Workaround for highlight blinkies in stills mode

Thank you tajohnson. One of the biggest reasons I'm sad about this site going away is the access I had to much smarter people than myself.
 
Thanks for the picture control, however it doesn't seem to work with my Z5 using a UniWB white balance.

At first I thought the picture control wasn't working at all because no dark spots were shown even when overexposing like crazy.

I then switched to a standard white balance and voila! I can see clipped areas.

Is it an issue with the Z5 only?
 
The picture control is linear from 0 to 254, and at 255, drops back to 0. The result is that when a pixel saturates, it becomes black in the viewfinder.
Very cool workaround.

How would this provide significantly different exposure from Highlight-weighted metering? More control over what blows out?
 
The picture control is linear from 0 to 254, and at 255, drops back to 0. The result is that when a pixel saturates, it becomes black in the viewfinder.
Very cool workaround.

How would this provide significantly different exposure from Highlight-weighted metering? More control over what blows out?
As I describe in my books, Highlight-weighted metering (HLM) does something unique, though you can fall prey to assumptions using it. With a broad area of brightness in a scene, HLM will place that area at middle gray, which actually underexposes white.
 
Last edited:
Thank you for the nice solution to a missing tool that Nikon still does not seem to want to provide.

Added it to my Z9 and it worked straight.

Added it to my D810 and D4 too by copying the curve from your file in Nikon picture control on my PC, applied the curve to a picture profile for each camera model and exported the file once for the D810 and the other for the D4.
 
Glad I could help. I've been using it for years now and it's an integral part of my process.
 
As I understand things even in cameras that provide blinkies by default, what they are reporting is not true saturation of the raw photosite but an upper saturation limit for the camera manufacturer's target jpeg. In other words, If we took your sample image above in to raw digger would we see the same level of over-exposure? A useful experiment would be to adjust the exposure until it just starts to show clipping on the LCD and take that in to Rawdigger or some other raw analysis tool to see whether any overexposure is observed. If none is seen then it can be assumed that the 255 point does not represent raw photosite saturation. I hope I am making some sense (I ask because my understanding of the topic is sketchy).

Thanks
OOf yes. This has been a pain for ETTR photographers forever. I went down an ETTR rabbit hole a while back - just for my own edification I don't shoot that way, and in the extreme photographers were taking photos of like... inverted colors from the camera on their computer monitors in order to create custom white balances that roughly treated each RGB channel equally so that the target JPG preview would show closer to accurate for when each channel would blow out.

The method in this thread + that would probably be the best way to deal with it - but then you'd be looking at something nothing like what the end result is, but if your goal is to maximize sensor information - the rabbit hole is the way to go.
 
Somehow I'm just seeing this thread now. Fantastic work Todd! And beatboxa's earlier work with profiles as well.

I saw a few posts about whether the 255 level accurately represents the raw clipping level, including for particular color channels / various WB levels. Stumbling around inside the Picture Control Utility 2 app I found it has a preview feature that not only lets you load your own NEF to test against the profile but it also lets you change the exposure and WB against that loaded image and see the profile respond to that as well. This can be used by photographers to calibrate Todd's work against their own raw images to see how closely it's matching to raw clipping levels, esp if you triangulate with RawDigger for the NEFs you load to test against the profile.

Here's a screenshot of an image I loaded in Picture Control Utility 2, with the adjustment dialog open as well:

Sample of loaded NEF tested against Todd's blinkie profile

Apologies if this has already been mentioned in the thread...
 
Last edited:
Thanks for this workaround!

But all google drive links under this topic are not available now.

Could you please upload or share again? Or sharing instructions about how I can create the configuration my self.

Thank you very much.

------

Edit: Oh I found Nikon Picture Control Utility 2.

Let me try!
 
I did some house cleaning and accidentally deleted the google drive folder that I linked to in the original post. I have restored it at the following link:

https://drive.google.com/drive/folders/1_3agtPdZAaHA3TOvF-Qs4Yq3KFga5f5g?usp=sharing

-Todd
Thank you for your work!

I tried a lot yesterday. Nikon's tool was too bad to edit the curve 😅

Then I opened the exported curve file, and finally cracked it!

b046e3dab398474994c8f88bb2214b98.jpg

Here, the 0x14 in row 00000080 indicates 20 (0x14=20) points on the curve.

And next are the point coordinates. It is represented by big endian double float numbers.

For example, in row 000001B0, 0xE0DFDFDFDFEF3F is 254/255 😅

With this way, I created 20 points with the hex editor manually, which is very accurate!



dd8202381d71417593cad0d6f201643b.jpg
 
I agree, the Nikon Picture Control Utility isn't designed for creating picture controls like this one. I used https://nikonpc.com/ which allows/requires specification of exact values for each control point.
 
I agree, the Nikon Picture Control Utility isn't designed for creating picture controls like this one. I used https://nikonpc.com/ which allows/requires specification of exact values for each control point.
The curve is only a visual aid. I learned the camera ignores the curve and only looks at a 256-entry LUT generated from the curve. With the LUT you can directly set values without having to fight with the cubic spline interpolation of the curve or be limited to the 18 values in the curve. You can read the details at my post linked below, including some picture controls I created by directly manipulating the LUT:

https://www.dpreview.com/forums/post/67163910
 
Last edited:
Just wanted to say THANK YOU for developing this picture profile. Not having highly clipping warning in photography was driving me crazy on my new Nikon Zf. This is not elegant, but fairly simple, and more importantly - works!!
 
Just wanted to say THANK YOU for developing this picture profile. Not having highly clipping warning in photography was driving me crazy on my new Nikon Zf. This is not elegant, but fairly simple, and more importantly - works!!
I'm sorry if I missed something here. I realize it's an old thread, but a few comments here are new. When this post first came out we didn't have the Z8 or Z9. Didn't my Z7II have blinkies (clipping warning)? My Z8 definitely has that feature. But I have to review an image to see the blinkies work. Or are you talking about in live view mode? Thanks.
 
Just wanted to say THANK YOU for developing this picture profile. Not having highly clipping warning in photography was driving me crazy on my new Nikon Zf. This is not elegant, but fairly simple, and more importantly - works!!
I'm sorry if I missed something here. I realize it's an old thread, but a few comments here are new. When this post first came out we didn't have the Z8 or Z9. Didn't my Z7II have blinkies (clipping warning)? My Z8 definitely has that feature. But I have to review an image to see the blinkies work. Or are you talking about in live view mode? Thanks.
Yes, it is about the highlight warning in live view mode.

The jpeg file will have the imprint of the overexposed blinkies- so only for those who shoot Raw +/- jpeg



NEF vs JPEG
NEF vs JPEG



--
——————
Pankaj
https://www.dpreview.com/galleries/1160557041/albums/chopta-birding-uttarakhand-india
 
Last edited:

Keyboard shortcuts

Back
Top