GH2 ISOs to avoid...Test results!
- Thread starter dvbrother
- Start date
Lucas Adamson
Well-known member
That's pretty surprising in a way, but not also. THe technology isn't hugely advanced from one camera to the next - merely compromised in a different way, to achieve different qualities, plus some lessons learned, but the GH1 is better in some ways, which is only a consolation to me if anything, as the GH2 is more practical in MANY MANY ways.
Thanks for the link to the GF1 noise test, shaveblog. To my eyes it looks like ISO 500 is nearly as clean as 400, and that ISO 640 is a usable maximum. I'd really like to know what's causing the dramatic increase in noise at ISO 800, it's just as bad as 1600.
I have both GF1 and GH1 cameras, and the GF1's low-light noise is noticeably worse than the GH1. As the image sensors are also made differently, I wouldn't assume that the GH1's ISO noise levels follow the same patterns as the GF1.
I have both GF1 and GH1 cameras, and the GF1's low-light noise is noticeably worse than the GH1. As the image sensors are also made differently, I wouldn't assume that the GH1's ISO noise levels follow the same patterns as the GF1.
bradleysteenkamp
Member
Good gf1 test, really falls over the edge at iso 800!
Yes, ISO 500 is only slightly noisier than 400, but it's been my experience that 400 is where the GF1 really maxes out for video. 500 is pushing it and 640 can be made to look acceptable with Neat Video but it still won't look as clean as I like. 800 is a disaster, as you say.
I love my GF13 and your Fast Action AVCHD and Low Light MJPEG settings get the very best out of it, but I accept that it falls apart at high ISOs for video. It's a tradeoff I'm willing to make for all the other things I like about the camera. I do plan to pick up a GH2 body this year, and that should bring a world of improvement.
I love my GF13 and your Fast Action AVCHD and Low Light MJPEG settings get the very best out of it, but I accept that it falls apart at high ISOs for video. It's a tradeoff I'm willing to make for all the other things I like about the camera. I do plan to pick up a GH2 body this year, and that should bring a world of improvement.
Thanks for the props, shaveblog. I've found the GF1 ideal for making test shots to evaluate the framing and lighting of a scene. I use my Reliable In-Camera Playback Patch on the GF1, which allows me to instantly review the test videos, even ones taken in 1920x810 anamorphic mode. Then I mount a GH1 with the Fast Action Patch on the lens and take the high quality shot. What's convenient is that my test shots remain on the SD card in the GF1, while the GH1 contains only the actual video takes. In addition, I can pass the GF1 to the Director for evaluation of the shot while I'm setting up the GH1 for the actual take. This helps compensate for the lack of video monitoring and playback of files on the GH1. (I actually prefer this style of shooting over instant review of video footage. Reviewing the footage of each take can really slow down a video shoot.)
In addition, the GF1 serves as a backup for the GH1. If the GH1 ever malfunctions for any reason, I can put its SD card with the Fast Action Patch in the GF1, reprogram it on the spot, and continue with the shoot. It's always a good idea to have a fallback plan in place.
Last edited:
stonebat
Veteran
That sounds plausible. If that applies on GH13, 100/160/320/640/1250 would be a good set. Since 1250 is not so good in GH13, a practical set would be 100/160/320/640.
Lucas Adamson
Well-known member
I don't think we can assume, can we, which iso settings are the "native" settings, and which are these 1/3 over or under digital quasi-settings.
Surely the solution for the GH1/ GH13 is switching off the 1/3 stop iso settings (switchable only in stills mode, by the way, but still applies to video). There is a reason that's a menu option - it's clearly digital manipulation to make exposure easier. I will now deal with this by going back to the whole stops option and use nd filters from now on to achieve my exposure correction - at least it remains analogue then, and will result in a cleaner image.
Sage
Well-known member
Just seeing this thread for the first time (been busy)
To throw my two cents in, from a similar test thread it was found that gh2s may have huge difference in iso per make.
I decided to use 160 and 400 on my gh2, with any occasional 640 for grainy look. In general I think that 160 and 400 are quite trusty across gh2s, with 250/500 showing the worst performance.
To throw my two cents in, from a similar test thread it was found that gh2s may have huge difference in iso per make.
I decided to use 160 and 400 on my gh2, with any occasional 640 for grainy look. In general I think that 160 and 400 are quite trusty across gh2s, with 250/500 showing the worst performance.
Great post. That sounds like a pretty reliable way to shoot on a budget. I found some scenarios where even the GH2 will give the memory card write speed error last night. Hopefully, we can find enough situations to force Panasonic to issue new firmware. A hacked GH2 would be a dream come true.
Lucas Adamson
Well-known member
So, returning to the idea, suggested by dvbrother, that 1/3 stop increments are digital +1/3 and -1/3 gain versions of the analogue whole stop iso settings, if I set my GH1 to whole stops instead of 1/3 stops, I get 100, 200, 400, 800, 1600 iso. Surely these are the clean, un-doctored, native settings?
Then there's the idea that -1/3 gain, can improve blacks and clean the image by reducing gain, so by this reckoning 160, 320, 640 & 1250 would be good to use too. By the same rationale, 125, 250, 500 & 1000 are the bad ones, featuring the gain boost, and to be avoided if at all possible.
Is that about right? It seems to be mostly borne out by the OP's GH2 video. Where's Barry Green when you need him?