Screen Cezanne Users Unite

[Peter De Smidt]

Welcome, Christian! And congratulations on choosing a Cezanne and a Cambo SC 4x5.

Yes, as far as I know, there is only an approximation of spi by dividing the front to back size of the crop into 8000 to get a purely optical number of samples, i.e. one without any interpolation. Note, though, that there is higher sampling resolution in the the other axis, due to the precision of the stepper system. In their tests, The Seybold found the the Cezanne resolved close to 6000 spi, the limit of their test.

[Christian-NRW]

Thanks Peter, good to know. I own a Silverfast USAF test target and will run some tests with different crops and orientations to see what the effective resolution is like in relation to the crop size.

I found Jim's great maintanance guide: http://www.thepwvhiline.com/FT5000_Clean_Up.jpg

I noticed the two white stripes left and right to the grid-transparancy. Any idea what they are for? They are in my cezanne as well, but I though this was something custom made by the pre-owner.

By the way, is it public wisdom that there is a german operation manual for Color Genius as well as the Cezanne hardware freely available as PDF under http://www.screen-dsd.de/fileadmin/Daten_Dainippon/pdf/ ? The german Color Genius manual is especially interesting as it appears not to be an 1:1 transtation but rather an interpretation written by an experienced user. If there is interest, I can summarize a certain amount of information.

Christian

[Peter De Smidt]

Jim's guide is terrific. Sorry, I don't know about the two white stripes. Thank you for the tip on the German manuals. I'm sure that someone down the road will find that useful.

[coisasdavida]

I managed to borrow 4 different manuals (start-up manual 4th ed, users manual 3rd ed, fit manual and precautions and restrictions 2nd ed).
I'm scanning them in order to have access to them after I return them.

I think I don't need this manual about FIT (monitor calibration software), does anyone use this and would like a scan of this?

[Christian-NRW]

Personally, I don't need it. but I think we should try to collect as much information as possible and distribute it within this group of users, so that the later availablitity is ensured.

I just found out that it is not a good idea to adjust the output profile in the fine tuning dialog. While one can set the range from 0-255 within the main fine tuning dialog, setting it for the output profile using the appropiate button and saving it in the respective file leads to being restricted to 8bit output, even if the modified profile originally is of 16bit nature. I have found no way out except to choose a different 16bit profile again.

I'm using the std16bit profile now that does not seem to be in the working profile directory by default. I copied it from the general profile directory. Output looks fine, d_range is below Hasselblads X5 judging by my samples, but not by much.

Christian

[coisasdavida]

I made PDFs of the three manuals and have them in my dropbox, if you need them, I can send a link in a PM.

[Christian-NRW]

Thanks coisasdavida, interesting read.

I've worked a bit on scanning negatives and found a workflow that seems to render quality results. I have not read about it here, so I'd like to put it up for discussion:

Instead of using the gradation curve, I invert the use of the HD and SD option.
For a new film, I choose a daylight photo with white and black parts (best case: shot of a test chart) and begin by setting HD(!) to 4.20(!) and SD(!) to 0.00(!). This will already invert the image, but channels are not equalized yet and there is exposure missing.
Then, I set color pickers on a black and on a white part of the image. I begin by first adjusting the SD values so that the black values of all channels are equal. I usually do not aim for 0 but for 10. The target figure is always the right figure of the color picker as the output is organised as: (original color / target color). The values of the white parts of the image are altered as well, which is why one should start by adjusting the SD values in the first place.
After that, I am changing the HD values to equalize the channels. There are two options right now: equalize all channels to the highest number of the three channels (which, from my memory, should be the blue channel), and strech the histogram in Photoshop afterwards. Or raise all channels to an equal value of about 250.

This way, I got very realistic color out of my negative scans and have not observed the severe noise near white tones I saw when scanning "as is" and converting afterwards with ColorNeg.

Christian

[Corran]

Interesting, though I'm not really following you. I'll have to try it next time I'm scanning to get a better idea of what you are saying.

Any chance you could post some scans and what film you used?

[Peter York]

Christian, my method is similar. I invert a color negative using SD for the highlights and HD for the shadows. I then alter the SD and HD values by about .1 to prevent clipping. I do not equalize the RGB values, however, as I assume each has their own cutoff level. In earlier scans I was encountering a weird, repeating pattern of lines, but I think this is related to errors I had made with my grafmatic holders in the past (not pushing the septum back in). My most recent color negs have been fine.

Diallo's book on B&W distinguished between hardware-level and software-level adjustments. My guess is that setting SD and HD this way on the Cezanne is a hardware-level adjustment (rather than software-level, where the scanner pulls an image and then alters the values in software). If this is correct, then results should exhibit less noise because you are not mashing the histogram like ColorNeg.

[Corran]

Speaking of this - I need to try some different ways of scanning color negatives, because the Cezanne is giving me really bad banding obvious in blue skies. There will be anywhere between 3-10 faint orange strips running through the sky that looks like the orange mask is still slightly there. It's only on negative film and only in blue skies (or only noticeable then?).