Scanning Ektar

Bormental said:

Helge, it really is the same process, see my comment above. You're right, the paper doesn't "divide" anything. The division happens as the light goes through the negative in an enlarger.

Click to expand...

Then it should be the same for the scanner.
Light is light.
A simple alpha blend of the appropriate cyan colour or just the equivalent channel adjustment should be enough.
That is not devision or multiplication, just subtraction/addition.

Found this post by PE:
https://www.photrio.com/forum/threads/orange-mask-on-ra4-paper.149017/#post-1952050
Apparently the paper does do some compensation.

Adding and subtracting filtration in an optically based system works the same as dividing one layer by another in a software layer based system.
I assume that this is due to how hues are recorded digitally.

Helge said:

Then it should be the same for the scanner.
Light is light.
A simple alpha blend of the appropriate cyan colour or just the equivalent channel adjustment should be enough.
That is not devision or multiplication, just subtraction/addition.

Found this post by PE:
https://www.photrio.com/forum/threads/orange-mask-on-ra4-paper.149017/#post-1952050
Apparently the paper does do some compensation.

Click to expand...

What Ron states is spot on the money (even in Photoshop) - the problems come from what scanners and cameras can do to otherwise innocent negatives - but even then it can be worked round.

Sample & divide before inversion is essentially doing nothing more than sticking a very specific colour of filter in the virtual light path - there are many other ways to the same end - via corrections that end up at the same point on a curves layer before inversion. The headaches appear when light sources wander from they should be tuned to for the system & under the surface DSP rears its ugly head.

Helge said:

I assume that this is due to how hues are recorded digitally.

Click to expand...

Well, the obvious thing is that division guarantees you won't go below zero The less obvious thing is that a digital color is defined by the ratio of the chosen primary colors, not by their absolute values (that's luminosity). Thus, to preserve the ratios you divide/multiply, you don't subtract.

Actually I do not use "sample and divide" method. I use (simpler for me) approach. I sample the mask, then I take red/green and red/blue ratios and multiply blue and green channels by them, effectively giving blue&green channels the same boost as does RA4 paper.

MattKing said:

I assume that this is due to how hues are recorded digitally.

Click to expand...

Bormental said:

Well, the obvious thing is that division guarantees you won't go below zero The less obvious thing is that a digital color is defined by the ratio of the chosen primary colors, not by their absolute values (that's luminosity). Thus, to preserve the ratios you divide/multiply, you don't subtract.

Actually I do not use "sample and divide" method. I use (simpler for me) approach. I sample the mask, then I take red/green and red/blue ratios and multiply blue and green channels by them, effectively giving blue&green channels the same boost as does RA4 paper.

Click to expand...

I don't know how you managed to quote me but attribute it to Helge, but I'm not offended and I hope Helge isn't either!

I just had a glass of my late Sunday red. Apologies if I clicked on the wrong quotation button. The last thing on my mind is offending anyone. Sending virtual hugs to both you and @Helge !

Lachlan Young said:

What Ron states is spot on the money (even in Photoshop) - the problems come from what scanners and cameras can do to otherwise innocent negatives - but even then it can be worked round.

Sample & divide before inversion is essentially doing nothing more than sticking a very specific colour of filter in the virtual light path - there are many other ways to the same end - via corrections that end up at the same point on a curves layer before inversion. The headaches appear when light sources wander from they should be tuned to for the system & under the surface DSP rears its ugly head.

Click to expand...

Oh, I think this is just Photoshop lingu.
Been some time since I used PS, and then I used it quite differently than this.

Many ways to the top of the mountain, but over them all shines the same moon, and all that.

I played around with conditional blending to achieve different effects in PS. But this is just a rose by another name (alpha blend/or just blend) AFAICS.
Of course you can’t just add or subtract bit ranges.
You’d just very quickly end up clipping or going into crushed black.
As a given you’d have to normalize, and that is almost always achieved by either dividing or multiplying.
But the real math/end result performed is a simple additive/subtractive blend.

I get ninety percent of the way there by just sticking cyan Wratten filters in front of the light source. I haven’t perfected it though.

That’s why it would be interesting to know whether the paper performs some function that isn’t “simply” the same as what the filter pack tweaks and finishes?

MattKing said:

Correct.
I'm not sure though that sampling in the rebate area provides the necessary correction.

Click to expand...

That has been my experience. I deal with it by simply white balancing it out in raw sample land before the inversion. If you're in photoshop, the easiest way to do that is via the divide method as described previously as photoshop does not give you an easy way to perform a raw white balance. Different way to get there, but effectively does the same thing at the end of the day.

Adrian Bacon said:

That has been my experience. I deal with it by simply white balancing it out in raw sample land before the inversion. If you're in photoshop, the easiest way to do that is via the divide method as described previously as photoshop does not give you an easy way to perform a raw white balance. Different way to get there, but effectively does the same thing at the end of the day.

Click to expand...

You can also white balance it out with curves pre-inversion in Photoshop too - if you don't mind working with the clipping warnings on the individual RGB channels.