Interesting thoughts, Richard. While I do think that Cinelerra could use a deal of refactoring and general cleaning up, I very much agree that a completely new base is not necessary. A lot of work has been put into what's already here, and it would be a shame to lose it.
Your three suggestions --particularly 1 and 2- are absolutely precise; a quality deinterlacer and a decent color grading solution are the primary lackings in open source video software. I've particularly been missing SyntheticAperature since making my switch to Linux. I've managed to rig a basic 3-way color corrector in Blender's compositor, though the workflow is absolutely horrific. Anyway, I work professionally in film and video and have a fair amount of experience with various color correction tools --SyntheticAperature, NUKE, IFX Piranha- so if anyone's up for the task of implementing an open source color correction app/plugin, I would be thrilled to provide my thoughts and experience on how such tools work. Several months ago actually, I wrote a very general spec for a 3way cc: http://fouressence.net/3way/DescriptiveLanguage.pdf If interest starts to pick up on this topic, I'll be thrilled to delve more deeply into the workings of color correction tools. Thanks Richard for bringing these points up. Definitely worth some thought. -=Derek > Hi, > > I just read through these plans for an API to the cinelerra > functionality, and I'd like to share my views a little. > > First of all, I do believe that everyone should hack whatever he wants > to, as long as he does it for fun. ;-) > > Ok, now what I don't like. In my opinion there are already plenty of > options for doing scripted or programmatical video editing. I remember > that I did that a long time ago using shell and imagemagick. And even > today, I find it fairly easy to hack up simple video tools using > libquicktime and gavl. I agree though that this might require a little > creativity and programming experience, but I guess you'd need that too > for any vision of libcinelerra. So a duplication of efforts here IMHO. > On the other hand I do like to imitate and recreate other peoples > software too, so if you feel you can make a difference, go ahead and > do it. :-) > > On the other hand, I do have a couple of ideas that I feel would not > only make a difference, but could be fun too. > > This is just I list of random stuff that I think is missing in the > free and open source video editing space, and that would be useful to > a number of projects, including, but not limited to cinelerra. > > 1) > Image-based Autodetection for Interlaced Video and 3:2-Pull-Down Sources. > required Skills: > C or C++ > Image Processing and Analysis Algorithms > Tasks: > create a simple set of libs and commandline tools for the purpose named > above, > in a fashion that they can be reused by a number of diverse open-source > video-editing and conversion applications. > Why: > While converting between different formats is use case that is widely > implemented in the open source space, there are limited tools for > "professional" formats, that is formats used by people how produce > media contrary to consuming media. > Interlaced And Telecined Video is such a category that is mostly > neglected by common tools, and even if implemented there are plenty of > occasions where users could screw up, so strong autodetection features > would certainly be useful. especially because not all devices and > formats apply flags media files to specify exactly what kind of > source is used. (Example: Camcorders exporting 24p as Telecine, or > distingishing between 25i and 25p DV-Files) > > 2) > Color Corrector Reverse Engineering: > required Skills: > Solid Understanding of Color Correction and LUTs > Access to various Color Grading Apps > Knows how to use test images and histograms to conclude from > UI-Interaction to > color computation. > No Programming required. > Tasks: > Use, analyse and document commen User-Interfaces and commerzial > tools, with the aim to recreate the user interface, to make it easier > for users to switch to other tools and platforms. Remember: The > algorithms for color transformations are mostly trivial, it's the more > advanced user interface that provides access to these features. > Imagine if you would combine a selection of different cinellera > plugins for brightness, gamma, colors, saturization into a single > plugin with an interface using colorwheels and all the stuff the > windows-kids like. > > 3) > Interoperability Engineering: > required Skills: > Experience with SDKs and Plugin-Development on Popular Commercial NLE > Software > C/C++ > Tasks: > Port Plugins and Technology developed in the Tasks above as Open-Source > and Free Software to other Platforms, Applications and > Plugin-Specifications, > to promote Interoperability, Free Software, and to help people switch to > Free > Software by giving them a preview on the available Technology. And to > promote > the idea and friendliness of Free Software in Communities around non-free > Applications. > Example: Exactly reimplement a color-corrector plugin for a > commercial/non-free video editing app in open-source, and then port > that open-source plugin to the plugin-interface of the non-free App, > and then step by step replace a non-free up with FLOSS plugins, until > it's only a hollow and undermined platform, that can be replaced by a > free and open source plugin host. Sneaky, subversive and fun approach > to FLOSS Market Domination IMHO. ;-) > > These were just a couple of my ideas, I think I do have some more of them. > ;-) > > Have fun, > -Richard > > > > > > > -- > Are you teaching the What and the How but without the Why and the When? > > _______________________________________________ > Cinelerra mailing list > [email protected] > https://init.linpro.no/mailman/skolelinux.no/listinfo/cinelerra > _______________________________________________ Cinelerra mailing list [email protected] https://init.linpro.no/mailman/skolelinux.no/listinfo/cinelerra
