Is Colorperfect worth the cost?

[hoffy]

Hey Folks,

While I love taking colour, I have always struggled to get it right. I see Colorperfect mentioned in a few places, but have balked at the cost.

I did down load the trial yesterday, but couldn't get a real good feel for it (those grid lines are VERY annoying!).

I suppose I am after any testimonials from PH3 members. Is it worth the cost?

Cheers

 

[adelorenzo]

It does a nice job on the color conversion but I don't find it gets me any better results than I can get through other means. Plus it adds another step in an already tedious scanning workflow.

I've bought just about everything out there to try to get a decent film scan. Epson flatbed, Nikon coolscan, Vuescan, Colorperfect, Better Scanning holders, wet mounting supplies.... My scans are still terrible so at some point I need to chalk it up to operator error.

 

[hoffy]

Thanks for your reply. My colour scanning sucks as well. In the end, it often turns into throwing mud on the wall and seeing what sticks.

 

[Born2Late]

I tried color printing about 25 years ago and had an terrible time getting the color correct. It wasn't until my first real vision exam several years later that I found out that I was color blind. Oh well, I like B&W better anyway.

 

[Poisson Du Jour]

It's not used and has never been used in my lab-based workflow. Nothing more is needed than a colourimetrically-correct scan and profiling ready for the target printer (broadly). Additional software thrown into the workflow means additional time, integration and complexity and the risk of errors that are not easily reversed. This is not the way to go when the colour process is effectively, really, a simple one, just made more and more difficult by people imagining it can be mastered by adding more and more software. That's wrong.

 

[Les Sarile]

I've bought just about everything out there to try to get a decent film scan. Epson flatbed, Nikon coolscan, Vuescan, Colorperfect, Better Scanning holders, wet mounting supplies.... My scans are still terrible so at some point I need to chalk it up to operator error.

Which Coolscan+software are you using?

I have over 45,000 frames of various film brands and types, automatically scanned with no pre or post using Coolscan 5000/9000 and Nikonscan 4 that good to go. I also have a few thousand scans from various other desktops (flatbed and film), mini labs (Agfa, Fuji and Noritsu) as well as Imacons for comparisons.

This one is a frame of Kodak Gold 100 automatically scanned with the Coolscan+Nikonscan as well as Noritsu.

With regards to colorperfect, I've never used it.

 

[Lee Rust]

I now use a DSLR for 'scanning' negative film and ColorPerfect makes color reversal quick and consistent... much more so than the alternate methods I've tried. The wide variety of preset film type profiles come in handy too. Although most of the advanced control parameters are quite beyond my comprehension, the program seems worth the cost just for reversal alone. Life was much easier back in the day when I had a (working) LS8000 and NikonScan software that managed the color issues quite nicely.

 

[WilmarcoImaging]

Does anyone know the principles of operation or basic algorithms used by minilab scanners like the Pakon, Noritsu or Frontier? Seems like those concepts could be outlined into straightforward postprocessing steps for color negative scanning.

 

[Les Sarile]

Does anyone know the principles of operation or basic algorithms used by minilab scanners like the Pakon, Noritsu or Frontier? Seems like those concepts could be outlined into straightforward postprocessing steps for color negative scanning.

Given the example I posted above from a Noritsu, it is probably not a good algorithm to implement.

I am eager to try out the new Nikon D850 that has a built-in color negative to positive conversion routine. I am hoping it will be as competent as the Coolscan+Nikonscan as a potential replacement. Granted it will not have the magical ICE dust and scratch removal of the Coolscan+Nikonscan combination.

 

[NortheastPhotographic]

I think people who do DSLR duping use Color Perfect to varying levels of success. I'd rather just get a Coolscan V, or 5000 and scan with Vuescan. Frankly though the best color negative scans I do come from the Fuji Frontier SP-3000. Even at it's lower pixel resolution and bit rate the scans are just fantastic. 35mm comes in at 19mp and that equates to about a 12x18 at 300ppi. It's just about perfect for the format if you ask me.

 

[hoffy]

Does anyone know the principles of operation or basic algorithms used by minilab scanners like the Pakon, Noritsu or Frontier? Seems like those concepts could be outlined into straightforward postprocessing steps for color negative scanning.

The question that I was afraid to ask......

Please correct me if I am wrong, but when it came to mini labs, I thought that the process of reviewing the colour balance and then setting up a scan to be printed was in a way similar to doing an optical print - based on setting the values for C,Y and M. Is that correct? How does that work? Did each lab need a base set of profiles for each film?

 

[Lachlan Young]

It's often simpler to do it by hand once you know what needs to be done. Main thing is getting a 16-bit positive scan/ image with some negative rebate without any clipping whatsoever (desktop flatbeds are not up to this in my experience owing to Dmax shortcomings - high end CCD or PMT scanners are really kind of necessary, also possibly one of the few things DSLRs are good for) and working in ProphotoRGB or similar, sample the rebate, create a new layer filled with the sampled colour, set to 'divide' which will remove the orange mask just about perfectly. Then you can flatten, do your inversion, create some curve layers to set the black and white points individually for RGB using the clipping warnings, do some tonal adjustments, fine colour correction etc & in less time than it's taken me to write this, you'll have a very good looking file that'll show you what film is really capable of. There are several ways of doing this, but the main thing is sampling & dividing out the orange mask before inversion.

 

[jim10219]

I tried Colorperfect based on the many recommendations out there on the internet and gave up on it. It's presets didn't give me good results. It's manual controls were unnecessarily complex and difficult to use. I eventually went back to using the Photoshop Curves tool. I can get excellent results in very little time with the Curves tool. But I will warn you, that tool took a lot of time and experience to get good at using it. It was extremely difficult at first, but eventually, I developed a feel for it, and now it doesn't take me long at all to color correct (or color match) a photo. The advantage to using the Curves tool so often is that I can now color correct just about anything in a very short time, not just color negative film scans. It's one of those skills that comes in handy a lot more often than you would think, once you've developed it.

I find the Curves tool to be much more intuitive than the Colorperfect tools. The Curves tool is simple, graphical, and extremely robust. It just takes a bit to put it all together in your mind to where you don't even have to think while using it. It's kind of like driving a car with a standard transmission. It's difficult to learn, but not to use once you get the hang of it.

 

[Ted Baker]

I have been working on my own solution recently, hopefully be ready for wider release shortly. If would be keen on doing a few inversions for if people are happy to send me files (3 or 4). Via dropbox or google drive.

The scan can be from a DSLR (needs to be a a raw file) or scanned as a positive in a film scanner with no adjustments and no clipping, or a raw scan in vuewscan/(HDI silverfast) PM me if your interested.

 

[hoffy]

Sure, why not

 

[ITKI]

Yes, will PM you..

 

[Ted Baker]

Does anyone know the principles of operation or basic algorithms used by minilab scanners like the Pakon, Noritsu or Frontier?

I would like to but no luck, I have a had a good look inside the config, for the Pakon, but nothing really meaningful other than few LUTs, I doubt the source code even exists any more. The current state of this technology is a real shame IMHO. There is a lot more public information in the cinema world, I guess the vendors are much less frightened of sharing that information, as those vendors are more interested in making movies than selling film.

 

[WilmarcoImaging]

Ted thanks for the comment. I would think in this hybrid world we live in, that Kodak would provide this information. It would facilitate film sales imho, because it would make things easier for the hybrid crowd to realize high quality results. Who/what businesses would be harmed if Kodak published information on how to scan C-41 negatives, for best digital display quality?

 

[Poisson Du Jour]

I would think in this hybrid world we live in, that Kodak would provide this information. It would facilitate film sales imho, because it would make things easier for the hybrid crowd to realize high quality results. Who/what businesses would be harmed if Kodak published information on how to scan C-41 negatives, for best digital display quality?

Scanning and post-production has nothing to do with Kodak, other than to make passive commentary that an emulsion is optimised for scanning (not entirely as vague as it sounds). Production of images and prints from scans is a professional qualification in many modern labs, requiring an understanding of colourimetrics, the interplay of the media, monitor, printer and calibration of each and all to work as a whole.

The "hybrid crowd" are well versed (and have been for decades) on how to produce quality printing from scans.

 

[Ted Baker]

Who/what businesses would be harmed if Kodak published information on how to scan C-41 negatives, for best digital display quality?

Maybe digital cameras, after all they invented them They are certainly missing a trick IMHO. I find it hard to believe with the combination of immense intellectual assets, and bright sparks working there, that they could not whip up a few white papers on the topic...

 

[Lachlan Young]

Maybe digital cameras, after all they invented them They are certainly missing a trick IMHO. I find it hard to believe with the combination of immense intellectual assets, and bright sparks working there, that they could not whip up a few white papers on the topic...

The answer is, I suspect, very simple: consumer grade scanners cannot meet the necessary baselines of accurate focus, optical systems free of gross aberrations & a Dmax capable of handling most negative stocks (never mind transparency) to make it sufficiently worthwhile writing up relevant approaches. Those who use high end CCD & PMT scanners have largely worked out suitable methodologies on their own - and it should be remembered that those machines came about in an era of transparencies being the dominant medium for colour repro.

Datacine and similar cinema film scanners are all professional grade machines and meet the baseline quality necessities as a matter of course, thus writing up protocols for scanning negative stock (the dominant cinema origination medium at the time) was worthwhile.

That said, I don't doubt that an amazing mass produced film scanner handling up to 8x10 could be built using the best of today's technology, but probably not for less than a middling four figure (USD) sum per unit.

 

[kruiwagen]

I've used it for a year now and I think it's worth it. Although it is more useful when scanning negatives. Choose the profile from the dropdown list and the colors are usually accurate. It's a lot faster this way then adjust the colors by hand for each scan. Especially if you're scanning multiple rolls of film. Color reversal scans usually need some tweaking beceause slide scanning always leads to color casts (at least for me). I'm scanning with a Pacific Image PrimeFilmXA and use Vuescan.