IQ180 and ISO question

[Wayne Fox]

In a recent thread about EttR at LuLa, one of the techier posters mentioned that increasing ISO when shooting a CCD based camera won't yield an image much different than just underexposing it . Underexposing and adjusting in C1 will result in an image of about the same when processed correctly in C1, as opposed to boosting the data in camera with the ISO. This surprised me a little, but I've never really thought about it.

I've done a few tests, and anything larger than the 2.5 exposure adjustment limit gets challenging, but basically files can be processed to be virtually identical and possibly the ISO 35 shots, even though more than 2 stops underexposed might be yielding slightly cleaner files. Boosting in camera ISO does still provide better camera previews if you are reviewing them for composition or focus, and it certainly is faster to post process.

Just curious as to habits of those using MFDB. If you need a slightly faster shutter speed which will result in underexposure, do you boost the ISO or just shoot it?

 

[Thierry]

hi Wayne,

You will find some information about that in the following thread, a few years ago:

ISO 800 - some Sinar eMotion 75 samples

Thierry

In a recent thread about EttR at LuLa, one of the techier posters mentioned that increasing ISO when shooting a CCD based camera won't yield an image much different than just underexposing it .

 

[Wayne Fox]

I glanced at the thread. I guess it's not exactly the same question ... in the example you are already cranking the ISO up to 800 and still underexposing. My question is more like if I"m at ISO 35 and 2.5 stops underexposed, what is the common practice ... go ahead and capture it or move the ISO to 200 leaving shutter/aperture unchanged (which is what I've always done). This delivers a "correctly" exposed image, but certainly doesn't put more light onto the sensor so I'm still not sure about which would deliver the optimum capture. Does the back's firmware massage the data a little better before it gets to C1 than doing all of it in C1?

 

[Shashin]

That is a tough question because of the unknowns in the signal processing. I shift the ISO on the assumption that the binning of the signal into 14-bit image file will be better--the number conversion has to take place somewhere and I think from the original signal would be the best place to start.

I have no evidence of my assertion. I have never really tested it. But I know it is harder to process shadows with an "underexposed" file and I seem to get better results by increasing ISO. All of this is informal (I only have myself to please).

 

[Thierry]

Yes Wayne, I had set the ISO to 800, but then some "clever" members found out from the provided DNG files, that the back was actually simply under-exposing from the standard or nominal ISO and this under-exposed signal was then "pushed" by the amount of 3 f-stops by the electronics. In other words the histogram is simply moved to the right.

I don't know how the current backs handle ISOs over the nominal sensitivity, but most of the backs up to the 39 MPx use to simply push the signal afterwards. It could well be that the IQ series do it differently, e.g. amplifying the signal before recording.

Thierry

... I guess it's not exactly the same question ... in the example you are already cranking the ISO up to 800 and still underexposing ...