ISO rating of the AF100 sensor

R

roro20

Well-known member
If the native ISO rating of the AF-100's sensor is 400, would it be correct to assume that the cleanest image (not taking into consideration scene file adjustments) would come when the camera is set at an ISO of 400.

I am basically trying to figure out what my best setting for the "L" ISO/gain setting should be.
 
I have found 200 to be the cleanest. I did a lot of testing today. Different lenses and NanoFlash/CF card/uncompressed SDI capture to computer using AJA KONA.

Very interesting results.
 
i too am interested in this. Because if it's anything like the 7D, then every increment of it's native ISO will be the cleanest.
 
I will post video to Vimeo and comments too. I am still evaluating the settings myself, basically to decide how I am going to use this really affordable possible revolutionary camera. I think Pany has really changed the landscape, and I think others are scrambling. Congrats Jan and the team.

But what you have to bear in mind is I am viewing this on a calibrated Pro Monitor using HDSDI into a Kona card through M100 and I am watching Tektronix hardware scopes at the same time. This is my standard setup for editing and setting up cameras. And there is no way that putting up video on Vimeo or stills will show what I am seeing. I never trust what I see on any computer screen or even high end consumer Plasma or LCD monitor/TV.

Basically my initial findings are Cine gamma V lower detail -3 Luma -2 more details later a few other tweaks I don't remember exactly. I am still working out details.

Lenses make a huge difference. Different lens, different contrast, feeling, and tone. I know this sounds simplistic but, it is wonderful. Very much like evaluating still camera gear.

As far as my initial observations on NF vs Native SDHC. I usually shoot NF at 100Mb/s Long GOP. I find very little difference between that and 280 IF. THis is what I used in my initial tests. The quality of the native recording is very good, I would say it rivals the EXcam codec at 35mb/s, this is at first glance only, but that is impressive, seeing how this is 24mb/s. The uncompressed and NF footage has more subtle colors and a slightly less contrasty look, I like this look. It does not jump out at you, but it is definitely there in the subtle gradients. An untrained eye may actually prefer the native footage because it seems more contrasty and saturated. I think there may be some enhancing going on here. But when you start to "push and pull" the image a bit the NF/uncompressed 422 is definitely better.

Sorry to tease here but I want to make sure my findings are correct. So please bear with me. I also have a lot of other work to do and I do have to play Hockey tomorrow. These are only my preliminary notes really.

The bottom line so far. I am very happy with this camera.

PS. the LCD does lean toward green-yellow, but it is not bad, and you get used to it quickly. Not long ago we only had BW monitors on "pro-cams". Come to think of it we used to use hand held light meters a while ago, actually I still do to figure out lighting.

Sorry for the ramble. Thinking out loud here.
 
Last edited:
Are you also evaluating the dynamic range at lower ISO? I would think that since you are negatively gaining the results, that lower values will always show less noise, however, after you move below the native ISO of the sensor, you start to loose dynamic range.
 
grimepoch said:
Are you also evaluating the dynamic range at lower ISO? I would think that since you are negatively gaining the results, that lower values will always show less noise, however, after you move below the native ISO of the sensor, you start to loose dynamic range.
Click to expand...

At this point I was using 200 400 800 on a scene with both blown out light through a window and a very supply lit still life setup with a color chart porselin faced doll and various other items including a tape measure for distance to imager, for setting whiteness marks on my lenses. And I found both dynamic range and noise to be better at 200. These are only preliminary findings on DR watching the scopes closely, and noise (grain crawl) visually on the monitor. Even 800 is not that bad but 200 is definitely the cleanest, grain vice.
 
Good to hear Olof, will look forward to your further tests. I've been shooting at 400 mostly and have been happy with the degree of noise/grain if any. If 200 gives it a little more DR all the better. Very curious..
 
real native iso

real native iso

It would be surprising if the af100 sensor native iso was 400. native means no amplification of the signal from full well, lowest noise and widest DR. if it's gh pedigree then 200 iso is the figure, as that's the sensitivity the sensor was designed for. unless panny did some serious reengineering here.
no wonder avc 24mbit is similar to EX 35mbit. the avc mpeg4 coding efficiency is 50% better than mpeg2.
the interesting question is how af100 gets to hd full raster from the big photo sensor. apart from the optical filter does it pixel skip? downsampling? how?
 
ive done some test too and from what i see is theres no native ISo. i see linear increase of noise from 200-800. 1250 is still clean for me. havent tried higher yet
ill try to do some test more

at 1250 or higher, the dynamic range does go down.
 
Last edited:
You must log in or register to reply here.
Back
Top