DIY 31 Megapixel Enlarger

[travelight]

I just bought a 7” 10k screen to put in my 5x7 durst. I’m working on some software to control it and doing the coding on my MacBook Pro. I won’t use the Mac in the darkroom, but would like to use the new screen for testing/debugging my work.
Has anyone managed to create custom resolutions for one of these screens on a Mac? Mine is showing it as a virtual screen when I attach it, which appears to prevent the Mac from applying any custom resolutions even if I could make one.

 

[Ethan Brossard]

I just bought a 7” 10k screen to put in my 5x7 durst. I’m working on some software to control it and doing the coding on my MacBook Pro. I won’t use the Mac in the darkroom, but would like to use the new screen for testing/debugging my work.
Has anyone managed to create custom resolutions for one of these screens on a Mac? Mine is showing it as a virtual screen when I attach it, which appears to prevent the Mac from applying any custom resolutions even if I could make one.

I have not found any way to run these screens on mac or windows well. I'd suggest getting a cheap raspberry pi to run the screen using Linux (it's what they were designed for).

 

[Ethan Brossard]

I was thinking of picking up a 16k display and the Raspberry Pi 5 and dive into this rabbit hole myself
I was wondering if you would mind sharing some info about what OS image/config you used to get the display running on the Pi @Ethan Brossard ?

I am running the Pi 5 with the standard Raspberry Pi OS. There is one config file I had to modify to get the display to run on it, I believe it is /boot/firmware/cmdline.txt I can't remember the exact code I added to it, and the pi is in the darkroom at work currently, but I'll try to remember to get it soon.

There are a few things to keep in mind about the 16k display:
- It only has a bit depth of 3, so can only display 8 tones of gray. This leads to an extremely posterized image if you want to use it straight out of the box.
- The way it reads data, to get a group of 8 grayscale pixels, you have to encode them properly into a single 24-bit RGB pixel
- The pixels aren't square, so you need to give it a "stretched" image in order for the image to render properly
- As best as I can tell, the display functions sort of like two 8k displays next to each other, and where they meet in the middle you can get some weird overlapping / glitching if you don't modify the image you send the display properly

I'm working on software to get around all those issues, and it works, but at the moment its very much not user friendly and likely wouldn't be usable by anyone but me. Hopefully in the next couple months I'll have a user friendly version ready.

Once it's ready, I'm planning to go into production with it on a very small scale. I've already had a few people reach out to me about purchasing it. My initial plan was to build a negative carrier unit which could be installed in existing 8x10 enlargers, but Beseler has a patent from several years ago for a digital enlarger negative carrier that fits into existing enlargers. I'm not fluent enough in legalese to know if it would prevent me from selling a digital negative carrier, but there's a chance I'd have to sell it as a whole enlarger rather than just a negative carrier.

 

[michaelbirkmose]

Received my 6.8" 9K display a few days ago and put together a PWM driver in Python (might build a lower-level solution later). I’m running it strictly in 1-bit mode. Initial results look promising.

For a quick low-tech test I used a laptop screen as the light source and contact-printed through the LCD. The prints came out surprisingly blurry — probably because the crystals sit inside the LCD casing, and I also didn’t use a sheet of glass on top to ensure proper contact. Even so, the basic proof of concept seems solid.

I coded it so exposure is controlled via a LUT: not by changing the input image, but by assigning exposure slots per grayscale value and modulating them through timing.

I’ve signed up at a community darkroom and plan to give it a more “real” test in a week or two. For now I’m running with the protective peel still on the LCD, just to avoid too much handling. For those who’ve worked with these displays: how fragile are they in practice? Do they scratch easily, or are they manageable with reasonable care? Is it possible to safely clean them with alcohol or similar solvents once the protective film is off?

All in all, this feels like a really fun project. Thanks to @avandesande for sharing the original idea!

 

[michaelbirkmose]

Managed to get something on paper using a Durst 1200 - looks promising!

 

[Ethan Brossard]

Managed to get something on paper using a Durst 1200 - looks promising!

looks great!

 

[Glauber]

I’ve spent the summer putting my machine together and, after finally getting through the software side, disaster struck on the very first full assembly test and the eDP connector on my 8K 10.1" screen broke.
Fortunately, all the solder pads on the board survived, so I’ve ordered a new connector. For anyone who might need the info, it’s a 20455-040 type. I’ll try to solder it back and see if the screen can be revived, but I’m also planning to order a spare soon, maybe even jump up to the 16K model this time.

 

[reneboehmer]

Just ordered the 14k 10-inch monitor and driver, I was thinking of maybe also using it to build a sensitometer which automatically exposes custom step wedges on paper and film.

 

[michaelbirkmose]

@Ethan Brossard I’m still in the early stages of experimenting, so I’ve been using the mono LCD with the protective film still on. I was curious—since it looks like you’ve already built a carrier platform - what’s your experience handling the display once the film is removed? Is it extremely delicate (easily scratched, fingerprints are permanently embedded, etc.), or can it be handled more normally - cleaned, dust removed, without much risk of damage?

 

[Ethan Brossard]

@Ethan Brossard I’m still in the early stages of experimenting, so I’ve been using the mono LCD with the protective film still on. I was curious—since it looks like you’ve already built a carrier platform - what’s your experience handling the display once the film is removed? Is it extremely delicate (easily scratched, fingerprints are permanently embedded, etc.), or can it be handled more normally - cleaned, dust removed, without much risk of damage?

I hold it by the edges and avoid touching the image area, but dust and whatnot is not too hard to clean off. Scratches will obviously show up in all your prints, but its fairly resistent to them.

 

[Graham06]

I have not yet removed the film. I removed my name that had been written on the screen to prevent fraud with alcohol. I am getting good results with the film in place so I am leaving it on for now. I'm sure I'll have to re-calibrate things if I eventually take it off.

 

[Ethan Brossard]

I have not yet removed the film. I removed my name that had been written on the screen to prevent fraud with alcohol. I am getting good results with the film in place so I am leaving it on for now. I'm sure I'll have to re-calibrate things if I eventually take it off.

I began testing with the film in place, and did not notice any changes in print exposure after removing it. The main reason I removed it is that dust sticks to the film more than the screen itself (and the film is harder to clean)

 

[michaelbirkmose]

Last night I did another test printing session, using the display’s grayscale mode (as opposed to running it in 1-bit mode) with straight exposures (no modulation). In areas with less tonal variation, I noticed faint brighter lines running vertically when the display is in landscape orientation.

The protective film is still on, but I doubt that’s the cause. I’m wondering if these lines are inherent to the display itself, or if they’re related to how grayscale mode works. They seem to repeat roughly every 20 pixels or so.

 

[michaelbirkmose]

Last night I did another test printing session, using the display’s grayscale mode (as opposed to running it in 1-bit mode) with straight exposures (no modulation). In areas with less tonal variation, I noticed faint brighter lines running vertically when the display is in landscape orientation.

The protective film is still on, but I doubt that’s the cause. I’m wondering if these lines are inherent to the display itself, or if they’re related to how grayscale mode works. They seem to repeat roughly every 20 pixels or so.

Went back this morning to test - this time using pure 1-bit mode, and time modulation - do not observe the lines then. Must be something about how the display is refreshed when running in 8-bit, that causes this.

 

[reneboehmer]

I have read online that the transparent state of the monitors is about 5% transmittance. I also see that it is mainly optimized to block 405nm Light. This will most likely introduce some very uncontrollable spectral issues when working with RA4, as mentioned somewhere in this thread. My idea was to use it as a shutter as a custom step wedge in a diy sensitometer. This seems hard though because based on my guesswork the needed illuminance range of about 3.0 log H is hard to obtain, the simulated DR must therefore also be 3.0. Even when taking the time axis into account, this might not be possible, the contrast ratio of the monitor is only 300:1. For typical silver halide papers, about DR 1.1 is good and easily achievable, as you guys are proving here. This is really exciting, and I can't wait to test everything once my monitor arrives.

 

[Ethan Brossard]

I have read online that the transparent state of the monitors is about 5% transmittance. I also see that it is mainly optimized to block 405nm Light. This will most likely introduce some very uncontrollable spectral issues when working with RA4, as mentioned somewhere in this thread. My idea was to use it as a shutter as a custom step wedge in a diy sensitometer. This seems hard though because based on my guesswork the needed illuminance range of about 3.0 log H is hard to obtain, the simulated DR must therefore also be 3.0. Even when taking the time axis into account, this might not be possible, the contrast ratio of the monitor is only 300:1. For typical silver halide papers, about DR 1.1 is good and easily achievable, as you guys are proving here. This is really exciting, and I can't wait to test everything once my monitor arrives.

RA4 should be doable- albeit a pain in the ass to calibrate. The monochrome displays have no color cast that I can see, so with three exposures for red green and blue light one could create a color print after linearizing each dye layer in the emulsion.

[edit: though I have not tested it, I would not be surprised if there is more than 5% transmittance for light further into the visible spectrum]

 

[AndrewBurns]

[edit: though I have not tested it, I would not be surprised if there is more than 5% transmittance for light further into the visible spectrum]

Yeah it's my understanding that the transmission of the screens drops as wavelength drops, so for my UV projector I'm using a 380nm light source through one of these screens and the manufacturer said that 'a bit less' would make it through compared to 405nm.