Flicker on top line after render with slow motion

2G wrote on 12/15/2008, 8:43 PM
Quite often, but not always, after rendering a project to AVI or M2T, there is a very pronounced flicker on the top line of the output. I've seen it for quite some time, both in standard def and high def, and over several versions of Vegas, but have just endured it. It only happens on slo-mo, but not on every slo-mo segment I have. On the sections of the video where it occurs, it's pretty obvious.

I'm suspecting it may have something to do with matching slo-mo frames to the actual frames per second of the output. If that is the case, is there some sort of formula to calculate the precise slo-mo percentage increments so that the flicker will not occur?

If I'm completely off base on my assumption, have others encountered this same problem with flicker? How do I work around it?

Comments

ScorpioProd wrote on 12/16/2008, 9:56 AM
You're probably just seeing interlace flicker slowed down, which would make it more noticable. Since you mention you're actually "seeing" the top line, I figure you are seeing the video letterboxed, right?

Try turning on the "reduce interlace flicker" switch for the clip involved. Or you could add a very slight gaussian blur on the vertical, the smallest possible, 0.001.
2G wrote on 12/16/2008, 10:05 AM
Thanks for the info. Not sure what you mean by 'seeing the video letterboxed'. I'm outputting HDV 16:9. But I used to see this when I was running standard 4:3 DV as well. I guess you could say it is slightly 'letterboxing' in that the flicker is a darker line at the very top all the way across.

I'll try the 'reduce interlace flicker' and see if that helps. I'll also try the gaussian blur suggestion.

Thanks again.

2G wrote on 12/16/2008, 10:15 AM
Two more questions...

First one is primarily curiosity... not every slo-mo clip has the flicker. And on some clips it 'appears' to be more prominent than on others. Just trying to get a better grasp on what is happening... but is interlace flicker not always constant (given the same camera, settings, etc.?) Or could it have something to do with the percentage of slo-mo? Again, not a big deal if the suggestions fix it. Just trying to figure it out.

Second question, more important.... what is the down side of turning on "Reduce Interlace Flicker". I figure it there wasn't a down side, the default would be 'on'. My initial tendency would be to turn it always on anything slo-mo. But I don't want to needlessly lose some quality somewhere else just by being over-zealous with the flicker reduction option.

Thanks again.
johnmeyer wrote on 12/16/2008, 12:03 PM
I think the flicker (or lack of flicker) is going to depend primarily on the slow-down ratio. If you choose a common integer divisor (like slowing it to 1/2 or 1/4), I would expect you would NOT see any flicker).
ScorpioProd wrote on 12/16/2008, 7:13 PM
I mentioned letterboxing since I wasn't sure how you were sure you were seeing the "top line" flickering unless it was in an easier location to see. When I've seen it happen playing slow motion on a 4:3 monitor letterboxed, it can be very easy to see since the "top line" is right there.

The downside of "Reduce Interlace Flicker" is that it has to do a slight softening to work its magic. But it is necessary at times, I use it on slow motion, and it is actually turned on by default in Vegas for all CG text generation, and I leave it on there.
LReavis wrote on 12/17/2008, 12:09 PM
you might want to skip Vegas for Slo-Mo work; on another thread, I read that MVtools works with AVIsynth 2.5x and that it interpolates each frame to do 'tweening, much like animation programs make animations by creating new frames based on a few keyframes. I haven't tried it, but before I went to HDV I used a Canopus tool that came with my ancient DV capture card from them, and it produced amazing Slo-Mo. Hopefully, these free tools mentioned above would do the same.
colins wrote on 12/17/2008, 1:08 PM
Have you tried right clicking on the clip - go to Pan Crop and choosing '16:9 Widescreen TV aspect ratio'. That worked for me once.