Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion. Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules. |
29th January 2012, 01:45 | #14082 | Link |
Registered User
Join Date: Jan 2009
Posts: 1,368
|
it was off-sync from the start afaik. I'll do a new run later today to check if it was a fluke or if i can reproduce it.
i'll do it in the following modes. full disc BD5 (pal enabled though afaik that shouldn't matter should it, as it will be a bd structure) movie only bd-5 (pal enabled) movie only bd-5 (pal disabled) I'll use the same disc (pillars of the earth blu-ray tv series). I'll keep you posted. |
29th January 2012, 10:32 | #14083 | Link | |
Registered User
Join Date: May 2006
Posts: 3,997
|
Quote:
Same as Ch3vr0n I have PAL output enabled (BD-RB does a nice pitch correction ). I don't remember if desync happens in every case. IIRC some came out in perfect sync. Here my ini: Code:
[Options] ENABLE_TEST=1 BLANK_THRESHOLD=3600 #FIXED_CRF=20 QUICK_CRF=27 #TWEAK_PASS_ONE=--ipratio 1.4 --me umh --b-adapt 2 --subme 8 #TWEAK_PASS_TWO=--ipratio 1.4 --me umh --b-adapt 2 --subme 8 TWEAK_PASS_ONE=--ipratio 1.4 --me umh --b-adapt 2 --subme 8 --tune film TWEAK_PASS_TWO=--ipratio 1.4 --me umh --b-adapt 2 --subme 8 --tune film #TWEAK_PASS_ONE=--ipratio 1.4 --me umh --b-adapt 2 --subme 8 --tune film --ref 6 #TWEAK_PASS_TWO=--ipratio 1.4 --me umh --b-adapt 2 --subme 8 --tune film --ref 6 #TWEAK_PASS_ONE=--ipratio 1.4 --me umh --b-adapt 2 --subme 8 --tune animation #TWEAK_PASS_TWO=--ipratio 1.4 --me umh --b-adapt 2 --subme 8 --tune animation B_PYRAMID=1 SAMPLE_GROUP=2400 #SAMPLE_SIZE=48 #DEINTERLACER_TYPE=4 ALTERNATE_PAL=1 WEIGHTP=2 DEFAULT_LANG=deu EXTENDED_GOP=1 KEEP_MBTREE=1 QUICK_USE_QUALITY=1 SHOW_ENCODER=1 FORCE_ENCODE=0 SECONDARY_USE_QUALITY=1 HC_PROFILE=BEST RESIZE=0 MODE=2 AUDIO_AMPLIFY=1.0 AUDIO_TO_KEEP=deu;eng;ger; SUBS_TO_KEEP=deu;ger; SD_CONVERT=0 COLOR_BOOST=0 RESIZE_1080=1 DTS_REENCODE=1 AC3_REENCODE=0 AC3_640=0 KEEP_HD_AUDIO=0 AVCHD=0 REMOVE_WORKFILES=0 AUDIO_TRACK_LIMIT=1 SUBTITLE_TRACK_LIMIT=0 CUSTOM_TARGET_SIZE=4483 VERBOSE_STATUS=1 ONEPASS_ENCODING=1 TARGET_SIZE=8032 QUICK_EXTRAS=1 PRIORITY_CLASS=1 ENCODE_QUALITY=2 AUTO_QUALITY=0 DEINTERLACE=0 AC3_192=0 FULL_AVCHD=0 REMOVE_OUTPUT=0 OPEN_GOP=1 USE_FILTERS=1 MOVIE_ONLY_LOOP=0 BDMV_CERT_ONLY=0 USE_LAVF=0 IVTC_PULLDOWN=0 ASSUME_DVD_PAL=1 VERSION=0.39.0.7 MOVIEONLY_TYPE=0 ALTCRF=24 ALTMETHOD=0 SD_TO_1080=0 CONVERT_WIDE=0 ALT_TARGET=1000 AUTO_BLANK=0 UNMASK_CHAPTER=0 ENABLE_BLANKING=1 COMPLETION_BEEP=0 ALTAUTOCROP=0 AVSFilter01=f:r:Sharpen(0.15) [Paths] ....... |
|
29th January 2012, 18:26 | #14084 | Link |
Registered User
Join Date: Dec 2002
Posts: 1,022
|
I wonder what the Japanese authoring shops do differently. I bumped into another Japanese BD (Goemon) which gives BD-RB a headache. Trying to do a simple movie-only backup resulted in audio playing a second late from the start. The original looks like standard 1080p23.976 but it does have a PiP video commentary track on it.
|
29th January 2012, 19:36 | #14085 | Link | |
Registered User
Join Date: May 2006
Posts: 3,997
|
Quote:
Sync is off when "Assume PAL for DVD output" is enabled Sync is perfect when "Assume PAL for DVD output" is disabled. Not sure now if this has been the case with previous releases of BD-RB. I am pretty sure having done similar backups before with PAL output enabled, without sync issues. The sync issue happens when making a DVD backup from a previous BD5 backup (which had the original DTS converted to AC3 448 in my cases). |
|
29th January 2012, 21:44 | #14086 | Link | |
Moderator
Join Date: Oct 2001
Posts: 20,975
|
Quote:
|
|
29th January 2012, 22:09 | #14087 | Link | |
Moderator
Join Date: Oct 2001
Posts: 20,975
|
Quote:
I have a DVD-9 backup of "Rise of the Planet of the Apes" laying around somewhere -- I'll find it and run against that for testing... Last edited by jdobbs; 29th January 2012 at 22:18. |
|
30th January 2012, 00:07 | #14088 | Link | |
Registered User
Join Date: May 2006
Posts: 3,997
|
Quote:
The framerate of the source (=BD5 movie-only backup) is 23.976 fps. Please note that the problem exists only when converting a BD5 backup to DVD. When I convert from the original Blu-ray disc to DVD the sync is ok. |
|
30th January 2012, 00:17 | #14089 | Link | |
Registered User
Join Date: May 2006
Posts: 3,997
|
Quote:
Anyway, as I mentioned I can always make a DVD from the original BD source rather than from its BD5 backup. So not too much to worry about. |
|
30th January 2012, 00:18 | #14090 | Link | |
Moderator
Join Date: Oct 2001
Posts: 20,975
|
Quote:
I just finished "Rise of the Planet of the Apes" from the BD-9 backup win ALTERNATE_PAL=1. It also was converted to AC3 5.1 from the original. It has perfect sync throughout. It's possible you may have a CODEC issue... but don't know for sure. I'd suggest you uninstall/reinstall BD-RB, HAALI, AVISYNTH, and FFDSHOW. Last edited by jdobbs; 30th January 2012 at 00:21. |
|
30th January 2012, 00:32 | #14091 | Link |
Moderator
Join Date: Oct 2001
Posts: 20,975
|
BD Rebuilder v0.40.01 (beta)
I've updated the first post of this thread with a new version of BD-RB (v0.40.01). Changes for this release:
Code:
- Corrected an issue related to the program map and remuxing of DTS Express and IGS streams. - Modified ALTERNATE routines so that MKV and MP4 (non-iPOD/iPAD) formats will keep HD audio when "Keep HD Audio" and "ALTERNATE/Intact Audio" are both enabled. - Modified ALTERNATE AVS creation to prevent unnecessary inclusion of resize commands. This can speed up some encodes. - Added routines to include alternate angles in the "Other Movie-Only Playlist" selection dialog. - Reverted the recommended HAALI media splitter to v1.9.42.1. Some systems seem to have issues with the later version(s). - Added a new (experimental) multi-processing mode for extremely fast computers. Divides encoding tasks among multiple instances of X264. Most useful on systems that don't use 100% of the available processor time. See the link in HIDDENOPTS.TXT for more details. - Updated the included version of MKVMERGE.EXE to a newer release (v5.2.1.0). - Updated the included version of X264.EXE to the latest release (r2146). - Updated the included version of X264-64.EXE to the latest release (r2146). - Other minor corrections and cosmetic fixes. Last edited by jdobbs; 30th January 2012 at 01:10. |
30th January 2012, 04:12 | #14093 | Link |
Registered User
Join Date: Jan 2010
Location: USA, Oregon
Posts: 791
|
I set mine for the same. I'll likely run an encode tonight, then run the same encode again during the day tomorrow. I imagine a super fast computer would be a 6 or more core computer(or intel extreme). I imagine Jdobbs has at least a 4 core. Which I consider fast. However not SUPER fast Can't wait for a 10 core. I'm waiting to upgrade til those come out.
|
30th January 2012, 04:45 | #14094 | Link |
Moderator
Join Date: Oct 2001
Posts: 20,975
|
I'm thinking specifically of the folks who have posted here that their computer doesn't run at 100% processor utilization during encoding. As I mentioned in the changelog and HIDDENOPTS.TXT, a little more detail on the multiprocess setting is available at this link.
I'm running a Phenom II X4 945 @3.5Ghz -- but my system has always run at 100% during pass 2 so I get nothing there -- and although pass 1 typically doesn't use it all, it also doesn't leave a lot on the table. So I pick up a little on pass 1, nothing to write home about... but I'm thinking others will see much better improvements. More importantly I'm trying to step out in front of all the really fast processors that are coming down the road. This multiprocessing feature was a lot of work, especially in trying to create frame-accurate segments in an environment that doesn't seem to care about frame accuracy. I had to do a lot of head banging -- so I'm hoping it doesn't disappoint. Last edited by jdobbs; 30th January 2012 at 04:59. |
30th January 2012, 05:20 | #14095 | Link | |
squadraphonic PSN
Join Date: Jul 2010
Location: Australia
Posts: 6
|
Quote:
|
|
30th January 2012, 07:17 | #14096 | Link | |
Registered User
Join Date: Aug 2005
Posts: 16,267
|
@jdobbs
Quote:
|
|
30th January 2012, 09:17 | #14097 | Link |
Registered User
Join Date: Jan 2010
Location: USA, Oregon
Posts: 791
|
I'm already impressed Jdobbs I'm seeing much more processor usage already, and i've barely began the test. Even if it isn't what I expect, I'm already impressed. You're definitely getting some support in the next 7 - 14 days. Can't wait to see the difference it's made.
__________________
Only one rooster, need be in the hen house... |
30th January 2012, 13:32 | #14098 | Link |
Registered User
Join Date: Dec 2002
Posts: 1,022
|
Dual Xeon with 12GB of RAM here. I set multiprocess to 3 right away to see what happens. During pass 1 CPU usage was usually somewhere around 20-30%. With multiprocess=3 it's now hovering at 75-80% and BD-RB reports "FPS: 381". Now, before anyone soils their pants I have to admit the file it's processing is only 480i
Pass 2 is now commencing. CPU usage jumped to 100% and BD-RB reports 117 fps. This is still on a 480i file which is being deinterlaced and re-encoded by BD-RB. Memory usage is only 2,10 GB. CPU use for the three instances of x264 is 45%, 33% and 22%. The instance with the highest CPU % is using 350MB of memory, and the other two use 320MB each. After ~75% of file had been processed one instance of x264 finished its work and quit. With only two x264 instances running FPS jumped to 133. This suggests it might be a good idea to run more x264 instances during pass 1 than pass 2. For example, 4 instances during pass 1 for maximum CPU usage, then only 2 instances during pass 2. Main movie file processing begins in a few minutes. I'll update this post then. BD-RB splitting main movie file (35GB) for 3-way encoding. Looks like a 10-minute job, approximately. (Used the waiting time wisely by making a donation.) The splitting process made me think about using an SSD drive to see if it affects performance during splitting phase. OK, here we go. Source: AVC 1080p/23.976 BD-RB set to Highest quality with "--tune film" tweaks in the .INI file. PASS 1 x264 instances #1: 20-30% CPU #2: 19-25% CPU #3: 18-23% CPU Total CPU use 59-73% Total RAM use 4,37 GB Speed: 2,61x FPS: ~63 RAM use has increased during the first few minutes from 800MB per x264 instance to 1,3GB per instance. Total RAM use has risen from below 4GB to 4,8GB. update: Pass 1 is now 56% complete. CPU usage hovers around 65-70% and each instance eats up 1,4GB of RAM. Total RAM use is now 5,15GB. PASS 2 at 10% completion x264 instances #1 at 50% CPU #2 at 15-20% CPU #3 at 30-40% CPU Instances #2 and #3 use 1,1 GB of RAM each, #1 uses 0,1GB more. Total CPU use 100% (Explorer goes like molasses) Total RAM use 4,4 GB Speed: 0,71x FPS: ~17 (clear improvement over the usual 11-12 fps) at 84% completion: 2 instances of x264 running #1: CPU 55%, 1,2GB of RAM #2: CPU 45%, 1,2GB of RAM Speed: 0,77x FPS: ~18,5 Total RAM use 3,62GB Last edited by colinhunt; 30th January 2012 at 19:34. |
30th January 2012, 14:11 | #14099 | Link |
Registered User
Join Date: Jan 2009
Posts: 1,368
|
I'm game for trying out the new multiprocessing on my quadcore. I'm curious if there will be a time difference between LAVF & multiprocessing. I'll let ya know. Oh and sry i havent had time to do that audio sync issue. That will be done later this week. Oh and ya might wanna change the year in the startup image Still lists 2011
** edit: looking nice so far ** Source: Friends with benefits on 157.415 frames @ 18.425kbs @ MPEG-4 (AVC) Mode: movie & menus quality: 1 pass auto mode Pass 1 3 instances of x264.exe #1: 30-40% #2: 26-35% #3: 24-35% All 3 using atm approx 650 - 670MB ram each Total CPU use: 100% Total ram use: 4.40-4.81GB atm 12% into the Speed: 1.44x FPS: ~34 ** edit 25% into the build, ram use increased to approx 800MB each and a total of 5.41 ** Last edited by Ch3vr0n; 30th January 2012 at 15:40. |
30th January 2012, 15:53 | #14100 | Link | |
Moderator
Join Date: Oct 2001
Posts: 20,975
|
Quote:
Unfortunately I can't change the number of X264 instances between passes 1 and 2 -- because X264 needs them to be identical. It would also require another split -- which would also add time. |
|
|
|