Silverfast SE 48bit HDR neg output question

[Tim Lookingbill]

With Silverfast SE HDR 48bit RAW neg output what is SF referencing to neutralize the orange mask since the selected Negafix profiles aren't applied in this mode with this version of SF. Was wondering why scans viewed in PS appear bluish/purple instead of the familiar orange color you get scanning negatives as positives. Is a generic profile used and if so what's it based on?

I'm using SF version E-6.4.4r2 with a refurbed Epson 4870, Mac OS 9.2.2.

[Kirk Gittings]

Sounds like you are not really getting a raw scan. It should be orange. The only thing that should apply in Raw is the basic scanner IT8 calibration. Something somewhere is set to some automatic interpretive function.

[Tim Lookingbill]

Thanks, Kirk.

I don't think I want to turn it off whatever it is. It's giving me some real good scans with very little work required to get a scene correct response. All I have to do is invert with the scanner profile assigned and adjust only the highlite sliders for each RGB channel in levels to set color temp and density. Then on some frames maybe a tweak to the middle slider and a slight increase in green shadow adjust with the bottom left slider.

I can then assign other wide gamut working spaces like NTSC or JoRGB to fix clipping the SF embedded scanner profile causes. I've had prints of some Kodak HD400 scanned and printed on my local Noritsu minilab to compare against another devices rendering of color and the SF version is far better and more realistic. I don't check the color against the minilab scans, only their prints which have more down to earth color intensity and look closer to the SF scans.

I've never had a scan that took well assigning a wide gamut worker space before and it's left me wondering if I'm doing things right or maybe my color judgement is off.

I'm not using an it8 calibration since they're negatives. The EpsonScan version of the same negatives require far more work to get the same results than what I have to do using SF.

[Tim Lookingbill]

I might have an idea what's going on after retracing my search steps over at the Silverfast forum yesterday. There's a lot to digest over there because of the many versions and upgrades you have to rumage through.

I remember reading a thread dated around 2003 or maybe later about when they first implemented Auto Tolerance default slider setting preset a little to the right. It was mentioned that later versions of SE would have this feature embedded internally preset the same way but without the ability to gain access for customizing. Maybe that's it.

[Kirk Gittings]

But if you are scanning them as positives, Sf is utilizing some general default IT8 trans. profile to calibrate itself with. Raw does not mean an uncalibrated scanner it means an unadjusted scan.

You could simply scan them in as unclipped negatives in one of the large color spaces allowing only for the mask to be filtered out.

I'm still not clear as to why you are taking this approach. I still think it is far better to adjust the basic color and inversion in SF, which is designed to do this. What does this approach accomplish? I tried many approaches to inverting color negatives in PS, but never got as good a product as I did once I learned to use SF to its full extent.

[Henry Ambrose]

What Kirk wrote.

You seem to be doing a lot of extra work of no benefit.

Using minilab prints for reference seems unlikely to help much. Assuming you have calibrated your monitor and your color settings are correct you should be able to see on screen what your colors look like and get that result in your prints.

[Tim Lookingbill]

I was using the prints as a color temp reference for outdoor light not actual color for color matching. Getting reasonably accurate color temp of midday sun lit subjects editing negatives can be quite difficult for me at least because of the negative's global cast that messes with my eyes. Negafix previews for Kodak UC & HD400 film gives somewhat a mud green cast and I can only output in 24bit and I have to select Kodak Royal Gold 400 Negafix profile to get the best previews.

No problems matching screen to print. Just wondering why my 48bit HDR scans are coming out already neutralized. The histograms for these scans in PS show all three channels pretty much lined up squashed and positioned to the left with the red channel the most diminished and the green and blue pretty much the same size. Somewhat dim previews with a blue/green cast when inverted.