View Bug Details
ID | Project | Category | View Status | Date Submitted | Last Update |
---|---|---|---|---|---|
0001030 | DCP-o-matic | Bugs | public | 2016-12-13 09:37 | 2018-10-17 20:10 |
Reporter | BoD | Assigned To | carl | ||
Priority | normal | Severity | major | Reproducibility | always |
Status | closed | Resolution | fixed | ||
Platform | 32 bit | OS | Windows | OS Version | XP |
Summary | 0001030: In ver 2.10.3 is DCP very bad. | ||||
Description | Problem is show on picture attached. | ||||
Tags | No tags attached. | ||||
Branch | |||||
Estimated weeks required | |||||
Estimated work required | Average | ||||
|
|
|
What type of source is this from? |
|
Convert progress is OK, convert time is OK, only output DCP is from many square blocks with bad color. |
|
I am do not testing 2.10.2. Ver 2.10.3 is bad, if I return to 2.10.1 all is OK. |
|
Can you send me the video file again? |
|
Yes, video is in Your e-mail box via https://www.wetransfer.com. |
|
I can't reproduce this on Linux; both 2.10.1 and 2.10.3 look the same. I'll try on Windows XP. |
|
BoD, you must accept the idea that Windows XP already Rest In Peace, you should use more actual OS. IMHO |
|
In ver. 2.10.2 is all OK. Ver. 2.10.3 and 2.10.4 is with problem. |
|
In 2.10.6 problem continue. |
|
In 2.11.10 problem continue. |
|
Can you double check again that 2.10.2 works and 2.10.3 does not work? There is only a tiny difference between them. Perhaps you could uninstall DCP-o-matic, try 2.10.2, then uninstall that and try 2.10.3? |
|
Yes, exactly. Now I have not 2.10.3 version yet. But I have 2.10.2 and all is OK. All past version mke unserviceable picture (vide 2016-12-13 09:37 post picture). |
|
Heureka !!! (Sorry for my English :-)) I am made video with framerate 25 fps. This video is for download on adress If I test this video with program MediaInfo, this program show framerate 25 fps. It is OK. Now I add this video to DCP-o-matic. Program analysing this video and show me 50 fps! This is Error! On page DCP is 50 fps too and exporting DCP file is no good picture. If I manual setting from 50 fps to 25 fps, exporting DCP is OK. All is on added picture Err_fps_analysing.jpg. |
|
Which version of DCP-o-matic are you using? I just downloaded this video, and both my players as well as DCP-o-matic 2.10.5 and 2.11.40 indicate it as being 25fps, see screenshot. Maybe something went wrong with your download. How Do you download this video at all? Do you use a video download helper? Or did YOU upload it to that location from your own m2ts source? Because, when I download it, it comes out as MP4, so we can not replicate your issue without access to the exact source file you are using. It seems to be camera footage, and the 25/50 fps misinterpretation could be an interlaced/non-interlaced misinterpretation. Please, don't reconnect to 'old bugs' with new issues, as it doesn't help Carl cleaning up. If you are not sure wether your issue actually qualifies as a bug, first take it to the DCP-o-matic forum. There are also more people possibly being able to help you by replicating the issue. If you are sure you do have a bug, please open a new one.
|
|
I use DCP-o-matic ver. 2.11.40 now. I connect to my old problem, because it is my old problem. In DCP-o-matic ver. 2.10.2 is this video detect OK - 25 fps. Video is from editing SW and I always using this M2TS is container. Codec is MP4. |
|
Can you provide this video in the original m2ts form that creates the problem (e.g. WeTransfer)? Please post a public WT link here, do not email it. Are you still on XP?
|
|
|
|
Yes I still on XP SP3. |
|
Okay, the M2TS file is correctly detected as being 25fps in both DOM 2.10.5 and 2.11.40 on my Mac. It is clearly interlaced, that is, 25i, (or 50i by old school style). Media Info also detects it is interlaced. When I create a DCP from it in 2.10.5 and 2.11.40, I see no issues. I can only guess your version of DCP-o-matic wrongly detects it as 50p.
I can only assume that the version of ffmpeg that compiles for your XP/32 version has trouble to detect the correct frame rate/interlacing property. This may be due to an actual bug in ffmpeg, or your file. What camera has been used for this file? |
|
If You need DCP-o-matic ver. 2.10.2 for XP 32 bit, I have older versions. I my send You. |
|
Yes, looks like an interlacing misinterpretation: My ffprobe: Your ffprobe: |
|
Yes, ffprobe in DCP-o-matic ver. 2.10.2 for XP 32 bit - 148 kB |
|
size doesn't matter - what does ffprobe 2.10.2 say for fps, tbr and tbc?
|
|
ffprobe from DCP-o-matic ver. 2.10.2 for XP 32 bit: |
|
Yup, there it is: 2.11.40 XP32: 25 fps, 50 tbr, 90k tbn, 50 tbc 2.10.2 XP32: 25 fps, 25 tbr, 90k tbn, 50 tbc It's possible that this issue only occurs with a certain type of interlaced sources, or with all interlaced sources. You could try to feed e.g. a plain vanilla 576i MPEG2 from a DVD to 2.11.40 and show us these values. Carl has to judge wether this justifies working on the XP version again. I was and am still using DOM on a very outdated Macbook under snow leopard until I recently bought a new (second hand) machine, and I was happy for every version that still supported this old machine. However, I also told Carl that if it would mean additional effort to keep DOM compatible with 10.6.8 snow leopard, he should give up on it. Let's face it, DOM does heavy work during the J2C conversion, and you need faster up to date machines to bring the conversion times down. At some point, you have to admit it doesn't make sense to put extra effort into supporting these older systems. WIN XP32 is probably the worst OS for running DCP-o-matic also because of the memory limitation. I'd say, prepare for giving up XP32... your machines may probably still run WIN7 nicely. I do have some XP32 machines myself - but not for running up-to-date software, but explicitly for running old software dealing with old hardware. I would never expect anyone to keep a piece of software compatible with these systems. |
|
The same video in MPEG2: All is OK From ffmpeg 2.11.40 XP32 |
|
That seems to come out correctly. So it doesn't seem to be a general interlace detection issue. When you load this MPEG2 video into DOM GUI - does it say 25fps or 50fps in the content tab?
|
|
In DOM 2.11.40 XP32 GUI say 25 fps. |
|
Okay, time for Carl to step in (or not). Carl, how many downloads do you count for the XP versions?
|
|
This is my archive :-) DCP-o-matic 1.76.0 32-bit Installer.exe Now I must go to work with my camera :-) |
|
What camera is it?
|
|
It is Camcorder Panasonic :-) |
|
Does it have an option to shoot noninterlaced?
|
|
Oddly my development version of DoM sees that m2ts as 50fps. |
|
Yes, but my m2ts is 25i fps. And 25i fps (interlaced) is important for my and my friend. |
|
Weird. I'll give it a try in WIN7 Carl, do you know which of these various parameters actually trigger the fundamental frame rate detection? From the ffprobe logs, it seems it is tbr, but is that true? If BoD is only experiencing this with footage from his own camera, there seems to be at least a workaround for him. But if your development version shows the same behaviour, there is probably something that actually needs to be taken care of. While we're at it - what about https://dcpomatic.com/mantis/view.php?id=1130 ? ;-) My Sony camera has a 1080i mode as well and creates M2TS - let's see.
|
|
Okay - there is clearly something wrong. I downloaded the m2ts file again, and: WIN7-64 - DOM 2.11.40 - detects 50fps, 50 tbr in ffprobe I know I wrote different above, but I made a mistake - when I use the same project/metadata folder between testing 2.10.5 and 2.11.40, DOM reuses a previous info file to display content properties for the same content. So, when 2.10.5. detected the content as being 25fps, that was stored in the info file. When I opened the same project with 2.11.40, and added the same content again, there was no new analysis being performed, but it just looked up the 25fps from the 2.10.5 info.
|
|
Still present in 2.11.42 |
|
Yes, problem continue. |
|
In 2.11.50 problem with fps (25i fps) still present. |
|
Could be an ffmpeg issue. We still don't know. As DCP-o-matic detects other interlaced videos correctly, it may well be that your camera does something strange. I took a 1080i50 video with my Sony cam. When I open it on my Mac, it comes up as BDMV, and I can only export it to a MOV with Quicktime player. This will be detected as 25i by DCP-o-matic. ffprobe 25 fps, 25 tbr, 90k tbn, 50 tbc
|
|
My video "Test_HD_5sec.m2ts" is no from camera, but is from edititing sw Canopus EDIUS. Now I test another video from my broadcast camera Panasonic. It is in "2017_08_26 • 15_41_32_ffprobe.log" file. FFprobe has problem with correct detect video. DOM 2.10.2. present all videos OK (but video from broadcast camera FFmpeg detect as 25.25 fps too, but 25 tbr - "25.25 fps, 25 tbr, 90k tbn, 50 tbc". It is in "2017_08_26 • 15_41_32_ffprobe_DOM_2_10_2.log" file. 2017_08_26 • 15_41_32_MediaInfo.txt (2,849 bytes)
General ID : 1 (0x1) Complete name : D:\Prac\16_9\2017_08_26 - 01 PT • Hospic Music Fest 1\2017_08_26 • 15_41_32.MTS Format : BDAV Format/Info : Blu-ray Video File size : 3.35 MiB Duration : 2s 358ms Overall bit rate mode : Variable Overall bit rate : 11.9 Mbps Maximum Overall bit rate : 18.0 Mbps Video ID : 4113 (0x1011) Menu ID : 1 (0x1) Format : AVC Format/Info : Advanced Video Codec Format profile : High@L4 Format settings, CABAC : Yes Format settings, ReFrames : 4 frames Format settings, GOP : M=3, N=12 Codec ID : 27 Duration : 2s 320ms Bit rate mode : Variable Bit rate : 11.2 Mbps Maximum bit rate : 16.8 Mbps Width : 1 920 pixels Height : 1 080 pixels Display aspect ratio : 16:9 Frame rate : 25.000 fps Color space : YUV Chroma subsampling : 4:2:0 Bit depth : 8 bits Scan type : Interlaced Scan type, store method : Separated fields Scan order : Top Field First Bits/(Pixel*Frame) : 0.216 Stream size : 3.09 MiB (92%) Audio ID : 4352 (0x1100) Menu ID : 1 (0x1) Format : AC-3 Format/Info : Audio Coding 3 Mode extension : CM (complete main) Format settings, Endianness : Big Codec ID : 129 Duration : 2s 432ms Bit rate mode : Constant Bit rate : 256 Kbps Channel(s) : 2 channels Channel positions : Front: L R Sampling rate : 48.0 KHz Frame rate : 31.250 fps (1536 spf) Compression mode : Lossy Delay relative to video : -80ms Stream size : 76.0 KiB (2%) Test_HD_5sec_MediaInfo.txt (2,419 bytes)
Hlavní ID : 0 (0x0) Complete name : D:\DCP\Video test\Test_HD_5sec.m2ts Format : BDAV Format/Info : Blu-ray Video File size : 12,3 MiB Duration : 5s 578ms Overall bit rate : 18,5 Mbps Maximum Overall bit rate : 18,1 Mbps Video ID : 4113 (0x1011) Menu ID : 1 (0x1) Format : AVC Format/Info : Advanced Video Codec Format profile : High@L4 Format settings, CABAC : Ano Format settings, ReFrames : 2 frame Format settings, GOP : M=3, N=18 Codec ID : 27 Duration : 5s 0ms Bit rate : 17,0 Mbps Width : 1 920pixely Height : 1 080pixely Display aspect ratio : 16:9 Frame rate : 25,000 fps Color space : YUV Chroma subsampling : 4:2:0 Bit depth : 8 bits Scan type : Interlaced Scan order : Top Field First Bits/(Pixel*Frame) : 0.328 Stream size : 10,4 MiB (84%) Audio ID : 4352 (0x1100) Menu ID : 1 (0x1) Format : AC-3 Format/Info : Audio Coding 3 Mode extension : CM (complete main) Format settings, Endianness : Big Codec ID : 129 Duration : 5s 24ms Bit rate : 384 Kbps Channel(s) : 2 kanály Channel positions : Front: L R Sampling rate : 48,0 KHz Frame rate : 31,250 fps (1536 spf) Stream size : 236 KiB (2%) |
|
These 'broken' framerates usually occur only with very short videos. Somehow the frame rate is not just read from the header, but there is also a calculation from total length vs. number of frames. I have seen that happening with a short WMV from powerpoint as well (30.0003 fps). It is weird that frame rate detection seems to be so ambigous in the digital (ffmpeg) world. One would think that there was a single solid/trustworthy base frame rate & and interlaced/progressive indicator, but that 25 fps, 50 tbr, 90k tbn, 50 tbc is quite confusing...
|
|
Hmm. I find more videos on my disc that are reported as 25fps by VLC and MediaInfo, but 50fps by DCP-o-matic. I wouldn't care for the occasional false detection, as long as user can easily correct this in DCP-o-matic. The trouble is - beginners will probably not notice this issue, will then be reminded to create SMPTE DCPs, and will probably create 50fps or 60fps DCPs without knowing that they won't play on most systems... If SMPTE is preselected in prefs, DCP-o-matic will silently create 50 or 60fps DCPs...not good.
|
|
Can I get hold of an example or two? |
|
Yes, send me Your mail addres to bod@bod.cz an I send You via WeTransfer two example. |
|
Thanks. I hope I don't live to regret this, but 17f9254f68fabfe5d8136d256b69a65fc58724ee should help. |
|
Did something about frame rate detection change between 2.10.5 and 2.11.x? WIN7-64 - DOM 2.11.40 - detects 50fps, 50 tbr in ffprobe
|
|
I have only 2.10.2. |
|
Could you try 2.11.54? |
|
Yes, DOM 2.11.54 is OK. |
|
Turns out that this bug report was caused by the same problem: https://dcpomatic.com/mantis/view.php?id=1140 The same content file was interpreted as 25fps by 2.10.5, but as 50fps by 2.11.27 - the DCP created with 2.10.5 played okay, 2.11.27 created a 50fps DCP, which did not play on the target system. Hence I think DCP-o-matic should always issue a warning when non-24fps content is created.
|
|
Since BoD prefers XP (even XP32) - the new FFMPEG v4 seems to have abandoned XP support - will that hit DCP-o-matic XP versions?
|
|
Probably. It depends whether they've broken it or just stopped supporting it. |
Date Modified | Username | Field | Change |
---|---|---|---|
2016-12-13 09:37 | BoD | New Bug | |
2016-12-13 09:37 | BoD | File Added: 2_10_1-3.jpg | |
2016-12-13 09:47 | carl | Note Added: 0001567 | |
2016-12-13 09:47 | carl | Assigned To | => carl |
2016-12-13 09:47 | carl | Status | new => feedback |
2016-12-13 10:14 | BoD | Note Added: 0001568 | |
2016-12-13 10:14 | BoD | Status | feedback => assigned |
2016-12-13 10:15 | BoD | Note Added: 0001569 | |
2016-12-13 10:16 | BoD | Note Edited: 0001569 | |
2016-12-13 10:24 | carl | Note Added: 0001570 | |
2016-12-13 10:24 | carl | Status | assigned => feedback |
2016-12-13 10:45 | BoD | Note Added: 0001571 | |
2016-12-13 10:45 | BoD | Status | feedback => assigned |
2016-12-13 14:10 | carl | Note Added: 0001572 | |
2016-12-14 06:09 | Igor.Voyt | Note Added: 0001573 | |
2016-12-15 15:57 | BoD | Note Added: 0001577 | |
2017-01-13 09:00 | BoD | Note Added: 0001595 | |
2017-06-08 05:38 | BoD | Note Added: 0001702 | |
2017-06-08 09:32 | carl | Note Added: 0001703 | |
2017-06-09 07:10 | BoD | Note Added: 0001709 | |
2018-01-12 13:04 | BoD | File Added: Err_fps_analysing.jpg | |
2018-01-12 13:04 | BoD | Note Added: 0001958 | |
2018-01-12 13:24 | Carsten | File Added: Bildschirmfoto 2018-01-12 um 14.16.27.png | |
2018-01-12 13:24 | Carsten | Note Added: 0001959 | |
2018-01-12 13:26 | Carsten | Note Edited: 0001959 | |
2018-01-12 13:27 | Carsten | Note Edited: 0001959 | |
2018-01-12 13:34 | BoD | Note Added: 0001960 | |
2018-01-12 13:39 | Carsten | Note Added: 0001961 | |
2018-01-12 13:41 | Carsten | Note Edited: 0001961 | |
2018-01-12 13:42 | Carsten | Note Edited: 0001961 | |
2018-01-12 14:02 | BoD | Note Added: 0001962 | |
2018-01-12 14:03 | BoD | Note Added: 0001963 | |
2018-01-12 14:25 | Carsten | Note Added: 0001964 | |
2018-01-12 14:25 | Carsten | Note Edited: 0001964 | |
2018-01-12 14:32 | Carsten | Note Edited: 0001964 | |
2018-01-12 14:36 | BoD | File Added: capture_scr.jpg | |
2018-01-12 14:36 | BoD | File Added: ffprobe.log | |
2018-01-12 14:36 | BoD | Note Added: 0001965 | |
2018-01-12 14:45 | Carsten | Note Added: 0001966 | |
2018-01-12 15:08 | BoD | Note Added: 0001967 | |
2018-01-12 15:33 | Carsten | Note Added: 0001968 | |
2018-01-12 15:42 | BoD | File Added: ffprobe_2_10_2.log | |
2018-01-12 15:42 | BoD | Note Added: 0001969 | |
2018-01-12 15:49 | Carsten | Note Added: 0001970 | |
2018-01-12 16:05 | Carsten | Note Edited: 0001970 | |
2018-01-12 16:09 | BoD | File Added: ffprobe_mpeg2.log | |
2018-01-12 16:09 | BoD | Note Added: 0001971 | |
2018-01-12 16:29 | Carsten | Note Added: 0001972 | |
2018-01-12 16:33 | BoD | Note Added: 0001973 | |
2018-01-12 16:37 | Carsten | Note Added: 0001974 | |
2018-01-12 16:55 | BoD | Note Added: 0001975 | |
2018-01-12 17:07 | Carsten | Note Added: 0001976 | |
2018-01-12 17:17 | BoD | Note Added: 0001977 | |
2018-01-12 18:04 | Carsten | Note Added: 0001978 | |
2018-01-12 23:57 | carl | Note Added: 0001980 | |
2018-01-13 06:43 | BoD | Note Added: 0001981 | |
2018-01-13 12:50 | Carsten | Note Added: 0001982 | |
2018-01-13 15:24 | Carsten | Note Edited: 0001982 | |
2018-01-13 16:04 | Carsten | Note Edited: 0001982 | |
2018-01-13 16:11 | Carsten | Note Edited: 0001982 | |
2018-01-13 16:27 | Carsten | Note Edited: 0001982 | |
2018-01-13 17:27 | Carsten | Note Added: 0001983 | |
2018-01-13 17:59 | Carsten | Note Edited: 0001983 | |
2018-01-14 02:09 | Carsten | Note Edited: 0001983 | |
2018-01-15 17:24 | Carsten | Note Added: 0001993 | |
2018-01-15 19:58 | BoD | File Added: ffprobe_2_11_42_test01.log | |
2018-01-15 19:58 | BoD | Note Added: 0001996 | |
2018-02-08 07:08 | BoD | File Added: ffprobe-2.log | |
2018-02-08 07:08 | BoD | Note Added: 0002086 | |
2018-02-09 02:14 | Carsten | Note Added: 0002094 | |
2018-02-09 08:05 | BoD | File Added: 2017_08_26 • 15_41_32.jpg | |
2018-02-09 08:05 | BoD | File Added: 2017_08_26 • 15_41_32_ffprobe.log | |
2018-02-09 08:05 | BoD | File Added: 2017_08_26 • 15_41_32_ffprobe_DOM_2_10_2.log | |
2018-02-09 08:05 | BoD | File Added: 2017_08_26 • 15_41_32_MediaInfo.txt | |
2018-02-09 08:05 | BoD | File Added: Test_HD_5sec_ffprobe.log | |
2018-02-09 08:05 | BoD | File Added: Test_HD_5sec_MediaInfo.txt | |
2018-02-09 08:05 | BoD | Note Added: 0002096 | |
2018-02-09 13:33 | Carsten | Note Added: 0002099 | |
2018-02-09 13:35 | Carsten | Note Edited: 0002099 | |
2018-02-10 12:36 | Carsten | Note Added: 0002124 | |
2018-02-10 15:56 | carl | Note Added: 0002125 | |
2018-02-10 16:06 | BoD | Note Added: 0002126 | |
2018-02-10 21:40 | carl | Note Added: 0002127 | |
2018-02-10 23:17 | Carsten | Note Added: 0002129 | |
2018-02-11 07:38 | BoD | File Added: DOM_2_10_2_32-bit_ffprobe.log | |
2018-02-11 07:38 | BoD | File Added: DOM_2_11_52_32-bit_ffprobe.log | |
2018-02-11 07:38 | BoD | Note Added: 0002131 | |
2018-02-11 16:03 | carl | Note Added: 0002133 | |
2018-02-11 18:31 | BoD | Note Added: 0002134 | |
2018-02-11 19:41 | carl | Status | assigned => resolved |
2018-02-11 19:41 | carl | Resolution | open => fixed |
2018-02-13 02:02 | Carsten | Note Added: 0002159 | |
2018-04-26 11:29 | Carsten | Note Added: 0002401 | |
2018-04-26 11:29 | Carsten | Note Edited: 0002401 | |
2018-04-26 11:34 | carl | Note Added: 0002402 | |
2018-10-17 20:10 | carl | Status | resolved => closed |