Thanks! I might check that out… I wonder how long Nikon is (or was?) providing the free fix for.
If you’ve made it this far you hopefully don’t have an affected copy, but you never know I guess.
1 Like
Since the moment I realised what a scene referred workflow can do for my image processing I have slowly but steadily switched to what I call “optimal ISO exposure” and “iso invariant processing”.
Note: this technique has been established for fast-paced environments where taking a second picture is often not possible while light conditions might change equally fast. For slow-paced situations I chimp my histograms just like everyone else.
From the photonstophotos charts I determine the real steps in ISO - where shadow boost happens and dynamic range is reduced. For example on my D500 these steps are iso100 and iso400 and that’s it. It does make no difference if you take a picture with iso400 and -4 stops or iso 6400 and no under exposure. From the RAW it all comes out the same - almost. Because with iso6400 you have just thrown away 4 stops of extra highlight dynamic range. Oops! Yes, it feels weird to correct exposure in darktable with settings like +5 but it works very very well.
My D3500 is the exact opposite: every stop of ISO both yields a shadow boost but also reduces the dynamic range. With that camera using the lowest possible ISO without introducing artefacts through motion while exposing exactly for the highlights gives the best quality.
In general I have my cameras set somewhere between -1/3 to -2 exposure compensation when using auto-exposure.
Here’s my chart with some cameras I need to be able to work with:
The LowLight number is where the dynamic range goes to or below 6.5 stops as defined by Claff.
3 Likes
I would really appreciate manufacturers catching up with their own products with better auto ISO and metering options for this approach. They only seem to care about your JPEG.
1 Like
I think Nikon broke their JPEG-focus a bit with their Highlight-Weighted Matrix metering mode. The do qualify its use in “JPEG terms”, stating it should only be used in certain situations based on departure from middle-gray JPEG metering. Me, I use it all the time now, because it protects the highlights I care about AND the sensor captures enough shadow data to let me pull it out of the depths.
Kinda sidetracked a bit on the dynamic range discussion, which I’m really ok with. Just wanted to recognize all the PlayRaw posters; yep, while the “Image Requires No Processing” it definitely takes well to all sorts of processing! Even the abstraction renditions have given me pause to think. Thanks all…
1 Like
…and, in a nod to some processing, here’s my rendition, same processing as post #1 except I used my SSF camera profile to improve the color saturation and added a bog-standard Duiker filmic curve:
4 Likes
Good one Glenn!
Ted
1 Like
Y’know, I’ve been all wrapped around the lens correction thing, thinking about the trades between all the lens design things, and your image just points out that a well-done blur on the appropriate subject just sets a mood… like it a lot, thanks.
Thanks Ted, but I can’t really take credit for too much here. Even the “added” processing is plain vanilla defaults…
[quote=“ggbutcher, post:1, topic:41646”]
… and the rendition I’m posting Requires NO Processing, beside the requisite black-subtract, white balance (with as-shot multipliers), and a decent demosaic …[/quote]
Almost an invitation to some, eh?
Yeah, you put a PlayRaw out there, and expose yourself to the whims of others. It’s a bit intimidating, but I’ve learned something from every one I’ve posted.
2 Likes
This is my first try with dt 4.6. So far, I haven’t tried anything different from before. I tried to kick it up just a little while keeping everything natural looking.
DSZ_0445.NEF.xmp (8.6 KB)
6 Likes
It’s amazing what can hide in those shadows. It’s a friendly ladybug. 
Thank you for that lovely image to play with.
And I have to say it is pretty much perfect out of the box. One can interpret but then the image will just draw you back to square one and remind you of the beauty of taking a picture and just let it be. Except for the cute little discovery I didn’t come up with anything different enough worth showing.
3 Likes
“Nice” is in the eye of the beholder.
Most look rather far removed from reality to me. Or represent different variaties of those flowers (they do exist in many colours…).
If you prefer a spectacular image, that’s fine (but don’t present it as “that’s how the flowers look”).
Some prefer more subdued colours, and that can even depend on situation/mood.
But if you need an exact recording, most here are “not so good” (and that’s not an esthetic judgment): e.g. some sites/apps allow you to identify flowers from photos. Inaccurate colours are not helpful there (to say the least).
To be fair, the original renderning wouldn’t be ideal either for that use: too dark 
1 Like
It looks like a really bleeping good painting by a Master, and that’s a good thing. You have to really look at it, rather than have it jump out and throttle you.
I think that of this PlayRaw: