forwarding the message to the list. since this will go to a public list, I have to add that those scripts - qith quite some others - will soon be published properly in a github, cleaner and revised, so use these ones here for testing purposes (even though they are fully functional).
cheers, flavio ---------- Forwarded message ---------- From: flavio soares <qazav...@gmail.com> Date: 2013/1/24 Subject: Re: elphel and audio recording To: andrey <and...@elphel.com> Cc: Olga Filippova <o...@elphel.com>, Sebastian Pichelhofer < sebastian.pichelho...@gmail.com>, "Oleg K. Dzhimiev" <o...@elphel.com> Ok, guys, So, new tests. Now I've joined the scripts. This is what I'm doing: My Elphel mount point is /var/hdd. Inside it there is a folder called "scripts" in which all these go (the first 4 scripts). I set up the camera using the graphical interface; I run camogm_start.sh to make things ready for recording; Then I run "record_camogm.sh test3_" (with a prefix) to start recording. And stopRecording_camogm.sh to stop. The record script is calling audio recording. I am recording with the same settings as the film: 2304x960 JP46, only with a stronger compression of 90% instead of 94-96% I was using, for testing purposes. Once recorded, I scp the video and the audio files to my computer. Inside a folder, I put the video file and run the 01_Transcode_MOV........sh to convert them to DNGs and create the JPEG proxies for editing -- please use the file I'm sending because you can simply just go adding more and more .mov files to this root folder and it will run just fine, not adding up work as the one you might find on the "Editing Scripts" folder I had sent. Now my tests have been pointing that by calling arecord before camogm, the files pratically keep sync. I have been registering floating differences of -2 to +2 frames of accuracy, which is irrelevant. In any case, the files are *very easily* syncable: the delay happens at the beginning of the record, not at the end (the stop is being very effective for both), so to sync those two files, in case of any delay, we just use the end of the video and audio as reference in Blender or Cinelerra. I was trying to adjust this sync by using 'sleep', but in Elphel I believe the sleep version doesn't accept floating numbers, only integers. Could you please check if you register the same results there? ps.: now the audio script is bound to find supposedly any USB Audio device, not just my Shure X2u adapter, so it should work out of the box for any USB mic, I guess. ps2.: to run the 01_transcode_mov.......sh script, you must have 'parallel' compiled. ps3.: if you want, I can do a screencast of the second part here, so it gets easy to see what I'm doing to play and sync the videos. cheers! flavio
camogm_start.sh
Description: Bourne shell script
record_camogm.sh
Description: Bourne shell script
stopRecording_camogm.sh
Description: Bourne shell script
01_Transcode_MOVs_to_DNGs_and_JPG70_FLIP_OK_SHRINK_3_WITHCOUNTER.sh
Description: Bourne shell script
_______________________________________________ Support-list mailing list Support-list@support.elphel.com http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com
