View Full Version : Whoever's got a GH-1, do me a favor.

06-03-2009, 11:37 AM
Do some ISO tests really quick with pulldown.

1. 1080/24P/FHD
2. Set ISO Value
3. Make sure you're properly exposed in whichever way you can but do NOT adjust ISO.
4. Record and Pan
5. Do your Pulldown
6. Let me know what your results are.

I think my camera just got back today, I can't do it now but I think I've got a new theory.

Low ISO's on this camera are noisey as hell. IN fact, you shouldn't use the low ISO's. Most of the time, a low ISO introduces software driven artifacts that you DO NOT WANT. It's the same with RED, MKii and I know for a fact that low ISOs are a no no on the GH-1, I just never got to test out which ones were proper.

Now, how many people with the GH-1 actually have/had Polarizers and ND filters? I'm almost certain that 90 percent have been relying on the ISO's lower values to control light...

But I have ND filters and Polarizers, and I use them ritually to attempt to stay at a 320 ISO.

I used a higher ISO for the action stuff because it was night and low light... maybe there's a chance that the low ISO's are also causing an issue when dealing with the footage?

This isn't fact, but Commander, Sonic and whoever else has one, try it?

06-03-2009, 10:58 PM
It is true that I have only noticed the problem at low ISOs, 100, 200. And I have never taken my Hoya UV filter off.

I doubt it's the cause, rather the low ISOs allow as much detail as possible to be recorded without noise reduction smearing, so the high detail pans are what cause the break up.

I will do a test later, with high ISOs outside, although monsoon like rain conditions today in Taipei!

Also it'll be interesting to try shooting some less detailed scenes outside to see where the threshold is that the break up occurs... it'll be interesting to find that the cause is indeed the amount of detail in a scene, or something else...because we've by no means confirmed it yet.

06-03-2009, 11:10 PM
Actually, the low ISO's were what I noticed to be the worst looking on the GH-1 AND the MKii. It's also the worst on RED.

They may seem like they record more detail but that certainly isn't the case. You should run some tests for sure.

John Caballero
06-03-2009, 11:21 PM
The HVX200 ISO by example is supposed to be rated at 320 which is indeed higher than the 200 commander shot the highest at in his first tests. Keep working on all the possibilities and calculations, you'll become legends when you discover the best formula!

06-05-2009, 01:50 PM
QTI, Commander (you've already tested), Iilya -- enable 1/3 incremental ISOs and try them in a panning situation.

Note which ones you use. By Commander's test, if my theory has any validity, then 160, 320, 640, and 1250 are gonna be the winners.

Here's commander's post in that big thread:


Is it just under-exposing or what? It could be the issue.