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

[runswithsizzers]

I do option number 3, which isn't really something you can do unless you write some code, but I take the raw scanned image and turn it into a floating point positive image that is in the native Adobe LR DNG format. From there, you can simply import it into LR and treat it just like any other raw file from a digital camera. Bit of a best of both worlds, if you will. No need to keep multiple versions around if you don't want to. All of your edits and everything goes into the DNG and stays with it.

Aside from the basic premise - that I should learn how to write software code (not going to happen) - I think I am seeing a workflow advantage to option #3.

I've been looking at file sizes associated with both Negative Lab Pro (NLP) and ColorPerfect (CP), as well as the DNG files made with Adrian Bacon's proprietary software.

If I photograph a 35mm negative with my Fuji X-T1 (16MB, APS-C sensor) I get a RAW file of about 33-34MB, and resolution of 4896 × 3264 pixels.

If I import my 34MB RAW file into LIghtroom and process it with NLP, the file size doesn't change - until I want to convert it to a Positive TIFF. Which I might want to do, so I can conveniently use Lightroom tools for post-processing (as opposed to the limited tool set provided by the NLP plugin). Creating a Positive TIFF with NLP results in a 92MB copy in addition to the 34MB RAW file - for a total 126MB per image.

The ColorPerfect workflow wants me to use a utility called "Make TIFF" to convert my Fuji RAW files to linear 16-bit TIFFs, which are then processed by the ColorPerfect plugin in Photoshop. In the process, my 34MB RAW files become 92-98MB TIFFs. Unlike NLP, which results in both the RAW file and the TIFF being stored in Lightroom, Workflow #2 using CP results in only the TIFF as the working file to be exported from Photoshop and stored in Lightroom; I can archive the RAW elsewhere, or not, as I choose.

Option 3 is to have Simple Photography Services (Adrian Bacon in Petaluma, California) process my negatives and make "scans" using his proprietary software. The DNG files I got back from Simple Photography Services are 44-59MB each, with a resolution of 11,700 x 7,800 pixels (EDIT: that can't be right; something wrong here?). The DNG files are positive RAW files which can live in Lightroom, and benefit from Lightroom's non-destructive workflow, just like my Fuji RAW files. And no need to buy or install additional plugins. (Installing ColorPerfect requires downloading 4 different software items from 3 different websites.)

Letting Simple Photography Services (SPS) do the scans and color inversion is the most attractive workflow, and gives me more resolution and smaller file sizes - but what about results? I am still in the process of evaluating results. My initial impression is that NLP and SPS both do a good enough job of getting close to where I want my images to be, but some manual color corrections are still needed. As for ColorPerfect, I am still struggling with the interface. It remains to be seen if I will ever understand how to use the plugin well enough to produce samples for comparison.

 

[dkonigs]

Aside from the basic premise - that I should learn how to write software code (not going to happen) - I think I am seeing a workflow advantage to option #3.

Option 3 is to have Simple Photography Services (Adrian Bacon in Petaluma, California) process my negatives and make "scans" using his proprietary software. The DNG files I got back from Simple Photography Services are 44-59MB each, with a resolution of 11,700 x 7,800 pixels. The DNG files are positive RAW files which can live in Lightroom, and benefit from Lightroom's non-destructive workflow, just like my Fuji RAW files.

I would love this option, if I could actually run that software myself. TIFF files always seem excessively enormous (even versus DNG or any other raw format), and this would routine of stuff just seems excessive. In an ideal world, I'd just take the TIFFs from my scanner and feed them through some batch-processing program to get inverted DNG output. I'd then just pull those directly into Lightroom and throw away the TIFFs.

I do actually know how to write software, so it is theoretically a project I could undertake. However, I don't really have any meaningful image processing experience. The problem here is that I'd have to figure out how to duplicate all the work SPS/NLP/CP have put into understanding image processing/inversion/colors, and the only thing I'd really add to it was changes to the input/output sides to better fit my use cases.

 

[Adrian Bacon]

Aside from the basic premise - that I should learn how to write software code (not going to happen) - I think I am seeing a workflow advantage to option #3.

I've been looking at file sizes associated with both Negative Lab Pro (NLP) and ColorPerfect (CP), as well as the DNG files made with Adrian Bacon's proprietary software.

If I photograph a 35mm negative with my Fuji X-T1 (16MB, APS-C sensor) I get a RAW file of about 33-34MB, and resolution of 4896 × 3264 pixels.

If I import my 34MB RAW file into LIghtroom and process it with NLP, the file size doesn't change - until I want to convert it to a Positive TIFF. Which I might want to do, so I can conveniently use Lightroom tools for post-processing (as opposed to the limited tool set provided by the NLP plugin). Creating a Positive TIFF with NLP results in a 92MB copy in addition to the 34MB RAW file - for a total 126MB per image.

The ColorPerfect workflow wants me to use a utility called "Make TIFF" to convert my Fuji RAW files to linear 16-bit TIFFs, which are then processed by the ColorPerfect plugin in Photoshop. In the process, my 34MB RAW files become 92-98MB TIFFs. Unlike NLP, which results in both the RAW file and the TIFF being stored in Lightroom, Workflow #2 using CP results in only the TIFF as the working file to be exported from Photoshop and stored in Lightroom; I can archive the RAW elsewhere, or not, as I choose.

Option 3 is to have Simple Photography Services (Adrian Bacon in Petaluma, California) process my negatives and make "scans" using his proprietary software. The DNG files I got back from Simple Photography Services are 44-59MB each, with a resolution of 11,700 x 7,800 pixels. The DNG files are positive RAW files which can live in Lightroom, and benefit from Lightroom's non-destructive workflow, just like my Fuji RAW files. And no need to buy or install additional plugins. (Installing ColorPerfect requires downloading 4 different software items from 3 different websites.)

Letting Simple Photography Services (SPS) do the scans and color inversion is the most attractive workflow, and gives me more resolution and smaller file sizes - but what about results? I am still in the process of evaluating results. My initial impression is that NLP and SPS both do a good enough job of getting close to where I want my images to be, but some manual color corrections are still needed. As for ColorPerfect, I am still struggling with the interface. It remains to be seen if I will ever understand how to use the plugin well enough to produce samples for comparison.

just as a note, I recently updated the hardware I use to use the canon 90D with the Sigma 70mm art macro lens, so scans now have a raw resolution of 30+MP instead of the old ~24MP raw resolution of my old setup. The samples in the DNG are also written as 16 bit floats.

for color a lot of it is pretty subjective. I generally try to deliver a relatively flat and colormetrically neutral image, however if someone wants to send me a lot of film, or has some specific changes in terms of color that they’d prefer, I have very fine grained control over the process and am willing to make custom profiles for users.

 

[GLS]

The ColorPerfect workflow wants me to use a utility called "Make TIFF" to convert my Fuji RAW files to linear 16-bit TIFFs, which are then processed by the ColorPerfect plugin in Photoshop

FWIW, using MakeTiff is not a necessity. I have used the ColorPerfect plugin directly on RAW files processed via either Adobe Camera RAW or Capture One, and the results are excellent.

As for ColorPerfect, I am still struggling with the interface. It remains to be seen if I will ever understand how to use the plugin well enough to produce samples for comparison

I found this instruction video useful for doing the bare minimum of processing to get the basic inversion:



He is using a faw file from a scanner, but the principals are the same.

 

[runswithsizzers]

just as a note, I recently updated the hardware I use to use the canon 90D with the Sigma 70mm art macro lens, so scans now have a raw resolution of 30+MP instead of the old ~24MP raw resolution of my old setup. The samples in the DNG are also written as 16 bit floats.

for color a lot of it is pretty subjective. I generally try to deliver a relatively flat and colormetrically neutral image, however if someone wants to send me a lot of film, or has some specific changes in terms of color that they’d prefer, I have very fine grained control over the process and am willing to make custom profiles for users.

Thanks for the update. I'm looking forward to seeing the results from your new setup when I finish the roll of Portra 160 that is in the camera now.

But I'm confused about the numbers used to describe image resolution. The last DNG files I got back from you have a creation date of Jul 31, 2019. So that would be from your previous "24MP" system, right? When I look at the EXIF data in Lightroom, it shows the dimensions as "12048x8044" pixels cropped to "11445x7630" But if I multiply 11445 x 7630, I get 87,325,350 which is about 87MP, right? Shouldn't a 24 MP image be something more like 6048 x 4024 pixels? What am I missing?

And what are the pixel dimensions and DNG file sizes of your newer "30+MP" files?

 

[Adrian Bacon]

I would love this option, if I could actually run that software myself. TIFF files always seem excessively enormous (even versus DNG or any other raw format), and this would routine of stuff just seems excessive. In an ideal world, I'd just take the TIFFs from my scanner and feed them through some batch-processing program to get inverted DNG output. I'd then just pull those directly into Lightroom and throw away the TIFFs.

I do actually know how to write software, so it is theoretically a project I could undertake. However, I don't really have any meaningful image processing experience. The problem here is that I'd have to figure out how to duplicate all the work SPS/NLP/CP have put into understanding image processing/inversion/colors, and the only thing I'd really add to it was changes to the input/output sides to better fit my use cases.

you’d be amazed at what you could do with a copy of the DNG spec and libtiff, both of which are free, assuming you’re willing to put in the time to read the spec and write the code to use libtiff. Libtiff can write DNG files, since DNG files are actually tiff tiles. The only major difference is instead of having a full rgb for each pixel, the image data is the raw CFA data, and there’s no embedded ICC profile because it’s raw data. There are a couple of tiff tags that are specific to DNG files, but libtiff can write them, so it’s no big deal.

 

[Adrian Bacon]

Thanks for the update. I'm looking forward to seeing the results from your new setup when I finish the roll of Portra 160 that is in the camera now.

But I'm confused about the numbers used to describe image resolution. The last DNG files I got back from you have a creation date of Jul 31, 2019. So that would be from your previous "24MP" system, right? When I look at the EXIF data in Lightroom, it shows the dimensions as "12048x8044" pixels cropped to "11445x7630" But if I multiply 11445 x 7630, I get 87,325,350 which is about 87MP, right? Shouldn't a 24 MP image be something more like 6048 x 4024 pixels? What am I missing?

And what are the pixel dimensions and DNG file sizes of your newer "30+MP" files?

the new setup happened in December 2019, gotta love them specials.

in terms of resolution, what you are seeing in LR is a rendered preview, DNG files let you specify whether to scale the preview up or down, or not at all. I tell it to scale it up 2x, as it makes it easier to spot dust out. When you export it, you should set the dimensions to your desired output resolution. LR will always render an export directly from the raw data. If you don’t specify an output resolution when you export, LR will default to the preview resolution.

the same goes for cropping, LR lets you specify a crop window. I generally have it set to crop a small portion of the image as a safety buffer, as sometimes it’s hard to see the edge of the frame and you end up with a small sliver on the edge. If you want to get that small image area back, just go into the develop module ibmLR, click on cropping, and it’s all there, so you can change it any way you want. I can alternatively set the safety margin to zero upon DNG creation if you ask for that ahead of time. I have a handful of presets in my code to deal with scanning frames that don’t match up to the aspect ratio of the sensor in the scanner. One of the presets is no crop at all.

the new image sizes are 6960x4640 raw, and the DNG files as they would be in the downloaded zip file you get from me are ~65MB each.

 

[runswithsizzers]

FWIW, using MakeTiff is not a necessity. I have used the ColorPerfect plugin directly on RAW files processed via either Adobe Camera RAW or Capture One, and the results are excellent.

I found this instruction video useful for doing the bare minimum of processing to get the basic inversion:



He is using a faw file from a scanner, but the principals are the same.

Thanks for that link. It was helpful for me to see that demonstration about how to use the ColorPerfect plugin.

But, as you say, that video is starts with a scanner file, a 16-bit untagged negative TIFF in Photoshop. As I understand it, the problem for me, starting with a digital camera RAW file, is that Photoshop always opens a RAW file with Adobe Camera Raw, which unavoidably applies a color profile to the image. I thought the reason for ColorPerfect wanting the user to use the Make TIFF utility is to avoid having ACRaw convert the image to a color profile before the ColorPerfect plugin can be opened(?)

I have tried it both ways - starting with a linear TIFF and starting with the RAF file - and so far, I'm not getting very good results from ColorPerfect, either way.

When you start with a camera RAW file, do you make any adustment in ACR before switching to Photoshop and opening the ColorPerfect filter?

 

[GLS]

I thought the reason for ColorPerfect wanting the user to use the Make TIFF utility is to avoid having ACRaw convert the image to a color profile before the ColorPerfect plugin can be opened(?)

I have tried it both ways - starting with a linear TIFF and starting with the RAF file - and so far, I'm not getting very good results from ColorPerfect, either way.

When you start with a camera RAW file, do you make any adustment in ACR before switching to Photoshop and opening the ColorPerfect filter?

I use Lightroom for my image organising and metadata tagging. My film "scans" are done via my D810, shot in RAW (i.e. NEF files). For my C41 frames the only edits I generally perform on the NEF files in Lightroom are as follows:

- Using the white balance tool to pick a part of the blank film, which cancels out the orange mask. I'm not sure if this is entirely necessary, as I have performed tests in ColorPerfect (CP from now on) with and without doing this, and it gives very similar results either way. Perhaps it can tell when this has been done. Either way, it is part of my routine and I stick to it.

- Use the automated chromatic aberration removal tick box.

- Cropping out the extraneous parts of the digital frame, leaving only the actual film frame area (this is important for CP to work correctly).

The resulting processed RAW files are then opened directly in Photoshop, which keeps them in the RAW NEF format with the above non-linear edits applied. As it is still a RAW file, no embedded colour profile therefore exists. You can check this is the case, as when you then invoke CP there should be a colour space drop-down menu in the top right of the dialogue, so you can choose one to work in from that point on. If you are paranoid about this whole process then just choose ProPhoto RGB as a working colour space; it is so huge that I can't see how doing so would affect the processing of CP at all.

 

[runswithsizzers]

[...]
The resulting processed RAW files are then opened directly in Photoshop, which keeps them in the RAW NEF format with the above non-linear edits applied. As it is still a RAW file, no embedded colour profile therefore exists. You can check this is the case, as when you then invoke CP there should be a colour space drop-down menu in the top right of the dialogue, so you can choose one to work in from that point on. If you are paranoid about this whole process then just choose ProPhoto RGB as a working colour space; it is so huge that I can't see how doing so would affect the processing of CP at all.

Not sure if I'm using the right terminology - and I hesitate to discuss something I don't completely understand - but, in an attempt to clarify ...

I believe we are talking about two different kinds of color profiles:
A. First, there is the color space, such as ProPhoto RGB, and Adobe RGB, and sRGB, etc. An image may, or may not, be tagged as being in one of these color spaces - which are designed to help different devices show colors consistantly in a color managed workflow.

B. Then there are color profiles like Adobe Color, Adobe Standard, Adobe Landscape. If you use some other RAW converter, the names (and profiles) may be different, but some kind of intrerpretation of the RAW data is necessary before a RAW image can be displayed on your screen. If a RAW image is displayed in Lightroom's Develop module, or the Adobe Camera RAW module in Photoshop, you will see a list of available "Profiles" (see below). The options on the list will vary, depending on what kind of RAW file it is. For example, my Fuji RAF files can be opened with an "Astia" or "Velvia" profile, but those two are not available for my Pentax DNG files. But, significantly, "No Profile" is not an option - if the user does not choose a color profile from the drop down list, the default profile will be used.

It is not 'paranoia' that is motivating my questions, but rather the desire to test the ColorPerfect software as the developer (Christoph Oldendorf) intended it to be used. In the Read Me file for the Make TIFF software, Oldendorf says,
"Please understand that we dislike what Adobe Camera Raw (ACR) et al. do to color even more than you do. Rest assured that we do not let Adobe DNG Converter touch the color."

And on <his website> he says,
"Our auxiliary program MakeTiff converts raw camera images into linear Tiff files. This is required for PerfectRAW because it is impossible to open raw photos with Adobe Photoshop [...] without using the host's raw converter and ruining the color."

Considering the developer's strong feelings about avoiding ACR, I feel like I should follow the author's intended workflow if I am to make a fair evaluation of the ColorPerfect plugin. After I understand how ColorPerfect supposed to work (if that ever happens), then I will be in a better position to modify the process to best fit my needs.

EDIT: By the way, the developer of Negative Lab Pro has his own solution for avoiding the use of the ACR color profiles when converting RAW files - his plugin adds his own Profile "Negative Lab ..."

 

[NortheastPhotographic]

Still though, DSLRs aren't really an optimal tool for other than with B&W negatives. They are fast though.

People say this but I think generally they just haven't optimized their workflow. A camera like the Panasonic S1R in pixel shift mode makes a true RGB capture and increases resolution with multishot. Couple that with a good lens and a high CRI light source and the results are as good or better than high end scanners. Drum scans are probably better when printing beyond 44" but I would put my scans up against any CCD scanner. I can make files in 1-2 captures that print at either 360 or 720ppi in my 24" Epson printer. They are highly detailed and have exceptional dynamic range, plus no banding and extremely low noise. And this is the first gen S1R!

The only thing I miss is ICE, but Creo, Imacons, Drums don't have that either. And unlike Imacons I can wet scan easily.

 

[GLS]

I believe we are talking about two different kinds of color profiles

I am referring to the colour space (Adobe RGB etc, or "A" in your post), not the profiles used by Lightroom/Adobe Camera RAW. I could be wrong, but as I understand it the latter are essentially instructions for how the RAW converter should render the image preview and/or final written image, but are nothing more than that until you save the file in a non-RAW format.

I agree the terminology isn't helpful, but the terms colour space & colour profile are used interchangably in many cases. Photoshop itself refers to them as both spaces and profiles, depending on the context. Regardless, by all means try the MakeTIFF program and see how you get on. I am simply saying that in my estimation it is not a requirement, and to bear in mind that the makers of ColorPerfect are naturally going to try and stress the indispensable nature of their own suite of tools over those from third parties.

 

[Marameo]

Does NLP just use the Curves pane and Saturation in the HSL pane to achieve its magic?

As I understand, the white balance is done just to remove the orange. Also, in the Camera Calibration pane a custom profile is used instead of the "embedded".

Exposure, Contrast and other tonal sliders are not used if I am correct.

Basically, everything happens in the Curve Pane.

How about sharpening?

Thanks

 

[runswithsizzers]

Does NLP just use the Curves pane and Saturation in the HSL pane to achieve its magic?

As I understand, the white balance is done just to remove the orange. Also, in the Camera Calibration pane a custom profile is used instead of the "embedded".

Exposure, Contrast and other tonal sliders are not used if I am correct.

Basically, everything happens in the Curve Pane.

How about sharpening?

Thanks

The Negative Lab Pro plugin has some basic sharpening settings which the developer describes in the <NLP Users Guide> as:

Sharpen: Leave as Set
new in v2.1) This will leave sharpening at whatever the user or lightroom as set sharpening at. Essentially, NLP just lets you manage it. This is the new default sharpening.

Sharpen: off
All sharpness settings are zeroed out. This is useful sometimes if you are working with scans that had sharpening already applied in their original software.

Sharpen: Lab
This produces beautiful, soft sharpening that is brilliant for skin tones and gives subjects an almost 3D appearance, while minimizing noise and grain. Great for portraits, or fine-grain films, like Portra. It is based on the default output of Fuji Frontier lab scanners. The sharpening scheme focuses on edges, with a mask to prevent sharpening unwanted grain or noise.

Sharpen: Scanner
This produces gritty, textured sharpening, that accentuates grain and noise, and is popular in lomographic photography circles. Great for street photography, black and white, or anything were you want a gritty vibe.

NOTE: Both of these sharpening schemes are based on RAW dslr scans, and the effect may vary if using a non-RAW or flat-bed scanned negative.​

---
Leave as set (for me, the Lightroom Default settings)

Lab - he mentions a "mask" for this setting, but unless there is something going on outside of the Lightroom Detail tab, it uses a combination of Color Noise reduction and reducing the Sharpening Detail rather than using Lightroom's Masking slider. I sometimes use this one as a starting point, but my camera copies often benefit from increasing the Radius to 1.4-1.6.

Scanner

 

[Adrian Bacon]

Does NLP just use the Curves pane and Saturation in the HSL pane to achieve its magic?

As I understand, the white balance is done just to remove the orange. Also, in the Camera Calibration pane a custom profile is used instead of the "embedded".

Exposure, Contrast and other tonal sliders are not used if I am correct.

Basically, everything happens in the Curve Pane.

How about sharpening?

Thanks

NLP does all of its magic using the tools in the develop module. It just presents a user interface to make it a little easier as all the controls are otherwise inverted (because you’re working on a negative image).

there is technically nothing wrong with this approach, and if it works for you and you’re happy with the results, then that’s what you should use.

personally, I would rather do all the magic (if you will) before it makes it into LR and present LR with essentially a colorimetrically correct raw color positive image. When done that way, all the tools in the develop module (including white balance) actually work as intended and you can treat the image just like any other raw image that you shot with any other digital camera.

It mostly comes down to personal preference.

 

[PhilBurton]

NLP does all of its magic using the tools in the develop module. It just presents a user interface to make it a little easier as all the controls are otherwise inverted (because you’re working on a negative image).

there is technically nothing wrong with this approach, and if it works for you and you’re happy with the results, then that’s what you should use.

personally, I would rather do all the magic (if you will) before it makes it into LR and present LR with essentially a colorimetrically correct raw color positive image. When done that way, all the tools in the develop module (including white balance) actually work as intended and you can treat the image just like any other raw image that you shot with any other digital camera.

It mostly comes down to personal preference.

I haven't followed this thread too carefully, because I will be using a Nikon 5000 scanner for all my color negs. While the Nikon scanner can product a "NEF" file, it's kind of bogus, and not a real NEF. As a practical matter, this scanner produces TIFF files. So my question is, will NLP be effective and useful in Lightroom (latest subscription version) with TIFF files of color neg scans? I realize that a TIFF file has less flexible editing options than a RAW, positive or negative.

 

[Marameo]

personally, I would rather do all the magic (if you will) before it makes it into LR and present LR with essentially a colorimetrically correct raw color positive image.

That indeed would be great and I wonder if it is doable in the DCP camera profile without coding or using tiff/dng libs.

Is it possible to see some flat-bed scanned negative (Vuescan DNGs) conversions made with NLP? All I've found is about RAW dslr scans.

 

[dkonigs]

NLP does all of its magic using the tools in the develop module. It just presents a user interface to make it a little easier as all the controls are otherwise inverted (because you’re working on a negative image).

So I recently decided to go and take a look at how NLP actually works. I don't think it really depends on the innards of Lightroom much at all, beyond being a useful framework to plug some LUA scripts into (so that it doesn't need to be a standalone application). What it actually appears to be doing, is using several calls into bundled copies of ImageMagick and ExifTool (and probably parsing the output as well) along with some bundled ICC profiles. I have no idea how much logic its actually performing between these calls, but it should absolutely be possible to simply write a script that mimics them.

 

[Marameo]

Yet, every change applied to the image thru the interface is reflected in the develep module. Maybe ImageMagick is used to create the final tiff file?

 

[dkonigs]

Yet, every change applied to the image thru the interface is reflected in the develep module. Maybe ImageMagick is used to create the final tiff file?

As I dig deeper into attempting to analyze its behavior, I think its actually a mixture of things. Its clearly doing a series of expensive ImageMagick operations that seem to return various bits of information about the image (guessing its white/black/grey points or similar) while doing its thing, but it may also be using LR itself to apply the final adjustments based on that information.

 

[Marameo]

I think its actually a mixture of things. Its clearly doing a series of expensive ImageMagick operations that seem to return various bits of information about the image (guessing its white/black/grey points or similar)

I think ImageMagick is doing auto color correction and returning RGB coordinates that are applied in LR.

 

[Adrian Bacon]

That indeed would be great and I wonder if it is doable in the DCP camera profile without coding or using tiff/dng libs.

Is it possible to see some flat-bed scanned negative (Vuescan DNGs) conversions made with NLP? All I've found is about RAW dslr scans.

My tool doesn't work on anything but bayer array data. Whenever I have to scan something with a flatbed, I make a vuescan raw TIFF scan and run it through another tool that I wrote that converts the full RGB data to RGGB bayer data in a TIFF that I then feed into my processing code.

 

[NortheastPhotographic]

I haven't followed this thread too carefully, because I will be using a Nikon 5000 scanner for all my color negs. While the Nikon scanner can product a "NEF" file, it's kind of bogus, and not a real NEF. As a practical matter, this scanner produces TIFF files. So my question is, will NLP be effective and useful in Lightroom (latest subscription version) with TIFF files of color neg scans? I realize that a TIFF file has less flexible editing options than a RAW, positive or negative.

The newer versions of NLP (all free upgrades) have focused a bit on users with normal scanners. I believe you just have to output a gamma 2.2 TIFF and of course scan as a positive. Then in NLP you tell it it’s a scanner file. The results folks post in the Facebook group look great.

 

[dkonigs]

The newer versions of NLP (all free upgrades) have focused a bit on users with normal scanners. I believe you just have to output a gamma 2.2 TIFF and of course scan as a positive. Then in NLP you tell it it’s a scanner file. The results folks post in the Facebook group look great.

Yeah, the output looks decent. The problem is that the workflow kinda sucks. Here's why...

ColorPerfect (the other option here) wants a flat scan at gamma 1.0
NLP, meanwhile, wants a flat scan at gamma 2.2, which you then pass through their "Tiff Scan Prep" module that creates a second output file (also as an enormous TIFF) that NLP can process better. (and these get stacked in Lightroom in an awkward way)

Its actually pretty easy to take that "prepared" TIFF and simply Invert (and do some small adjustments) in Photoshop to get halfway decent results.

In any case, what I think I really want is a completely standalone utility/script that does the same function as "Tiff Scan Prep" and stores its output as a DNG file. I can then basically throw away these original off-the-scanner TIFFs and import those (slightly smaller) DNGs into Lightroom for the rest of the NLP processing.

I have also looked at exactly what "Tiff Scan Prep" does, and it seems to be 2 calls to ImageMagick and one housekeeping call to ExifTool. It really shouldn't be hard at all to write a script that glues this together and gives the behavior I want. I might just do that before the next time I'm ready to try using NLP for a bunch of color scans.

 

[PhilBurton]

Yeah, the output looks decent. The problem is that the workflow kinda sucks. Here's why...

ColorPerfect (the other option here) wants a flat scan at gamma 1.0
NLP, meanwhile, wants a flat scan at gamma 2.2, which you then pass through their "Tiff Scan Prep" module that creates a second output file (also as an enormous TIFF) that NLP can process better. (and these get stacked in Lightroom in an awkward way)

Its actually pretty easy to take that "prepared" TIFF and simply Invert (and do some small adjustments) in Photoshop to get halfway decent results.

In any case, what I think I really want is a completely standalone utility/script that does the same function as "Tiff Scan Prep" and stores its output as a DNG file. I can then basically throw away these original off-the-scanner TIFFs and import those (slightly smaller) DNGs into Lightroom for the rest of the NLP processing.

I have also looked at exactly what "Tiff Scan Prep" does, and it seems to be 2 calls to ImageMagick and one housekeeping call to ExifTool. It really shouldn't be hard at all to write a script that glues this together and gives the behavior I want. I might just do that before the next time I'm ready to try using NLP for a bunch of color scans.

dkonigs,

Thanks for this detailed reply. So, IF, IF, IF I had my druthers,

• I hope that I can establish just one workflow to process all my C41 and older C22 negatives, allowing for differences in emulsions. What I mean is scanner settings, TIFF output file, final TIFF (or DNG?) file and its filename that serves as the "negative" for Lightroom work.
• I would have a purely Lightroom workflow. Yes, I do have Photoshop as part of my Adobe subscription plan, but I try to reserve Photoshop only for "final touchups," etc.
• I hope that this "script" would be something simple in Windows. And that it can batch-process a bunch of files. My scripting skills are currently limited to cmd file. Yes, I guess I could learn PowerShell if I absolutely had to, or else a third-party scripting language.

Long, long story short, I haven't started my "big scanning project" yet. (Never mind exactly why. It just is.) So this is the right time to get the color negative workflow nailed down. I just want to say how useful and helpful this entire thread has been.