ChartThrob V1.01

[Bjorke]

Originally Posted by Keith Taylor

My chart is printing (in palladium) with a nice range of tones except for the top two lightest bars. When I scan it and rerun ChartThrob, it suggests I increase the exposure to compensate, but the tones in the darker areas would block up and I'd have the same problem in reverse, right?

Ah, okay, I think I get it. Thanks — it's tremendously helpful when great printers like yourself, Bill, David (everyone here, in fact) tell me what you're trying, needing, and expecting. A real gift to me, since my own printing skills are only so-so and my pt/pd experience rather thin. Thanks very much!

ChartThrob actually expects that the first print being scanned will have loss at both the highlight and shadow ends of the scale. That's why the sample chart runs from full 100% white to full 100% black -- there's an assumption that the computer file, and the resulting negative, will have a greater range than the contact print.

When ChartThrob looks at the scan of that printed chart, it looks for those blocked-up ranges and uses them to know where your print dmin and dmax are, then builds the curve so that subsequent pictures will hit those in a way that corresponds to the computer file's blacks and whites. The "Suggestion" box you saw is just the program telling you that you have more highlight loss than shadow loss, that it might be possible to get a little more range.

A good way to test the result (I can't believe I didn't put this in the instruction webpage, DUH!) is to take the original uncorrected chart, apply the generated curve to it, and print again. If things are working properly, then you should get a full range from the new curve-corrected chart.

(As for the guy in the snap, it's my buddy Michal Daniel (or here too) - among other things he shot all the big images that decorate the new Guthrie Theatre)

[BillSchwab]

Kevin,

Didn't want you to think I forgot. Got a heavy printing load right now and am having little time to give it another try. But, I did realize one thing. I have all color controls shut-off in my version of CS, but not in the new CS2. Once I changed that all looked different. Still haven't had the chance to do another test neg though. I'll let you know what happens. No sense in sending you test results of my mistakes.

B,

[Keith Taylor]

Originally Posted by Bjorke

A good way to test the result (I can't believe I didn't put this in the instruction webpage, DUH!) is to take the original uncorrected chart, apply the generated curve to it, and print again. If things are working properly, then you should get a full range from the new curve-corrected chart.
(As for the guy in the snap, it's my buddy Michal Daniel (or here too) - among other things he shot all the big images that decorate the new Guthrie Theatre)

That makes complete sense, I'll give it a try. The new Guthrie is an amazing place and in fact I'm working on a neg of the building right now! I love the way they've incorporated your friend's photos into it.
Keith.

[SanFranciscoNudes]

Well, I have some mixed results. I first printed a new cyanotype using the ChartThrob curve. The highlights were a bit blown out - not awful but definitely not where they should be.

I printed a Vandyke and it was just plain awful - very washed out.

However - I was never totally convinced I was getting a "good" scan, both in terms of the gamma being set correctly and in terms of the white point being set correctly.

I then decided to just take a photograph of the test charts just to see what would happen. I have zero idea if this even should be legit, but bear with me. I generated curves, which were qualitatively the same except towards the high end where they were substantially different. And applying them to a positive and just eyeballing the result, they both were much closer to what I would have expected. Basically if I swap between scanner curve and the photo curve you can see the problem areas are the ones that change and they change in the right direction.

I printed one negative for each process and hope to print them tomorrow. I had a little unrelated excitement where I broke my coating rod (sigh) but fabricated a new one in the lab today. It seems OK but we'll see. I'll post results when I have them.

In the meantime, obviously if the images are good I'll go with the new curves, but "should" this work? If you take a picture in RAW mode, and set the white and black point manually when converting them, should the resulting image be linear?

[SanFranciscoNudes]

Judging from one wet print from each process the results look very nice - nice enough I made a test chart with platinum toned vandyke. More in a day or two when everything's dry...

[tsobota]

Kevin,

Yesterday I installed ChartThrob. It promptly bombed when analysing the scan of a print.

Looking at the code, I see the following in lines 874 and following:

pColour.rgb.red = findPV(doc.channels["Red"].histogram);

pColour.rgb.green = findPV(doc.channels["Green"].histogram);

pColour.rgb.blue = findPV(doc.channels["Blue"].histogram);


Sadly and surprisingly, this will not work in international versions of Photoshop. Some brilliant mind in Adobe decided to translate the color names, so for my Spanish version "Red", "Green" and "Blue" don't work, but instead "Rojo", "Verde" and "Azul" do.

So I just changed the color names to their indexes:

pColour.rgb.red = findPV(doc.channels[0].histogram);

pColour.rgb.green = findPV(doc.channels[1].histogram);

pColour.rgb.blue = findPV(doc.channels[2].histogram);


and it worked perfectly.

May I suggest this change ?

Tom Sobota
Madrid, Spain

[Bjorke]

Wow Tom, that is pretty bent. I'm glad that you found a solution quickly, as I'm far from home and unable to do any ChartThrob tweaking this week.

I can't for now imagine why this fix wouldn't be fine, I'll check on localization of the color names too.

kb

[SanFranciscoNudes]

Originally Posted by SanFranciscoNudes

Judging from one wet print from each process the results look very nice - nice enough I made a test chart with platinum toned vandyke. More in a day or two when everything's dry...

Just to follow up on that, I've made some more prints with each process and they all look very nice. I just printed negatives off of ChartThrob curves for platinum toned vandyke and Ziatype, I'll hopefully comment in a few days on how those went.

[Helen Bach]

I'd also like to thank Kevin for one of the most useful pieces of software for the creation of digital negatives, and for showing the potential for javascript in CS2.

Best,
Helen

[SanFranciscoNudes]

Originally Posted by SanFranciscoNudes

Just to follow up on that, I've made some more prints with each process and they all look very nice. I just printed negatives off of ChartThrob curves for platinum toned vandyke and Ziatype, I'll hopefully comment in a few days on how those went.

Very happy with both, as it turns out!

If you want to see a Ziatype, look here: (warning - contains nudity) http://www.eljay.org/ebay/ziatype5.jpg