An alternative to Negative Lab Pro and Lr has to exist (C-41 reversal and orange mask removal)?!

[Nelari]

I don't use the PL inline help, but i have just checked on my system, and it works, ie. clicking on top menu bar "?" > "Hilfthemen" , but shortkey call F1 doesn't (could be configurable in Wine settings).

When I click '?' in the menu or hit the F1 key, what comes up is a window with two columns, just like what you show. Oddly, the left-hand one (navigation) works as it should, but the right-hand one stays blank. That is, no help text appears. This is not helpful.

OK, I sent an email to PhotoLine support.

 

[removedacct2]

When I click '?' in the menu or hit the F1 key, what comes up is a window with two columns, just like what you show. Oddly, the left-hand one (navigation) works as it should, but the right-hand one stays blank. That is, no help text appears. This is not helpful.

OK, I sent an email to PhotoLine support.

it may be that in your WINE settings, the HTML rendering engine used by Microsoft to display help books is missing. The name of this wine component is "wine-gecko". Typically it is downloaded and installed the first time that "winecfg" is run: a small window pops asking authorization (Wine Gecko Installer) , same for the other component called "Wine Mono". I guess this is automatic in main distributions like Debian. What is your Linux flavor?

 

[Nelari]

OK, I sent an email to PhotoLine support.

The answer was that the "help system is something Windows specific":

It is possible, that Wine is not able to display this correct.
You can use the PDF manual from
www.pl32.com/plbin/manual.zip

But this is contradicted by @antonio_b who showed us a screencap with the help system working, in fact, under Wine.

This is odd. Does it have something to do with language? @antonio_b seemed to be using German.

 

[removedacct2]

When I click '?' in the menu or hit the F1 key, what comes up is a window with two columns, just like what you show. Oddly, the left-hand one (navigation) works as it should, but the right-hand one stays blank. That is, no help text appears. This is not helpful.

also it may give debugging hints to run the help from the command line.
In a default installation, the PhotoLine help files are in
$HOME (ie. your home directory typically /home/"you"/) /.wine/drive_c/Program Files/PhotoLine/hlp/
the files are: PhotoLine.chm and subdirectories for languages (Chinese, Deustch, English, Francais, Italian).
you can run the help book with the WINE hh.exe command like this:

wine /home/<you>/.wine/drive_c/windows/hh.exe PhotoLine.chm

 

[removedacct2]

The answer was that the "help system is something Windows specific":

But this is contradicted by @antonio_b who showed us a screencap with the help system working, in fact, under Wine.

This is odd. Does it have something to do with language? @antonio_b seemed to be using German.

yes PL support ie correct the help system is Windows specific, and as told in my previous comment, in WINE it is implemented by the "wine-gecko" package, which may be missing on your Linux.... That may be the reason.

 

[removedacct2]

Does it have something to do with language? @antonio_b seemed to be using German.

yes I use Photoline in its native german, but when I switch to english, help system works equally well. So again, it may be an incomplete WINE installation.

 

[Tim Gray]

Interestingly, while playing with DSLR scanning and stitching large images together with Affinity, I discovered that the "Auto Colors" option under Affinity did a pretty accurate job of fixing colors, either pre- or post- inversion, at least on Portra 400 film.

I’m fully convinced that most of these programs are just doing variations of auto color. It works surprisingly well, and if you have a full range negative, it makes sense if you think about it.

 

[Nelari]

it may be that in your WINE settings, the HTML rendering engine used by Microsoft to display help books is missing. The name of this wine component is "wine-gecko". Typically it is downloaded and installed the first time that "winecfg" is run: a small window pops asking authorization (Wine Gecko Installer) , same for the other component called "Wine Mono". I guess this is automatic in main distributions like Debian. What is your Linux flavor?

The flavor is Ubuntu.

I checked that the PhotoLine help files are installed. And there is a gecko directory in both windows/system32 and windows/syswow64.

I ran winecfg just now but didn't see anything related to gecko. I can't remember being ever asked whether to install it.

 

[urnem57]

....sheesh....I’m glad I have a Mac......way too complicated.....lol.....carry on......

 

[urnem57]

No doubt. As a famous sound engineer once said, “Understand what’s going on, don’t just be a knob twirler” Joel Grimes just put out a video on his process to convert color files to spectacular b/w. It’s on YT. But it’s like anything with photography, there really isn’t only one right way to do things.

 

[Nelari]

it may be that in your WINE settings, the HTML rendering engine used by Microsoft to display help books is missing. The name of this wine component is "wine-gecko". Typically it is downloaded and installed the first time that "winecfg" is run: a small window pops asking authorization (Wine Gecko Installer) , same for the other component called "Wine Mono".

I had another reply from PhotoLine support. I was told to do the following:

wget http://dl.winehq.org/wine/wine-gecko/2.47.1/wine-gecko-2.47.1-x86_64.msi
wine msiexec /i wine-gecko-2.47.1-x86_64.msi

This is way over my head, but it has to do with Gecko, right? Anyway, it helped, and PhotoLine's help now displays correctly. That's just as well, because I really need it.

On the other hand, i had the wild idea to try this on a Mac computer, too. But here, I can't figure out how to tell PhotoLine where the ColorPerfect plugin resides. Oh well. Macs are different.

 

[removedacct2]

I had another reply from PhotoLine support. I was told to do the following:
wget http://dl.winehq.org/wine/wine-gecko/2.47.1/wine-gecko-2.47.1-x86_64.msi
wine msiexec /i wine-gecko-2.47.1-x86_64.msi
This is way over my head, but it has to do with Gecko, right? Anyway, it helped, and PhotoLine's help now displays correctly. That's just as well, because I really need it.

yes that's it: the wine-gecko component, which as said earlier is an HTML engine used by Windows for displaying the helps. It could had been installed directly from the package/software management of your Linux distribution also. What Linux flavour do you use?

that said, the help system is only for Photoline itself. ColorPerfect comes with its own help (you clik "help" top right, the main pane becomes a text display, you click or activate whatever button or menu of the right pane and explanation displays in main one.). Otherwise the videos are well done: https://www.youtube.com/c/Colorneg/videos

On the other hand, i had the wild idea to try this on a Mac computer, too. But here, I can't figure out how to tell PhotoLine where the ColorPerfect plugin resides. Oh well. Macs are different.

I don˝t run MacOSX by now, but what I recall is that on Windows one has to look for the ColorPerfect.8bf file, but on MacOSX for ColorPerfect.plugin. Otherwise from CP site (https://www.colorperfect.com/ColorPerfect/ReadMe-file-for-Intel-Macs/)
Open the ColorPerfectIntelMac2.25.dmg file in a new Finder window by double clicking it. Locate the two plug-in files ColorPerfect.plugin and ColorPerfect64.plugin.
The two plug-ins will need to be copied into one of PhotoLine's Plug-ins folders. It is best to open another Finder window to locate the desired folder and to place the plug-ins by drag and drop. The Plug-ins folder is any folder of your own choosing which you have already declared and activated or will declare and activate using the "Manage Plug-ins" entry in PhotoLine's Filter menu.

 

[Tim Gray]

I am fairly certain that NLP and Negmaster do not. Their authors have explained the mechanics on a couple of occasions. C41 films have a well-known and documented sensitivity offsets for their YCM layers, trying to "guess" color balance via auto-colors would be stupid.

Not here to get in an argument, but 'auto-colors' has several different options, and the one where you maximize contrast on individual RGB channels, or on darkest/lightest colors actually do a very good job (I prefer find dark & light colors by default, but will use the per channel setting sometimes). Yes, tweaks are required afterwards.

The 'well documented offsets' is taken care of in the raw developer when you white balance on the rebate. NLP and Negmaster actually don't do anything at the step - Lightroom and ACR do the work unless I'm mistaken.

After one negates the orange mask in linear raw space (white balance on the rebate), you have an image where the lightest color in the picture is the rebate. That is nominally black - one end of your auto colors. If you have a 'full range' negative, as I stated in the original post, you probably have something close to a true white highlight in the original scene. This is your nominal white - the other end of your auto colors. When you use auto colors (in PS with 'Enhance per channel contrast') your are setting each RGB curve to go from the nominal black (which is neutral because you white balanced on it) to the nominal white. This has the effect of also setting the RGB channel slope (gamma) to an approximately correct level. It really, shockingly, does give pretty good results on FULL RANGE negatives that have had their base white balanced away in a raw tool. If you don't have a full range negative, the 'variations on auto color' algorithms can't set the end points correctly, and you get funny results like the snow scene image posted somewhere in this thread.

This doesn't mean there aren't tweaks required, or that NLP or Negmaster are worthless. I bought Negmaster - it's a nice tool and it looks like the author put in a bunch of work. I'm sure I'll pick up NLP as well. NLP is the same - there are a lot of nice features that they added in and worked on.

From reading what Adrian is doing, I do think he is taking a different approach and using preset gamma curves and CMY layer offsets instead of an 'auto' calculation of color range. Fix the black point and the slope, and let the highlights end up where they will instead of set the two endpoints, and derive the slope. And a lot of meticulous color space math.

Those tools (NLP in particular) do have a lot of functionality after the mask negation (done in raw editor), inversion, and the basic color correction. Those are useful tools for but in my mind are secondary to the first pass 'default' inversion and color correction. And certainly there can be emulsion specific tweaks that one can bake in due to the behavior of specific films. I find Portra 400 needs a lot less work than Ektar or Portra 800.

But hey, just my opinion.

 

[Tim Gray]

To illustrate how effective auto colors can be, here's an example.

1. Sony A7rII scan of negative in ACR
2. Scan white balanced on film base in ACR
3. Image in PS with an Invert adjustment layer
4. Curves layer with where 'auto' has been run (Find Dark & Light colors, no 'snap neutral midtones'

After that, I usually apply slight preset contrast tweak with a level layer by moving the midpoint slider. I also have some layers (not active) to adjust midtone white balance and a slight S shape contrast curve, but those are best after adjusting the auto-color adjustment if needed, which as stated above, is needed for more difficult films or shots.

 

[Tim Gray]

Three more images.

1. I am not an NLP pro - here's the default NLP conversion (from the demo) using 'LAB standard' tone setting, Auto-Neutral WB and no other adjustments.
2. Negmaster inversion. Default inversion with NM ACR profile applied.
3. Negmaster ACR profile, PS inversion/color balance with auto colors.

As I also stated prior, the NM profile does have some tweaks baked in which in my opinion to help achieve less color crossover in the highlights than using no custom profile in ACR. You can see it when you apply it in ACR - the highlights get bluer. I'm sure one could recreate similar adjustments, but probably worth the money right there. The NM profile works fine with your own PS inversion/color balance routines.

 

[urnem57]

Agreed. My experience so far is showing me that there is a big savings in time to at least get from the raw camera scan file to a starting point. I am not going to work where the sliders and editing adjustments are reversed.

 

[Nelari]

yes that's it: the wine-gecko component, which as said earlier is an HTML engine used by Windows for displaying the helps. It could had been installed directly from the package/software management of your Linux distribution also. What Linux flavour do you use?

I use Ubuntu. Now, with the help issue solved, I got so far as to shoot a roll of Fuji Pro 400H and develop it. This was my first time developing color negative film, and I approached it with some trepidation. But, well, it was no big deal with the Cinestill Cs41 pack. The negatives look OK, but the positives I get when converting with ColorPerfect/VueScan/RawTherapee do not. I wish I knew whether it's the developing or the converting that is the problem. Of course, I get terrible colors almost every time I try to convert a negative...

(I developed the film myself because the local shops that still offer this service are getting fewer and fewer, and farther and farther.)

 

[removedacct2]

I use Ubuntu. Now, with the help issue solved, I got so far as to shoot a roll of Fuji Pro 400H and develop it. This was my first time developing color negative film, and I approached it with some trepidation. But, well, it was no big deal with the Cinestill Cs41 pack. The negatives look OK, but the positives I get when converting with ColorPerfect/VueScan/RawTherapee do not. I wish I knew whether it's the developing or the converting that is the problem. Of course, I get terrible colors almost every time I try to convert a negative...

(I developed the film myself because the local shops that still offer this service are getting fewer and fewer, and farther and farther.)

I develop everything myself since long, only four or five places in whole Norway doing it and postal fees alone are just not worth it.
I always get good results (not final, this may need more work) unless i use stretched developer (over the limit of recommended sessions) in which case there's overall color shift because the depletion/exhaustion.

i have shot only little Fuji 400H but it went ok each time.
In Vuescan if "source" is set as "color negative" bits per pixel to "48 bit RGB" , and "Color" as "Auto" or "White balance" or "Neutral", typically you will see an ok positive rendering, not at all optimal/final in many cases, will need adjustments, but not terrible.

 

[Tim Gray]

I'm confused here, probably because I know Lightroom pretty well, but not ACR or Photoshop (where I am a true noob). If you do the actions in Photoshop, isn't that where the inversion takes place? So why is ACR involved?
Do you have a link for the source for any of these actions?

Phil, I promised you I'd package up my actions and write up my process. I make no claims to be a professional at PS, action writing, photography, or even writing, so no promises are being made. In fact, I can't even vouch for the quality of the work I do for a living, which is very far away from photography While I think most of what I do in ACR/PS isn't anything fancy for B&W negatives, it's similar to the concept of operations for color negatives, hence the two stage process.

Actions are here: https://125px.com/articles/photography/digital/b-w-film-scan-processing/BWactions.zip

Write up is here: https://125px.com/articles/photography/digital/b-w-film-scan-processing/

 

[warden]

Phil, I promised you I'd package up my actions and write up my process. I make no claims to be a professional at PS, action writing, photography, or even writing, so no promises are being made. In fact, I can't even vouch for the quality of the work I do for a living, which is very far away from photography While I think most of what I do in ACR/PS isn't anything fancy for B&W negatives, it's similar to the concept of operations for color negatives, hence the two stage process.

Actions are here: https://125px.com/articles/photography/digital/b-w-film-scan-processing/BWactions.zip

Write up is here: https://125px.com/articles/photography/digital/b-w-film-scan-processing/

Thanks for that Tim!

 

[PhilBurton]

Thanks for that Tim!

Me too. Thanks!

 

[Tim Gray]

No problem. I write this stuff up so I remember what I was doing

Updated the actions to remove a dependency that wasn't included, and added an alternate action that adds a bigger layer mask for landscape/portrait 35mm images with borders.

 

[pwadoc]

They do not. Two parallel lines will remain parallel in another coordinate system. If the gamut is big enough, you are not "deforming" anything, you are simply changing labels on X/Y/Z. If I measure your height in inches vs centimeters you're not going to become taller or shorter. And if I do this in 3D, the relationship between your arms and legs will not change just because their coordinates are in inches now.

That's exactly what colorspace transforms are: rotating (if the image fits within both) or rotating + scaling down (when it does not fit into the target space).

On your visualization, just pretend that the size of the space (gamut) is bigger than your monitor can show, then moving your sliders will not move a single pixel. And this is indeed what you will see by converting a photo from one wide colorspace to another, while staring at it on a sRGB monitor.

What Adrian is saying, is that there's another transformation taking place which is not a trivial matrix multiplication, and I will be reading up on that.

So having thought about this for a while, you're totally correct that color space transformation does not deform the space, which should mean that the curves should remain parallel, which is kind of confusing. I actually did a test to confirm what I've observed, that the curves do not actually remain parallel when scanned. I took the reference test strip from my C-41 process control strips and scanned it, and without any other processing, white balanced on the strip. You can see the result below:

Here's what happens if I white balance on the film base:

Here's what happens if I white balance on medium grey:

It's pretty clear in this example. Whatever point you white balance on throws the rest of the image off in ways that no linear adjustment can correct. You need a curve to correct this issue. You could derive that curve by sampling each step on the wedge and getting a set of points which would bring the entire image into alignment, which is I believe the process that Adrian Bacon has described.

This has been driving me a little nuts, because as you pointed out, if the response curves are parallel on the film, they should be parallel when you scan them as well, but they clearly aren't. Additionally, using RGB lights to scan fixes this problem. My setup is undergoing surgery right now so I don't have an example image to upload, but if you scan that same step wedge with RGB lights, composite the channels and white balance, the entire thing is grey. So something is going on here that I'm missing, and I was totally baffled. Earlier today however, I came across this document which I believe explains what's going on. The response curves of film are actually only linear in certain bands. In certain parts of the wavelength, they go nonlinear. This isn't a problem in RA-4 printing, because RA-4 paper isn't sensitive in the bands where this nonlinearity occurs, the paper is actually sensitive only in pretty narrow bands for each channel. When you scan a negative with white light, your camera is picking up the entire spectrum, including the areas where the negative channels are non-linear, and that is what causes the problem above.

 

[removedacct2]

shoot a roll of Fuji Pro 400H and develop it. This was my first time developing color negative film, and I approached it with some trepidation. But, well, it was no big deal with the Cinestill Cs41 pack. The negatives look OK, but the positives I get when converting with ColorPerfect/VueScan/RawTherapee do not. I wish I knew whether it's the developing or the converting that is the problem. Of course, I get terrible colors almost every time I try to convert a negative...

well, this would be a fork into another thread.
You can scan one of the frames as "image" in "raw" format, save it as DNG and load it to some free site, so we could have a look at it. I can run it into Rawtherapee, NLP on LR, CP on PL, etc.
Set Vuescan as follow:

input tab: Media as "image" bits/pixel "48 bit RGB", resolution whatever you want, not too low though (here is just my Reflecta ProScan 10 set at 5000dpi)

the Filter tab: don't set "restore colors" "restore fading" no Grain reduction, no sharpening, so Vuescan won't touch colours at all. Infrared cleaning you can use some it's just for any dust, but with freshly processed film kept safe in sleeves there will be little need, anyway it won't affect colours.

in the Color tab: no color balance, no film type it's useless on Vuescan so just leave it as default "generic". In that shot there's an ICC profile loaded from calibrating my scanner, but you don't need either, just let the field default.

in Output tab: set the folders for output, chose "raw file" "48 bit RGB" and "raw DNG format"

when you do a preview with these settings, the main window will show indeed a transparency scan of the negative itself, not a positive rendering:

on the "Media" menu of the "input" tab, if you change the media type from "Image" to "color negative", the preview will change to the positive rendering done by Vuescan. It gives an idea of what it could be:

but in order to debug your issue, just do the Scan with "Media" set as "image", ie. the former "negative" preview.
With the Proscan 10T @5000dpi this 35mm color frame gives a ~180Mb raw file.
Load the file to some free hosting site and give us the link.

 

[Tim Gray]

This isn't a problem in RA-4 printing, because RA-4 paper isn't sensitive in the bands where this nonlinearity occurs, the paper is actually sensitive only in pretty narrow bands for each channel. When you scan a negative with white light, your camera is picking up the entire spectrum, including the areas where the negative channels are non-linear, and that is what causes the problem above.

Not knowing what software you are using or how you are white balancing, this could be a linear color space thing. A lot of photo software is just not set up to operate properly in linear space. It really creates a lot of problems for this type of manipulation. RAW developers do, which is why they are effective at negating the film base. You might be able to trick PS into working correctly, but I'm not sure.

Or I could be totally wrong