Digital Negatives for Alternative Printing

[L Gebhardt]

Larry,

How do you output to the image setter? I assume you "print" to it directly from your computer. Do you use any software such as 'Rastus' to prepare the image data? I'm definitely envious of your new "printer".

Don Bryant

Don, it is not an image setter, but rather a continuous tone printer. An image setter makes up an image by printing dots as either on or off. If you want to make an image with 256 gray scales you need a 16x16 grid of dots to represent 1 pixel, so the image setter's actual resolution is only 1/16th of its' stated resolution for gray scale images. The lightjet can vary the intensity of each laser over every pixel, so it has a true resolution of 2032 (I actually suspect it is lower due to light bleed in the film).

It came with some software called Pictographics System Manager that runs on a PC. This will print tif formatted images. It also handles color calibration and lets you apply curves to the image before printing. It will also RIP postcript files, though I really have no use for that feature.

I got the thing because I want to print color separations for dye transfer printing. The fact that I can also make digital negatives for contact printing is a very nice bonus. I also want to try making lo contrast negatives for Ilfochrome printing without contrast masks. If I had found a later model printer it would have been able to direct print onto color and black and white paper as well.

[eyaniv]

Wow,

This is right out of my league right now but sounds interesting.

Yesterday I purchased an Epson 1280 for the my negative output. It will have to do for now.

Next week I will be buying the Nikon Coolscan V ED for my input. Other than a calibration tool, my digital negative system is complete. Now I just have to learn the methods of work.

Ehud
circlesofconfusion.ca

[JeffD]

Sandy,

Here is something I have never understood. If inkjet output devices can handle 720, 1440, or more dots per inch, I wonder why we limit our scans to only 360 samples per inch? Is this because we humans supposedly can't resolve more resolution than 360? If we can scan at 720, or higher, is there any downsize to going above 360 that you know of? Maybe 720 is overkill, but if we have enough cpu power and memory, doesn't seem like it would hurt.

Also, I read in another thread your experiments with Ansco 130 (I think referred to as Developer 130 at Photog's Formulary). You standardized on two minutes of development, using AZO. I noticed the instructions I received with my first kit suggested 1 minute of development. Just curious what prompted a longer development time for you.

Thanks for your posts... they have helped me out on a number of occasions.

It depends on the format you are scanning and the output size. A good rule of thumb is to scan at 360 dpi at the final print size. That means if you scan a 4X5" negative at 1200 dpi you should be able to print at about 3.3X the size of the original, that is, about 13 X 17" in size. Scan resolution of 1200 dpi is inadequate for 35mm and for medium format unless you plan to make very small prints.



Sandy

[Dave Miller]

Sandy,

Here is something I have never understood. If inkjet output devices can handle 720, 1440, or more dots per inch, I wonder why we limit our scans to only 360 samples per inch? Is this because we humans supposedly can't resolve more resolution than 360? If we can scan at 720, or higher, is there any downsize to going above 360 that you know of? Maybe 720 is overkill, but if we have enough cpu power and memory, doesn't seem like it would hurt.



Thanks for your posts... they have helped me out on a number of occasions.

We scan for an output of 360dpi since this is the maximum most printers can provide.

[JeffD]

We scan for an output of 360dpi since this is the maximum most printers can provide.

It has been many years since I had an inkjet printer that can only do 360 dpi.
Epson quotes the 1280 as follows:

Maximum Resolution (dots per inch)
* Up to 2880 x 720 dpi

[L Gebhardt]

It has been many years since I had an inkjet printer that can only do 360 dpi.
Epson quotes the 1280 as follows:

Maximum Resolution (dots per inch)
* Up to 2880 x 720 dpi

There are places on the web that will explain it better than this, but here is my short version. Injet printers print an image pixel as a series of smaller dots. The smaller dot resolution is what is advertised. Internally most printer drivers scale the image to a fixed resolution, probably to make the driver easier to write. For Epson this is suspected to be 360ppi (pixels per inch) and 300ppi for HP and others. These 360ppi images are then turned into a dot pattern that has 720x2880 dpi (or what ever resolution you set in the driver). Think of this as very similar to the halftoning that happens in a printed magazine. Look under a loupe to see the details and it becomes clearer.

Other printers such as dye sublimation have a one to one mapping of image pixels and output dots.

[sanking]

Sandy,


Also, I read in another thread your experiments with Ansco 130 (I think referred to as Developer 130 at Photog's Formulary). You standardized on two minutes of development, using AZO. I noticed the instructions I received with my first kit suggested 1 minute of development. Just curious what prompted a longer development time for you.

Since the issue of print resolution with Epson printers has already been addressed I will simply add that I have made tests with image files at 360, 480 and 720 dpi and there was no difference, at last in resolution, in final image quality. So in practice you scale the image file to 360 dpi at the printing size and set your printer to maximum resolution, or best photo resolution (which is usually the same). Image files with resolution greater tha 360 dpi are simply larger and take up more space but do not increase final print quality on Epson printers.

As for development times with Ansco 130 I simply find one minute too short. The image is still changing density rather quickly at this point and I find it difficult to judge by a safelight the exact moment when it looks best. By two minutes the rate of density increase has slowed down a look, which makes it much less critical when I pull the print from the developer.

Sandy

[Dave Miller]

It has been many years since I had an inkjet printer that can only do 360 dpi.
Epson quotes the 1280 as follows:

Maximum Resolution (dots per inch)
* Up to 2880 x 720 dpi

Sandy has addressed this apparent paradox, but I would go a little further and suggest that for normal printing there is nothing to be gained by printing at a resolution setting greater than 240dpi. However for the purposes of producing digital negatives it is worth using a setting of 360dpi just to be on the safeside. I base this on my experience with a couple of Epson 1290's, their 2100 may offer better quality.

[JeffD]

That makes sense, I guess. I used Piezography for a while, and they advocated, at least with their earlier product, sending the image file to the printer at up to 1440dpi resolution, if you can. Their explanation was that their proprietary driver could handle that resolution, and apparantly didn't utilze the same dither pattern as the Epson driver. From comments I have read, the 720 or 1440 file resolution was noticably better than a 360dpi file, using this system, but I never made comparative tests. I usually would scan 720 pixesl per inch...

Thanks for the comments on the Ansco 130- I am trying to develop a standard workflow, to begin the digi-neg calibration, and this was one step that I wasn't sure what I would standardize on.

[sanking]

I usually would scan 720 pixesl per inch...

As for the actual scan itself I usually do it at the maximum dpi I can record on CD for a file. For example, I scan 5X7 negatives in 16-bit at 3200 dpi. This gives a huge file, over 600 mb, but one that I can store on one CD.
What I want to do is get enough information from the original scan so that I never have to scan the image again.

Now, for processing the image I will down-size to 360 dpi at the printing size. This reduces the file size a lot. For example, down sizing from 3200 dpi at 5X7" to 360 dpi at 10X16" reduces the file size from 600+ mb to less than 100 mb.

But if I ever decide that I want to make a larger image I can always go back to the original scan and downsize as necessary, always keeping resolution at a minimum of 360 dpi.

Sandy