Scanning Color Negatives

[rgacpa]

I've been reading up and doing a little testing on c-f system's ColorNeg and ColorPos plugins for Adobe PS (PC only). Does anyone have any experience using these plugins? I'm not sure I want to spend the $ but would like anyone else's opinion to consider. Info can be found at:
http://www.c-f-systems.com/Phototips.html
I have no association with this product; just trying to achieve the perfect scan!
Bob

 

[dmr]

May I ask what is deficient in your current scans that you think this product will correct?

I'm not trying to be difficult, I just question the necessity of all kinds of plug-ins and extra software when it seems the normal methods should work.

 

[rgacpa]

May I ask what is deficient in your current scans that you think this product will correct?

I'm specifically asking about color negatives.

Both pieces of software that I currently use (Nikon Scan and Vue Scan) have algorithms for interpreting the data scanned from the emulsion. With Nikon I choose 35mm Negative - Color. With ViewScan I choose Transparancy, Generic. When I open up these scans (in CS2) these negatives require color correction and level adjustments (beyond those done pre-scan) This c-f software seems to have built algorithms specific to film type.

So perhaps I'm over complicating this but I believe every negative emulsion or combination of emulsions is different. Compound that with the fact that I don't have a standard to calibrate the interpretation from, and I'm relying on either the software or my memory of the scene to interpret what my image looks like. I cannot look at the negative and say, no the sky isn't the same color blue as I see on my screen so I need to adjust my blue curve...

With transparancies I don't have this problem because a transparency is WYSIWYG. The blue on my screen should match the blue I see on my light table.

So that's why I'm searching for a more accurate way to interpret the data my scanner (Nikon 8000) reports from the scan of the 3 or 4 emulsion layers (depending on the film being used).

Please let me know if I'm way off base, in your opinion. and what you would suggest.
Thanks,
Bob

 

[Ray Heath]

g'day rg

though i'm not consersant with neg scanning i would suggest that you do have to interpret every image individually

whether the image comes from tranny, neg, file, print, stubby finger dipped in cow shit or whatever you have to decide how to present the final version, not some software/algorithm/computer thingy

it's your work, work at it

 

[jd callow]

Ray,
There are redundant, mechanical operations which need to be done for all scanned images that have little to do with "working at it."

Bob,
I went to the web site and it appears to be some interesting software. I hope that you receive some responses. I'd be interested in how well it works. The software doesn't look as if it is going to colour correct for you, but convert from neg to pos and apply an initial optimization in a more accurate manner than most other software (or at least this is what I'm lead to believe by reading their site). Colour correction as well as density and to a far lesser degree contrast is or should not be memory based nor should it require a 'go by' (as in the trany). A competent or experienced person should be able to achieve good density and colour balance by simply looking at the screen/test print. Over time you’ll learn to recognize the need for more or less m, or y not only in your own work but in any print or reproduction. This includes machine prints which I suspect is the kind of automation you are after.

 

[Ray Heath]

Ray,
There are redundant, mechanical operations which need to be done for all scanned images that have little to do with "working at it."

Bob,
I went to the web site and it appears to be some interesting software. I hope that you receive some responses. I'd be interested in how well it works. The software doesn't look as if it is going to colour correct for you, but convert from neg to pos and apply an initial optimization in a more accurate manner than most other software (or at least this is what I'm lead to believe by reading their site). Colour correction as well as density and to a far lesser degree contrast is or should not be memory based nor should it require a 'go by' (as in the trany). A competent or experienced person should be able to achieve good density and colour balance by simply looking at the screen/test print. Over time youll learn to recognize the need for more or less m, or y not only in your own work but in any print or reproduction. This includes machine prints which I suspect is the kind of automation you are after.

i disagree with most of this

optimum image adjustment, preperation, output and presentation must be done on experience based guidelines, not by preset software controlled manipulation

in this way the artist/photographer/image maker builds his knowledge and experience and 'works at' creating his own imagery

 

[rgacpa]

Well I tried the sofware and was not impressed.
As I stated in my opening question, I'm simply looking for something to accurately translate my color negative, when it's scanned, to the proper positive colors. This sofware is not it.
The workflow that seems to be best right now is the Nikon Scan software with adjustment of levels and curves. It seems to produce the most accurate (to my eye) interpretation of the color negative to the image I remember.
At some point I'll photograph a color card or gray card in the first frame before beginning a series, and use that as a reference. My guess is that will provide even greater accuracy.
My thanks to all those who took the time to respond to my initial question.
Best of light,
Bob

 

[Lee L]

Bob,

This is yet more software and another bit of hardware, but it might help with some of your goals, not in scanning per se, but in color matching after scanning.

Picture Window Pro (www.dl-c.com) for Windows has the capacity to correct the scanned image of a Gretag Macbeth Color Checker (a long time industry standard). It displays a grid that you stretch and skew to overlay the image of the chart, then makes corrections knowing the color standards of the chart and how each of the 24 color patches should look. The software allows a download for a 30 day trial. The Color Checker is designed to simulate common colors and the inks are designed for lowest possible metamerism (appearing to shift color with changes in lighting) so it's well suited to the task.

I've also seen some collapsible frame panels (same makers as studio light control panels) that have white, medium gray, and black sections that might accomplish the same thing in a simpler fashion, and might work well enough with your current scanning setup without extra software. Some are small enough to carry around for a quick reference shot in difficult lighting.

Some carefully chosen paint chips, laminate samples, matt board or similar might be pressed into service to allow your software some standard white, medium gray, and black points for color balancing. Watch out for specular reflections if you go this route. Matte is better. Fotowand also makes some nice standard targets sized for carrying along in a camera bag.

Lee

 

[Lee L]

Forgot to mention: I'm a little confused about your use of Vuescan. Are you using color transparency settings for scanning negatives? Vuescan has a number of presets for color negative films.

Lee

 

[rgacpa]

Lee,
Thanks for you input.
I have a Gretag card; came with my EyeOne Photo. Just didn't have it with me on my trip (shot 8 rolls of Fuji Pro160H).
For the software I referenced in my first post, you have to scan to a RAW file in VueScan. If I want to use VueScan without that software, the best film selection seems to be color negative generic. Works OK, but takes forever to scan at 16X (2.5hrs per frame for 35mm....).
With Nikon Scan, it takes about 15min per scan and the results are fine; as good as VueScan for me.
I'll look into the software you recommend, but I'm thinking my EyeOne Photo came with the same capability. Your just supposed to use it with your digital camera. Same technique; shot a frame with the color tile and use that to white balance the remaining images shot in that light. Should work with scanned film I would imagine.
Again, many thanks for taking the time to chime in,
Bob

 

[blumesan]

From a purely theoretical perspective it seems unlikely to me that any software can (automatically) do what you are asking. Consider that the same negative, scanned on 3 different high quality scanners, will more than likely produce three different results, none of which may be true to the original scene. How can the software make the necessary corrections knowing only the type of film used?

Now if you could generate a profile, specific to your scanner and for each type of film you use, then perhaps you could achieve your goal. I am aware that VueScan does provide something like this but I have not tried it. I have also not found any glowing reviews from those that have.

So at the end of the day YOU have "to interpret the data my scanner (Nikon 8000) reports from the scan of the 3 or 4 emulsion layers (depending on the film being used)." The only success I have had is to open the image in PhotoShot and get to work.

Cheers/Mike

 

[rgacpa]

The only success I have had is to open the image in PhotoShot and get to work.

Cheers/Mike

Mike,
You are absolutely correct and that is exactly what I am doing. Here are 7 of my favorites (out of about 330) that I like best. I have 2 more to work on.
Dead Link Removed
Thanks for your input. You couldn't be more right. I hope you like the results.
Bob

 

[gr82bart]

Hmmm...I'm not sure, but I think this Plug-In is a bust. The only proper way to scan a negative or anything for that matter is to calibrate the scanner with an IT8 target, create an ICC profile for it, and then scan the doc using the ICC profile. I don't see how this plug-in can do that without a reference.

Hope that helps.

Regards, Art.

 

[donbga]

Hmmm...I'm not sure, but I think this Plug-In is a bust. The only proper way to scan a negative or anything for that matter is to calibrate the scanner with an IT8 target, create an ICC profile for it, and then scan the doc using the ICC profile. I don't see how this plug-in can do that without a reference.

Hope that helps.

Regards, Art.

Art,

There are no IT8 targets made for color negative materials or negative materials for B&W.

From the Monaco/XRite website:

Can I Profile for Color Negatives?

Q: Can I create a profile for color negatives.
A: Sorry, no. Profiling by definition is intended for positive materials only. It's goal to make original copy, photographs, monitors, printers, proofing systems, and presses visually match (within the physical limitations of each device). Visit the ICC (International Color Consortium) website for more details.

Don Bryant

 

[Bob Carnie]

I agree with Mike here,
I use a Imacon Flextight and even though there are profiles for each film on our lookup tables, these are only estimated starting points for the emulsions and 3 times out of 10 we get lucky with the profile , 7 times out of ten we must make adjustments for the negative.
As we process C41 here our own film process to scan seems more predictable , but with foriegn film *processed not by us* it is more difficult.
Ultimately we set the scanner to standard negative and then make adjustments with the tools in Flextight.
Transparancies are not as difficult as negative as every one states.
We have had to make specific profiles or mandatory adjustments for Pyro Black and White negative as there are no look up tables for trix in pyro.
For us , the negative scan is the most likely to pass our doors and potentially over time the most troublesome. Today we are finding our clients worried about negative stability*rightly so* and are scanning more.
So the problem gets larger. fading film or emulsion layers, film that has not been profiled , film that has been processed properly or stored incorrectly. I could go on with the problems associated with making a good scan .
Personally I use one film , one shot C41 process, and scanning important images as soon as I can.
As time goes on I believe better methods or software will emerge to help this problem.

From a purely theoretical perspective it seems unlikely to me that any software can (automatically) do what you are asking. Consider that the same negative, scanned on 3 different high quality scanners, will more than likely produce three different results, none of which may be true to the original scene. How can the software make the necessary corrections knowing only the type of film used?

Now if you could generate a profile, specific to your scanner and for each type of film you use, then perhaps you could achieve your goal. I am aware that VueScan does provide something like this but I have not tried it. I have also not found any glowing reviews from those that have.

So at the end of the day YOU have "to interpret the data my scanner (Nikon 8000) reports from the scan of the 3 or 4 emulsion layers (depending on the film being used)." The only success I have had is to open the image in PhotoShot and get to work.

Cheers/Mike

 

[jonathanbennett120]

rgacpa, it seems like you have a bigger issue to deal with in your scanning than color correction. You have some beautiful photographs there that are marred by thin lines that pass across the width of your images? It's not noticable in the waterlilly shot, but most of the others, especially the mountain coast shot and the background of the pink flower closeup shot have a very distinct distracting pattern from the scanner?!? Doesn't that bother you? I'm sure it's not on the negative and it can't be intentional? Maybe I missed discussion of that issue. If so, I'm sorry to re-tread.

J

 

[jd callow]

I have also tried using the prebuilt film profiles with no joy. My experience is:

• you need to calibrate the scanner. Most scanners have a target that sets focus and measures how accurately it is capturing colour. This way the scanner removes much of its own bias via a scanner profile and assures a sharper scan.
• many scanners have windows or tubes that need to stay clean otherwise they produce artifacts (generally lines on the vertical axis)
• some scanners can acquire and send the image data faster than the machine can process it which can also produce more artifacts (generally lines on the horizontal axis)
• Scan the image in excess of the final resolution (in a perfect world you'd scan at max res)
• Set the scanner to the edge of the histogram (or not), but don't try to colour correct, fix contrast, or density. This is not foolproof or the only way to do it. With some scanners you can set multiple exposures and combine them later or set one exposure and aim it toward your intent... The bottom line is that the best scanning software is only ok at colour correction.
• Do all colour correction, contrast and density adjustments in a program built for it such as Photoshop.

Within this list are things that can be done automatically (such as batch scans with minimal optimization) or with actions (adjustment layers with auto colour settings) that can reduce the time spent doing the redundant.

 

[rgacpa]

rgacpa, it seems like you have a bigger issue to deal with in your scanning than color correction. You have some beautiful photographs there that are marred by thin lines that pass across the width of your images? It's not noticable in the waterlilly shot, but most of the others, especially the mountain coast shot and the background of the pink flower closeup shot have a very distinct distracting pattern from the scanner?!? Doesn't that bother you? I'm sure it's not on the negative and it can't be intentional? Maybe I missed discussion of that issue. If so, I'm sorry to re-tread.

J

Yes. I'm trying to figure out where these came from. My guess is my scanner is getting old. The images posted have had some adjustments made to them (as this whole thread is about), so tonight I am going to look at the unadjusted scans to see if the lines are present. If they are, I'll need to open up the scanner and clean it.
Bummer...
Any suggestions on trouble shooting/cleaning a Nikon 8000ED would be appreciated.
Bob

 

[frugal]

With Vuescan, I just go through the process of locking the film base colour in my batch scanning. If memory serves, you select a frame with the most clear film base in the image (or you could just scan a clear frame for that matter. I'm using a Canon 9950F which is a flatbed so I can see all the frames I'm going to scan at once which makes this easier. You then go through a process of previewing that image, lock the exposure and then I think you preview again and lock the film base (it's been awhile since I've done it).

I then batch scan all the images, I find the colour isn't great but locking the exposure and film base colour means I end up with a scan where there's no clipping in the shadows or highlights. I do have to do some work with the colour but that's frequently on an image by image basis anyway.

 

[jonathanbennett120]

I can't say for sure, but in my experience those lines can be washed off the scanner glass with windex. Assuming this is some sort of a glass topped scanner? I've never seen inside one of hose enclosed film/slide scanners before.

I had a cheap HPscanner (until yesterday! I just got my Microtek 9800XL with TMA) that I used for 6 years and after maybe 4 years I noticed that "banding" happening in flat areas. It seems to coincide with the movements of the bulb as it passes over the glass. I thought it was a pattern that was "baked" into the glass and couldn't be removed, but it was very easy to clean off and they make it pretty easy to get flatbed glass off for cleaning.

Good luck with that. Great photos.

 

[rgacpa]

Jonathan,

It appears that this problem has been noted as an issue with the Nikon Coolscan 8000ED (I'm not sure about the 9000). The solution is to scan using the "Fine Scan" mode in the Nikon Scan software. This uses only one of the 3 LED's in the scanner to perform the scan. The scan resolution, dMax, and resulting file size are the same; it just takes 3 times as long. I rescanned the "Maui - Window" photo last night and the lines are gone. Done deal! I'm really glad I don't have to break into it to clean it or send it off to Nikon. I'll re-post the new scan tonight (1/31).

That said, it sure would be nice to have a better scanner (Imacon Flextight for example), but choosing between parting with $10k and waiting an extra half hour or so is an easy decision for me.

BTW, you have a very sharp eye for having noticed those lines on the jpeg file!

And thank you for your kind comment about the photos.
Bob

 

[Robert Budding]

I've had good success with the CF System COlorNeg plug-in. But I also have color darkroom experience, so using CC filters is second nature.

I've started a Flickr user group to share tips and tricks here:

http://flickr.com/groups/977912@N23/

 

[Kino]

I don't see anyone addressing what color space everyone is working in; that is a big factor in properly interpreting what a negative should look like on a display and IF it will translate to the desired target display properly.

Adobe and all other prosumer software writers are now in the throes of implementing and debugging colorspace profiles and how they transcode from one to another; woe be to me, it gives me a headache!

I hope you are all ready for the ride!

 

[nsouto]

Yes, indeed.
I'm using Adobe RGB (1998) simply because it's by far the best profile for my HP8750.

And I wish someone could tell me which fields do I have to set in the EXIF info to get browsers and image management software correctly identify that profile. I use exiftools.

I've tried to extract the info from an image file of a dslr set the same way, but no joy. There is something proprietary (any wonder? It's Adobe...) going on with that profile that just thwarts all efforts so far. And if any of the intervening workflow products eliminates the profile info from the scanner, I'm cooked!