Something I've wondered about shots like this: can we consider there to be a sort of inverse relationship between how much you expect to have to crop and how high an ISO you're willing to go to on a photo?
For instance, this photo has only a very small crop from 19.4 to 15.5 MP to get rid of a little of the edges, but even at ISO 10000 it looks decent:
View attachment 91756
A larger crop is this one, going from 45.7 MP to 18MP.
View attachment 91757
Even for ISO 11400 it's really pretty good after denoising.
Here's one brought from 45.7 to around 9MP:
View attachment 91760
This one actually took a few different tries with denoising and sharpening to get where it is. In other words, even though it was only at ISO 2200 versus the higher ISOs of the other photos, it took much more effort to get looking decent. Is the much larger crop the reason why?
Then there is one like this, which went from 45.7MP all the way down to 3.6!
View attachment 91758
I think from a distance it looks okayish, but it really starts to look pretty lousy if you look at all closely.
By a crop from 45.7 to 2.5 MP, even at ISO 360 what little noise exists seems to big large enough to eliminate any detail that may have otherwise been there:
View attachment 91761
This isn't actually half bad considering the extreme crop and may even look slightly better than the eagle in all honesty, but really there is very little detail there if you look. This isn't even a photo I'd ever try to really use or save but I just posted for the sake of the question.
Is this idea of judging the workable ISO based on the subject size in frame a good one? We'd all like to avoid cropping of course, but if we decide to do so, is it a good practice to try to keep the ISO lower to make the crop work as well as possible?