Thank you, you have confirmed my previous post. Take ISO out of the equation, it doesn't matter. Just set whatever shutter speed and aperture you need to get the shot and ignore ISO. You'll underexpose and get a dark preview but when you edit and brighten in post, the image will look about the same as it would if you boosted ISO. Or if it makes you feel better to boost ISO, do so but it doesn't matter.
These threads have pretty much all been red herrings, for practical purposes the only thing changing the ISO changes is the brightness of the preview on the back of the camera.
p.s.
It's funny how the key thing that seems to get overlooked in these threads in that the ISO knob only effects the read noise component. And that is drowned out in modern cameras by shot noise, so the effect of the ISO knob in modern cameras in mostly an illusion (potentially a tiny noise improvement for the perfectionist). Just as DPR have said for years.
Since I don't consider myself an expert, I will launch into details.
First and second generations of Nikon DSLRs (up to D3) were nearly ISO-invariant, raising ISO setting helped very little with read noise, half a stop at best. Learned it the hard way, and rather late, on my second year with D2X. Clipped a lot of highlights and missed a lot of opportunities because of my ignorance.
pretty much my conclusion from my test shots tonight. BUT i have to use iso all the time as i shoot live monitors. just tested AF and the boost on the a74 is amazing and iso makes no difference.
i will also comment that the colour tones of the underexposed 100iso lifted were more saturated and pleasing even though the noise was the same. good test and we both have learn someting new or at least confirmed some common misconceptions. modern cameras are awesome.
I understand. Sometimes circumstances force us to work a certain way and that's fine. But it is funny that what seems like hundreds of posts have been written about this and that, and it turns out that for most purposes, either approach is fine and worrying about the effect of the ISO knob on image quality is pointless. It works pretty much the same, either approach!
Adjusting the ISO for the optimum for your particular camera can improve the SNR if that is how it is designed. But the shot noise increases in lockstep with the signal, so raising ISO does nothing to improve this. It only works on read noise. And read noise is much lower these days. You might even think of it as negligible. Jim Kasson reckons you can squeeze a little more out of an A7riii by setting the optimal ISO, but it's such a small improvement, the more casual among us might consider it not worth the effort of thinking about it. But if you are a good craftsman you can eek out a tiny bit more SNR.
I'm such a bad craftsman, I set my expectations lower. If I shoot to within about a stop of optimum, I'm delighted.
The downside of raising ISO is it increases the risk of clipping. Although I'm not sure what the practical circumstances would be where it is a real issue. If I'm struggling for light, it's unlikely I'll overexpose. I can see why it is a bad idea to do what my wife did, shooting her safari at ISO 4000 in bright light. A combination of underexposure and risk of clipping is not good.
To be honest, in good light you could just about train a monkey to take a nice looking shot with any modern camera in Auto mode. It's in low light situations where I find ETTR and maximising the exposure* within artistic constraints is most beneficial for minimising visible noise. In good light, if i'm 1/2 to 1 stop under optimal exposure** I find it generally has little affect on the final image quality after processing the raw data. With raw data I find I can be a little lazy sometimes when fine tuning exposure*.
* exposure - amount of light that struck the sensor per unit area while the shutter was open
** optimal exposure - the maximum exposure* within dof and motion blur requirements without clipping important highlights.
*** under exposed - more exposure* could have been added with the DOF and blur constraints still being met without clipping important highlights.
The ISO 100 one looks brighter. Is that what you mean? If so, just reduce the settings in your raw editor until it looks the same. You can set the brightness in post to anything you want.
isnt that weird. last night i could see different noise patterns in the black background between the 2 images ,but this morning it is not visible due to ambient light hitting my monitor.
