Vuescan & IT8 targets & colour casts

[dave_osborne]

Thanks Roger... yes I'm happier now I have a workable workflow at least - even if it is a bit more convoluted and slower than it used to be.

Very pleased with the eventual results though - my first step in photoshop was always colour correction, and in the 30+ scans I did today I haven't had to do that once :)

 

Would be nice to get it scanning right straight from the slide mind you.

[mendel_leisk]

Dave, sorry for overlooking your last paragraph in your initital posting.

 

For me, step one is outputting raw files of everything, *including* the it8 target. From then on, everything is Vuescan scan-from-disk, including the scanner profiling process. So my output is like your "tiffB", I think. This is a very convenient approach.

 

Regarding the naming conventions in scanner profiling process, the Vuescan helpfile is *very* confusing. You can name the files anything you want, and have them saved (or sourced) anywhere. Also, it is a good idea to fill in something descriptive in the description field. If you happen to use the profile in PS, that is what it will show.

 

I actually create my raws through my Minolta scanner's software, but that's another story. The outcome is essentially the same. I scan both the regular slides, and the target, with auto exposure off, and all exposure settings "centered".

 

One caution, if you get it into your head to output a Vuescan raw from a raw (needs to be enabled in Prefs tab), do be careful. The safest way is to have it set save "at scan". If you set it to save "at save" (at completion of scan), *and* have an ICC profile being applied in the color tab, your second raw will *not* match the first (it should match): it will have the profile applied (contrary to Vuescan Help file). Further caution, be careful to output your new raw somewhere other than the source raw directory, or atleast with different name. And set "warn on overwrite".<div></div>

[dave_osborne]

I am now well and truely cross-eyed after trying to work out what the difference is when scanning raws created during scan and at save time.

You're right there is a difference... a very small one but it's there. As to which is correct I have no idea!

I guess it's a cleaner flow if I just output 64bit RGBI files at scan and then scan & filter them.

 

The final colour results still aren't spot on after some scrutiny - some of the grey boxes in the final tiff (after scanning the IT8 slide) are definitely not r=g=b.. but is better than before so it hasn't been a waste of time!

 

Thanks everyone for your help.

[roger_smith4]

The truth is some of the grey boxes in the IT8 target aren't R=G=B. Did you look at the IT8 data sheet (try opening it with wordpad or another program) to see the actual values?

[User_1577653]

"Jeff.. interesting point about the IT8 mask.. how accurate does it have to be?"

 

If I'm not mistaken, Ed Hamrick once said that Vuescan looks at what's in the center 80% of each "box", when creating the profile. So the alignment shouldn't need to be hyper accurate.