P: Ability to invert negative scans to positives (color and black-and-white)

[jeffk5545]

I would dearly like to see the Lightroom 4 Beta team implement an additional feature in the final release. That feature would be the ability to take a camera+macro lens image of a B&W negative -- essentially a camera-based scan of a negative -- and invert the negative image to a positive image at the beginning of the development process in such a way that the resulting sliders in the LR4 Develop Module would not operate in reverse. As I understand it, this capability exists in Photoshop, but I don't own Photoshop. I do own Photoshop Elements 9, but that program only supports an 8-bit workflow, not 16-bits per channel, and round-tripping between LR & PSE9 requires the reimportation of a TIFF file that is more than twice the size of my NEF RAW files. Since this programming wizardry already exists in Photoshop, I would think that it would be a relatively simple matter to transfer and adapt that code for LR4 -- but then, I'm not a programmer, so what do I know...

I've been digitizing 40-year-old Kodachrome slides from my Peace Corps days in Africa, using a 55mm Micro-Nikkor (macro) lens, coupled to a Nikon ES-1 Slide Copy Attachment, and even on a D300s body, I can get truly excellent results. I can't wait to continue that work using the pending 36 megapixel Nikon D800 body with an upgraded f/2.8 macro lens (mine is the old 55mm f/3.5 design). I really, REALLY want to be able to camera-scan my many B&W negatives without having to generate huge intermediate TIFF files.

You can respond to this request by emailing me, Jeff Kennedy Thanks, in advance, for taking the time to review and consider my request. I LOVE Lightroom 3, and from what I've seen, I'm going to love LR4 even more. I REALLY appreciate the effort that Adobe takes to solicit input from the photographic user community.

BTW, if the feature I request *can't* be implemented right away, could the LR support team provide detailed, interim instructions as to how to use the "backwards" sliders, and in what sequence? That would be very much appreciated. I'm sure many older LR users have considerable analog image collections that they would like to digitize, and doing so in-camera is both 1) of surprisingly high quality, 2) MUCH faster than using flatbed scanners and 3) of much higher quality and resolution than flatbed scan and MUCH cheaper than professional drum scans.

[johnrellis]

Thomas Geist wrote, "Gunter, just invert the curve and you're done."

There are two serious issues with this:

1. After inverting the tone curve, the other sliders work backwards, which is confusing if you also work with positives.

2. Due to the non-linear nature of the Develop module's image-adaptive controls, inverting the tone curve gives much different results than doing the inversion in Photoshop. See this thread for more details: https://forums.adobe.com/message/8067...

[mfraser3]

Two simple requests for Lightroom:

1. An 'Invert' command. Many of us use Lightroom to manage our film scans, and currently have to go through Photoshop to invert (CMD-I) to process our negatives. Staying within LR for this would be hugely helpful (reversing the RGB curves works, but results in all of the sliders working backward).

2. A per-channel 'Auto' button for RGB curves. One of the key parts of converting colour negative film scans to positive is to correct for the orange film mask, and this involves maximizing per-channel contrast (and snapping the neutral mid tones). This currently requires a roundtrip into Photoshop.

Adding these two features would make Lightroom THE preeminent tool for processing film scans (other RAW processors do offer per channel contrast maximization, but not invert). Photoshop would still be used for other important tasks, but the primary workflow could be entirely within Lightroom.

[jimvk222]

it would/will be great. the work rounds are arduous. thanks

[FeedbackCommunityMember]

we want an Invert! not have to do this "workaround"

[Keith Reeder]

"We" DON'T want an invert. Maybe YOU do...

[ohnoezitasploded1]

Yup. This seems like a simple fix that would be hugely helpful for my workflow

[FeedbackCommunityMember]

Can you please add a simple "Invert Colours" function to the develop module? I take a lot of Solargraph images which I then scan as a paper negative and its frustrating to load them into Photoshop only to invert colours - since LR develop module has much more intuitive funcionality for the remainder of my processing

[henriko72124252]

I too would love an invert feature when using my DSLR for scanning film negatives into raw files for later edits all within LR. As pointed out many places, inverting the tone curve is not that useful, as many other treatment controls are then inverted, and making contrast adjustments in tone curve is also still inverted.

The solution to export and reimport tiffs with an invert curve preset is noted though as a workaround. But I like to have only my raw files if possible, nothing else - hence the wish for an invert switch in the develop module. Thanks.

If anyone else could use this feature, remember to vote at the top of this page.

[kbarre]

The only current solution for editing negatives is to invert the curves, export the images, and then re-import into Lightroom. But this isn't possible with RAW files. It would be awesome to be able to select an "Invert" button, and then have all other functions of Lightroom work exactly the same, so that high quality capture of negatives using a macro lens would be possible.

[RoyBDigipix]

I'm looking at digitizing thousands of color negatives to digital files via a high-rez camera.  I would love to have an invert color option within Camera RAW.  What better way to preserve old film that with a RAW digital file! Even better with a feature to eliminate the color cast of the film.  I think many PS users would embrace it.  I could see this being a huge draw for entry level users as well, as many are trying to figure out what to do with old negatives that are deteriorating.

[Jack_2]

Take into account that 35mm. color negatives of 100-200 ASA films have an effective resolution of not more than 2000x3000 pixels. Color slides of the type of Kodachrome 25 have higher numbers but not much more than approx. 4000x6000 if professional lenses have been used.These numbers were found by the Kodak people that researched the resolution necessary to digitize color negatives when they were doing the concept research for Photo CD scanners in the early nineties. 
You can use high-rez cameras or scanners, but you are not capturing pictorial information, just patterns of grain and unsharp airy-discs.......In microscopy this is called empty magnification. Here I would call it empty resolution.

[timreeves]

4k x 6k is 24mp
So to effectively scan at this resolution on a modern digi cam with a cfa we need a minimum of triple that resolution.

Most modern digital cameras are around 24mp, so 3 to 4 shots per frame stitched with a perfect lens should get you close to kodaks figures.

We still need an invert function, and this request is almost as old as lightroom.

[TheDigitalDog]

Jack, keep in mind that newer PIW's from Kodak produced larger scan sizes than 2Kx3K. I know, I used to run one. At the time, an 18MB scan (max resolution from the image pack) from the original PIW was a lot of data for desktop machines. Kodak up'd that for good reason. As someone that also ran a couple drum scanners (ScanView, Howtek), there was plenty of reasons we scanned above that so called 'effective resolution' you state. In those old days, it was common to have to output retouched images to an LVT 4x5 for repro. You didn't do that from a PhotoCD scan at the so called effective resolution you state if you wanted the best quality output the technology of those days provided. 

[Todd Shaner]

 *Jack Klaber said: You can use high-rez cameras or scanners, but you are not capturing pictorial information, just patterns of grain and unsharp airy-discs.......In microscopy this is called empty magnification. Here I would call it empty resolution.

Jack, I think the primary objective being discussed here is using camera capture to expedite capturing film images (<1 sec. vs minutes). I also agree with *Andrew Rodney that the effective resolution of fine grain 35mm negative and slide emulsions is a lot higher than the numbers you stated. My results using camera film capture with a 5D MKII 21 mp camera produce images from 35mm Fujicolor Super G100 color negatives and Kodachrome 64 slides that look near identical in detail sharpness to normal camera images shot with the same lens and camera. Using an even higher megapixel camera would be beneficial especially if capturing larger format film images (645, 4x5, 8x10).

BTW- A similar post was merged here that no longer pulls up doing a search. It might be of interest:

https://feedback.photoshop.com/photoshop_family/topics/color-invert-for-negative-film-in-camera-raw

[Jack_2]

Andrew, I was a Kodak PCD & PIW technician and account manager and still have all the manuals and tech. sheets at hand.
The Pro PCD was a 4x5" scanner that could scan 135 and 120 film at 4x6k max. resolution. The 4x5" sheet film scan was done at 4096x5230 resolution, the 135 (24x36) was done at 4096x6144 resolution while the 120 (6x4.5) was done at 4096x3072 resolution.
To scan a color negative from a Kodacolor 400, 200 or 100 film with 4x6k resolution was futile and just created a big 64BASE image that had no more usable information than the 2x3k 16BASE. This is what I called "effective resolution".
But EKTAR and KODACHROME 25,40 and 64 35mm. slide films could definitely profit from a PRO scan at 4x6k !! 

[TheDigitalDog]

We are in agreement that 35mm PCD pro scans were higher resolution than 2kx3k for good reason.

[Jack_2]

Tim, there is no need to scan three times the same image to offset the CFA of a DSLR. This would only be effective if you had a monochrome FPA  sensor with 4x6k pixels and then take 3 exposures using 3 appropriate color filters.
The same would be true for a monochrome linear CCD scanner.  

[Jack_2]

Indeed Todd, expediting capturing film images (without compromising the information quality) is of prime importance, especially when thousands of negatives/slides must be processed.....
Be advised that up to 4x5 a 4x6k scan is sufficient to capture the effective information in the sheet film. Technical camera's use a bigger diffusion circle (airy disc) and thus, the spatial resolution of sheet film is lower than 135 and 120 film. 
8x10 sheets I would preferably scan with a transparency scanner such as f.i. the Epson V850 Pro Scanner. Although it is much more time consuming, I don't believe you have thousands of 8x10 sheets waiting to be digitized. Setting up a professional light table to photograph 8x10's is no sinecure..... 

[ohnoezitasploded1]

The point of using a DSLR versus a flatbed scanner isn't to improve the quality of the resulting scan, the point is that using a DSLR is multiple times faster than a flatbed.  We are asking for a modification to ACR/Lightroom to support this workflow because a lot of people working in mixed film/digital environments are using it.

So a discussion of the theoretical resolution of film is off topic, there are many other threads on the internet where this subject can be hashed and rehashed and rerehashed.

[2Pop]

We have had the opportunity to compare digitalization of a batch of  Portra 400 color negative on a Heidelberg Tango scanner and our PhaseOne system and the results are - in our opinion - far better with the digital back. Not speaking of the speed, digitalizing with a modern medium format camera gives you a RAW file (with all post-prod ajustements available) that a conventionnal scanner cannot produce

[stevel24076854]

Maxpierson, 

Could you go over this camera procedure again.  I guess I could put one of my 5 cameras to use.  Like Jack's, my v850 is a good one but I agree it might be a tad slower than your camera procedure.   So now I am interested in your speedier concept, rerun it?   Earlier, I was sold on my flatbed procedure and my v850 does a fantastic job with transparency, slides, neg's and anything we can throw at it that requires dual light scanning.   It has all adapters and include (Jack's) 120, 4x5 and 8x10 formats.  The adapters for that scanner will do 12 slides and 12 neg's at a time.   Plus, aside from the v850, I have a fully automatic multi-threading slide scanner that zooms through a much larger batch, (while I am restoring the last batch) and I don't have to shuffle the cards.  But if there is a faster rate of speed via camera, I heard it before, but I want a rerun if its procedure - if you don't mind going over it once more.   Is there an adapter for the camera lens?  It must do one at a time but if there is an adapter, then it's an easy change maybe - I'm picturing it like a shuffle of a cards, maybe?    As for the camera RAW thing, it can be run through a DNG converter to change it to a TIFF, the diff being 16 bit converted to 8 bit, ready for printing.   

Steve Lehman, mcse

[stevel24076854]

Maxpierson, 

Could you go over this camera procedure again.  I guess I could put one of my 5 cameras to use.  Like Jack's, my v850 is a good one but I agree it might be a tad slower than your camera procedure.   So now I am interested in your speedier concept, rerun it?   Earlier, I was sold on my flatbed procedure and my v850 does a fantastic job with transparency, slides, neg's and anything we can throw at it that requires dual light scanning.   It has all adapters and include (Jack's) 120, 4x5 and 8x10 formats.  The adapters for that scanner will do 12 slides and 12 neg's at a time.   Plus, aside from the v850, I have a fully automatic multi-threading slide scanner that zooms through a much larger batch, (while I am restoring the last batch) and I don't have to shuffle the cards.  But if there is a faster rate of speed via camera, I heard it before, but I want a rerun if its procedure - if you don't mind going over it once more.   Is there an adapter for the camera lens?  It must do one at a time but if there is an adapter, then it's an easy change maybe - I'm picturing it like a shuffle of a cards, maybe?    As for the camera RAW thing, it can be run through a DNG converter to change it to a TIFF, the diff being 16 bit converted to 8 bit, ready for printing.   

Steve Lehman, mcse

[stevel24076854]

Vbus,
Anytime a drum scanner is used, it's much better than a typical digital scanner.   

[2Pop]

Well I disagree with that. Modern medium format camera (Hasselblad or PhaseOne) with high-end macro-lens put the Tango on his knees...

[Todd Shaner]

Steve, at the link to the other post I provided a Dropbox link to an article Mark Segal and I co-authored on Luminous Landscape. You can download the article and a detailed instructions PDF along with other materials for processing camera capture color negative film files. I've included a 35mm Fujicolor Super G100 color negative captured with a Canon 5D MKII 21mp camera for comparison.

Camera film capture using a diffused light source provides better definition and lower grain than using a film scanner. This is covered in the LULA article with a picture example using a Plustek 7600i scanner and Canon 5D MKII 21mp camera with Sigma 50mm macro lens.

https://www.dropbox.com/sh/u5pnzbesvqen1hv/AABZyaWAn33jCwEinkT45iFAa?dl=0