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

[michaelbothager]

Do you know what you actually get with NLP?

The Facebook group [https://www.facebook.com/groups/negativelabpro/] has a lot of discussions of actual use of NLP, I suggest that you join and read/ask questions. I guess it is better to ask people who are using the plugin.

 

[Lachlan Young]

Thanks for the advice. Much appreciated.

Do you know what you actually get with NLP?
I just watched Matt Days video on it and camera scanning again.
There seems to be much the same work in it as you described above.
Is it really just a hundred dollar orange mask remover?

It's got more in common with minilab software in terms of how it works - ie lots of preset options to mix & match & less reliance on the end user having a good sense of colour & how to correct it. I have described the basics of how to get an inverted and coherently colour correctable file by manual means. From there, you need to know what you are doing with curves, masks, sharpening, de-noising etc. His system is intended to produce a baked-in product at the end of a bulk batch. It's like the difference between a high throughput minilab (his system) and a custom exhibition printer (what my working methods are intended towards producing). And yes, the sample, fill, divide seems to be what's going on at the outset. I should also add that manually controlled inversions are highly reliant on user skill - minilab type setups like the one in the plugin being described are intended to minimise operator errors. They aren't bad if you don't mind the films all being largely forced down the same hosepipe!

 

[Cholentpot]

I use a LR and PS combo. Where we getting PS and LR for $10 a month?

Also it's partly automated for me at this point using PS actions. For B&W it goes pretty fast if the negs are clean. I can do a full roll of 36 in under an hour. Color a bit longer...

 

[calebarchie]

If I was a business, I would be cautious of what I say on the web especially to potential customers or market-base - have seen bad things happen with labs over here in OZ.

Bests

 

[jtk]

There are many of these (digicam copy of color negative) on youtube that take you through each of their processes and they all confirm what I already know from my own efforts and that they are not universally applicable to each frame of color negative film. And of course on top of that you still have to deal with dust and scratch removal.

Fortunately I have been using a Coolscan 5000/9000 with Nikonscan and over 40,000 frames of various films later, I have not found anything close to this combination when it comes to accurately rendering color negatives - as well as all other film types. I am hoping that Nikon has instilled into their D850's built-in color negative conversion the same accuracy as Coolscan+Nikonscan combination but I have not yet tried it myself to compare it. But of course this will still not have the unequaled ICE built-into the Coolscan+Nikonscan combination.

 

[jtk]

Exactly. People who can afford to keep shooting film can afford to pay a good lab to scan it...if they've missed out on Nikon V.

 

[Cholentpot]

Exactly. People who can afford to keep shooting film can afford to pay a good lab to scan it...if they've missed out on Nikon V.

Or you find a way to do it really really cheaply. You'll need to embrace the failures and unpredictability but you can really get the price of everything down to less than a .50 a roll. Shot, developed and scanned. Maybe $1 for color.

 

[StepheKoontz]

Grain is not objectionable and with fine grain film it’s not even noticeable at normal viewing distances at poster size.

If that was true, no one would ever use larger formats.

 

[Eric Hiss]

The other option to look at is the ColorPerfect software: https://www.colorperfect.com/

This is a great software in terms of both results and control and is the one I currently use. However I have to admit, I've not spent enough time to really learn how to use all their tools. I have found the colorperfect (color neg) plugin can be used to batch run a folder full of scans with a photoshop droplet and it seems to handle small differences in exposure with no problems. I don't think you can get around having to spend a little time on each 'scan' if you want the best results though.

 

[Adrian Bacon]

How do the DSLR scanners here reverse their colour negatives?
I don’t want to pay extortion prices for Light Room, for features I’ll never use and certainly not for the necessary Negative Lab Pro on top of that.

Fiddling with reversed curves in PS or other image editors, is not a long term viable option and doesn’t get you results that compares well to a wet optical print.

There has to be a simple piece of software that does the relatively simple job of reversing the orange mask and with profiles for common negative films.
Only I can’t seem to get a clear answer when I ask google.

I used to use Vuescan with the incredibly clunky import option, but I’m really tired of that as you have to work too much for every frame.

Advice would be much appreciated.

Negative Lab Pro is one option, and works fine.

For my own setup and for all the work I do for customers, I use a DSLR and Macro lens in a custom copy stand jig designed for mass scanning. I wrote my own software to do the conversion. I've never released my software to the general public simply because I make a significant part of my living from using it to do customer work, however, having looked at NLP, I can tell you that it easily provides results equal to if not better than NLP and predates NLP by a couple of years. When NLP first came out, I took a look at it and decided that I'd rather keep doing my software simply because my software runs outside of LR and generates floating point DNG files that can be ingested by pretty much any other software that supports DNG files. NLP operates inside of LR and basically uses LRs Develop Module tools to get to a reasonably correct positive image. That can and does work, but in all honesty, that's not what those tools were really designed for.

A few other benefits of my software over NLP is when you're in LR, it behaves exactly as if you had shot the picture in raw with a digital camera, meaning the exposure slider is correct and accurate, the color temperature slider is correct and accurate, so daylight film is 5500K, Tungsten film is 3200K, etc. If you shot your daylight film indoors in tungsten lighting, it's a simple matter of changing the white balance to 3200K in the develop module and your color goes back to normal, or if you shot your daylight film in really overcast weather, just change the color temperature slider 6500-8500K or whatever looks the best for the conditions. The LR Develop Module works exactly the same way it would with a digital camera and is very accurate. You can just add additional edits using the Develop module just like with any other raw file.

My software internally operates in 80 bit floating point color and uses the ACES 2065-1 color profile and stores the raw floating point values in the DNG in that color space and embeds a color matrix into the DNG it generates so that LR can convert the samples to its internal ProPhoto color space. It can output either 16 bit floats, 32 bit floats, or gamma corrected 16 bit ints as native CFA samples with the correct metadata so LR knows what to do with it. Essentially, my software looks like a digital camera to LR. It takes the raw scans from all the film types I support and conforms them to look as if it was shot by my software. Since I process and scan film for other people, my software is also designed to natively work within a twin-check sticker film tracking regime, and is designed to bulk process raw scans to DNG files. In short, you get your DNGs from me and import them into LR or your image editor of choice and carry on as if you shot it digitally. For the end user, it couldn't be simpler, for me, it has been quite a bit of work, however, the results are excellent. Most images I've uploaded to my media gallery and to my Flickr stream have been run through my software if anybody wants to see it's output.

I'm not knocking on NLP... It can be a great solution for a lot of people, and in pretty much every way is a big step forward in film scanning. It's just not what I would do if it were me.

 

[Eric Hiss]

Negative Lab Pro is one option, and works fine.

For my own setup and for all the work I do for customers, I use a DSLR and Macro lens in a custom copy stand jig designed for mass scanning. I wrote my own software to do the conversion.

I agree with you about being free from LR - I mean I do use LR but then again I use digital backs to 'scan' with and LR doesn't do as nice a conversion with those anyway and LR doesn't work with some file formats at all. I had been using a CF-528 multishot / microstep back with a Rollei 6008AF and Rodenstock APO Rodagon lens for my scanning work suspended over a Kaiser light table but I had found the set-up to be very sensitive to focus. Now I'm using a Creo iQsmart 3 which is slower but easier for larger film. I'm not all that happy with the color from the iqsmart sometimes though.

 

[Helge]

If that was true, no one would ever use larger formats.

Of course it depends on the effect you want to achieve.
Medium format is undeniably higher resolution and lower on grain.
There is of course also a multitude of ways to filter your 35 mm frames. Both in a dark room and on the computer.

 

[Adrian Bacon]

I agree with you about being free from LR - I mean I do use LR but then again I use digital backs to 'scan' with and LR doesn't do as nice a conversion with those anyway and LR doesn't work with some file formats at all. I had been using a CF-528 multishot / microstep back with a Rollei 6008AF and Rodenstock APO Rodagon lens for my scanning work suspended over a Kaiser light table but I had found the set-up to be very sensitive to focus. Now I'm using a Creo iQsmart 3 which is slower but easier for larger film. I'm not all that happy with the color from the iqsmart sometimes though.

I use a strobe diffused through a custom diffuser with enough separation between the film and the last diffuser stage that its completely out of focus and therefore a very smooth and diffuse light source. The strobe gives me very consistent and repeatable results, and I don't have to worry too much about light contamination because I'm operating at f/11 to f/16 at 1/250, which is easily several stops above the ambient room light. If the strobe don't fire, you get a black frame, when it does fire, you get the absolute minimum amount of anything that might introduce vibration induced blur. 35mm film can sometimes be a bit finicky with focus, especially if it's got a strong curl along it's longitude, but for that, I just go up to f/16-f/22 and turn the strobe power up. There's a little added diffraction, however, I'm using a very expensive macro lens, and with it, my setup is over 4000dpi for 35mm film. Jpegs scaled down to even a quarter of the resolution of the native sensor resolution have a *crazy* amount of fine detail. I'd rather have it in focus, than go for maximum sharpness and struggle with having frames in focus. For 120, there's close to 2 feet between the film and sensor. F/11 is plenty of DOF there. For me, going the DSLR route with custom software is a simple matter of speed. I can buzz through a freshly dry roll of 36 exposure film and be looking at the entire roll in LR in less than 15 minutes, with little to no retouching except for a spot of dust/fiber here or there, or the occasional scratch, and at most, some basic Develop Module touch ups if that.

 

[ludwighagelstein]

Negative Lab Pro is one option, and works fine.

For my own setup and for all the work I do for customers, I use a DSLR and Macro lens in a custom copy stand jig designed for mass scanning. I wrote my own software to do the conversion. I've never released my software to the general public simply because I make a significant part of my living from using it to do customer work, however, having looked at NLP, I can tell you that it easily provides results equal to if not better than NLP and predates NLP by a couple of years. When NLP first came out, I took a look at it and decided that I'd rather keep doing my software simply because my software runs outside of LR and generates floating point DNG files that can be ingested by pretty much any other software that supports DNG files. NLP operates inside of LR and basically uses LRs Develop Module tools to get to a reasonably correct positive image. That can and does work, but in all honesty, that's not what those tools were really designed for.

A few other benefits of my software over NLP is when you're in LR, it behaves exactly as if you had shot the picture in raw with a digital camera, meaning the exposure slider is correct and accurate, the color temperature slider is correct and accurate, so daylight film is 5500K, Tungsten film is 3200K, etc. If you shot your daylight film indoors in tungsten lighting, it's a simple matter of changing the white balance to 3200K in the develop module and your color goes back to normal, or if you shot your daylight film in really overcast weather, just change the color temperature slider 6500-8500K or whatever looks the best for the conditions. The LR Develop Module works exactly the same way it would with a digital camera and is very accurate. You can just add additional edits using the Develop module just like with any other raw file.

My software internally operates in 80 bit floating point color and uses the ACES 2065-1 color profile and stores the raw floating point values in the DNG in that color space and embeds a color matrix into the DNG it generates so that LR can convert the samples to its internal ProPhoto color space. It can output either 16 bit floats, 32 bit floats, or gamma corrected 16 bit ints as native CFA samples with the correct metadata so LR knows what to do with it. Essentially, my software looks like a digital camera to LR. It takes the raw scans from all the film types I support and conforms them to look as if it was shot by my software. Since I process and scan film for other people, my software is also designed to natively work within a twin-check sticker film tracking regime, and is designed to bulk process raw scans to DNG files. In short, you get your DNGs from me and import them into LR or your image editor of choice and carry on as if you shot it digitally. For the end user, it couldn't be simpler, for me, it has been quite a bit of work, however, the results are excellent. Most images I've uploaded to my media gallery and to my Flickr stream have been run through my software if anybody wants to see it's output.

I'm not knocking on NLP... It can be a great solution for a lot of people, and in pretty much every way is a big step forward in film scanning. It's just not what I would do if it were me.

Can you at least give us a few results of your wonder software to compare ? An image of your digitization setup would be quite nice also

 

[Dani]

Here you can see samples of his software https://www.flickr.com/photos/tags/simpleimagetools

I'd take pleasing and easy workflow any day of the week. Shooting tethered into LR and one stop workflow with pleasing results will beat anything out there.

 

[Photo Engineer]

If that was true, no one would ever use larger formats.

What about tilt and swing, extended bellows and other "conveniences" offered by the LF cameras?

PE

 

[StepheKoontz]

What about tilt and swing, extended bellows and other "conveniences" offered by the LF cameras?

PE

Most of these, other than shift, are needed to deal with the inherent shallow DOF larger formats have. I sure wouldn't choose using my 4X5 over a 35mm with a ext tube for "convenience".

 

[Adrian Bacon]

Can you at least give us a few results of your wonder software to compare ? An image of your digitization setup would be quite nice also

Wonder software? lol... there's nothing wonder about it... In fact, it's shockingly simple and straightforward... anybody who knows how to write basic C or C++ code and has figured out how digital color spaces works can do it. I'm actually amazed more people haven't done it yet. The software isn't without value, but there's also a lot of value in the knowledge of how to take a raw image of a film negative and how to programmatically turn it into a positive image that conforms to a given digital color space. Most of my "aha!" moments actually came from reading the Adobe DNG spec, which is available for free. Once you have the samples inverted and linearized, they're actually not really any different from raw samples of an image you actually took with a digital camera, and the DNG spec lays out how to deal with them so you can put them into a DNG file.

I'm not going to share customer work, and have already pointed out where you can see sample output from film I've shot myself and run through the setup, and I've already described in broad strokes the physical set up of the scanner so I won't be disclosing finer detail as it's something that I've spent a fair amount of time and energy refining over the years, and it has potential market value. The same goes for the software. I see no value in releasing it for free only to have a huge amount of un-paid time sucked up supporting other people who want to use it but don't know how to get it to work and would rather you just get it to work for them so that they can use it for free. You can argue that I'd be doing the community a solid if I did that, however, I just don't have that kind of time and I have bills to pay.

 

[Photo Engineer]

Adrian, as one who has written quite a few programs in C, C++, Assembly and other languages, and one who has worked with color space for product design, this is NOT easy or straightforward to the average person. This is daunting. Go easy on the average guy here.

After all, if such things were easy, I would challenge you to design a color film - and process.

PE

 

[jim10219]

Take a deep breath Helge. You're too tense.

 

[Adrian Bacon]

Adrian, as one who has written quite a few programs in C, C++, Assembly and other languages, and one who has worked with color space for product design, this is NOT easy or straightforward to the average person. This is daunting. Go easy on the average guy here.

After all, if such things were easy, I would challenge you to design a color film - and process.

PE

Point taken. It’s all too easy to forget that everybody else is at their own knowledge level.

 

[Helge]

Most of these, other than shift, are needed to deal with the inherent shallow DOF larger formats have. I sure wouldn't choose using my 4X5 over a 35mm with a ext tube for "convenience".

Not true. That might be the origin of tilt, but it has so many artistic uses.
Want a slight tele effect on top of your frame to enhance clouds? Presto!
Want the whole forrest floor in sharp focus, while the treetops near to far are out? There you go.
For any shift you do, you will almost always want to control focus either by aperture, which can go very tight without refraction blur on LF, or by slight tilt.

 

[Photo Engineer]

And there you go! That was included among the "other conveniences" that I noted above. Thanks Helge.

PE

 

[Helge]

Take a deep breath Helge. You're too tense.

Thanks for the friendly advice. I’ll try that.
In the meantime see if you might need to tighten up. You can be too loose you know.
In what state does patronizing lines like that usually work?

 

[Eric Hiss]

I use a strobe diffused through a custom diffuser with enough separation between the film and the last diffuser stage that its completely out of focus and therefore a very smooth and diffuse light source. The strobe gives me very consistent and repeatable results, and I don't have to worry too much about light contamination because I'm operating at f/11 to f/16 at 1/250, which is easily several stops above the ambient room light. If the strobe don't fire, you get a black frame, when it does fire, you get the absolute minimum amount of anything that might introduce vibration induced blur. 35mm film can sometimes be a bit finicky with focus, especially if it's got a strong curl along it's longitude, but for that, I just go up to f/16-f/22 and turn the strobe power up. There's a little added diffraction, however, I'm using a very expensive macro lens, and with it, my setup is over 4000dpi for 35mm film. Jpegs scaled down to even a quarter of the resolution of the native sensor resolution have a *crazy* amount of fine detail. I'd rather have it in focus, than go for maximum sharpness and struggle with having frames in focus. For 120, there's close to 2 feet between the film and sensor. F/11 is plenty of DOF there. For me, going the DSLR route with custom software is a simple matter of speed. I can buzz through a freshly dry roll of 36 exposure film and be looking at the entire roll in LR in less than 15 minutes, with little to no retouching except for a spot of dust/fiber here or there, or the occasional scratch, and at most, some basic Develop Module touch ups if that.

Adrian,
Thanks for the info on your process. I think with a smaller sensor camera the focusing would be easier. I've often wondered how well I would do with a Sony A7r3 or other similar camera that can do the pixel shift. With my Rollei 6008AF + CF528 I need to get obviously close to 1::1 magnification to 'scan' a 6x6 negative. I do see such a big improvement in tonality and detail with the multishot vs single shot that I wouldn't want to do it otherwise. I have tried using the bowens Illumitran (look it up if you aren't familiar - it's an older negative copy stand and light) but didn't find that much improvement over the Kaiser light table. I do use with the light table a second diffuser also far away from the film. I hold the film with a bowen negatrans which makes advancing from one frame to the next convenient and also holds the film flat. But as I wrote, mostly now I am going to the iqsmart 3 even though I do believe the 6008AF can yield a better result than the iqsmart when done properly. The focus at 1::1 magnification isn't trivial and is time consuming. btw - I see that you're just up in Petaluma so perhaps we'll cross paths at some point.