Default), 1=AAC (Note: Audio bitrate less than 384Kbs implies stereo downmix)
# 2=[Intact audio -- MKV only]
# cType= Container/limits type, 0=iPod (mp4), 1=MKV, 2=iPad (mp4), 3=DVD-5, 4=DVD-9, 5=MP4 6=iPad 2+ (MP4)
Note from the ALTERNATE.TXT type descriptions that aType=2 only applies to MKV output. I can't remember why -- but there was a reason for that. I'd have to go back through my notes to see why. Your output container type is set for MP4.

The names of subtitle types change depending upon the input container type which makes it confusing when you look at them with MediaInfo. BD-RB remuxes non-MKV files to MKV during import so they can be consistent (and I don't have to create an infinite list of synonyms). I'm not that familiar with tx3g, except I believe it is used in MP4 files. For some reason I seem to recall the use of XML for that type -- which might mean it isn't currently supported at all. So I assume you ran MediaInfo against an MP4 file? BD-RB currently supports import of PGS, ASS, and UTF8 (SRT) subtitle types (identified in an MKV container as S_TEXT/UTF8, S_TEXT/ASS, and S_HDMV/PGS).

I'll have to look around for an MP4 file with that subtitle type and see what equivalent type name is (assuming it exists) when remuxed into an MKV container.

[bbcby]

Hello everyone.
How to get bitrate from DTS-HD file?
MediaInfo don't show, only core.

[tebasuna51]

How to get bitrate from DTS-HD file?

Average bitrate = FileSize in Kbits / Time to play in seconds

[bbcby]

Ok, thanks. And the Peak Bitrate?

[tebasuna51]

I don't know for what you need the data, but you need a tool than read all the frames to reach the max frame length.

MediaInfo can't read all frames for speed, you can try with LeeAudBi6 (http://forum.doom9.org/showthread.php?p=1522330#post1522330), is slow because need read all the frames, and need some operations to obtain the bitrates.

A example of relevant info in the LeeAudBi report:
File ........: D:\Test\AudioD\Samples\dts\dtsma\Hd-maH.dts
Size ........: 4705772 bytes

----------------------------------------- First Frame Info
...
Number of PCM Sample Blocks .: 15 ( 512 samples/frame)
Primary Frame Byte Size .....: 2011 ( 2012 bytes/frame)
Core Audio Samp. Frequency ..: 13 (48 kHz)
...
--------------------------------------------- Revised Info
...
Duration ..........: 14,037 seconds. ( 0 h. 0 m. 14,037 s.)
Master A. min./max.: 68 / 6456 (HD-MA subframe bytes)
------------------------------------------------- End Info

Average bitrate = (4705.772 * 8) Kb / 14.037 s. = 2681.925 Kb/s.

Core (Primary frame) bitrate = (2.012 * 8) Kb / (512 / 48000) s. = 1509 Kb/s.

Peak (Primary frame + Max. subframe) bitrate = ((2.012 + 6.456) * 8) Kb / (512 / 48000) s. = 6351 Kb/s

[EDIT]
With standard values (512 samples/frame, 48 kHz)

Bitrate in Kb/s = 0.75 * FrameSize in bytes

@jdobbs
Maybe these posts must be in Audio Encoding subforum (https://forum.doom9.org/forumdisplay.php?f=11)

Feel free of move them if you want.

[cartman0208]

@cartman0208
# aType=Output Audio Type, 0=AC3 (Default), 1=AAC (Note: Audio bitrate less than 384Kbs implies stereo downmix)
# 2=[Intact audio -- MKV only]
# cType= Container/limits type, 0=iPod (mp4), 1=MKV, 2=iPad (mp4), 3=DVD-5, 4=DVD-9, 5=MP4 6=iPad 2+ (MP4)
Note from the ALTERNATE.TXT type descriptions that aType=2 only applies to MKV output.

Sorry, I didn't notice that, my mistake :o
But the MKV setting also reencodes audio:
[00011]
caption=MKV, IntactVideo, NO_RESIZE, IntactAudio
aType=2
cType=1
vFormat=8
vBitrate=0

The reencode seems to happen during import. And despite the log showing "keeping original audio" the reencoded stream is muxed


The names of subtitle types change depending upon the input container type which makes it confusing when you look at them with MediaInfo. BD-RB remuxes non-MKV files to MKV during import so they can be consistent (and I don't have to create an infinite list of synonyms). I'm not that familiar with tx3g, except I believe it is used in MP4 files. For some reason I seem to recall the use of XML for that type -- which might mean it isn't currently supported at all. So I assume you ran MediaInfo against an MP4 file? BD-RB currently supports import of PGS, ASS, and UTF8 (SRT) subtitle types (identified in an MKV container as S_TEXT/UTF8, S_TEXT/ASS, and S_HDMV/PGS).

I'll have to look around for an MP4 file with that subtitle type and see what equivalent type name is (assuming it exists) when remuxed into an MKV container.

You're right, thats from an MP4-file created with the newest product from the RedFox team.
Unfortunately only embedded timed text is currently available (I don't like having extra subtitle files)
I just uploaded a file to the known location ;)

[sieve]

Immediately after the encode, look at the command line in LASTCMD.TXT. See if a strange frame rate is somehow being used (--fps xx.xx in the command line).

Here is the LASTCMD.txt right after encode. Did not see a fps. Timestamp 01:47:

"C:\BD_Rebuilder\tools\x264.exe" "C:\WORKING\WORKFILES\VID_00263.AVS" --preset slow --bluray-compat --ref 4 --b-adapt 1 --b-pyramid none --weightp 1 --qpmin=0 --bitrate 28074 --level 4.1 --qpfile "C:\WORKING\WORKFILES\VID_00263.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 "C:\WORKING\WORKFILES\VID_00263.AVS.264.stats" --pass 2 --output "C:\WORKING\WORKFILES\VID_00263.AVS.264"


And here's the related log file:

[04/13/21] BD Rebuilder v0.61.21
[20:14:27] Source: ROCKY_HORROR_WW_00800
- Input BD size: 26.60 GB
- Approximate total content: [01:38:40.914]
- Target BD size: 22.95 GB
- Windows Version: 6.2 [9200]
- MOVIE-ONLY mode enabled
- Quality: Highest (Very Slow), Two Pass
- Decoding/Frame serving: DirectShow
- Audio Settings: AC3=0 DTS=0 HD=0 Kbs=640
[20:14:27] PHASE ONE, Encoding
- [20:14:27] Processing: VID_00260 (1 of 3)
- [20:14:27] Extracting A/V streams [VID_00260]
- [20:15:04] Reencoding video [VID_00260]
- [20:15:04] Keeping original video (no reencode)
- [20:15:04] Processing audio tracks
- Track 4352 (eng): Keeping original audio
- [20:15:04] Processing: VID_00261 (2 of 3)
- [20:15:04] Extracting A/V streams [VID_00261]
- [20:15:39] Reencoding video [VID_00261]
- Source Video: MPEG-4 (AVC), 1920x1080
- Rate/Length: 23.976fps, 106,957 frames
- Bitrate: 28,451 Kbs
- [20:15:39] Reencoding: VID_00261, Pass 1 of 2
- [21:19:49] Reencoding: VID_00261, Pass 2 of 2
- [01:45:01] Video Encode complete
- [01:45:01] Processing: VID_00263 (3 of 3)
- [01:45:01] Extracting A/V streams [VID_00263]
- [01:45:07] Reencoding video [VID_00263]
- Source Video: MPEG-4 (AVC), 1920x1080
- Rate/Length: 23.976fps, 5,935 frames
- Bitrate: 28,074 Kbs
- [01:45:07] Reencoding: VID_00263, Pass 1 of 2
- [01:47:26] Reencoding: VID_00263, Pass 2 of 2
- [02:03:20] Video Encode complete
[02:03:20]PHASE ONE complete
[02:03:21]PHASE TWO - Rebuild Started
- [02:03:21] Rebuilding BD file Structure
[02:04:09] - Encode and Rebuild complete
- WORKFILES folder removed.
[02:04:09] JOB: ROCKY_HORROR_WW finished.

[bbcby]

I don't know for what you need the data.

BD Rebuilder calculates this somehow to avoid buffer overflow.

[MrVideo]

Here is the LASTCMD.txt right after encode. Did not see a fps. Timestamp 01:47:
Take a look at the contents of this file: C:\WORKING\WORKFILES\VID_00263.AVS

[sieve]

I had "delete workfiles" checked in settings. I'll have to reenable that and redo the encode. Stay tuned. And thx!

[MrVideo]

I had "delete workfiles" checked in settings. I'll have to reenable that and redo the encode. Stay tuned. And thx!
The AVS filename should be different.

[jdobbs]

The reencode seems to happen during import. And despite the log showing "keeping original audio" the reencoded stream is muxed Yes, the audio is being reencoded during import. That's because it is EAC3 (not standard AC3). It does that because EAC3 is not supported in the BD standard as a primary audio stream.

When you create the ALTERNATE output, the audio stream existing in the pseudo-BD structure (AC3) is the one that is "kept intact".

[sieve]

Take a look at the contents of this file: C:\WORKING\WORKFILES\VID_00263.AVS

Here is the contents:
#Created by BD Rebuilder - v0.61.21
DirectshowSource("C:\VIDEO\ROCKY_HORROR_WW\BDMV\STREAM\00263.m2ts", fps=23.976, framecount=5935, audio=false)
ConvertToYV12().AssumeFPS(24000,1001)

[MrVideo]

Here is the contents:
If you were to code wrap the text, instead of quote wrap it, it would show up in this quoted post.

That said, 23.976 fps is being passed to the x264 encoder. The resulting C:\WORKING\WORKFILES\VID_00263.AVS.264 is not (according to mediainfo)?

[sieve]

That said, 23.976 fps is being passed to the x264 encoder. The resulting C:\WORKING\WORKFILES\VID_00263.AVS.264 is not (according to mediainfo)?

I think the below indicates it is passing 23.976 FPS to the file (VID_00263.AVS.264)?

Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L4.1
Format settings : CABAC / 4 Ref Frames
Format settings, CABAC : Yes
Format settings, Reference frames : 4 frames
Bit rate mode : Variable
Bit rate : 28.1 Mb/s
Maximum bit rate : 35.0 Mb/s
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate : 23.976 (24000/1001) FPS
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Progressive
Bits/(Pixel*Frame) : 0.565
Writing library : x264 core 158 r2984 3759fcb
Encoding settings : cabac=1 / ref=4 / deblock=1:0:0 / analyse=0x3:0x113 / me=hex / subme=8 / psy=1 / psy_rd=1.00:0.00 / mixed_ref=1 / me_range=16 / chroma_me=1 / trellis=2 / 8x8dct=1 / cqm=0 / deadzone=21,11 / fast_pskip=1 / chroma_qp_offset=-2 / threads=12 / lookahead_threads=2 / sliced_threads=0 / slices=4 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=1 / constrained_intra=0 / bframes=3 / b_pyramid=0 / b_adapt=1 / b_bias=0 / direct=3 / weightb=1 / open_gop=0 / weightp=1 / keyint=24 / keyint_min=1 / scenecut=40 / intra_refresh=0 / rc_lookahead=24 / rc=2pass / mbtree=1 / bitrate=28074 / ratetol=1.0 / qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / cplxblur=20.0 / qblur=0.5 / vbv_maxrate=35000 / vbv_bufsize=30000 / nal_hrd=vbr / filler=0 / ip_ratio=1.10 / aq=1:1.00

[MrVideo]

I think the below indicates it is passing 23.976 FPS to the file (VID_00263.AVS.264)?
Frame rate : 23.976 (24000/1001) FPS
Yep. So I am confused as to why you are having issues.

[sieve]

Thanks for assisting with the diagnosis. I guess this one won't be solved.
Hopefully it is just an anomaly related to this specific blu-ray. I'll report back if this happens with the next one I do.

[gamete]

Is it better ac3 or eac3 ?
Thanks

[jdobbs]

Is it better ac3 or eac3 ?
ThanksIt depends on the usage. If you are trying to create a blu-ray disc, EAC3 without core audio is not compliant for the primary stream. That's because EAC3 (DD+) support is optional in a blu-ray player, (AC3 support is required). The way the standard ensures support is by requiring EAC3 to have standard AC3 embedded in the stream as well (core audio).

The main advantage to EAC3 is its ability to support more channels. Standard AC3 has a limit of 6 channels (e.g. 5.1). EAC3, for example, could support 7.1. EAC3 also supports higher data rates.

In BD-RB, since all sources are either imported or come directly from a BD source, an imported EAC3 source would almost always be converted to AC3. There are several reasons -- including the requirement for a "core" audio to be included -- which pretty much never exists in an import source.

Now... which is better? Since Dolby's double blind tests show that a 5.1 AC3 stream encoded at 640Kbs is indistiguishable from an uncompressed source (to the human ear) -- it is debatable whether EAC3's higher bitrates actually provide any useful value. If you need 7.1 for some reason, then maybe... but since a sound mixed on both the front and rear channels of a 5.1 source gives the appearance of a side channel -- it's also debatable as to whether 7.1 is anything more than a marketing gimmick. Others disagree on that -- so what I've stated is nothing more than opinion.

[MrVideo]

It is also a means by which Atmos is provided.

[RetsimLegin]

After it fails, go to the BD-RB folder and open LASTCMD.TXT. Then open a DOS windows and cut/paste the command from LASTCMD.TXT to the DOS window. Let it attempt the encode, and post the command along wth any error messages here so we can take a look at what is happening.


Just as an aside, why use MultiAVCHD when you can import the MKV files directly with BD-RB?I'll try this and your other suggestions and report back later. Thanks for your input.

To answer the "aside" - partly familiarity as I used MultiAVCHD for some other projects that didn't need re-encoding and/or weren't SD content before finding BDRB; and coupled with that - the disc menu is (I think) somewhat more configurable. Indeed the only time I attempted to get BDRB to build its own menu I got a nice background and title card music, but no actual text or highlightable/selectable controls. That was a while back and I no longer have details of what it was I was doing at the time.

[tebasuna51]

Is it better ac3 or eac3 ?
Remember forum rule 12:
Do not ask "what's best" because this question cannot be answered objectively. Each and everyone has their own view about what's best in a certain area.

Like jdobbs say, in order to BD rebuild, the support of EAC3 is optional in BD players then in this area select AC3 can be the best option.

Out of BD rebuild there are other opinions:

1) About more channels than max 5.1 supported by AC3:
If you need 7.1 for some reason, then maybe... but since a sound mixed on both the front and rear channels of a 5.1 source gives the appearance of a side channel -- it's also debatable as to whether 7.1 is anything more than a marketing gimmick.
It is my opinion: for surround 2D is more than enough 5.1 channels, 7.1 is usseless.

2) About bitrate I agree with jdobbs when say:
Since Dolby's double blind tests show that a 5.1 AC3 stream encoded at 640Kbs is indistiguishable from an uncompressed source (to the human ear) -- it is debatable whether EAC3's higher bitrates actually provide any useful value

But, in a general way of course, EAC3 is better than AC3 because:

1) Support not only more channels than AC3 but also surround 3D with Atmos, not allowed for AC3.

2) The encoder is more efficient and we can obtain more quality with less bitrate, like we can see in a multichannel test (https://tech.ebu.ch/files/live/sites/tech/files/shared/tech/tech3324.pdf).

The DD+ (EAC3) 448 Kb/s is comparable to DTS 1500 like AC3 640

Then, if device player support all EAC3 features, EAC3 is better.

[RetsimLegin]

@jdobbs

Confirmed using GeForce experience that driver is current - NVIDIA studio Driver v462.31 (14 Apr 21) was installed just prior to these tests.

Tests using a previously attempted single .mkv as the source.

1: add NVENC_NO_FRAMECHK=1 to .ini file at the last line BEFORE [paths]
Launch BDRB and backup.
Outcome: NVENCC prematurely terminated (at ~79%)
(Encoded 72102, expected 90127)

Just in case it was a typo
2: Exit BDRB
add NVENCC_NO_FRAMECHK=1 to .ini file at the last line BEFORE [paths]
Launch BDRB and backup. (Do not resume job; empty working folder)
Outcome: NVENCC prematurely terminated (at ~79%)
(Encoded 72102, expected 90127)

3: revert above change. Add IMPORT_VFR_SCAN=1 to .ini file at the last line BEFORE [paths]
Launch BDRB and backup. (Do not resume job; empty working folder)
Outcome: NVENCC prematurely terminated (at ~79%)
(Encoded 72102, expected 90127)

4: After the last fail (i.e with the last change to .ini still in place), copied & pasted the contents of LASTCMD.TXT into an elevated command prompt

LASTCMD.TXT
==========


"C:\Users\nigel\Downloads\BDRebuilder\tools\nvenc\nvencc.exe" --avhw --fps 29.97 -i "C:\USERS\NIGEL\VIDEOS\NVENCC TEST\IMPORTS\GFU_FOR_NVENCC_TEST_COREL\BDMV\STREAM\00000.m2ts" --codec h264 --preset performance --bluray --qp-min 0 --vbr 2327 --aq --aq-temporal --keyfile "C:\USERS\NIGEL\VIDEOS\NVENCC TEST\WORKFILES\VID_00000.CHP" --sar 10:11 --aud --pic-struct --vbv-bufsize 13000 --max-bitrate 15000 --gop-len 30 -o "C:\USERS\NIGEL\VIDEOS\NVENCC TEST\WORKFILES\VID_00000.AVS.264"


=========
DOS output
=========

Microsoft Windows [Version 10.0.19042.906]
(c) Microsoft Corporation. All rights reserved.

C:\windows\system32>"C:\Users\nigel\Downloads\BDRebuilder\tools\nvenc\nvencc.exe" --avhw --fps 29.97 -i "C:\USERS\NIGEL\VIDEOS\NVENCC TEST\IMPORTS\GFU_FOR_NVENCC_TEST_COREL\BDMV\STREAM\00000.m2ts" --codec h264 --preset performance --bluray --qp-min 0 --vbr 2327 --aq --aq-temporal --keyfile "C:\USERS\NIGEL\VIDEOS\NVENCC TEST\WORKFILES\VID_00000.CHP" --sar 10:11 --aud --pic-struct --vbv-bufsize 13000 --max-bitrate 15000 --gop-len 30 -o "C:\USERS\NIGEL\VIDEOS\NVENCC TEST\WORKFILES\VID_00000.AVS.264"
--------------------------------------------------------------------------------
C:\USERS\NIGEL\VIDEOS\NVENCC TEST\WORKFILES\VID_00000.AVS.264
--------------------------------------------------------------------------------
NVEncC (x86) 5.15 (r1658) by rigaya, Sep 12 2020 14:50:44 (VC 1900/Win/avx2)
OS Version Windows 10 x64 (19042)
CPU 11th Gen Intel Core i7-1165G7 @ 2.80GHz [TB: 4.69GHz] (4C/8T)
GPU #0: GeForce GTX 1650 Ti (1024 cores, 1485 MHz)[2147483.64]
NVENC / CUDA NVENC API 10.0, CUDA 11.2, schedule mode: auto
Input Buffers CUDA, 20 frames
Input Info avcuvid: MPEG2, 720x480, 30000/1001 fps
Vpp Filters copyDtoD
Output Info H.264/AVC high @ Level auto
720x480p 10:11 29.970fps (30000/1001fps)
Encoder Preset performance
Rate Control VBR
Multipass none
Bitrate 2327 kbps (Max: 15000 kbps)
Target Quality auto
Initial QP I:20 P:23 B:25
QP range I:0-51 P:0-51 B:0-51
VBV buf size 13000 kbit
Lookahead off
GOP length 28 frames
B frames 3 frames [ref mode: disabled]
Ref frames 3 frames
AQ on(spatial, temporal, strength auto)
Others mv:auto cabac deblock adapt-transform:auto bdirect:auto aud pic-struct

encoded 72102 frames, 1953.45 fps, 2337.80 kbps, 670.47 MB
encode time 0:00:36, CPU: 11.9%, GPU: 49.8%, VE: 90.2%, VD: 96.0%
frame type IDR 2583
frame type I 2583, total size 84.34 MB
frame type P 18028, total size 375.14 MB
frame type B 51491, total size 210.98 MB

[jdobbs]

3: revert above change. Add IMPORT_VFR_SCAN=1 to .ini file at the last line BEFORE [paths]
Launch BDRB and backup. (Do not resume job; empty working folder)
Outcome: NVENCC prematurely terminated (at ~79%)
(Encoded 72102, expected 90127) For this to have any impact you would have to import the file again.
2: Exit BDRB
add NVENCC_NO_FRAMECHK=1 to .ini file at the last line BEFORE [paths]
Launch BDRB and backup. (Do not resume job; empty working folder)
Outcome: NVENCC prematurely terminated (at ~79%)
(Encoded 72102, expected 90127)You added it with a syntax error. It should actually say NVENC_NO_FRAMECHK=1 (you added an extra "C"). If that flag is set, BD-RB doesn't even do the check for premature termination.

[RetsimLegin]

For this to have any impact you would have to import the file again.
Tried this again with a fresh import of the same .mkv
Same outcome.



You added it with a syntax error. It should actually say NVENC_NO_FRAMECHK=1 (you added an extra "C"). If that flag is set, BD-RB doesn't even do the check for premature termination.Sure - as I say that was a "just in case" of a typo. Test 1 had it as you quote.

Dumb thought: I edited the .ini file after first making a backup copy of it in the same folder (with a different name). I'm going to do 1 and 3 again after first putting the copy somewhere else entirely. Back in a jiffy.

[RetsimLegin]

Repeat test 1

This is the .ini file

================================

[Options]
VERSION=0.61.0.21
ENCODER=1
MODE=3
ENCODE_QUALITY=5
ONEPASS_ENCODING=2
AUTO_QUALITY=0
AUDIO_TO_KEEP=eng;
SUBS_TO_KEEP=all
SD_CONVERT=0
OPEN_GOP=0
RESIZE_1080=0
RESIZE_1440=0
RESIZE_720=0
DEINTERLACE=0
SD_TO_1080=0
IGNORE_3D=0
CONVERT_WIDE=0
DTS_REENCODE=0
AC3_REENCODE=0
AC3_640=1
AC3_192=0
KEEP_HD_AUDIO=0
AUDIO_DRC=0
DECODER=4
AVCHD=1
REMOVE_WORKFILES=0
REMOVE_OUTPUT=0
USE_FILTERS=0
BDMV_CERT_ONLY=1
IVTC_PULLDOWN=0
ASSUME_DVD_PAL=0
FRIMSOURCE=0
COMPLETION_BEEP=0
OUTPUT_SBS=0
NEROAAC=0
SUPTITLE=0
PGSTOSRT=0
AUDIO_TRACK_LIMIT=1
SUBTITLE_TRACK_LIMIT=1
CUSTOM_TARGET_SIZE=1000
NVENC_CAPABLE=1
TARGET_SIZE=43000
PRIORITY_CLASS=2
NVENC_NO_FRAMECHK=1
[Paths]
WORKING_PATH=C:\USERS\NIGEL\VIDEOS\NVENCC TEST\
SOURCE_PATH=C:\USERS\NIGEL\VIDEOS\NVENCC TEST\IMPORTS\GFU_FOR_NVENCC_TEST_COREL\


============================

This is the log

============================

[04/17/21] BD Rebuilder v0.61.21
[14:03:38] Source: GFU_FOR_NVENCC_TEST_COREL_00000
- Input BD size: 1.94 GB
- Approximate total content: [00:50:07.237]
- Target BD size: 0.98 GB
- Windows Version: 6.2 [9200]
- MOVIE-ONLY mode enabled
- Quality: Good (Fastest, BD-25+), ABR
- Decoding/Frame serving: NVENCC
- Audio Settings: AC3=0 DTS=0 HD=0 Kbs=640
[14:03:41] PHASE ONE, Encoding
- [14:03:41] Processing: VID_00000 (1 of 1)
- [14:03:41] Extracting A/V streams [VID_00000]
- [14:03:51] Reencoding video [VID_00000]
- Source Video: MPEG-2, 720x480
- Rate/Length: 29.970fps, 90,127 frames
- Bitrate: 2,327 Kbs
- [14:03:51] Reencoding: VID_00000, Pass 1 of 1
- Error: NVENCC prematurely terminated.
- Encoded: 72102, expected: 90127
[14:04:29] - Failed video encode, aborted

[gamete]

Thanks @jdobbs and @tebasuna51

[mikeonecl]

i make two processing, one with nvenc activated and the second with x264. On nvenc, there are glitches on some frames.
but on x264 is working fine.

can you check this??

i make on Sonic The Hedgehog bluray to bd25. i check others conversion make with nvenc and all have frame glitch.

[LowDead]

i make two processing, one with nvenc activated and the second with x264. On nvenc, there are glitches on some frames.
but on x264 is working fine.

can you check this??

i make on Sonic The Hedgehog bluray to bd25. i check others conversion make with nvenc and all have frame glitch.

Update to latest nVidia drivers and try again. If it still doesn't work there could be other problems with your system. But if you only get errors with nvenc it sounds to me that it is graphic card related or maybe a harddrive giving up, but that is more far fetched according to your error..

//LD

[CraigWally]

i make two processing, one with nvenc activated and the second with x264. On nvenc, there are glitches on some frames.
but on x264 is working fine.

can you check this??

i make on Sonic The Hedgehog bluray to bd25. i check others conversion make with nvenc and all have frame glitch.

If you have quality on Highest change to High (Default) and then report back

[jdobbs]

@RetsimLegin

Based on what I'm seeing in this post (https://forum.doom9.org/showthread.php?p=1940965#post1940965), it appears that NVENCC is finding some indication of an end-of-file before all the frames are processed. I'm at a loss as to why it is stopping (with no error) -- unless maybe there is an H264 end indicator in the stream.

[RetsimLegin]

@jdobbs
Curious. To recap on how consistent this has been
First failure: a full BD layout (compiled using MultiAVCHD from MakeMKV files). The first title is a ~90 minute NTSC SD item. Failed on the first title ~75%.
(Note that the same project encoded throughout just fine using the much slower X264/5 encoder)
Second failure: single file import of the same ~90 minute title .mkv. ~75%
Third failure: single file import of a different title from the same series ~44 minutes ~79%
(various attempts at other settings like deinterlace as described earlier)
Fourth failure: single file import of a ~45 minute title from an entirely different disc set/series .mkv also NTSC SD ~79%
Fifth failure: same source as 4, this time extracted as a .mpg using an entirely different app. ~79%


If you think it may be informative, I could try (for example)
- remuxing (tsMuxeR or AVIDemux) any of the above before trying to encode
- yet another different title from yet another source
- something PAL format
or anything else you think may help narrow it down.

[jdobbs]

@jdobbs
Curious. To recap on how consistent this has been
First failure: a full BD layout (compiled using MultiAVCHD from MakeMKV files). The first title is a ~90 minute NTSC SD item. Failed on the first title ~75%.
(Note that the same project encoded throughout just fine using the much slower X264/5 encoder)
Second failure: single file import of the same ~90 minute title .mkv. ~75%
Third failure: single file import of a different title from the same series ~44 minutes ~79%
(various attempts at other settings like deinterlace as described earlier)
Fourth failure: single file import of a ~45 minute title from an entirely different disc set/series .mkv also NTSC SD ~79%
Fifth failure: same source as 4, this time extracted as a .mpg using an entirely different app. ~79%


If you think it may be informative, I could try (for example)
- remuxing (tsMuxeR or AVIDemux) any of the above before trying to encode
- yet another different title from yet another source
- something PAL format
or anything else you think may help narrow it down.All I can say for sure is that it appears to be something specific to your system, as others aren't experiencing it. But right now I have no idea what it is.

[cartman0208]

@jdobbs
Curious. To recap on how consistent this has been
First failure: a full BD layout (compiled using MultiAVCHD from MakeMKV files). The first title is a ~90 minute NTSC SD item. Failed on the first title ~75%.
(Note that the same project encoded throughout just fine using the much slower X264/5 encoder)
Second failure: single file import of the same ~90 minute title .mkv. ~75%
Third failure: single file import of a different title from the same series ~44 minutes ~79%
(various attempts at other settings like deinterlace as described earlier)
Fourth failure: single file import of a ~45 minute title from an entirely different disc set/series .mkv also NTSC SD ~79%
Fifth failure: same source as 4, this time extracted as a .mpg using an entirely different app. ~79%


If you think it may be informative, I could try (for example)
- remuxing (tsMuxeR or AVIDemux) any of the above before trying to encode
- yet another different title from yet another source
- something PAL format
or anything else you think may help narrow it down.

Could you monitor the free disk space of all of your disks (not just the one with working directory) during the encode? Also keep an eye on your GPU memory (via W10 Task manager).

[RetsimLegin]

@cartman0208
Yes, probably. But it's perhaps worth repeating (as regards both of those factors)
I have tried a ~90 minute title and several ~45 minute titles, and the point at which each failed was not the same in data terms; it was similar in percentage complete terms i.e. the longer title encoded close to twice as much duration (and data) as the shorter one.

[RetsimLegin]

All I can say for sure is that it appears to be something specific to your system, as others aren't experiencing it. But right now I have no idea what it is.
I used tsMuXeR to remux one of the previous test files
a) leaving the frame rate at 29.97 as per source
b) changing the frame rate to 23.976 and remove pulldown (it is a film source).
Reverted the .ini file to remove the no frame check flag.

Outcome:
a) failed at the same place. I did look at the imports folder > {title} > BDMV > STREAM and the entire title was there at its oiriginal size and MPEG2 codec and played back OK in VLC.
b) re-encoded the file fully. But interestingly (?) it left the frame rate at 23.976 which (as far as I know?) isn't valid for SD content in H264 on BluRay.

[cartman0208]

@cartman0208
Yes, probably. But it's perhaps worth repeating (as regards both of those factors)
I have tried a ~90 minute title and several ~45 minute titles, and the point at which each failed was not the same in data terms; it was similar in percentage complete terms i.e. the longer title encoded close to twice as much duration (and data) as the shorter one.

If there are no findings, and Jdobbs has no objections:
You could try to replace the nvenc-files in the BD_Rebuilder\tools\nvenc folder with the most current ones from here (https://github.com/rigaya/NVEnc/releases)
You should use the 32bit-version.
Meanwhile there's like 15 updates between the most current and the version used in BDRB.
I didn't read all the patchnotes, but maybe there's something in there that might help.
Make a backup of the old folder, of course.

[RetsimLegin]

Could you monitor the free disk space of all of your disks (not just the one with working directory) during the encode? Also keep an eye on your GPU memory (via W10 Task manager).
Disk space - >600gb free throughout.
GPU memory - hardly visible at the bottom of the graph.
However I did note that - with automatic quality settings, the Video encode and Video decode graphs hovered around 80% until it got to the point of failure (always exactly the same # of frames in the same title) when it briefly shot up to 100% before abruptly droping to zero as it terminated.

[RetsimLegin]

If there are no findings, and Jdobbs has no objections:
You could try to replace the nvenc-files in the BD_Rebuilder\tools\nvenc folder with the most current ones from here (https://github.com/rigaya/NVEnc/releases)
You should use the 32bit-version.
Meanwhile there's like 15 updates between the most current and the version used in BDRB.
I didn't read all the patchnotes, but maybe there's something in there that might help.
Make a backup of the old folder, of course.

Thanks for this suggestion. Outcome:
1: It ran the GPU at a lower rate (~55%) than previously, but
2: As before, shot up to 100% at exactly the same frame number as before and then terminated.

(Arrgghh!)

[cartman0208]

Thanks for this suggestion. Outcome:
1: It ran the GPU at a lower rate (~55%) than previously, but
2: As before, shot up to 100% at exactly the same frame number as before and then terminated.

(Arrgghh!)

Hmm, ok ... if it is always the same frame ... then there might be something off with the source... I'm not that much into video encodeing but maybe there are tools to analyze that exact frame (maybe also the one before and after) and see if there is something suspicious

Also, if possible, you could try the encode with the same source on a different PC

[kpic]

Sorry to resurrect a question from several pages ago but in that question I had asked about UHD as ISO (so no compress through BDRB) or putting the UHD title on BD-50 media & JDobbs responded;

Yes. You can either set FORCE_NOENCODE=1 in the INI file, or create a custom output size and make it larger than the input. If there is no need to recompress, the BD-RB will keep it intact.

Just a warning, though. The maximum encode bitrate for a BD-66 or BD-100 is higher than a BD-50. I haven't personally had an issue with it... but it's possible there could be playback issues if the bitrate goes higher than acceptable for BD-50, assuming the source came from a BD-66 or BD-100.

So that second part had me wondering; The maximum encode bitrate for BD-50 is the BR spec of (I think 48 MBs all-in?) or at least I assume that is its limit or slightly higher so when I go to run my UHD title through BDRB for compressing is there a way to tell BDRB to specifically limit any peak bitrate to " this amount "?

I know it will happen somewhat just by compressing but for my test title going from the source 54gb to 43gb the peak bitrate might still be too high for BD-50 media.

[mikeonecl]

[QUOTE=CraigWally;1941117]If you have quality on Highest change to High (Default) and then report back[/QUOTE

I tested on High Quality and NVENC the glitch disappear.

any reason to fail on highest quality?

[LowDead]

[QUOTE=CraigWally;1941117]If you have quality on Highest change to High (Default) and then report back[/QUOTE

I tested on High Quality and NVENC the glitch disappear.

any reason to fail on highest quality?

Are the glitches on the main movie or the extras? If on extras, is it 1080i/p?

//LD

[RetsimLegin]

Hmm, ok ... if it is always the same frame ... then there might be something off with the source... I'm not that much into video encodeing but maybe there are tools to analyze that exact frame (maybe also the one before and after) and see if there is something suspicious

Also, if possible, you could try the encode with the same source on a different PC
Thanks for your further input.

I only the have one box with a suitable NVidia card. But using either the same box and the CPU (X)H264 encoder or a different box (no NVidia) and X264 - encodes fine (but slower).

I wouldn't know how to analyse the source beyond sticking it into an editor and looking at the frames - and there is nothing to see there.

[mikeonecl]

[QUOTE=mikeonecl;1941426]

Are the glitches on the main movie or the extras? If on extras, is it 1080i/p?

//LD

on main movie

[CraigWally]

[QUOTE=CraigWally;1941117]If you have quality on Highest change to High (Default) and then report back[/QUOTE

I tested on High Quality and NVENC the glitch disappear.

any reason to fail on highest quality?

I have no idea why, but that is what happens on all my encodes of BR using Nvenc (UHD is fine on Highest)
:confused:

[gamete]

Someone help me with this problem?

Here my error

https://forum.doom9.org/showpost.php?p=1937502&postcount=30428

[DoctorM]

I've run into something odd (I don't have the log any more to post).
I was building a movie only disc of the Kino Lorber release of Bodies, Rest and Motion.
The playlist has three video files, the first is KL's splash screen (which has no audio) , the second is the movie and the third seems to be blank.

The final disc has the audio and subtitles about 11 seconds early, which is the duration of the KL splash screen.
It looks like TSMuxer is the problem (not something that can be fixed in BDRB). The 3 .M2TS files are merged into a single file, but no delay is added to make up for the first segment having no audio or subtitles.

I'm not sure there is a work around for this except a full disc rebuild. Any suggestions?

[jdobbs]

I've run into something odd (I don't have the log any more to post).
I was building a movie only disc of the Kino Lorber release of Bodies, Rest and Motion.
The playlist has three video files, the first is KL's splash screen (which has no audio) , the second is the movie and the third seems to be blank.

The final disc has the audio and subtitles about 11 seconds early, which is the duration of the KL splash screen.
It looks like TSMuxer is the problem (not something that can be fixed in BDRB). The 3 .M2TS files are merged into a single file, but no delay is added to make up for the first segment having no audio or subtitles.

I'm not sure there is a work around for this except a full disc rebuild. Any suggestions?BD-RB looks for and should adjust for delayed audio when it starts at a different time than the video. But... if the playlist has multiple parts and the first M2TS in the playlist has no audio, it would not find a delta (no audio=no delta). I'll have to look at a way of catching that scenario and put a delay in equal to the first (or more) item(s) in the playlist.

I got a similar report on another title recently.

[DoctorM]

I tried using tsMuxerGUI directly and Eac3to to remux first and they both produce the same result. The only work around that occurs to me would be to remux the M2TS directly and recalculate the chapters from the playlist allowing for the splash screen... but at that point a menu/movie/blank extras would just be easier.