DIY 31 Megapixel Enlarger

[koraks]

You have to pay attention to developer age, how many hours ago it was mixed, how many prints have been made, developer temperature etc.

Not really since print development typically goes to completion. This means that whatever effects aging developer has, are mostly on curve shape and not so much absolute sensitivity. The implication is that working with aging developer would necessitate (if anything) adjustment of your calibration curve, not so much exposure!

using 'factorial development'

That'll work fine and avoid most problems. Exposure can remain unchanged.

Exposure can be 'calibrated' using some kind of light metering device acting on base board level; this can be a regular light meter, lux meter, a fancy enlarging meter etc. Take your pick.

 

[Ethan Brossard]

Here is a selection of 36x48" images printed using the 16k LCD as a negative. The sharpness is quite good (there are around 235 pixels per inch along the long edge, and 172 pixels per inch along the short edge; the image uses an 11264 x 6224 pixel area of the display). With enlargements of this size (just under 8x enlargement) and viewing distances of less than a foot, the LCD pixel circuitry becomes visible, but further than that it is invisible. Every three or four pixels there is a small opaque bit of circuitry which shows up in the print as a tiny dot lighter than the surrounding image. They are very small, and impercievable at any reasonable viewing distance, but I am working on ways to hide them. I am currently testing an assortment of cinema soften and blur effect filters with that goal. So far I've found that Tiffen's Soft/FX filter does a good job of reducing sharpness without a reduction in contrast, but I have yet to narrow down which strength will hide the dots without also creating a meaningful loss of resolution. Cine filters are irritatingly expensive. These prints are on Bergger CB Semi Gloss paper, and I've also noticed that this paper (perhaps because it is warmtone, or maybe because of Bergger's recipes) is slightly less sharp than my standard Ilford MGFB glossy. it isn't a loss of sharpness which would be noticable with anything other than a high definition resolution target, but it has the result of making the dots slightly less apparent.

 

[AndrewBurns]

Here is a selection of 36x48" images printed using the 16k LCD as a negative. The sharpness is quite good (there are around 235 pixels per inch along the long edge, and 172 pixels per inch along the short edge; the image uses an 11264 x 6224 pixel area of the display). With enlargements of this size (just under 8x enlargement) and viewing distances of less than a foot, the LCD pixel circuitry becomes visible, but further than that it is invisible. Every three or four pixels there is a small opaque bit of circuitry which shows up in the print as a tiny dot lighter than the surrounding image. They are very small, and impercievable at any reasonable viewing distance, but I am working on ways to hide them. I am currently testing an assortment of cinema soften and blur effect filters with that goal. So far I've found that Tiffen's Soft/FX filter does a good job of reducing sharpness without a reduction in contrast, but I have yet to narrow down which strength will hide the dots without also creating a meaningful loss of resolution. Cine filters are irritatingly expensive. These prints are on Bergger CB Semi Gloss paper, and I've also noticed that this paper (perhaps because it is warmtone, or maybe because of Bergger's recipes) is slightly less sharp than my standard Ilford MGFB glossy. it isn't a loss of sharpness which would be noticable with anything other than a high definition resolution target, but it has the result of making the dots slightly less apparent.

Very cool, good to see some properly large prints being demonstrated using digital negatives. One option you might have to eliminate the slightly lighter pixels is to apply an inverse mask to the image being projected, I do a similar thing with my UV contact printing setup to eliminate any uneven exposure effects from the light source (which is very uneven). The technique is used a lot in scientific imaging, called flat-field correction. In your case if for example you know that every third pixel you get one pixel with slightly lower transmission (resulting in a lighter spot on the print) you can compensate by making every other pixel slightly darker, so that the final projection is perfectly even. Sure, this has the unwanted effect of slightly increasing your exposure time as you're making the LCD more opaque on average, but they would be even exposures.

 

[Ethan Brossard]

Here is a view through the grain magnifier of the pixel array. It is not actually the pixel itself which is darker, but a small diamond shape which repeats every 9 pixels in a row, or in approximately a 3x3 grid. even on the big prints, I'd say each dot is no more than 0.2 millimeters in diameter, they really are tiny.

You're idea is interesting, but I'm not sure if it would be compatible with my software which currently drives the display. perhaps if I generated an image file in photoshop with slightly brighter pixels around the dots, I could reprogram the software to overlay it on the output files. I may explore that route if the analog filter path doesn't lead anywhere.

 

[AndrewBurns]

Yeah that's basically what I do, I have a python script that I use to pre-process my image (because I use a raspberry pi to control my screen). The script flips, inverts, applies my cyanotype adjustment curve, applies the flat field correction mask, compresses the RGB channels for the monochrome screen and saves the resulting image as a PNG. Then I transfer the modified PNG image to the raspberry pi and I have another simple script on there that just displays the image full-screen and turns on the UV light for the correct length of time.