I wonder if this old script still works (should make Cinelerra framelist for dir full of images) and if so - can it be included in 'contrib' folder of source tree (because hunting it via google or reimplementing each time one need to make 'list of images' input file for Cin... a bit counterproductive?) not sure about license - this is just support script, not linked to binary and not requred for main editing operations - but still useful IMO for working with image files generated by some other program
OK, so I went down the rabbit hole, just like Alice in Wonderland. The manual has the lines to a similar working script that is easy to access: https://cinelerra-gg.org/download/CinelerraGG_Manual/Image_Sequence_Creation... In executing mkframelist, which mostly works although /dev/stdin does not have permission by default (can just be easily changed to "cat -" instead), I still could not create a workable sequence list without doing additional editing. It has some advantages over the one in the manual, but I decided to slightly improve the one in the manual instead to make it clearer and more flexible. I checked this Manual change into GIT. So in summary, I do not think it necessary to include mkframelist (which is no licensing problem since the script says Creative Commons) because it is just as easy to reference the manual and the capability will not be lost. If you disagree, please just let me know. Thanks, Phyllis On Tue, Dec 7, 2021 at 9:05 AM Andrew Randrianasulu via Cin < [email protected]> wrote:
I wonder if this old script still works (should make Cinelerra framelist for dir full of images) and if so - can it be included in 'contrib' folder of source tree (because hunting it via google or reimplementing each time one need to make 'list of images' input file for Cin... a bit counterproductive?)
not sure about license - this is just support script, not linked to binary and not requred for main editing operations - but still useful IMO for working with image files generated by some other program -- Cin mailing list [email protected] https://lists.cinelerra-gg.org/mailman/listinfo/cin
Andrea, Latex problem in Auxiliaryprograms.tex that I tried several ways to fix. If you have any idea how to fix this, please do so. What happens is that "fi" in the image sequence example as shown here: if [ x$(dirname "$1") = x"$dir" ]; then f=./‘basename "$1"‘; else f="$1"; fi becomes a ligature, so that instead of 2 distinct characters of "f" and "i", it is squished together "fi". You do not necessarily notice this UNTIL you copy the text to a file so you can execute it and then it does not work correctly because it needs a real "fi" to complete the "if". So what I tried to fix it after looking it up on the internet was {}, "|, \/, \klern0pt -- none of which fixed it. On Fri, Dec 10, 2021 at 7:29 PM Phyllis Smith <[email protected]> wrote:
OK, so I went down the rabbit hole, just like Alice in Wonderland. The manual has the lines to a similar working script that is easy to access:
https://cinelerra-gg.org/download/CinelerraGG_Manual/Image_Sequence_Creation...
In executing mkframelist, which mostly works although /dev/stdin does not have permission by default (can just be easily changed to "cat -" instead), I still could not create a workable sequence list without doing additional editing. It has some advantages over the one in the manual, but I decided to slightly improve the one in the manual instead to make it clearer and more flexible. I checked this Manual change into GIT.
So in summary, I do not think it necessary to include mkframelist (which is no licensing problem since the script says Creative Commons) because it is just as easy to reference the manual and the capability will not be lost. If you disagree, please just let me know. Thanks, Phyllis
On Tue, Dec 7, 2021 at 9:05 AM Andrew Randrianasulu via Cin < [email protected]> wrote:
I wonder if this old script still works (should make Cinelerra framelist for dir full of images) and if so - can it be included in 'contrib' folder of source tree (because hunting it via google or reimplementing each time one need to make 'list of images' input file for Cin... a bit counterproductive?)
not sure about license - this is just support script, not linked to binary and not requred for main editing operations - but still useful IMO for working with image files generated by some other program -- Cin mailing list [email protected] https://lists.cinelerra-gg.org/mailman/listinfo/cin
On Fri, 10 Dec 2021, Phyllis Smith via Cin wrote:
What happens is that "fi" in the image sequence example as shown here: becomes a ligature, so that instead of 2 distinct characters of "f" and "i", it is squished together "fi". You do not necessarily notice this
Phyllis, try to make {f}{i} Perhaps this could make difference. And f{i} and {f}i should work as well. _______________________________________________________________________________ Georgy Salnikov NMR Group Novosibirsk Institute of Organic Chemistry Lavrentjeva, 9, 630090 Novosibirsk, Russia Phone +7-383-3307864 Email [email protected] _______________________________________________________________________________
On Saturday, December 11, 2021, Phyllis Smith via Cin < [email protected]> wrote:
OK, so I went down the rabbit hole, just like Alice in Wonderland. The manual has the lines to a similar working script that is easy to access: https://cinelerra-gg.org/download/CinelerraGG_Manual/ Image_Sequence_Creation.html
In executing mkframelist, which mostly works although /dev/stdin does not have permission by default (can just be easily changed to "cat -" instead), I still could not create a workable sequence list without doing additional editing. It has some advantages over the one in the manual, but I decided to slightly improve the one in the manual instead to make it clearer and more flexible. I checked this Manual change into GIT.
https://git.cinelerra-gg.org/git/?p=goodguy/cin-manual-latex.git;a=blobdiff;... may be also rename jpeglist to imagelist? And may be add user-selectable framerate...
So in summary, I do not think it necessary to include mkframelist (which is no licensing problem since the script says Creative Commons) because it is just as easy to reference the manual and the capability will not be lost. If you disagree, please just let me know. Thanks, Phyllis
On Tue, Dec 7, 2021 at 9:05 AM Andrew Randrianasulu via Cin < [email protected]> wrote:
I wonder if this old script still works (should make Cinelerra framelist for dir full of images) and if so - can it be included in 'contrib' folder of source tree (because hunting it via google or reimplementing each time one need to make 'list of images' input file for Cin... a bit counterproductive?)
not sure about license - this is just support script, not linked to binary and not requred for main editing operations - but still useful IMO for working with image files generated by some other program -- Cin mailing list [email protected] https://lists.cinelerra-gg.org/mailman/listinfo/cin
Andrea, Latex problem in Auxiliaryprograms.tex that I tried several ways to fix. If you have any idea how to fix this, please do so.
Sorry Phyllis, I don't think I understand the problem. I tried manually copying the script into a file (jpeglist.sh) and it works without error. I tried copying and pasting from the html page into a file and that works fine too. However, copying and pasting from the pdf brings an error on line 20 and doesn't work. Finally doing the copy and paste from the LaTeX code works fine. In short, only copying from the pdf produced by latex leads to the error. Doing a diff between the jpeglist.sh file produced from pdf and html does not show any difference (only indentation differences). The "listing" package should only show the code that is written in it without alterations. Let's hear Andrey if he can explain this behavior. I attach the 3 jpeglist.sh that I used. NOTE1: to the command "jpeglist.sh outfile infiles*.jpg" I have to prefix ./ otherwise it doesn't work (and I have to make it executable). NOTE2: The pdf manual splits the "code" block into 2 pages. Could this be the cause of the problem? NOTE3: even the "outfile" that came without error, do not work once imported into CinGG.
On Monday, December 13, 2021, Andrea paz <[email protected]> wrote:
Andrea, Latex problem in Auxiliaryprograms.tex that I tried several ways to fix. If you have any idea how to fix this, please do so.
Sorry Phyllis, I don't think I understand the problem. I tried manually copying the script into a file (jpeglist.sh) and it works without error. I tried copying and pasting from the html page into a file and that works fine too. However, copying and pasting from the pdf brings an error on line 20 and doesn't work. Finally doing the copy and paste from the LaTeX code works fine. In short, only copying from the pdf produced by latex leads to the error. Doing a diff between the jpeglist.sh file produced from pdf and html does not show any difference (only indentation differences). The "listing" package should only show the code that is written in it without alterations. Let's hear Andrey if he can explain this behavior. I attach the 3 jpeglist.sh that I used. NOTE1: to the command "jpeglist.sh outfile infiles*.jpg" I have to prefix ./ otherwise it doesn't work (and I have to make it executable). NOTE2: The pdf manual splits the "code" block into 2 pages. Could this be the cause of the problem? NOTE3: even the "outfile" that came without error, do not work once imported into CinGG.
well, last note a bit concerned - because user migh hope provided script work out of the box? did you looked into created list - was it using relative image pathes? or absolute? did changing them one way or another resulted in working list?
On Mon, 13 Dec 2021, Andrea paz via Cin wrote:
file and that works fine too. However, copying and pasting from the pdf brings an error on line 20 and doesn't work. Finally doing the
Copypasting from pdf is not always a good idea. Usually, when a pdf document contains one page (or less) of plain text, then copypasting works. But if the text is partly fed to the next page, something can fall in between (for example, page numbers) if there are some non-flat constructions (for example, tables, or floating objects with captions), the ordering of the parts of the text can be intermixed if some font switching takes place, then the driver which produces that pdf can also change (perhaps optimize in some way) the text layout And then, the copypaster can actually follow the ordering of the letters as they appear in the (complex structured) pdf file, which can easily differ from the order in which they appear on the screen. _______________________________________________________________________________ Georgy Salnikov NMR Group Novosibirsk Institute of Organic Chemistry Lavrentjeva, 9, 630090 Novosibirsk, Russia Phone +7-383-3307864 Email [email protected] _______________________________________________________________________________
@Andrew Sorry, I didn't put the list file. Changing the path to absolute, loading in CinGG works fine. Maybe you should have the script produce the absolute path? Note: I used images of resolution 1280*720 and others 720*1280 (vertical!). The script took only the size of the first image and then set it to all the others. @Georgy I think Phyllis' problem didn't show up in my tests: if I understand correctly, she used copy/paste from html.
On Monday, December 13, 2021, Andrea paz via Cin <[email protected]> wrote:
@Andrew Sorry, I didn't put the list file. Changing the path to absolute, loading in CinGG works fine. Maybe you should have the script produce the absolute path?
According to documentation it should do this if your output (listfile) not in the same diectory as input files. https://cinelerra-gg.org/download/CinelerraGG_Manual/Timelapse_Sequence_Imag... === If <path> is the same on both outfile and infiles, then file.jpg is created in the same directory as infiles, the directory contains the entire asset, and the file list uses relative paths; otherwise the file list contains absolute paths. ==== does it behave according to documentation?
Note: I used images of resolution 1280*720 and others 720*1280 (vertical!). The script took only the size of the first image and then set it to all the others.
well, this is good to know. So no crash even if image dimensions are different initially? @Georgy
I think Phyllis' problem didn't show up in my tests: if I understand correctly, she used copy/paste from html.
Feedback; and now I am moving on to other issues: *Georgy/Andrea *- you are right the ligature only shows up if I copy/paste from the pdf file which I should not have done. *Andrew* - changed in the Auxiliary program section to imagelist instead of jpeglist and noted that user can edit the file to change the frame rate or other values if needed. Tested Load and Save section description and made some changes for clarification there. It all seems to work as now described. *Problem* - somewhat confused when Andrea said he loaded images where some were 1280x720 and others 720x1280 and I even loaded different but similar image width/height images with no crash (65x65, 69x68, 70x58, 58x70, 63x63, 83x64). So there must be something else going on that is specifically causing the crash but the Freelancer is now going to look into that. Checked into the Manual GIT, minor fixes in these 2 sections.
====
does it behave according to documentation?
Note: I used images of resolution 1280*720 and others 720*1280 (vertical!). The script took only the size of the first image and then set it to all the others.
well, this is good to know. So no crash even if image dimensions are different initially?
@Georgy
I think Phyllis' problem didn't show up in my tests: if I understand correctly, she used copy/paste from html.
-- Cin mailing list [email protected] https://lists.cinelerra-gg.org/mailman/listinfo/cin
On Thursday, December 16, 2021, Phyllis Smith via Cin < [email protected]> wrote:
Feedback; and now I am moving on to other issues:
*Georgy/Andrea *- you are right the ligature only shows up if I copy/paste from the pdf file which I should not have done. *Andrew* - changed in the Auxiliary program section to imagelist instead of jpeglist and noted that user can edit the file to change the frame rate or other values if needed. Tested Load and Save section description and made some changes for clarification there. It all seems to work as now described. *Problem* - somewhat confused when Andrea said he loaded images where some were 1280x720 and others 720x1280 and I even loaded different but similar image width/height images with no crash (65x65, 69x68, 70x58, 58x70, 63x63, 83x64). So there must be something else going on that is specifically causing the crash but the Freelancer is now going to look into that.
Checked into the Manual GIT, minor fixes in these 2 sections.
THANKS! /me still builds obsoleted patched software and hunting for small enough m2ts streams with IGS substream/pid (for testing demuxer/muxer written 9 years ago for avconv/libav)
====
does it behave according to documentation?
Note: I used images of resolution 1280*720 and others 720*1280 (vertical!). The script took only the size of the first image and then set it to all the others.
well, this is good to know. So no crash even if image dimensions are different initially?
@Georgy
I think Phyllis' problem didn't show up in my tests: if I understand correctly, she used copy/paste from html.
-- Cin mailing list [email protected] https://lists.cinelerra-gg.org/mailman/listinfo/cin
participants (4)
-
Andrea paz -
Andrew Randrianasulu -
Georgy Salnikov -
Phyllis Smith