Photo_Smith
Well-known
I was thinking I'd pile everything into my van and bring it down. If we can't get it sorted I could dump it in the channel
Pete
I guess beach combers in Holland have to look out for V700 flotsam...
Before you do that PM me first and we'll try to see if anything obvious is getting in the system.
I will help, but like I say I'm more a film guy than a scanner guy possibly I'm just lucky :angel:
A cheap nasty scanner with Epson scan on a free out of date film on a winter yellow sun day and no I haven't balanced this one my daughter just wouldn't let me take any more...
colour test by Photo Utopia, on Flickr
Yes her jacket is 'purpleish' which is a technical colour from the Munsell system
Ranchu
Veteran
I believe it just shrinks saturation a la perceptual rendering intent. I haven't experimented with profoto and whatnot with the scanned files.
cabbiinc
Slightly Irregular
Photo_Smith
Well-known
Negatives have an orange mask which varies slightly with exposure, also the numbers will be different. You have to use a different method it is possible in theory but not in practice.
Just use Munsell or some other colour patch system to match input (scene) to output (print/screen/etc) a negative is half of the process so mid grey in the final print is negative (exposure) + Print (exposure) = 18% grey.
I think negative have a gamma of 0.7 and I think (from memory) the gamma target is 1.5 for the negative + print output.
Just use Munsell or some other colour patch system to match input (scene) to output (print/screen/etc) a negative is half of the process so mid grey in the final print is negative (exposure) + Print (exposure) = 18% grey.
I think negative have a gamma of 0.7 and I think (from memory) the gamma target is 1.5 for the negative + print output.
cabbiinc
Slightly Irregular
It doesn't matter. The whole point of the IT8 target for calibrating your scanner is that your scanner reads colors the way they are supposed to be read. Otherwise your scanner may introduce a color shift that you'll then need to compensate for. All mask color compensation is done in the computer. The inversion of the image is done in the computer. Why wouldn't you want to start with the most true color image to begin with? Not calibrating your scanner is like shooting with a colored filter on your camera all of the time. I just don't get why it would work with slide film yet can't with negative film.
Ranchu
Veteran
Then do it.
Photo_Smith
Well-known
Give it a try, first try to find a negative IT8 target but I think you'll draw a blank (that should ring alarm bells) so you'll have to shoot one yourself.
The standard IT8 won't have enough patches to describe the colour space, so you'll need to have a custom one made or find one will a larger than normal gamut.
The resulting profile will only be valid for precisely the same exposure conditions as the target capture, including lighting colour, lighting intensity and exposure.
If you go this route you'll need to make one IT8 for each 'scene value' that is each different colour temperature or lighting possibility, the work will be extreme; the accuracy relatively low-possible but not worth the effort.
This is why Vuescan and Silverfast profiles for negative aren't very good, just ballpark.
Your main problems will be:
The orange colour correction mask built-in to colour negative films is not only exposure-dependent and different from emulsion-to-emulsion type, but also highly non-linear relative to exposure. Being produced from up to 12 separate emulsion/dye layers. The colour space of c-neg film is not a simple 3-D product, and vastly more complex and less uniform (in 3-D terms) than the colour space of transparency slide film or a digital camera. To characterise this increased colour space complexity, a negative film target will therefore contain many more patches than a normal scanner target.
Not impossible just really tough and you won't get any better results than Kodak's 'scene matching' method where they shoot a known target under a given light and balance that on final output with a colourimeter.
Wburgess
Established
I can see what Cabbinc is saying, and I've had the same thought a few times.
If he's on the same lines as me, he is wondering if using a slide it8 target would help vuescan (or other) get a more linear scan of a negative, in the fact that the scanner now knows how it sees colour.
Not a profile to be used to get a positive result from a negative in a profile.
Photo_Smith
Well-known
I can see what he's saying too, correctly characterising the scanner is to be applauded, but it won't help you one jot scanning negatives because of the reasons mentioned.
Slides behave very differently from negatives they sit in a different colour space using the CMM from slides won't help that negative space become more linear-it is what it is.
So the question is how will measuring product A that is linear, help with linearisation and characterisation of product B which isn't linear. Like fixing jelly to a wall blindfolded.
Give it a try... slide and print it8 targets are cheap I have loads of them. Have a go, then report back your methodology.
I think you'll soon see why its impractical.
Slides behave very differently from negatives they sit in a different colour space using the CMM from slides won't help that negative space become more linear-it is what it is.
So the question is how will measuring product A that is linear, help with linearisation and characterisation of product B which isn't linear. Like fixing jelly to a wall blindfolded.
Give it a try... slide and print it8 targets are cheap I have loads of them. Have a go, then report back your methodology.
I think you'll soon see why its impractical.
k__43
Registered Film User
Since the title is scanning color (thus includes slides) I want to add my experience on IT8 proofing of my scanners. I used a kodak target from LaserSoft with the following results
The V500 output went from flat/dull to punchy with not much impact on the color balance
The Plustek 8200 went from too green to a bit too brown with Fuji films (so I used the Auto Levels setting to correct that) - it was ok with Kodak slides
I never scanned slides with the FS4000 so I've not done the procedure
The LS-8000 almost had no difference in the output
One inconvenience of using the IT8 proofing for the scanner is that I keep forgetting to switch that off when I go back to C41 .. and really messes up the colors there.
The V500 output went from flat/dull to punchy with not much impact on the color balance
The Plustek 8200 went from too green to a bit too brown with Fuji films (so I used the Auto Levels setting to correct that) - it was ok with Kodak slides
I never scanned slides with the FS4000 so I've not done the procedure
The LS-8000 almost had no difference in the output
One inconvenience of using the IT8 proofing for the scanner is that I keep forgetting to switch that off when I go back to C41 .. and really messes up the colors there.
Lancelot365
Member
Here's the link to the RAW scan, which I think I did correctly...
https://dl.dropboxusercontent.com/u/7030288/Linked/_scan problems/400H_test.tif
Got this directly from CP. The color is pretty nice to me.
visualbassist
Member
my biggest problem with colorperfect, is the lack of consistency. anyone else finds it hard to match two images?
cpc
Established
You can get consistency with locking exposure while scanning the images you want to be consistent, if your scanning software can do this.
Alternatively, you can use the "Previous" initial settings (clicking the Initial>Previous button which cycles through various initial settings -> yep, CP interface can be confusing) and CP attempts to mimic the look of the previous image you worked on. This doesn't always result in a perfect match but I find it is often reasonable.
Michalm
Well-known
Here is my take, scan inverted with negfix 8 and then I use free program Lightbox image editor which allows 16 bit non destructive Tiff editing .This so far worked better then lightroom and Color perfect for me with less fiddling involved.