Full Analog Prints With Digital Tweaking

[Mike1234]

I don't know if it's appropriate here but... this suggestion, a hypothesis really, will produce 100 percent analog prints (never any pixels). The only thing that could be called "digital" is the control of the bank of LED's serving as the enlarger light source.

How about a panel of LED's tightly packed together as the enlarger light source and with diffusion between it and the film and adjustable height so the diffusion is variable? Individual LED intensity could be computer controlled. This would, in effect, produce automatically dodged/burned prints with 100 percent accuracy/repeatablility and one could save the tweak for use anytime later. Of course, an interactive computer interface would be needed to accept human input for the adjustments. I don't see why this couldn't work for color as well as B&W and the color one could alter localized color, not just intensity... say you want to enhance a sunset while cooling down the forground colors. However, much smaller LED's would be needed as the spacing is a problem between all those color LED's. The color unit could also be used to make split contrast prints... infinity variable, actually.

BTW, this could also be used for contact prints.

No, I don't have the technical skills to build one. And, yes, it would be pricey.

EDIT, I read a post somewhere in which someone suggested using an LED display panel as the light source for contact prints. But I don't think anyone suggested making a "digital mask" for it. At any rate, that would be a fully analog process too unless one wants to pick nits at the light source.

Oh... and photoexpedition posted EARLIER the concept of using LED's as an enlarger light source.

 

[keithwms]

The closest thing to this that I can imagine is a high-resolution computer screen. You'd lay the paper atop the screen to expose it in contact. N.b. you can already do this, and by putting a diffusing medium between the screen and paper, you can reduce pixelation. In theory you could decrease the dot pitch to the point that yhe dots were invisible except under loupe.

I have also done 'hybralogue' prints by inkjetting a mask onto trad'l photo paper, which was then exposed, the ink cleared, and developed normally. No pixelation. I think I can clam that fairly because I can go to the resolution limit of the printer and there are no limits to the resolution of the file that I can send it. If I had an imagesetter then....

But.... the 'digital' part of this process, to which some would reasonably object, is the preparation of the image file. It is possible to prepare a fully analogue exposure device, as you note, but controlling it would be tricky: binning is what allows formerly continuous-tone images to be converted into compact files that you can easily read/write/store. You could get around this issue by not having a storage step at all: you'd just scan the neg and then print it immediately. This is basically how the first faxes worked... they can be fully analogue. The objected to be faxed was drum scanned and the (analogue) signal was transmitted, without binning, to a receiver which then inked up the image. This technology was actually used more than a century ago. There was also a spy-satellite technology that operated on similar principles and also, IIRC, delivered a fully analogue print to the ground after developing film onboard. Actually I am not sure if it is public record what exactly was transmitted... analogue or digital... but if it was analogue then there would be major encryption issues and backup/storage would be a problem.

 

[Mike1234]

Keith... WOW... that's too cool!! If I get into analog printing I'll have to buy me a really big hi-res monitor.

Still, the above suggestion eliminates the pixels completely and no analog scanning either. I didn't know there was such a thing, BTW. At any rate, the absence of pixelizing is why I felt it's okay to post it here. I hope I don't offend anyone.

 

[keithwms]

The pixelation really isn't the issue that should most offend the analogue mind, in my opinion. There are already scanners and printers available that can read/write at 4000+ dpi... way beyond what the eye can see and even beyond what many loupes can see. If and when I really need a digital file e.g. from a damaged, prized neg, I get an LVT (light valve technology negative) done; the LVT negs are so high res they can be enlarged substantially. I have enlarged some LVTs past 8x and still seen no pixelation.

So... I really don't think pixelation is the issue, you can tell me any dpi you want and I can tell you how to get it. It's the binning of the tones (and hence finite bit depth) that is the issue. That occurs at the analogue-to-digital conversion (ADC) step. It's truly no problem whatsoever to scan a neg in a fully analogue way with a drum scanner, and arguably, you can get more information out of the neg that way than any other method, including optical printing... but the devil in the details is what you do with all that information. That's where the ADC comes in and tempers flare

 

[David A. Goldfarb]

Along the same lines there have long been contact printers with something like 21 incandescent light bulbs that could be switched on or off individually for rough dodge/burn effects. Ansel Adams had a similar light source for his horizontal 8x10" enlarger, but I don't know if the bulbs were individually controllable.

 

[keithwms]

Along the same lines there have long been contact printers with something like 21 incandescent light bulbs that could be switched on or off individually for rough dodge/burn effects.

As I mentioned in the other thread I have one of those. It's a Morse contact printer with (IIRC) ~36 argon bulbs. I think it can cover 11x14. It has individual switches for each bulb. The clever thing is that the bulbs are on a concave surface... which takes care of light-source falloff.

 

[Mike1234]

The advantage to having an array of several hundred or perhaps thousands of tightly packed individually controlled light sources is the ultra high resolution of the semiautomated dodging/burning. In fact, if the film and light source are large enough, I theorize it would provide finer control than using even the most meticulously made custom cut dodging/burning tools.

 

[keithwms]

Mike, a much more practical tool would be an imagesetter-like device. You only need one LED for b&w (or three or four for colour) which then rasters over the paper with whatever resolution you desire. A lightjet comes quite close to this: it uses rastered lasers to print on trad photopaper. If you had a rastering laser head that projected the light through your neg onto your paper, voila. You could dodge and burn with ~1 micron precision. The contrast mask would, for practical reasons, be a digital file, and the resolution of that would limit the 'analogueness' of the final print.

The advantage of rastering is that you can easily surpass optical resolution with the rastering stage. I mean, I have stages in my lab with ~1 nm resolution. Okay they are expensive, I am just saying, even a cruddy old inkjet will give you 300 dpi at least. 1 micron stages are easy and fast. Nowadays, rastering is essentially limitless in its resolution, and if you dither on top of that, you can interpolate the digital signal and create a continuous tone image. You just have to remember that if you can take several finite-bit-depth signals and dither them to interpolate and rid yourself of any binning artifacts. Some hi-def TVs work on this principle: the individual pixels alternate between two or more levels to create intermediate tones and yoru eye can't see the transitions quickly enough to tell. Note that the high-def TVs actually have pitifully low pixel resolution!

 

[PeteZ8]

The monitor idea is actually fairly ingenious, especially for a contact print. The process would be quite simple (although for physical reasons, you may want to use an LCD, a CRT might be a bit heavy to set on top of a neg )

Make a low res scan of the negative and bring it into photoshop. On a separate layer, do all of your dodging/burning with a soft brush tool with low opacity. Then hide the layer that shows the image. What you are left with is a screen with varying density that would represent your dodge/burn patterns. Then invert the colors.

The trick would be controlling exposure time, as most monitors do not turn on and off instantly, even LCD's. A large darkslide would probably work, could be simple as suspending the monitor 1/4 inch above the glass and sliding a piece of posterboard or foam core out and back in.

To go one step further, you could also paint in different polycontrast colors if you are using such a paper; and achieve high degrees of localized contrast control.

 

[Mike1234]

The monitor idea is actually fairly ingenious, especially for a contact print. The process would be quite simple (although for physical reasons, you may want to use an LCD, a CRT might be a bit heavy to set on top of a neg )

Make a low res scan of the negative and bring it into photoshop. On a separate layer, do all of your dodging/burning with a soft brush tool with low opacity. Then hide the layer that shows the image. What you are left with is a screen with varying density that would represent your dodge/burn patterns. Then invert the colors.

The trick would be controlling exposure time, as most monitors do not turn on and off instantly, even LCD's. A large darkslide would probably work, could be simple as suspending the monitor 1/4 inch above the glass and sliding a piece of posterboard or foam core out and back in.

To go one step further, you could also paint in different polycontrast colors if you are using such a paper; and achieve high degrees of localized contrast control.

Yep... there you go. But again, the monitor idea isn't mine.

 

[keithwms]

My solution for the monitor on/off flicker problem is simply to use a lower-sensitivity paper.... or reduce the sensitivity via a thin 'neutral density' sheet (which could serve the dual purpose of depixelating the light source).

Note also that if you dither the screen (or the paper) even very slightly during exposure, the pixelation will be gone. Some high def monitors will dither for you. Dithering works as follows. Imagine two adjacent pixels, and suppose that you want to create a smooth gradient across them. You just vary the pixel values in time, and the average will be any intermediate value you desire. I am thinking that dithering is now so fast that one might be able to control exposures with it as well. I don't know how much residual luminescence is left on the screen, but my guess is that it is red shifted and thus wouldn't matter much for most papers.

I did also consider a mechanical stage, with which you would slap the print onto the monitor and then pull it back, controlling exposure via good old inverse square law. I estimate that repeatable exposures as short as a 1/100 sec or so are possible that way.

So... it's definitely doable. And of course the monitor controls all the d&b and contrast.

 

[ic-racer]

Even if you did get it to work, the results with a conventional film mask will probably be better.

 

[PeteZ8]

Even if you did get it to work, the results with a conventional film mask will probably be better.

Can you explain why you think this would be the case? What specific disadvantages would you see in this type of system, assuming the bugs were worked out of the exposure timing etc.

 

[RalphLambrecht]

Along the same lines there have long been contact printers with something like 21 incandescent light bulbs that could be switched on or off individually for rough dodge/burn effects. Ansel Adams had a similar light source for his horizontal 8x10" enlarger, but I don't know if the bulbs were individually controllable.

They were, by screwing them in or out.

 

[PeteZ8]

My solution for the monitor on/off flicker problem is simply to use a lower-sensitivity paper.... or reduce the sensitivity via a thin 'neutral density' sheet (which could serve the dual purpose of depixelating the light source).

Note also that if you dither the screen (or the paper) even very slightly during exposure, the pixelation will be gone. Some high def monitors will dither for you. Dithering works as follows. Imagine two adjacent pixels, and suppose that you want to create a smooth gradient across them. You just vary the pixel values in time, and the average will be any intermediate value you desire. I am thinking that dithering is now so fast that one might be able to control exposures with it as well. I don't know how much residual luminescence is left on the screen, but my guess is that it is red shifted and thus wouldn't matter much for most papers.

I did also consider a mechanical stage, with which you would slap the print onto the monitor and then pull it back, controlling exposure via good old inverse square law. I estimate that repeatable exposures as short as a 1/100 sec or so are possible that way.

So... it's definitely doable. And of course the monitor controls all the d&b and contrast.

Not sure any dithering would be necessary. As long as the monitor is not placed directly in contact with the film/glass, a short gap of 1/4-1/2 inch should provide enough diffusion of the pixels that it should not show any "pixelation" on the final print. I would venture that using a ground/etched glass for the contact glass may also help.

 

[keithwms]

Not sure any dithering would be necessary. As long as the monitor is not placed directly in contact with the film/glass, a short gap of 1/4-1/2 inch should provide enough diffusion of the pixels that it should not show any "pixelation" on the final print. I would venture that using a ground/etched glass for the contact glass may also help.

I'm not as worried about pixelation as posterization of the tone emitted by the screen. If the tone curve of the mask isn't smooth then the silver paper will emphasize that for sure. My point is just that with dithering any tone can be rendered, even in the very worst case of only 1-bit white or black screen pixels.

Anyway... it just has to be tried

 

[Mike1234]

Keith, I would really like to pick your brain. However, it would be a waste of your time. Mine brain has grossly lost its plasticity and short/long term memory retention... much like a computer with a damaged CPU, inadequate/damaged RAM, and failing HD......

 

[keithwms]

Why don't you just upgrade and reboot...

 

[Mike1234]

I wish I could, Keith. Maybe the reboot will come soon enough though.

 

[L Gebhardt]

I've been using a technique developed by Alan Ross (as far as I know) to make masks. Start with a low res digital scan. make selections and fill them in with black on a separate layer. Adjust the opacity to get the right density on the mask. Print on transparency film at actual size. Place above the negative with a thin diffusion layer between. All the benefits of the high tech enlarger approach, but with minimal cost (assuming you own a cheap scanner, printer and computer). You can also print with yellow or magenta to control contrast in small areas.

I had been making masks for a while with pencil on frosted mylar, but the computer makes this much more precise. It's amazing how well this works.

 

[Kirk Keyes]

I've been using a technique developed by Alan Ross (as far as I know) to make masks.

I've done this a couple times too, and I found it worked really well also.

 

[Bob Carnie]

Hi Larry

This sounds like a newer version of contour mapping technique that has been around for quite a long time. I have a book from the 70"s that shows this method.
Contour Mapping as described in a book I have is laying a sheet of frosted mylar directly over the negative and then cutting out other frosted mylar to the shape that needs to be dodged with a bit of practice a perfect balance would eventually be achieved. This was used for printing editions with precise repeatablity. Each cut out piece is laid on top of each other which looks like a contour map when finished.
The negative and mask is then put into the enlarger and a print is made with no dodging.

I have been following this thread and did not mention contour mapping as the OP observations lead to the possibliity of burning in as well as dodging which contour mapping or inkjet mask would not allow you to do. You still have to manually burn in the highlight areas for effect.
I guess you could make the print for highlight only region only and use the mask for dodging out everything else but I think this would not be practical.

I've been using a technique developed by Alan Ross (as far as I know) to make masks. Start with a low res digital scan. make selections and fill them in with black on a separate layer. Adjust the opacity to get the right density on the mask. Print on transparency film at actual size. Place above the negative with a thin diffusion layer between. All the benefits of the high tech enlarger approach, but with minimal cost (assuming you own a cheap scanner, printer and computer). You can also print with yellow or magenta to control contrast in small areas.

I had been making masks for a while with pencil on frosted mylar, but the computer makes this much more precise. It's amazing how well this works.

 

[Mike1234]

My choice of words was inappropriate. I should not have used the words "dodge" or "burn" because "masking" is closer to the correct terminology. The idea is to make precisely repeatable prints that don't require any physical intervention such as manual dodging, burning, masking, etc., so the methods L. Gebhardt and Bob Carnie outlined will produce similar results, especially Gebhardt's (Ross'?). The difference is that the controls go straught from the computer to the printer light source with no steps in between. Now that I've thought about it I realize it's a far-fetched idea... far too complicated and pricey to fool with. I feel silly mentioning now except this did bring attention to the Ross technique of masking of which I was unaware until now. I'll read up on it though. Sounds intriguing.

Digi-haters please stop reading... I've been using a similar method of masking on multiple layers for many years combined with multiple adjustment layers using curves, hue/sat, etc. One can create either hard-edged fills or feather any amount necessary. Black fills for burning and color fills for altering hue... some interesting results by using differing layer transparency types... sometimes useful for unhancing sunsets or other colorful areas. At any rate, I never thought of making optical masks this way for use in analog/optical printing. That's just kewl. The only down side I can see is losing the minute detail of the mask. It will take a little different approach than before... keeping the blurring effect in mind.

 

[keithwms]

I am not especially fond of the print-to-transparency step used by some methods. Yes it works, but it's quite pricey. Maybe I should try vellum.

 

[polyglot]

You can disassemble a cheap old LCD panel so that it's a transparent panel with no backlight or anythng behind it. You can then put that above the print; its height will mean it's just out of focus. That would allow you to do arbitrary contrast masking cheaply and repeatably since old 17" panels are regularly discarded, particularly when their backlight fails...