Thanks for reading, I’m glad you find my suggestions interesting. Since I am not a developer I am not quite sure if I am the right person to talk about these things at a technical conference… But I will post all my ideas and suggestions to this blog and try to talk to all those who are actually developing these tools. I think just the fact that the Lumiera developers found this blog entry hours after I posted it is a good sign! Some tools might not be in place yet, but I have no doubt that open-source can not only catch up with FCP but in the long run also can come up with a better product! If it works for 3D (Blender), why should it not work for non-linear video editing…?!

More info on the Libre Graphics Meeting 2008 here:
http://www.libregraphicsmeeting.org/2008/index.php?lang=en

This recent blog post by Dave Neary outlines some of the significant advances that have been made in color management on Linux as a result of the synergies at previous Libre Graphics Meetings.

http://blogs.gnome.org/bolsh/2008/04/08/libre-graphics-meeting-stories-colour-management/

The workflow integration between applications you have outlined would be a massive step forward for Linux NLE.

If there is anything you can do to promote these advances, Linux and FOSS users throughout the world would be most grateful.

that sounds very interesting…

Here a couple of thoughts to Cinelerra and how I would try to design a new version:

We had a quite detailed presentation of Cinelerra at the TOSMI training and my main criticism is that it did not have non-overlapping windows and felt very heavy and cluttered. I’m not a coder, but the way FCP seems to be done is also only by using some kind of fake non-overlapping windows system. It does not have this incredible genius fast way that the context sensitive Blender window system has: you just move the mouse over a window and without even clicking you can apply a one key shortcut (e.g. not ctrl-a but just “a” for “select all”). This simple interface feature is huge! And there are many things that FCP does not have that would only require to have a closer look at how people actually work with their tools and what more efficient ways there might be for doing things…

So this is something where an open-source alternative could possibly even *surpass* FCP:

• for a start not by trying to clone all the many little features but by making the interface better! You might find that *you can get rid of some of the tools that you thought you needed*…! This is what is wrong with 3ds Max: the programme is so terribly bloated and the interface so inconsistent and individual parts are not talking to each other so that they had to invent many, many sets of new tools and buttons where I have a single one in Blender…! Simplicity really is the secret.

(And since most people outside of the open-source world would not expect a slick and cool interface from an open-source app this would be a huge bonus when it comes to promoting it.)

Apple outdid Avid when it came to simplicity and ease of use, it was not just the huge prize difference. The problem was that Apple kept adding features to FCP but the interface now feels as outdated as the Avid Media Composer felt to me when I had a closer look at it back around 2000. And this is because the FCP interface never really adapted to the new features and starts feeling to be as bloated as iTunes feels today…

One reason why I started with Blender was because I got bored by what you could do with FCP: at one point I wanted to apply a set of filters to a music video – by exporting the video as a series of frames for use in the GIMP… In the end I had to give up because the only way of doing it would have been to do a lot of work by hand, frame by frame because:

- there is no way to use the GIMP’s filters inside of FCP (and vice versa)
- no way to combine the app’s features in a creative and modular way (maybe done like Blender’s Compositing Nodes system but only across a variety of compatible open-source apps that act like a plug-in networks for each other… Apple now has made first steps in that direction e.g. by allowing you to easily move documents between FCP and Motion – of course this is not a feature Apple tries to offers to other apps like the GIMP…)
- the GIMP itself has this rather hard to understand scripting language, at least if you’ve never seen anything like it it does look intimidating… But the main problem at the time was that it seemed impossible to simply have a Script-Fu that automatically opens images from one folder and later saves the filter applied images to another folder

• so some kind of inter-app functionality, a plug-in or even better a nodes system would be a great *long term* goal… And the step after that would be to allow to create real app mash-ups: take a tool from the GIMP and use it in the NLE – to create the perfect tool for the job…!

I know, that’s asking for *a lot*, but since I’ve seen Blender and started learning it 22 months ago I really feel like I have seen the future of how creativity tools will be designed… An NLE is of course not a 3D app, but I don’t see much innovation in FCP, it now feels more like a dinosaur from the last century (which it really is).

So thanks for your comment and I’ll drop by on that IRC channel tomorrow or so!

If you need more informations on Cinelerra, Lumiera or other open video tools you can join us on irc://freenode.org/cinelerra

Regards.

Roland (wildhostile)