[LAU] LAC: A musician's workflow slides?

[Robin Gareus]

On 05/15/2011 12:31 PM, andy baxter wrote:
> On 15/05/11 11:18, Jeremy Jongepier wrote:
>>
>> Hello Heikki,
>>
>> The tapes have been captured and I've started editing. So far kdenlive
>> doesn't like a system without PulseAudio or the avi files sitting on a
>> NTFS drive apparently so I didn't get very far yet. I'll continue
>> tomorrow.
> 
> I've just been editing the raw file of my talk from the website in
> kdenlive (just to clip it, adjust the volume and add an introductory
> slide). 

Hi Andy,

Clip/cut can be done easily using oggz-chop. It's a loss-less way to
edit ogg files.

However pre-pending an intro slide is more complicated than it ought to
be. OGG files with an ogg-skeleton track can simply be concatenated but
most video-players won't recognize that and some [ffmpeg] transcoding
magic is needed to make it compatible :(

>It had trouble using alsa until I worked out that you need to
> install 'libsdl1.2debian-alsa'. (in ubuntu, probably the same in debian).
> 
> Another gotcha that took me a while to work out is that the version of
> ffmpeg on ubuntu lucid won't read ogg theora videos properly, so you
> need to either use a newer distro, or compile your own ffmpeg from the
> website. (if you compile your own you need to add --enable-libtheora
> --enable-libvorbis to the configure flags to be able to export to ogg
> theora).
>
> One question I have is whether you have captured the screenshot and
> camera streams separately on disk to edit later? The raw video of my
> talk spends a lot of time faded to the screenshot.

Alas, no. We only had one DV scan converter and captured the output
after the video-mixer.
We do have raw DV files (better quality than the encoded OGG/Theora) for
each presentation. But they're in the order of 8-10 GB per presentation.
If you like to have those, get in touch with John Lato
via lac at linuxaudio.org.

best,
robin