Great conversion script !!
Almost got there. At virtualdub the output screen is sometimes black with the msg
Frame could not be processed 1st pass.
Did you have the problem too? What might be wrong?
thanks
otto_oz
Great conversion script !!
Almost got there. At virtualdub the output screen is sometimes black with the msg
Frame could not be processed 1st pass.
Did you have the problem too? What might be wrong?
thanks
otto_oz
Thanks ! There is an OLD thread on this forum, where a couple of openSUSE users (ken_yap and microchip if my memory serves me right) helped me create that one line ‘script’ (command).
I do have a problem that when the command is running virtual dub under wine, I can’t play with the PC in another desktop window. As soon as I do that, VirtualDub in wine gives a bad output. The wine/virtualdub combination insists on undivided attention. Could it be you were trying something in a different window in the PC at a different time ?
Also, when loading the video into VirtualDub, it may appear black but as soon as one moves the timeline bar in VirtualDub one can see the video.
I don’t know if both those cases address your question.
Hi oldcpu
The problem has something to do with the deshaker and/or settings. I need to run a few more tests to give specifics.
There is something you may like to know. Via handbrake I converted my old 4:3 videos to widescreen via cropping a proportion from the top and bottom. The trick was to find out the best settings.
http://thumbnails51.imagebam.com/14029/ea1f9a140285852.jpg](http://www.imagebam.com/image/ea1f9a140285852)
Based on my tests, cropping 62 from top and 80 from bottom was optimal. I would say it worked in 95% plus of all clips. The rest I did not worry about, the loss was only minor. The results where stunning. Considering the actual information loss due to cropping, the wide-screen view suggested (subjectively) more and better information.
cheers
otto_oz
EXACTLY what settings have you applied? I do not have that with the example settings I provided to you. I also apply other settings, but the ‘other’ settings really depend on the video characteristics. Different videos require different settings and this is noted on the deshaker web site. I don’t apply settings ‘bllindly’. I pick and choose my settings carefully BASED on the video and based on how I want to stabilize it, and based on limitations of the deshaker plugin.
I have not seen ANY free software that runs under GNU/Linux natively, nor under GNU/Linux under wine, come anywhere close to the capabilities of VirtualDub with deshaker for the majority of the videos that I have taken/processed. As noted, I don’t get the error you noted when I judiciously apply settings.
One can’t just apply ‘any’ setting in VirtualDub deshaker, but I assume the example settings I posted do not cause that error. Am I not correct ? Are you getting THAT error with the settings I noted on nominal videos (ie not ones that are taken in the dark … nor ones that pan wildly … ) ?
I re-checked…and I am using your settings. That does not mean I missed something.
Before I post more details I need to run a few more tests to be sure.
cheers
I suppose its possible the video from your camcorder is significantly different from mine. Does it work at all for any videos ? or only not work for some ?
Its possible the Source Pixel Aspect and Video Type are different for you camcorder.
I suppose its possible the video from your camcorder is significantly different from mine. Does it work at all for any videos ? or only not work for some ?
Its possible the Source Pixel Aspect and Video Type are different for you camcorder.
That may well be true. The camcorder records in h.264 format and has a CMOS sensor.
That suggest different pixel aspect.
I am trying currently different settings. I let you know how I went.
cheers
Hi oldcpu
I got it ! needs a bit more fine tuning…anyway
I altered your avi conversion script to full HD
for i in *.MP4; do ffmpeg -y -i "$i" -f avi -vcodec mpeg4 -b 8000000 -acodec ac3 -ab 128000 -s 1920x1080 YYYY_MM_DD_${i/%MP4/avi}; done
Since I have an HD camcorder, according to the de-saker web-site, I have to use square pixels.
ImageBam - Fast, Free Image Hosting and Photo Sharing
According to the de-shaker website one has to run “analyse” to make sure that no frame will be missed.
Then goto video=>filters, select deshaker, click 2nd pass and use your settings plus I used for “edge compensation”=> "adaptive-fixed.
Then save avi. the output was jerky, some fine-tuning needs to be done. However, I imported the clip into kdenlive, rendered it and the output clip was smooth/ viewable.
cheers
otto_oz
Excellent !!
Thanks for sharing your solution.
hi
just wonder if you managed setting slow motion on videos in kdenlive?
Yes, its pretty easy.
Drag a clip to the timeline. Select the clip in the timeline. Add effect > Motion > Speed … and then edit the speed from its nominal 100 to some number less than 100, … say to 10 for something really slow. The clip of course (for 10) will then be 10x longer in duration.
hi oldcpu,
thanks for your answear, currently I am nowhere near my opensuse but will check it later for sure.
best of luck
Hi oldcpu
Great hint. thanks !
Tested it. 10% is a bit slow and the result is not smooth. I tried 50% and 75%. 75% produced good results.
cheers
otto_oz
10% was an example - pick and choose what you want , whether it very slow for study, or smooth for a different impression.
Now reference kdenlive features, you really should be asking for advice here: Forums | Kdenlive
Hi
When loading low resolution clips eg. old home videos, I can move the time-line cursor and get a smooth picture sequence. Loading HD-clips (1920x1080) the response is sluggish to say the least, which makes sense since there is much more info to be processed.
What I am looking for is a setting give kdenlive more resources. re-nice the process does not make any difference.
thanks and cheers otto_oz
Solved: Enable proxy in project settings
cheers