One thing to consider is which of the two exposure workflows
Expose for end result
Expose for maximum dynamic range and snr
Is more likely to require further exposure tweaking and other more involved post processing. My thinking is that the second way of working is likely to involve a lot more tweaking whilst the former can be a pretty minimal affair as far as pp goes.
It would be interesting to hear peoples thoughts. If you’re doing further tweaks anyway the default exposure is pretty meaningless? Getting a good image out of the box is mostly useful for light and importantly fast editing I would think.
I do a lot of highlight protecting exposures but I primarily do that in manual and find that what the camera thinks is correct exposure is pretty meaningless in those circumstances. It doesn’t know what my subject is. At family gatherings and such I’ve started going for correct exposure in camera because of bulk and speed.
Exactly right! And this is why we can’t do this on a per-image basis but it has to be done on a per-camera basis! The result of your camera would then always be + ~1 EV instead of +0.5 EV.
It’s actually possible to then add camera metered auto compensation on top of that which might actually give you the correct result straight away if your camera wasn’t too confused from that backlit light.
It sounds like what I’m thinking about would indeed help you in your workflow which is encouraging. I’m thinking about downloading the full raw.pixls.us dataset and seeing if I can script the extraction of these per-camera-based middle grey normalization constants.
Oooh, that sounds nice! Have been thinking about getting myself an setup like that!
Incidentally, with a mirrorless camera taking bird pictures, I often use in-camera exposure boost to get a clearer image in the viewfinder.
With mirrorless, the viewfinder is just a small video screen, so when birds are in shade/hiding behind twigs in a bush it can be tricky to see what is going on. In that case, boosting in-camera exposure is not about ETTR, but more about checking the focus is on the bird not the background.
From my perspective, the JPG is only useful for filtering out obviously bad/outof focus images. DT exposure compensation just gives me a better starting point for adjusting the lighting etc on the raw image.
You may also try to deactivate exposure simulation in your viewfinder. I’ve got such option in Fujifilm X-T1 - it’s called “Preview picture effect: OFF”.
I don’t think my Canon R6 has this option, but just to be clear about my bird photo scenario - I want to have the exposure simulation in the viewfinder:
The tree is in full sunlight, but the bird is “inside” the tree, meaning it is sitting on a branch, in shade and likely a bit obscured by leaves or twigs.
Looking through the viewfinder (with no camera exposure adjustment), you see brightly lit leaves and twigs in front of the dim outline of the bird. It can be hard to see what is going on in the picture.
Autofocus (central spot only) may hit the bird or it may the surrounding twigs/leaves. To decide whether manual focus adjustment is needed, you want to see the bird clearly and this is where I will use the camera exposure wheel, usually to boost exposure by 1-3 Ev.
Once I get the raw image into DT, I am happy to auto-remove the exposure bias, and apply my standard scene-referred baseline preset
You want to clearly see what’s going on, but JPG-based preview is limiting you, because of smaller JPG dynamic range which the camera tries to simulate.
To counteract this, you use exp. comp. to brighten the view but risk highlight clipping.
Disable ()
The image is displayed at standard brightness, so it is easy to see. Even if you set exposure compensation, the image is displayed at the standard brightness.
You’ll se better, but the exposure remains intact.
This is interesting - I hadn’t seen or fully appreciated the Canon Exposure Simulation Option, but reading the manual carefully I understand what is happening.
Canon are offering good options but they are assuming that the image is easy to see when displayed at “standard” brightness ie with no exposure simulation.
In my bird-in-the-bush example, the standard viewfinder image is not much good. So I WANT to boost/bias the image I see in the viewfinder by enabling exposure simulation. However, I also want ETTR because the camera is very likely exposing for an area bigger than the bird and likely still underexposing the bird.
Therefore, when shooting with AV/Aperture priority
I enable exposure compensation in the camera,
I adjust the camera exposure wheel, but not so much I get clipping on the bird
In DT I enable auto-removal of camera exposure bias
I apply a standard preset (Exposure +1.5Ev, Filmic on, Color calibration etc etc)
As an extreme example, we were driving back through a gamepark towards our lodge, just after sunset. It was pretty dark, but we could see an animal moving on the road about 50m ahead. I put the camera onto the beast, but could only just about make out the markings of a giraffe. Push the exposure wheel up to +5Ev and incredibly the view finder image showed a big giraffe seemingly in full daylight. The ISO was something like 64000 and the final picture was a little grainy but still passable. Not a great way to get giraffe photos, but great for things like aardvarks and other night creatures.
)