acquisition frame rate at DVinfo.net

Bruce Watson · February 7th, 2012, 11:43 AM

I'm going to film a class. USA, NTSC. The target output is DVD, Blu-ray, and some web trailers. It's going to be filmed in a huge classroom that has hundreds of magnetic fluorescent lights. They won't let me use enough equipment to overcome the lights, so they are my major light source, 60Hz flicker, buzz, and all. Out of my control.

It looks to me like the common denominator is to capture in 1080/30p. DVDs require 30i, Blu-ray supports 30i, and the web supports 30p. I'm running Production Premium CS5, and I think Media Encoder should be fine in making all the transcodes required without inducing any serious judder (like translating from 24p which would give me some 2:3 pullup for the DVD, yes?). And, this should work well enough with the lights if I can use a shutter speed of 1/60th, which should minimize flicker from the fluorescents.

So 1080/30p? Or am I missing something?

Battle Vaughan · February 7th, 2012, 12:33 PM

Somebody check me on this, but seems to me a 1/60th shutter would probably be in synch with the pulsing and exacerbate the problem. Using a slower shutter would allow more than one cycle of the lights to be recorded per frame, overlapping the pulses. Or so I think. I'm relating this to trying to synch shutter with the image on a TV monitor, where you have to use a slow shutter to avoid catching the scan lines... or, maybe I'm missing something too....

Bruce Watson · February 7th, 2012, 02:52 PM

Quote:

Originally Posted by Battle Vaughan

Somebody check me on this, but seems to me a 1/60th shutter would probably be in synch with the pulsing and exacerbate the problem.

If I read this reference correctly, we want an even multiple of the mains frequency. So that the shutter opens at the same place in the mains cycle, and closes in the same place of the mains cycle, for every frame exposed. And this should eliminate flicker. So if the mains frequency is 60 Hz, we'd want either 1/60, 1/30, 1/20, etc. But I could easily be wrong.

Jay West · February 8th, 2012, 05:44 PM

Even multiples it is. 1/60th is a good rule o thumb for NTSC under florescents. About a year ago, I got a lucid explanation of the details of this from Chris Soucy. See this thread:

http://www.dvinfo.net/forum/canon-xh...ter-speed.html

Bruce Watson · February 17th, 2012, 01:26 PM

For the record... I did manage to get some time on location to run some tests. I can confirm that 1080/30p at 1/60 second does indeed work perfectly with the fluorescent ceiling lights. No buzz, no flicker, no strobing visible in the resulting video. Interestingly the white balance also took out nearly all of the green spike, so it's an easier color grade than I was expecting. All good things.

Just posting this in case anyone comes looking for this information in the future.

February 7th, 2012, 11:43 AM	#1
Bruce Watson Major Player Join Date: Mar 2010 Location: Raleigh, NC, USA Posts: 710	acquisition frame rate I'm going to film a class. USA, NTSC. The target output is DVD, Blu-ray, and some web trailers. It's going to be filmed in a huge classroom that has hundreds of magnetic fluorescent lights. They won't let me use enough equipment to overcome the lights, so they are my major light source, 60Hz flicker, buzz, and all. Out of my control. It looks to me like the common denominator is to capture in 1080/30p. DVDs require 30i, Blu-ray supports 30i, and the web supports 30p. I'm running Production Premium CS5, and I think Media Encoder should be fine in making all the transcodes required without inducing any serious judder (like translating from 24p which would give me some 2:3 pullup for the DVD, yes?). And, this should work well enough with the lights if I can use a shutter speed of 1/60th, which should minimize flicker from the fluorescents. So 1080/30p? Or am I missing something?

February 7th, 2012, 12:33 PM	#2
Battle Vaughan Trustee Join Date: Nov 2008 Location: Miami, FL USA Posts: 1,505	Re: acquisition frame rate Somebody check me on this, but seems to me a 1/60th shutter would probably be in synch with the pulsing and exacerbate the problem. Using a slower shutter would allow more than one cycle of the lights to be recorded per frame, overlapping the pulses. Or so I think. I'm relating this to trying to synch shutter with the image on a TV monitor, where you have to use a slow shutter to avoid catching the scan lines... or, maybe I'm missing something too....

February 8th, 2012, 05:44 PM	#4
Jay West Major Player Join Date: Mar 2010 Location: Red Lodge, Montana Posts: 889	Re: acquisition frame rate Even multiples it is. 1/60th is a good rule o thumb for NTSC under florescents. About a year ago, I got a lucid explanation of the details of this from Chris Soucy. See this thread: http://www.dvinfo.net/forum/canon-xh...ter-speed.html

February 17th, 2012, 01:26 PM	#5
Bruce Watson Major Player Join Date: Mar 2010 Location: Raleigh, NC, USA Posts: 710	Re: acquisition frame rate For the record... I did manage to get some time on location to run some tests. I can confirm that 1080/30p at 1/60 second does indeed work perfectly with the fluorescent ceiling lights. No buzz, no flicker, no strobing visible in the resulting video. Interestingly the white balance also took out nearly all of the green spike, so it's an easier color grade than I was expecting. All good things. Just posting this in case anyone comes looking for this information in the future.