This IS the Avid L2: RE: [Avid-L2] [FR] Facial Recognition for CC

Thursday, August 21, 2014

RE: [Avid-L2] [FR] Facial Recognition for CC

Sounds like something a good DIT could solve on the front end.

Dave Spraker

Principal, spraker.tv

Shared Storage Solutions | Consulting | Sports Audio

www.spraker.tv

Northwest Territory Manager, Western Rep Associates

Broadcast and AV Manufacturer’s Representative

www.westernrep.com

dave@spraker.tv
(503) 897-0250

From: Avid-L2@yahoogroups.com [mailto:Avid-L2@yahoogroups.com]
Sent: Thursday, August 21, 2014 3:42 PM
To: Avid-L2@yahoogroups.com
Subject: RE: [Avid-L2] [FR] Facial Recognition for CC

Hi Steve,

In my experience, when a clip has no tape name and when the master clip name is AA0002 (along with shots from 4 others interviews), those relationships cease to have any meaningful value at all. I find myself in source segment mode 99% of the time. There's got to be a better way (also, there used to be a better way).

Point taken about facial recognition but I think that's a problem that could be overcome with a bit of ingenuity.

Also, I'm sure the folks over in the forensic video department at Avid (if they still exist) would love that feature too

Warmly on a warm day,

Seth

To: Avid-L2@yahoogroups.com
From: Avid-L2@yahoogroups.com
Date: Thu, 21 Aug 2014 17:10:06 -0500
Subject: Re: [Avid-L2] [FR] Facial Recognition for CC

This basically exists in Symphony already. Not the facial recognition, but you can do it based on clip name, so anything with the same master clip name works this way. You can also base it on "tape name" which is a workaround you can use, so that any interview that you force to have the same tape name can have the same color correction applied across all clips. I've never done that. Essentially having the Symphony's ability to grade all the clips from the same tape was a great help, but now, with file-based media, this is essentially not as helpful, but it DOES exist already.

The facial recognition thing would not work even if it DID work. Think about it. You shoot an interview with interviewee #1 outside, then shoot another interview with the same person three months later in the studio. Do you want the color corrections to match? I know you kind of addressed that, but what makes far more sense that facial recognition is the way Avid Symphony already does these - based on clip name, master clip or tape name.

Steve

On Aug 21, 2014, at 3:20 PM, Seth Isaac Buncher seth_buncher@hotmail.com [Avid-L2] <Avid-L2@yahoogroups.com> wrote:

Hi y'all,

In the olden days, I used to color shows like lightning using tape-based color corrections. You know, you'd get 6 interviews, 2 cameras each and if the production people named their tapes well, half the show would be graded in a matter of minutes.

But of course now there's just Clip 1 of each interview, Clip 2 of each camera, and so on. So Source Tape, Source Clip Name, etc are completely useless. Now it's just saving a grade into a bin and applying it the next time you see it, and hoping you grabbed the right one.

I've thought about making a custom column in a bin and then color-coding segments based on that but it just seems like more work than it's worth. I'd be curious to know if y'all have other strategies for working with this problem.

Anyway, if Avid wanted to up the ante a bit on coloring, I would love to see them try to make a relationship between clips using facial recognition. Then we could apply a correction and have it apply to all other segments in the timeline that have a similar image. Ideally, there would be some kind of verify Yes or No on each segment. Heck, I'd bet the offliners would love to see color coding segments in the timeline based on the image.

My $.02.

Warmly on a warm day,
Seth