|
|
@fary54 Crashes are always bad, but I can not reproduce this. These are the steps I used to try to do so. What am I doing differently than you? 1) I am in Cut and Paste mode with Resources window set to Full Play on a Fedora distro 2) loaded a video with 2 audio tracks 3) cut out a part and made a clip 4) dragged that clip to the tracks after the original video 5) cut out a part of that clip to make clip 2 6) dragged that second clip to the timeline after the other clip 7) used the File pulldown to do a "Save as" - no crash; tried "Save" - no crash 8) did the above steps again and this time used File pulldown to "Save backup" - no crash; did Quit - no crash When it crashes, do you get a file on /tmp with the name of cinelerra_xxxxx.dmp? If so, could you send that to me? |
|
|
First of all, thank you for your quick response. Don't worry, cinelerra GG doesn't crash. What happens is that you lose the first clip "clip 1" during the rescue because it is replaced by "clip 2". As I can understand, you cannot reproduce this bug because you use the timeline as a buffer. So I modified your way of doing to reproduce this bug: 1) ok 2) ok 3) ok 4) Do not drag clip 1 to the timeline but enter inside that clip 1 (via openEDL menu) 5) Create clip2 from that clip 1 elements 6) Do not drag clip 2 to the timeline 7) Exit clip 1 (via close openEDL) and save So, schematically you get Project | clip 1 | clip 2 Try to reloading your job and you get Project | clip 2 So everything that was in clip 1 is lost and is replaced by clip 2. I can redo a new video if you want. |
|
|
Sorry, but somehow I missed the video and 2 session files (I feel like an idiot). GG sees the problem in your video and says it is a probllem because there is no place to put a Clip inside of a Clip. He can not think of a solution for this so there may just have to be a restriction error message generated so that at least the user knows there is a problem right away. More later. |
|
|
A fix has been checked into GIT. As it turns out, the clip was actually saved in the saved EDL, but it could not be reloaded when loading the project. I tested it and it seems to work now, but is is a little of an unexpected usage. |
|
|
@PhyllisSmith Frequently using this function, I am delighted to find that a solution will be found to resolve this problem. A big thank you for your quick and efficient action. |
|
|
@fary54 Can I ask you about the use cases you make of openEDL? I wanted to expand on the text in the manual, but I'm not an expert and can't say when it's useful to use this function. |
|
|
For me, openEDL is a revolution for cinelerra and has only advantages: editing speed, clarity, ease of finding a given moment, movement by block of subject (Nest to media). Why ? First of all, for information, I make one-hour films (film finished) using three different cameras. * A person uses camera 1 to film a general view of the subject (ex: a jazz group) * Another person uses camera 2 to film a detail of the subject (ex: saxophonist, trumpet player, ...) * camera 3 is a gopro. Placed on the windshield of our vehicle, it is used to record the movements between the subjects. So my final film consists of all the subjects separated by camera rushe 3. Rem: camera 1 and 2 are both filming the same subject so the number of hours for the rushes from these two cameras easily reaches 20 hours and I have to make a film of one hour maximum. How ? In fact, I use openEdl throughout the editing process and only use the timeline as my workspace. Let me explain 1) Pre-sorting of my rushes: I load all the rushes from camera 1 and I create a clip, I do the same with cameras 2 and 3. I watch each clip and sort the images to keep (so via opendEDL). I delete the bougers, the images not retained. Clips ------------------------------------------------------------------------ | | | Rushes camera1 (clip) Rushes camera2 (clip) Rushes camera3 (clip) 2) Extraction of the subjects I extract each common subject from the camera 1 & 2 rushes and create a clip (ex: a place, a city, a subject, ...). So I get as many clips as there are subjects (ex: Barcelona & Lisbon) Clips --------------------------------------------------------------------------------------------------------------------------- | | | | | Rushes camera1 (clip) Rushes camera2 (clip) Rushes camera3 (clip) Barcelone (clip) Lisbonne (clip) 3) Derush of the subjects I edit each subject (therefore via opendEDL). From two tracks 1 & 2 (camera1 & 2) I get a single track. Timeline (ex: Barcelone) --------------------------------------------------------- -piste 1:vidép > rushes of the camera 1 & 2 audio audio 4) Add sound effects and video/audio effects I add the sound effects (such as birds, fountain,...) and video / audio effect for each subject (so via opendEDL) 4) Compaction Each subject is converted to Nest to media, making it a group Clips Nest to media ---------------------------------------------------------------------- ------------------------------------- | | | | | Rushes camera1 (clip) Rushes camera2 (clip) Rushes camera3 (clip) Barcelone (nest to media) Lisbonne (nest to media) 5) Final assembly 5a) Import each subject (Nest to media) on the main timeline. Converting to Nest to Media makes it easy to move and position each subject. You can move one subject after or before another. Main timeline ------------------------------------------------------------------------------------------------------- Piste 1: Generic > Barcelone (nest to media) > Lisbonne (nest to media) > ... 5b) Insert the selected camera rushes3 between the different subjects Main timeline ---------------------------------------------------------------------------------------------------------------------------- Piste1: Generic > Barcelone (nest to media) > camera 3 > Lisbonne (nest to media) > ... 5c) Add background music and comment Main timeline ---------------------------------------------------------------------------------------------------------------------------- Piste1: Generic > Barcelone (nest to media) > camera 3 > Lisbonne (nest to media) > ... Piste2: Background music Piste3: Voice off ( comment ) 6) At this stage, each subject can still be edited and undergo any modification via openEDL (again ... and always) 7) Render your movie You understand that it is very difficult to set up such a project without using tools such as OpenEDl and Nest to media. These tools, for me, are essential and make them pillars for the development of cinelerraGG for important projects. Rem: If necessary, I am ready to make a short video to show you the progress of the process developed above. |
|
|
@PhyllisSmith Sorry, I see that my tables are not readable. Look at the Explanation.pdf for more readability |
|
|
Very clear explanation; thank you. If you make a video, we can upload it to the CinGG's Youtube channel. I would also like to put all the pdf in the manual, maybe in the appendix, as a separate section (title: example of advanced editing; or something like that). Phyllis, what do you think? |
|
|
@Andrea_Paz So glad to get an explanation of usage -- it gives gg more motivation to improve it when possible. I have not had a chance to read it thoroughly yet, but if Andrea goes ahead and converts it to latex, I will add it to the manual and check for english. |
|
|
@fary54 With Phyllis' consent, I wanted to ask your permission to put your pdf in the manual. If you want to make any changes, let me know. |
|
|
@Andrea_Paz No problem for me if it can enhance cinelerra GG, can interest new followers or improve the knowledge of the community. For my part, if that is clear enough for you, for the moment anyway, I have no changes to make to the pdf document. |
|
|
@fary54 I've been trying to rebuild your workflow in CinGG, but I've had a lot of problems. I enclose my version of your pdf. Since I still don't understand OpenEDL very well, I would like to ask you to correct it so I can put it in the manual. @Phyllis During my tests, I crashed when I tried to save ("s") while I was in OpenEDL. It is probably wrong to save in OpenEDL; however, I enclose the dump. |
|
|
@Andrea_Paz I send you the odt file modified by me as well as the explanatory videos. Indeed, it is excluded to save inside an openEDL but well outside of it (see video 2). This method works but is not viable in the long term (bug) video2.webm (2,725,792 bytes) video3.webm (474,942 bytes) |
|
|
@Andrea_Paz Being limited to video transfer (file size), the rest is here video4.webm (3,066,988 bytes) video5.webm (963,615 bytes) |
|
|
@Andrea_Paz And finally your odt file |
|
|
@fary54 Your tutorial has been included in the manual (Appendix D). Do you like it? If you want to make changes or corrections let us know. Thank you very much. |
|
|
A fix for the crash when saving inside of an OpenEDL layer has been checked into GIT. Both Andrea and Fary54 confirmed this. @Andrea_Paz -- if you do a rebuild, could you verify that it is now saving correctly? |
|
|
@Andrea_Paz Superb work. Rereading I find that there may be confusion about the understanding of the subjects. I therefore suggest that you make a small modification to the content of page 605 (see attached file). What do you think about ? @PhyllisSmith As far as I'm concerned, no problem to test but provided that I have the latest version A big thank you for having accepted my proposal and making it available to the community. |
|
|
@fary54 Thanks for the update that Andrea has passed along for the manual. We did not create a Debian 10 test build for the "save" option because I was hoping to get some kind of resolution for BT 0000430 (audio / OpenEDL update) and 0000440 (proxy format) first. But I have not yet convinced GG that these might need to be changed somehow -- especially 430. I do not think he will be able to change 440. |
|
|
@PhyllisSmith Ok, so right now I can't do the test as requested. For the resolution of the bugs in process, including 0000470, I am also awaiting the decision of GG. |
|
|
I don't really understand the management of multiple OpenEDL sessions (original bug); that's how it works for me: I create clip 1 and then I open it as openEDL. Clip 1 disappears from the Resources/Clips window but is open in the timeline. You can say that we are in layer 1. From this openEDL session, i.e. layer 1 I create a new clip that I call clip 2. This one appears alone in the Resources/Clips window. We can open a new openEDL session with this clip 2 and we will then be in layer 2. In the Resources/Clips window clip 2 disappears. Now I close layer 2; in the timeline layer 1 returns and in the Resources/Clips window clip 2 reappears. I close layer 1; the timeline returns to the original timeline and clip 1 and only clip 1 reappears in the Resources/Clips window. Even if we save, close and reopen the project, we maintain this hierarchy: Original Timeline --> Resources/clips: clip 1 Layer 1 Timeline --> Resources/clips: clip 2 Layer 2 Timeline --> Resources/clips: empty Is this behavior right? We'll never have clip 1 and clip 2 together? |
|
|
@Andrea_Paz I am not sure about the behavior, but will see if GG can explain it. |
|
|
@Andrea_Paz For me it is obvious since openEDL mission is to hierarchize places or subjects. This therefore brings several advantages. If I take your example: 1- Original Timeline -> Resources / clips: clip 1 --> Objective: Clearly separate the different subjects and places. 2- Layer 1 Timeline -> Resources / clips: clip 2 --> Objective: Retrieve from the Layer 1 timeline the clips to classify in another subcategory (ex: bonus, catalog, trash can, ...) Of course, as you say, we will never have clip 1 and clip 2 together because they are not on the same level of hierarchy. The goal here is not to mix all the clips on the same level (management impossible) but to keep a logic of hierarchy all along the edit until the final edit. If you need more information or details let me know. How did you do to save, close and reopen the project, while maintaining this hierarchy ? Because this is precisely the subject of this bug 0000470 (loss of hierarchy on reloading) that I can't do... Would the bug be resolved on a version of cinelerraGG other than mine (1: 5.1.20200630) ? |
|
|
Thanks for the clarification. Also for me to keep the hierarchy, so separate clips, it's right, but I wanted to be sure. I'm using the latest version, but it's from 20-07-05 (I think) that works well. Phyllis can you prepare a binary for fary54 (Debian 10)? If he also confirms the bug fix we can put a Resolved before the monthly release. |
|
|
@Andrea_Paz. You are right. i confirm the correct operation on the latest version of cinelerraGG. Superb work. Thank you very much to everyone involved in this file Andrea, can you update the manual page (see appendix). |
|
|
@Andrea_Paz. Sorry for the appendix, here is the actual version I planned to send you. manual.jpg (40,690 bytes) |
|
|
@fary54 OK, done! Thank you very much. |
|
|
@Andrea_Paz Is the manual correct now -- I have not carefully followed the notes? If so, I would like to close this issue. |
|
|
Here it is! Real-World.tex (7,566 bytes) |
|
|
@Andrea_Paz OK, thank you. |
|
|
Closing as the last thing to do to update the manual is now on the website. |
|
|
Will close later as it is better to mark Resolved first so people can see it and then close later (speculation). |
- Anonymous
