Reminder: By posting here you’re sharing your raw images / linear Tiff files and derivatives for others to download, edit, and re-post inside the forum.
Christoph Oldendorf / C F Systems may use originals and derivatives for demos, docs, and marketing of ColorPerfect.
Outside use by other users requires your explicit permission or a license you add. Make sure you own the photo and have consent if people are shown.

Getting exposure right for color negatives on Nikon Zf and comparing to Coolscan V

User avatar
AlexisMagni
ColorPerfect User
Posts: 44
Joined: Wed Aug 27, 2025 7:24 am

This post has been a very interesting read, and I have some questions now for Christoph :) Specially after I finished scanning a Color Film both using my Nikon CoolScan, and the new DSLR Scanning Rig I have.

- How can you have Clipped Highlights shown during Camera Histogram in capture, and confirmed by Lightroom as Clipped, and when using MakeTiff, the histogram gets shifter to the left and doesnt clip? This doesnt seem to make sense to me?

- I use the technique of setting Exposure using the Film Base, until it clips, then I move it a bit to the left to leave some headroom. Using the same technique, I find that the Raw Linear Tiff from the Nikon CoolScan looks much brighter in Photoshop than the MakeTif generated Linear from Camera, is there anything that changes between these two?

From my understanding, if it clips in normal histogram, it should show as clipping in Linear as well? You cant generate data that doesnt exist? Or can you?

EDIT: In order to give better understanding of what I am talking about, I have hosted two Linear TIFFs for you both to try.
The one with "IR" on the name is from my Nikon Coolscan Output as Linear Tiff by VueScan. The "PixelShift" is from my Nikon Zf Scan, going through MakeTiff.
https://we.tl/t-ySYxfiLEHc

Edit2: The Original DNG file from Nikon ZF before going through MakeTiff, it shows a very acceptable Histogram
https://we.tl/t-wQUgRQOXjO

Thanks!
Attachments
050925-HarmanPhoenixII-LeicaM2-IR-12.tif
(134.37 MiB) Downloaded 7 times
050925-HarmanPhoenixII-LeicaM2-PixelShift_12.DNG
(191.98 MiB) Downloaded 3 times
You can check my photos on Flickr:
https://www.flickr.com/photos/alec246_photography/
User avatar
robyferrero
ColorPerfect User
Posts: 151
Joined: Wed Aug 20, 2025 4:12 pm
Location: Italia

Of course, we'll wait for Christoph's technical comments.

In the meantime, we can discuss it as enthusiasts.

It could be that LR is clipping it due to a hidden default curve.

But actually, looking at the DNG histogram, which extends to the right,
it would suggest the MakeTiff histogram is much further to the right than it actually is. In my opinion, it should be almost halfway down the histogram.
User avatar
AlexisMagni
ColorPerfect User
Posts: 44
Joined: Wed Aug 27, 2025 7:24 am

robyferrero wrote: Sun Sep 07, 2025 11:16 am Of course, we'll wait for Christoph's technical comments.

In the meantime, we can discuss it as enthusiasts.

It could be that LR is clipping it due to a hidden default curve.

But actually, looking at the DNG histogram, which extends to the right,
it would suggest the MakeTiff histogram is much further to the right than it actually is. In my opinion, it should be almost halfway down the histogram.
Thanks for having a go! Exactly, I would expect this to show much further to the right in the linear histogram.

I have a shot of the Film Base on the other roll I took, that I used to adjust the exposure for the Film Scanning Process. I had the histogram showing as clipped, and I adjusted the shutter speed least the camera allows me, 1/3EV Minus. It could have not been clipped, but it visually was basically on the clipping limit, and I always leave some headroom.

This is the Lightroom Histogram of that Capture.
Screenshot 2025-09-07 at 21.08.26.png
Screenshot 2025-09-07 at 21.08.26.png (96.27 KiB) Viewed 711 times
[Full image link - opens in new tab]

And this is the PS Histogram after I have converted using MakeTiff to Linear! I mean, there has to be something wrong? I am 1/3 EV of Clipping, and the Linear is not even filling the first 1/4 of the total histagram? This is making me very confused! :?
You can check my photos on Flickr:
https://www.flickr.com/photos/alec246_photography/
User avatar
robyferrero
ColorPerfect User
Posts: 151
Joined: Wed Aug 20, 2025 4:12 pm
Location: Italia

Well, I'd ignore the LR/PS histogram and expose 1 and 2 stops to the right, until the MakeTiff linear histogram reaches the middle.

Anyway, have you tried calling Color Perfect with that low linear file?


Then, if anything, look at the PS histogram after Color Perfect.

In any case, always in the MakeTiff histogram, make sure that the only line running to the right doesn't clip.
Attachments
Schermata 2025-09-07 alle 23.15.12.png
Schermata 2025-09-07 alle 23.15.12.png (184.91 KiB) Viewed 703 times
[Full image link - opens in new tab]

C.Oldendorf
Developer
Posts: 167
Joined: Fri Sep 02, 2022 10:31 am
Contact:

That is one of the central sources of confusion in digital photography: we really want to know what the sensor is doing, but all we are shown is what the JPEG engine or live-view stream produces. That is the histogram we see; that is the source of the blinkies we may get as warnings. It can mask underexposure by up to 2 stops, or it can obscure overexposure in other situations.

The good thing with reproduction work is that even different color negative films often have similar film bases. If the base does not overexpose at a given aperture and light level from a fixed light source you use, you are fine. In practice, you can set the camera to M mode and stop worrying about it afterwards. Everyone will get this wrong at first, and results will be worse than they could be, which is why this is an important read.

Like I showed for the Canon 5D Mk II earlier:
in-camera the ideal exposure looks like a ruined image.
IMG_1634 +2 EW Camera JPEG.jpg
IMG_1634 +2 EW Camera JPEG.jpg (66.13 KiB) Viewed 699 times
[Full image link - opens in new tab]

But the sensor data of this capture is perfectly fine. Generally, I’d prefer to underexpose by another 1/3 stop—so +1.66 EV rather than +2 EV—because that small reduction doesn’t hurt anything and leaves a little headroom.
IMG_1634 +2 EW MakeTiff.jpg
IMG_1634 +2 EW MakeTiff.jpg (49.79 KiB) Viewed 699 times
[Full image link - opens in new tab]

ColorPerfect 3.0 in ColorNeg mode even tells you how much exposure you can safely add when you call it—so no histogram guessing is needed:
ColorPerfect_Headroom.jpg
ColorPerfect_Headroom.jpg (406.44 KiB) Viewed 699 times
[Full image link - opens in new tab]

Aim for the brightest of the color channels to come out around ~0.3, and let the others fall where they may. Do not experiment with large changes to the light source’s color; use something resembling daylight. When done correctly, merging two captures into one with higher dynamic range and detail is possible, but that is beyond the scope of this thread.

As for Lightroom: it seems to be a design goal to match the brightness of the camera JPEG when opening a RAW file. I stopped thinking about what they do and why long ago. ;-)
C.Oldendorf
Developer
Posts: 167
Joined: Fri Sep 02, 2022 10:31 am
Contact:

AlexisMagni wrote: Sun Sep 07, 2025 4:12 pm I mean, there has to be something wrong? I am 1/3 EV of Clipping, and the Linear is not even filling the first 1/4 of the total histagram? This is making me very confused! :?
There is nothing “wrong” here—comparing histograms across different encodings of the data is tricky. With linear data, the brightest value is white; half of that is just 1 EV darker. Half again is another EV darker. So at 1/4 of the total histogram you’re only at –2 EV relative to the maximum the sensor can capture. That’s exactly why we want to expose generously for color negative repros.

On the 5D Mk II, for example, you have 12-bit detail. That means:
  • 2048 values in the brightest stop
  • 1024 in the next stop down
  • then 512, 256, 128, 64, 32, 16, 8, 4, 2, 1, 1
By giving the negative 2 EV more light, the densest parts—where much of the detail is actually encoded—end up outside those regions with very little differentiation.
It also keeps the noise pushed further away from the important parts of the image.
User avatar
AlexisMagni
ColorPerfect User
Posts: 44
Joined: Wed Aug 27, 2025 7:24 am

Thank you first of all Christoph for taking your time to properly explain us how this works internally. This is immensely helpful.

What strikes me the most is that the range between full white, and halfway on a Linear Histogram is just 1 Stop? Wow... I had no idea it was like this.

I always set my camera in Neutral or Flat Profile, because I know how it's internal Histogram will be shaped by the used profile, and that is enough evidence to let us know what you get is not the true raw data. But I would have hoped at least using Flat or Neutral profile would give us confidence as much as possible how the dynamic range is looking in the raw.

So, now I have to figure it out a way to have a true to raw information when I am setting up the camera exposure. VueScan is golden, now I realize it shows you the Linear Histogram, and that makes me see exactly what is getting to the sensor. I wish there was a way to do that with digital cameras as well.

What worries me the most is how the Linear Histogram behaves. It will have a huge peak at the first 1/4, and then it goes so thin that I might not even realize if I am getting clipping or not? At least the Histogram in Photoshop is not clear to me, I dont know if there are ways to display better and clearer data from a Linear File.

And I learn there is this very useful feature in CP 3.0 now, showing how much Clipping we have left? I Wish we had an easier way to adjust exposure, as this would require to take a photo, convert to linear, open PS, Open ColorPerfect and adjust, every time I Scan my negatives, but if I dont find any other way, it will have to do it.

Another issue is how coarse the camera let's us adjust exposure. 1/3 EV minimum, with Linear Histogram, this moves very fast!
You can check my photos on Flickr:
https://www.flickr.com/photos/alec246_photography/
User avatar
robyferrero
ColorPerfect User
Posts: 151
Joined: Wed Aug 20, 2025 4:12 pm
Location: Italia

Keep a flat profile to make the camera's JPG histogram as close to the RAW as possible?
Maybe you're right; it's better than nothing. The rule of thumb is that you can clip the camera's histogram a little. But how much is unknown.

This rule of thumb applies to outdoor photography, whether you're shooting with a reasoned eye or using point-and-shoot techniques.

The alternative is to do test after test, comparing the camera's histogram with the RAW image on your computer, and in any case, you may never be able to accurately determine the histogram.

I think there's software that allows you to read the correct raw data, but I don't remember the name.

For reproducing color negatives, however, you could take three shots: N, +1, +2. This would save you from having to go back and forth between the camera and the computer to check the histogram.
I think the PS histogram is precise enough to read the MakeTiff linear file.

Just understand that exposure with MakeTiff increases, or, I'd say, doubles, like apertures:
If you're at 1/4 of the histogram, +1EV takes you to the middle. If you're in the middle of the histogram, +1EV takes you to the right.
Your example file requires +2EV.
But this +2EV, in this case, only applies to the reproduction of the color negative.
If your example file, with this MakeTiff histogram, were a normal photograph taken outdoors, then you would be underexposed by -1EV, so you would have to overexpose by +1EV.
And it's only the MakeTiff histogram you need to evaluate, otherwise it will confuse you.

If you try to process the same RAW with Color Perfect, but convert one with MakeTiff and the other with CameraRaw, these two will again look different. The same thing, in my opinion, happens if you exit Lightroom with one of those linear profiles.

The solution is always MakeTiff with its histogram.
Color Perfect is best designed to work with MakeTiff; it's true that it works wonders with other TIFFs, but nothing beats MakeTiff.
User avatar
AlexisMagni
ColorPerfect User
Posts: 44
Joined: Wed Aug 27, 2025 7:24 am

robyferrero wrote: Mon Sep 08, 2025 10:30 am Keep a flat profile to make the camera's JPG histogram as close to the RAW as possible?
Maybe you're right; it's better than nothing. The rule of thumb is that you can clip the camera's histogram a little. But how much is unknown.

This rule of thumb applies to outdoor photography, whether you're shooting with a reasoned eye or using point-and-shoot techniques.

The alternative is to do test after test, comparing the camera's histogram with the RAW image on your computer, and in any case, you may never be able to accurately determine the histogram.

I think there's software that allows you to read the correct raw data, but I don't remember the name.

For reproducing color negatives, however, you could take three shots: N, +1, +2. This would save you from having to go back and forth between the camera and the computer to check the histogram.
I think the PS histogram is precise enough to read the MakeTiff linear file.

Just understand that exposure with MakeTiff increases, or, I'd say, doubles, like apertures:
If you're at 1/4 of the histogram, +1EV takes you to the middle. If you're in the middle of the histogram, +1EV takes you to the right.
Your example file requires +2EV.
But this +2EV, in this case, only applies to the reproduction of the color negative.
If your example file, with this MakeTiff histogram, were a normal photograph taken outdoors, then you would be underexposed by -1EV, so you would have to overexpose by +1EV.
And it's only the MakeTiff histogram you need to evaluate, otherwise it will confuse you.

If you try to process the same RAW with Color Perfect, but convert one with MakeTiff and the other with CameraRaw, these two will again look different. The same thing, in my opinion, happens if you exit Lightroom with one of those linear profiles.

The solution is always MakeTiff with its histogram.
Color Perfect is best designed to work with MakeTiff; it's true that it works wonders with other TIFFs, but nothing beats MakeTiff.
That's great advice Roberto!

It is my early days of shooting with Linear Tiffs in mind, you of course have been doing that a lot longer, as you convert your Raws through the powerful PerfectRAW, which requires Linear files, instead of going to the usual Adobe solutions. In the process of making these softwares more user friendly, they stop showing the users crucial information, such as the ones we are talking about. By using a regular Histogram, depending on the Profile, Camera Brand, and even Photo Editor, you get different interpretations of the same Raw file, where in one place you would think you have blown your highlights, while in other you have headroom to spare.

This process has been confusing, but at the same time, very enriching to better grasp how digital files work beyond what histogram show us!

You give several ideas. Trying different exposures, and trying to correlate the in Camera Histogram to Linear is an interesting one, although this could be affected by several factors. But this would give a rough idea at least to work with.

About the bracketing exposures, it's something I definitely will be doing when shooting digitally. But for Film Scanning, this would probably get out of hand quickly :) 36 Frames to Scan, and as I use Pixel Shift to get better Signal to Noise Ratio, and Color Reproduction, each frame consists of 8 Pictures taken. So instead of 8X36 for a roll, we would be doing 3X(8X36), which would take longer, a ton of disk space etc. That's why I much prefer to find the right exposure first, and then I can just shoot away the roll and be done with it quickly!

I am gonna be doing more experimentations, I have already disassembled my Stand unfortunately, and it takes a bit of time to get everything to match correctly, but as soon as I do, more experimentations for sure!
You can check my photos on Flickr:
https://www.flickr.com/photos/alec246_photography/
User avatar
AlexisMagni
ColorPerfect User
Posts: 44
Joined: Wed Aug 27, 2025 7:24 am

In case of interest, I seem to have found an application that lets you inspect Raw Files in every way you could dream of, including Linear Histogram :)

https://www.rawdigger.com

There is a 30 Day Trial. From my comparison so far, It seem to match very well the Linear Tiff by MakeTiff after inspecting the Histogram on PS. It does however offer a lot more visualization options, including detailed max points for each color channel, which is a huge help. What I have yet to understand is what is considered to be Clipping in a 14Bit Raw File. It seem to be around 16000, but the file that have the Clipped Angel Wings I posted on the other section, and I know for a fact it is completely clipped, has a max point on every channel at 15375, which is lower than what theorically 14 bits should fit, which is 16,385 tones for channel. Maybe this is some headroom that depends on camera model?
You can check my photos on Flickr:
https://www.flickr.com/photos/alec246_photography/
User avatar
robyferrero
ColorPerfect User
Posts: 151
Joined: Wed Aug 20, 2025 4:12 pm
Location: Italia

AlexisMagni wrote: Mon Sep 08, 2025 1:23 pm About the bracketing exposures, it's something I definitely will be doing when shooting digitally. But for Film Scanning, this would probably get out of hand quickly :) 36 Frames to Scan, and as I use Pixel Shift to get better Signal to Noise Ratio, and Color Reproduction, each frame consists of 8 Pictures taken. So instead of 8X36 for a roll, we would be doing 3X(8X36), which would take longer, a ton of disk space etc. That's why I much prefer to find the right exposure first, and then I can just shoot away the roll and be done with it quickly!
Perhaps it would be worthwhile to have a low-quality, low-cost flatbed scanner.
Just to have all 36-38 frames in your digital archive.
Then you can make reproductions from color negatives with three different exposures only on the frames chosen for the portfolio.
User avatar
robyferrero
ColorPerfect User
Posts: 151
Joined: Wed Aug 20, 2025 4:12 pm
Location: Italia

AlexisMagni wrote: Mon Sep 08, 2025 2:03 pm In case of interest, I seem to have found an application that lets you inspect Raw Files in every way you could dream of, including Linear Histogram :)

https://www.rawdigger.com
There it is, RawDigger.
C.Oldendorf
Developer
Posts: 167
Joined: Fri Sep 02, 2022 10:31 am
Contact:

ColorPerfect_Headroom_Alexis.jpg
ColorPerfect_Headroom_Alexis.jpg (534.41 KiB) Viewed 678 times
[Full image link - opens in new tab]

I’m a bit puzzled by the complexity you see in getting exposure right for color negative repro shots. Once my rig is set up, I usually just do one of two things:

a) Since I found that on the Nikon Zf the internal metering doesn’t quite fill the sensor, I set exposure compensation to +2 EV and let the camera handle it.

b) Or I go fully manual: set up the light, pick the aperture I want for quality reasons (not too open, not so far stopped down that diffraction kicks in — usually f/5.6 or f/8), and start with an exposure that’s a bit on the short side. Then I take a shot of the space between two frames so the film base is included. There can’t be anything brighter on that roll than the base. I look at ColorPerfect’s headroom output for that test, adjust exposure to fill the headroom, and then shoot the whole roll with those settings. Different films are close enough in base density that if you stay about 1/3 EV short of actual saturation, the same setup works across a bunch of rolls too. You get the hang of it quickly.

In PS we'd see the same, roughly plus 2EV:
Sensor_histogram.jpg
Sensor_histogram.jpg (38.4 KiB) Viewed 678 times
[Full image link - opens in new tab]

Let's revisit this once you shot using the full sensor data range. We will retain film grain in the sky but 2 EV should do something.
coolscan_zf.jpg
coolscan_zf.jpg (739.69 KiB) Viewed 676 times
[Full image link - opens in new tab]

User avatar
AlexisMagni
ColorPerfect User
Posts: 44
Joined: Wed Aug 27, 2025 7:24 am

You know when you have been taught something your entire life, and a better understanding leads to make sense breaking that rule, and doing differently? It is how I feel about watching my Camera Histogram showing Clipping Data, and continue pushing because the Linear Sensor Data still have more than a stop of headroom :D It just feels wrong... I am sure this will go away as I get used to this practice!

Now, your CP 3.0 Headroom indication is very valuable! I didn't know that existed until you mentioned just recently, and will make this process a lot easier.

You also have a Nikon Zf if I understand correctly? If you Scan with +2 EV Exp Comp, you get optimal exposure for Scanning after MakeTiff? That would be nice to know. Even though it makes sense to optimize exposure for each frame, I think I prefer to set a fixed exposure using the Film Base, maximizing as much as I can from that sensor, and hoping all subsequent photos are using enough of the dynamic range for good quality.

I should setup these next days my scanning setup again, and then I will post the results using the improved exposure!

Thanks!
You can check my photos on Flickr:
https://www.flickr.com/photos/alec246_photography/
User avatar
robyferrero
ColorPerfect User
Posts: 151
Joined: Wed Aug 20, 2025 4:12 pm
Location: Italia

Looking at your histogram, I'd go with the camera's exposure, +2EV compensation.
Since you're testing on a single frame, it doesn't hurt to adjust +1EV and +2EV, but you'll see that +2EV will work fine.
C.Oldendorf
Developer
Posts: 167
Joined: Fri Sep 02, 2022 10:31 am
Contact:

AlexisMagni wrote: Mon Sep 08, 2025 5:57 pm You know when you have been taught something your entire life, and a better understanding leads to make sense breaking that rule, and doing differently? It is how I feel about watching my Camera Histogram showing Clipping Data, and continue pushing because the Linear Sensor Data still have more than a stop of headroom :D It just feels wrong... I am sure this will go away as I get used to this practice!
Yes, I know exactly what you mean. It feels very wrong at first to keep pushing exposure when the camera histogram is already showing clipping. But once you’ve seen how much headroom the raw sensor data still has, you realize the problem isn’t with the camera — it’s with what the camera shows you. The histogram and the blinkies come from the JPEG preview, not from the raw. They’re gamma-encoded, shifted by picture controls and white balance, so they warn you about “JPEG clipping,” not raw clipping.

A true raw histogram would solve this, and projects like Magic Lantern or CHDK proved long ago that it’s perfectly possible by reading the sensor buffer directly. But Nikon never gave us that, not even on their finest bodies. My D3X and D800 are perfect cameras in every way, yet they too only show the JPEG histogram. Your Zf behaves exactly the same. With color negatives it’s even worse: the orange mask throws the JPEG engine off so far that the histogram becomes actively misleading, and if you trust it, you’ll underexpose.

For scanning I go fully manual: expose against the film base, then check the headroom with ColorPerfect. Once that’s set, I can do the whole roll without second-guessing. And the point isn’t to chase that last 1/3 of a stop on the edge — it’s that if we don’t trust the linear data, we throw away two full stops of exposure we could have used. That’s four times as many samples captured in the brightest stops, where all the fine detail and smooth tonality lives. It would be a pity not to take that when the sensor gives it to us so easily.

In “normal” photography you don’t always have this headroom, so it’s good to learn where your raw data really sits. A simple test is to photograph an 18% gray card and then a white wall: the metering will place the gray where it thinks it belongs, traditionally both shots should come out roughly the same, and you can then see in the raw data exactly how far you are from the clipping point. Once you’ve done that experiment, it becomes much easier to trust the numbers instead of the JPEG preview.
User avatar
robyferrero
ColorPerfect User
Posts: 151
Joined: Wed Aug 20, 2025 4:12 pm
Location: Italia

Even though the file wasn't exposed correctly, fortunately we have ColorNEG to fix it.

The result still seems acceptable to me.
It may not have all the necessary information, especially in the shadows, but personally, I'd still print a file like this. It would be interesting to see what happens with a correctly exposed file.
Attachments
050925-HarmanPhoenixII-LeicaM2-PixelShift_12-MakeTiff_ColorNEG-E.jpg
050925-HarmanPhoenixII-LeicaM2-PixelShift_12-MakeTiff_ColorNEG-E.jpg (427.5 KiB) Viewed 649 times
[Full image link - opens in new tab]

050925-HarmanPhoenixII-LeicaM2-PixelShift_12-maketiff_colorneg-E.tif
(130.9 MiB) Downloaded 3 times
User avatar
AlexisMagni
ColorPerfect User
Posts: 44
Joined: Wed Aug 27, 2025 7:24 am

C.OLDENDORF wrote: Mon Sep 08, 2025 6:25 pm
A true raw histogram would solve this, and projects like Magic Lantern or CHDK proved long ago that it’s perfectly possible by reading the sensor buffer directly. But Nikon never gave us that, not even on their finest bodies. My D3X and D800 are perfect cameras in every way, yet they too only show the JPEG histogram. Your Zf behaves exactly the same. With color negatives it’s even worse: the orange mask throws the JPEG engine off so far that the histogram becomes actively misleading, and if you trust it, you’ll underexpose.
That was what I Was missing to understand this whole thing, and it just didn't make sense to my mind. Now, after going though a lot of articles i found online, I think I have a much better understanding of how a digital sensor works, and how manufacturers will interpret that data to the end user! This even opened new ideas for Digital Photography as well, as I will now take care to fully use the range of bits whenever possible to have the best image quality with the least amount of noise. So this journey really paid off! Appreciate all the information!

robyferrero wrote: Tue Sep 09, 2025 12:40 pm Even though the file wasn't exposed correctly, fortunately we have ColorNEG to fix it.

The result still seems acceptable to me.
It may not have all the necessary information, especially in the shadows, but personally, I'd still print a file like this. It would be interesting to see what happens with a correctly exposed file.
That looks beautiful! I have to confess, it was my first time shooting this new Harman Phoenix II Color Film, and probably the last. I didn't like how this new stock behaves overall.

This is among one of the few shots that I believe I can appreciate the colors. It will have different cast depending on the type of illumination, time of day, etc. It will very easily show halation, and with that, you loose all the colors, it all gets muted, and fast. Very unpredictable.

Next time I Scan these with the correct exposure settings I learned, I Will also give other color neg I like so we can compare the results too.
You can check my photos on Flickr:
https://www.flickr.com/photos/alec246_photography/
User avatar
robyferrero
ColorPerfect User
Posts: 151
Joined: Wed Aug 20, 2025 4:12 pm
Location: Italia

AlexisMagni wrote: Tue Sep 09, 2025 5:39 pm That looks beautiful! I have to confess, it was my first time shooting this new Harman Phoenix II Color Film, and probably the last. I didn't like how this new stock behaves overall.
It seems like a somewhat ambiguous film to me, too; it certainly doesn't look like Kodak Portra 160.
I don't know if you noticed, but I've also attached the TIFF file so you can view it in detail. I even dusted it off for you, free of charge :D
C.Oldendorf
Developer
Posts: 167
Joined: Fri Sep 02, 2022 10:31 am
Contact:

What threw me off was that blue. I’ll be interested to see whether that changes, or whether it’s part of the digital-camera-to-color-negative math in a way that’s too deeply built in to be mended.
Of course, one can click the Saturation button, set Spread to 10, pick some blue, and desaturate selectively—like in my screenshot at 66%.
These things are tricky. I grab them, assign my monitor’s color profile, then convert to sRGB so we get a rough impression of what I saw. Good enough for reference—but just saying, screenshots can be tricky color-wise.
robyferrero wrote: Tue Sep 09, 2025 5:55 pm I don't know if you noticed, but I've also attached the TIFF file so you can view it in detail. I even dusted it off for you, free of charge :D
I for one did not notice. Love it. So I'll do what I did on the Tiff ;) Edit: 75% turned out better on the Tiff I think.
With that I could live better in total I suppose. Looking at it from some distance, it's not half bad 8-)
Attachments
Blue_Sky_to_Saturation_75.CPMetaT
(2.93 KiB) Downloaded 21 times
C.Oldendorf
Developer
Posts: 167
Joined: Fri Sep 02, 2022 10:31 am
Contact:

robyferrero wrote: Tue Sep 09, 2025 5:55 pm It seems like a somewhat ambiguous film to me, too; it certainly doesn't look like Kodak Portra 160.
It sure is not...

HARMAN Phoenix II 200
HARMAN Phoenix II 200.jpg
HARMAN Phoenix II 200.jpg (47.29 KiB) Viewed 668 times
[Full image link - opens in new tab]

Kodak Portra 160 (2011)
Portra 160 (2011).jpg
Portra 160 (2011).jpg (109.99 KiB) Viewed 668 times
[Full image link - opens in new tab]

User avatar
robyferrero
ColorPerfect User
Posts: 151
Joined: Wed Aug 20, 2025 4:12 pm
Location: Italia

Actually, that blue bothers me a bit, too.
It can be toned down, as you said, by making local adjustments.
If we're talking about normal, general adjustments across the entire image, if we make the blue a little less electric, more yellow, the ground, which is already yellow, and even the wall, which is nice and neutral, becomes too yellow, and unpleasant from my point of view. Desaturation could certainly be enough, but I still find that blue electric. So for me, the solution is to make the sky yellow, but only the sky, and maybe a little desaturation for those dark, electric corners.
C.Oldendorf
Developer
Posts: 167
Joined: Fri Sep 02, 2022 10:31 am
Contact:

robyferrero wrote: Tue Sep 09, 2025 6:10 pm If we're talking about normal, general adjustments across the entire image, if we make the blue a little less electric, more yellow, the ground, which is already yellow, and even the wall, which is nice and neutral, becomes too yellow, and unpleasant from my point of view. Desaturation could certainly be enough, but I still find that blue electric. So for me, the solution is to make the sky yellow, but only the sky, and maybe a little desaturation for those dark, electric corners.
Well, why not. Like you said, at full resolution (edited to be oriented right, now that we've been told): What did I do on your TIFF?

In Photoshop: from the Select menu choose Color Range. Use the +Eyedropper, click and hold while dragging over the blue until we have it. Feather the selection by about 2px.
color_range.jpg
color_range.jpg (288.59 KiB) Viewed 667 times
[Full image link - opens in new tab]

In ColorPerfect: use Selection, and on the inner part of it click into a CC box and add +15Y (if you know you want +15Y). Otherwise, ring around until it looks right. Okay, out and back in Photoshop: Layer → Create New via Copy.
Then select a luminance mask (Windows shortcut: Ctrl+Alt+Shift+3).
luminance_mask.jpg
luminance_mask.jpg (344.72 KiB) Viewed 667 times
[Full image link - opens in new tab]

That means white is fully selected, black is not selected, and all brightness levels in between are weighted relatively.

Now apply ColorPerfect on that layer with that mask: This allows the darker blues to be desaturated more than the brighter blues. Faltten the image layers afterwards and we’re done.

Doing it takes about two or three minutes; documenting what’s done takes slightly longer. :D
User avatar
robyferrero
ColorPerfect User
Posts: 151
Joined: Wed Aug 20, 2025 4:12 pm
Location: Italia

Well, these are things I do more or less the same way in PS.
Last time, which was the other day, when you were showing the SmartClip selections, I tried making the selection and calling Color Perfect, but PS crashed.
I thought I was making some mistake, but then, doing something else, I forgot to tell you about it.
C.Oldendorf
Developer
Posts: 167
Joined: Fri Sep 02, 2022 10:31 am
Contact:

robyferrero wrote: Tue Sep 09, 2025 6:54 pm Last time, which was the other day, when you were showing the SmartClip selections, I tried making the selection and calling Color Perfect, but PS crashed.
It may be the one known but rare crash on macOs I'm still chasing. For crashes the logs you can display afterwards as text are super helpful. Anything that crashes is a must fix.
Post Reply

Return to “Samples and Challenges - Let Others Play With Your Negatives and Vice Versa”