View Bug Details
ID | Project | Category | View Status | Date Submitted | Last Update |
---|---|---|---|---|---|
0000872 | DCP-o-matic | Clean-up | public | 2016-05-09 21:20 | 2018-10-17 20:14 |
Reporter | rexxtype | Assigned To | carl | ||
Priority | high | Severity | minor | Reproducibility | always |
Status | closed | Resolution | no change required | ||
Platform | Dolby server | OS | Apache? | OS Version | ? |
Summary | 0000872: Glitch looking colored frames before first image plays | ||||
Description | v2.8.0 In two of the DCP's I have tested and made, (trailers for our movie theater), I have gotten an error of colored glitched bars and visual noise for some frames before the first actual .j2c image is played. Work at an indie movie theater using a 2k barco projector and dolby OS. worked fine for another trailer so 1/3 have NOT had this error and 2/3 HAVE. I recorded a video displaying the problem available at this link: https://drive.google.com/open?id=0Bxn0CQYT0ZJMX2haRThIZTJMc2s I have also included a zip of the dcp files but not any .mxf's or images. : Thanks for everything you do in making this kind of thing available for us!~ | ||||
Steps To Reproduce | Tested multiple times on cinema 2k projector. | ||||
Additional Information | from log: 5/2/2016 5:35:58 PM: DCP-o-matic 2.8.0 git 8e2c945bd4 using libavcodec 57.27.101, libavfilter 6.38.100, libavformat 57.27.100, libavutil 55.19.100, libswscale 4.0.100, ImageMagick 6.8.7-10 Q16 i686 2016-02-10 http://www.imagemagick.org, libssh 0.5.2/openssl/zlib, libdcp 1.3.3 git 0d69f5c68c metadata: | ||||
Tags | No tags attached. | ||||
Branch | |||||
Estimated weeks required | |||||
Estimated work required | Average | ||||
|
That kind of artifact is often seen between content with different frame rates. Do you only see it between a 24fps and a 25fps clip, or vice versa? I believe this is a projector limitation. |
|
Hmm. Both of the dcps with artifacts were single youtube video clips i had downloaded from an online converter. And they both had still title frames at the end. The working dcp waa the same as well. Will double check my timeline, but shouldnt all files have been conformed to whatever my export settings were out of premiere with the added still titles after the videos? I guess just trying to figure out ways to make sure it never happens in my workflow |
|
YouTube RIPS can be very strange. It may be that the codec is broken at the beginning or end. You may not notice it because of the complex nature of these codecs. It's very easy to check if you try with different source footage. If it only happens with this media, there is a problem within. You may try to trim it in DOM or externally to see wether these glitches are gone. You could also try MPEG Streamclip to trim and/or analyze time code errors.
|
|
Do we know yet what frame rates we are working with when these problems occur? |
|
I know that some Dolby servers show that issue with adjacent clips - e.g. green flashing. Sometimes related to mixed 24p/25p footage. A colleague with Dolby servers calls me every now and then about it. I guess we need to see some examples (source footage). It should also be possible to analyze the problematic DCPs, as all source-related/interframe codec issues are gone on that level, and one should quickly see wether any broken frames are now baked into j2c, or, if clean, are related to server/projector.
|
|
Yes the source video is indeed 25fps . Would you like me to give you any more information as well? And yes I would love to analyze the DCP to come up with solutions etc. So I'm assuming it would be wise to make sure every clip is conformed to the same frame rate? |
|
I would also like to add that this is not a high priority issue currently for me but am fascinated in understanding and troubleshooting the reproduction of these source files. |
|
It's not just the source video, it's the combination of the video with the flash and the one before it in the playlist e.g. if you have MY_GREAT_24FPS_DCP in the playlist, some systems will put in a glitchy frame or two just as the 24fps DCP finishes and the 25fps one starts. There may be nothing wrong with the DCPs. I'm just trying to eliminate this as a possibility, as there isn't much we can do about it from DCP-o-matic's end. It may be that the DCPs have glitches, which we can also investigate once this has been ruled out. |
|
Right this makes sense. The other video was 24fps in the dolby playlist we had built (i assume that is what you are talking about) |
|
OK. I'm not sure if the Dolby server suffers from this problem. Perhaps you could make a 25fps version of the "other" DCP (just change the frame rate in the DCP tab to 25 and re-make). Then see if you see a glitch when you play the two 25fps DCPs one after the other. |
|
I'm assuming this was sorted out, or at least the cause of it is now understood. |
Date Modified | Username | Field | Change |
---|---|---|---|
2016-05-09 21:20 | rexxtype | New Bug | |
2016-05-10 00:20 | carl | Note Added: 0001231 | |
2016-05-10 00:20 | carl | Assigned To | => carl |
2016-05-10 00:20 | carl | Status | new => feedback |
2016-05-10 20:12 | rexxtype | Note Added: 0001232 | |
2016-05-10 20:12 | rexxtype | Status | feedback => assigned |
2016-05-10 20:12 | rexxtype | Note Edited: 0001232 | |
2016-05-11 23:36 | Carsten | Note Added: 0001243 | |
2016-07-04 13:50 | carl | Note Added: 0001330 | |
2016-07-04 13:50 | carl | Status | assigned => feedback |
2016-07-04 13:59 | Carsten | Note Added: 0001333 | |
2016-07-04 18:50 | rexxtype | Note Added: 0001334 | |
2016-07-04 18:50 | rexxtype | Status | feedback => assigned |
2016-07-04 18:52 | rexxtype | Note Added: 0001335 | |
2016-07-04 19:09 | carl | Note Added: 0001336 | |
2016-07-04 19:13 | carl | Status | assigned => feedback |
2016-07-04 19:39 | rexxtype | Note Added: 0001338 | |
2016-07-04 19:39 | rexxtype | Status | feedback => assigned |
2016-07-04 19:54 | carl | Note Added: 0001339 | |
2016-11-02 12:45 | carl | Note Added: 0001518 | |
2016-11-02 12:45 | carl | Status | assigned => resolved |
2016-11-02 12:45 | carl | Resolution | open => no change required |
2018-10-17 20:14 | carl | Status | resolved => closed |