ISO 100 vs 200

[greypilgrim]

Hey all,
So, I have seen references here and there that say that ISO 200 is actually better than 100 for m4/3, but I haven't seen a good answer as to actually why this would be the case, and what the real image quality and or performance difference might be.

Anyone have a good answer and even better, also some examples that demonstrate this difference?

Thanks,

Doug

 

[Vivek]

I actually prefer ISO100 and use it whenever I can.

 

[greypilgrim]

Well, that is where I sit also, but I keep seeing these references to ISO 200 being the "native" ISO of the m4/3 cameras and that you "lose dynamic range" at ISO 100.

Doug

 

[Vivek]

Doug, I asked myself this question a while ago (about those chart pundits):

"How many snaps do they make other than that of charts and book shelves?"

I do not think I found a direct answer yet but sort of guessed that it is pretty minimal.

PS: If you want any worthwhile DR at all out of the NMOS sensor, get rid of the IR cut filter and shoot in IR.

 

[greypilgrim]

Well, I only have the one G1 at the moment, so I am not quite ready to go IR with it ...

But when I have a second m4/3 body, then I definitely have plans along those lines.

Doug

 

[Tullio]

The EP1 native ISO is 200 (when using AUTO). I took identical pictures with ISO 200 and forced 100 and indeed, the images at ISO 100 were much brighter and highlights were clipped a lot more than with ISO 200. However, if you set to ISO 100, lower the exposure by -0.3 or even -0.7 and reduce contrast by -1, the results between ISO 100 and 200 are very similar and you have the benefit of getting less noise.

 

[kds315]

I also try to use ISO100 as often as possible as I find it best.
I also underexpose -0.3/0.7EV usually.

 

[pellicle]

this may seem a silly answer, but if you photograph the same thing in the same lighting conditions at 100 and 200 (100% the same thing, not even move the camera 1deg AKA tripod) and you can't find any difference ... then perhaps that's good enough?

examine the image carefully, use the channel splitter in photoshop and look at R G and B individually, look into the shadows and look into the hilights, examine the histogram (Ctl + Tab between the images) .. if its not bunching up more then its likely to be all good.

 

[greypilgrim]

Silly? No. Just a reminder that I could actually do a test instead of posting ...

I guess I am just curious to see if there is some reality to this before I start pixel peeping which is something I try to avoid at all costs..

Doug