Desktop Vs. Imagesetter digital negatives

All,

I am experimenting, with help from a friend, making digital negatives for contact printing on silver gelatin paper. We are following Dan Burkholders book and updated Companion information. And we are working with several different materials including OHP film, mylar, and Epson backlit film. All are being printed on the Epson 7600.

But here is my question - Has anyone compared the quality of a desktop negative generated on one of these medium to those generated by an imagesetter? Dan Burkholder on his site suggests that none of these will be as good as an imagesetter negative for silver gelatin prints - but I would like to hear from anyone with personal experience.

I have prints from Lenswork, and those are intentionally generated from imagesetter negatives with an observable pattern so as not to be confused as originals. Of course I could do the experiment my self, having desktop and imagesetter negatives generated from the same file. However, I try not to reinvent the wheel every time.....

Thanks for your help.

Mike

From what I know imagesetter negs can be enlarged and desktop negs can only be contact printed.

steve simmons

I'd like to also hear of anyone's experiences with this process. Also, are there good online sites that have a concise discussion of the process and materials?

Michael, I have yet to see a sharp photograph on silver gelatin paper from inkjet negs. However some folks are making outstanding sharp prints alt process with the newer printers. I am currently working on a calibration procedure (using a densitometer and the yule nielson equation) for making stochastic negs on imagesetter, for silver gelatin papers, hopefully reducing test costs. I did not like halftones. I could see the pattern by eye even at 425lpi (3600dpi). Having said all that I find the imagestter negs to be good on silver gelatin paper but not as sharp as in camera negs. I do find the tradeoff tolerable as a salvage tool (hogged up negatives) and as a process for mass producing photographs. I'll post back if my tests are succesful in about 2-3 weeks. I have also found there are not that many places to get imagesetter negs made at a decent cost.

Steve,

Thanks for your reply. I think you were confused with imagesetter negatives and LVT (Light Valve Technology) generated negatives. The former would show pixels with enlargement. The LVT negatives are extremely high resolution, being generated on TMAX or Transparency or color negative film, and are capable of being enlarged. I have had two LVT negatives generated from digital files, but was not real pleased with the result. There are few labs capable of generating the LVT negatives.

Phil,

Can you elaborate on what the yule nielson equation is - and how you would use it?? What lab did you use for imagesetter negatives?

Thanks,

Mike

Michael, the yule nielson equation converts density to percent dot. There is a multiplier that compensates for a material's reflectiveness. When N=1 (for many graphics arts applications) the equation is same same as the murray-davies equation. These equations are used in graphic arts densitometers. There are a couple of these spreedsheets floating around. I got mine from Brooks Jensen at lenswork mag. and he advise N = 3 for silver gelatin papers. I could just tweak someone's curve for a given paper, however, I may have found a way to use a number of papers at min cost (2 tests each at about $30). Copygraphics is the bureau I have used for the imagesetter negs.

Right now I have been sending 600 dpi grayscale files (with the curve applied) and they have been applying the screen using rastus and then processing for 3600dpi output. No color management - no profiles! I have had streaking problems in the past and this is apparently a problem with the imagesetter at the higher dpi and Peter Ellzey is doing some workaround to solve the problem. I have tested for the best unsharp mask setting and found 50,10,3 to be the best. At 100,10,3 the images got "crispy" looking. Sharpen onscreen before applying these for the process.

Success. I can now construct a curve for a given paper with two imagesetter negs. But after the initial tests: initial test chart and then a neg to prove the curve construction, I'll only need to run a neg to prove the curve since I already have the test chart neg. The only time I'd have to print another test chart is if I used a different linescreen. We are now running a test using icefields. I am then going to compare a AZO print from a in-camera neg against a imagesetter neg.

Hi There,
I trying to make a hight resolution graphic to be use in etching process by using a hight resolution image on transparency. The size of the parts are very small around 015" diameter. We create the drawing in autocad and looks fine when zooming in but when we converted to bipmap or post script all the tiny details are gone. When saving the pic from autocad we set the output to be 4000 dpi x 4000 dpi that creates a large file but still the bipmap/ps files does not shows well as it shows on autocad. We could print directly from autocad but problem is you can not create negative graphic on autocad. Secondly, printing on regular laser will not give us a good resolution image. Third, imagesetter printers will not read cad files so the solution is to convert this cad file to either bipmap or post scripts. Any suggestions?



Luis..
Orthomechanics LCC
Orthomechanics.com

Conversion to PostScript allows accurate placement of lines, ensuring relative alignment of image elements on the final print. However, a regular imagesetter and output are probably closer to 2400 dpi or 2540 dpi output (depending upon manufacturer). You can get a potentially higher resolution with gravure, though unless you are doing a long run it would be very very expensive.

You might want to contact AGFA to see what they can produce. Currently AGFA have one of the smallest micron imagesetter capabilities, combined with a highly accurate line placement.

Ciao!

Gordon Moat
A G Studio

Gordon,
The problem is how can I send files to the imagesetter company with good resolution that shows the image exactly as we created in autocad ?

If we create a post script file 4096 dot per inch why lines .001" are lost when converting to PS or any other format?

.001" is 1 inch/1000 and the PS output image file was 1/4096 dot so .001" lines should be print easily.


Luis Hernandez
Orthomechanics LC