På Thu, 15 Nov 2012 05:07:42 +0100, skrev Basil Chupin
<[email protected]>:
On 15/11/12 02:59, Herman Robak wrote:
As far as I have understood, Audio Offset is for _calibration_,
and nudge is for _correction_.
Thanks for the response, Herman.
Yep. As I mentioned to Einar, I wonder why it afflicts Cinelerra when
all my tv programs, videos are all in perfect sync - or, at least, so
close to being perfect that any difference is not noticeable.
I don't know. Maybe Cinelerra has bugs that the other don't have.
Calibration is per computer/installation, and correction is
per project/format/track/clip.
Now, if only there was a way to reliably calibrate audio/video
synchronisation on a computer! I'd like a cheap gizmo, let's
call it a Synchrophone, that has a simple microphone and a fast
rate low-res sensor (one monochrome pixel might do).
I like this idea! :-) Like a clapperboard but only in the form of
software.
You sound like someone who knows something so can you explain this,
please :-) .
Uh-oh...
In the Cinelerra manual, page 23, instructions are given on how to
calculate the audio offset. The part which really has me confused is
this:
"Play the timeline from 0 and watch to see if the gradient effect starts
exactly when the audio starts."
It's a good idea to not use video and audio _files_, but rather use
generated sound and video, eliminating codecs and file I/O.
But it's still hard to align audio with video like that. The offset
is supposed to be constant, not proportional with playback speed, so
you can't check it at 1/4 speed.
When exactly does the audio start - is it when the audio level meter
appears on the right-hand side of the Compositor window or when the
actual *sound* is heard (to match the appearance of the synthesiser)?
When the sound is heard, obviously.
Synchronising sound with another sound is easier, and can be FAR more
accurate. Do you have a monitor that hums loudly when the screen is
bright? Or some way to bleed analogue video (VGA could do, HDMI or
DVI less likely) signals into one of the audio output channels?
Wild idea: Covering everything but Cinelerra's audio meters with
black windows, so quiet audio results in little or no screen buzz.
a) Does the buzz sound like a pre-echo? Audio is too late!
b) Does the buzz come like an echo? Audio is too early!
c) Do the audio and the buzz seem in phase? Bingo!
This could be used BOTH for the one-off Audio Offset calibration,
and for the asset-by-asset nudge.
(Hm, if this is workable, it might be worthy a test suite...)
--
Herman Robak
_______________________________________________
Cinelerra mailing list
[email protected]
https://init.linpro.no/mailman/skolelinux.no/listinfo/cinelerra
