[GRLUG] Status of things

[Michael Mol]

On Mon, Jul 13, 2009 at 3:28 PM, Ben Rousch<brousch at gmail.com> wrote:
> On Mon, Jul 13, 2009 at 3:02 PM, Michael Mol <mikemol at gmail.com> wrote:
>>
>> The only bit I think can be automated is the actual recording and
>> possibly scripting up a few intro frames based on form inputs. (Which
>> I will probably have the skill to do in PHP, by the end of this
>> particular work day.)
>
> It really doesn't take much time to make simple text-only title and
> attribute screens.

Which is why it would be relatively easy to automate. :)

> All I did with the video was crop it down to the
> presentation. The content after the presentation is audio-only, so I may try
> to extract that and upload it. Most of the time is spent waiting for the
> videos to render. Using Kino, it has to render to DV before editing, then
> you render to DV/XVid/whatever after editing. Even on my six year old
> computer, it was not disruptive to do other things while the video rendered.

If I were to do it my process would have been:

1 Open the file in AVIDemux
2 Crop off the leadup and tail
3 Save to a different format.
4 Notice that my video encode choices left irritating compression
artifacts or left the file too large
5. Jump back to step 3 until I eventually get past step 4 satisfied.

It's the encode selection and configuration that's the trickiest part.
 And AVIDemux on the distros I've used up to this point tend to be
unable to find and use audio codecs properly.  Getting the encode
selection and configuration right is what took me so darn long when
trying to encode the first meeting.  It's not something I enjoy
fiddling with.

As for the audio-only segments, those would probably be suitable as
short podcasts or some such.

>
>>
>> Actually obtaining the licenses from the involved parties is
>> necessarily a manual thing, and there's no reliable way to pull a
>> human being out of the editing side of things; We're talking about at
>> least three audio input channels, and which channel is most
>> interesting is a matter of subjectivity. (Even if you provide each
>> audio source as a separate stream, you still need a master mix track
>> since the vast majority of media players won't let you play mix the
>> audio streams during playback.)
>
> I have not dealt with multiple audio streams.

Were I to do it, I'd load them into Audacity and mix them as sounded
appropriate.  It might be fun to write an adaptive mixer, though,
letting you give priority weights to the different sources and
auto-mixes based on those weights.  But that's me talking about theory
and risking over-extending myself again. :-|

>
>>
>> But good source material is the first step. :)
>>
>> I might suggest using VNC server on the machine in question to pull
>> high-quality video source directly from any computerized presentation
>> if there's no human being in front of the camera, but that could be a
>> tricky thing to arrange, and would require a wired connection for
>> latency's sake.
>
> I think there is screen-casting software created just for this purpose.

Slick!


-- 
:wq