Minimum LED pulse-time for paper before failing reciprocity

If you have a LED head, you can control your LEDs by pulsing them (called PWM). Does anyone know how short these pulses can be before papers start failing reciprocity? I know film does fine with flash-times under 1 millisecond. My test of Ilford's MGRC Deluxe shows that 2.5 msec is fine. I'm wondering how short can that pulse-time be, because the shorter I can go, the more dimming I can get, which is useful for blue at low grades.

I just sent the following query to Ilford:
How short of an exposure would cause significant reciprocity-failure for MGRC Deluxe? Like many people, I have converted my enlarger to use LEDs, and their brightness can be controlled by PWM, which is a series of short bright pulses. My testing shows that a pulse-time of 2.5 milliseconds is fine. But how short can pulses be?

Mark Overton

Harman sells roll paper that is specifically designed for short duration exposure - MULTIGRADE RC EXPRESS PF is optimized for very short exposures.
https://www.ilfordphoto.com/multigrade-rc-express-pearl

I can't find MULTIGRADE RC EXPRESS PF on Ilford's web-site. But a search shows it's on their Canada site:
https://www.ilfordphoto.ca/product/multigrade-rc-express-pearl-8in-x-250ft-roll-so/
But the minimum size is a 250-foot roll. Too much for me. Thanks for the suggestion though.
Mark Overton

Why not just decrease the brightness of the blue LEDs? Or switch some of them off for low-contrast exposures? I did both when I built my LED head; it worked quite well for me. My problem was getting hard grades, though my negs rarely needed them. I should have used some shorter-wavelength violet LEDs... hindsight!

Use RC-filter (resistor + capacitor) to even out the pulses.

vedostuu said:

Use RC-filter (resistor + capacitor) to even out the pulses.

Click to expand...

That would introduce huge non-linearities which may or may not be a problem depending on application. It would be a nuisance in terms of accurate control in any case. If you do some simulations on a pulsed power source driving a LED with an RC filter to smooth things, you'll see what I mean. With a small C you'll just have a slightly smoothed out edge to the pulses, with a large C you just get a sawtooth. Hence, it won't solve the problem, and it'll create problems with non-linearity across different duty cycles.

The issue is a valid concern and accurate information for contemporary paper emulsions seems hard to come by. Datasheets don't list short-exposure reciprocity failure, and whatever information is available is mostly research done into film emulsions in the 1940s through 1970s. Inferring from this, exposures shorter than something like 1 ms may induce reciprocity issues, but there's a lot of assumptions involved.

If you're concerned about this, I'd suggest simply reducing the PWM frequency to 100-200Hz, which would effectively eliminate the issues as long as duty cycle isn't insanely low.

An RC-filter would be roughly equivalent to resistive dimming. My LED head uses BuckBlock LED-drivers, and as I suspect koraks knows, they become very nonlinear at dim levels using resistive dimming. In fact, their curve is flat at a minimum current, and then they simply turn off! Koracs, thanks for your reply. At 100 Hz, 4 stops of dimming would yield a cycle of 0.625 msec on and 9.375 msec off. If paper doesn't fail reciprocity at 0.625 msec, I'll do it. I'm thinking of trying 50 Hz. The danger is that movement of burning/dodging tools might produce choppy transitions, which I've seen with low frequencies.
Mark Overton

Well, in my current led head I use a frequency of 1.5Khz if memory serves and I've never noted any issues with unpredictable results on either color or b&w paper. I didn't explicitly test for reciprocity failure though.

MattKing said:

Harman sells roll paper that is specifically designed for short duration exposure - MULTIGRADE RC EXPRESS PF is optimized for very short exposures.
https://www.ilfordphoto.com/multigrade-rc-express-pearl

Click to expand...

Strange enough there is no data sheet available. (In the Multigrade RC data sheet it does not show up.)

I think you're misunderstanding reciprocity failure.

The long forms of the term are either LIRF (Low Intensity Reciprocity Failure), the one we encounter with long exposures, especially in large format and pinhole, and HIRF (High Intensity Reciprocity Failure), the opposite number that we rarely run into (it's mainly an issue for super high speed photography, like using a genuinely HUGE level of flash power to stop a bullet in flight). These long forms indicate the conditions in which reciprocity corrections will be needed: either very low or very high light intensity. Nothing is said about the length of the exposure.

What that means for your enlarging is that your pulse width doesn't matter in the least, as long as the intensity of the light is within the linear response portion of the paper's (or film's) sensitivity curve. From a more ordinary, film-oriented perspective, if you use enough flash power to get a proper exposure in, say, 50 microseconds (= 1/20,000), you'll need to correct with virtually any film, but if you give a 50 microsecond flash at a light level that requires you to repeat it fifty times to get the exposure you need (total exposure 2500 microseconds = 1/400) you'll need no correction (with most films).

Enlarging papers are optimized for much longer exposures than film, however, so if your blue LED would need to be "on" for a total exposure of, say, four seconds to give you the final exposure and grade you want, it doesn't matter if you give that exposure all in a lump, or as four hundred 10 millisecond pulses (or four thousand pulses of one millisecond, though pulses that short might reduce your total light emission due to ramp-up and ramp-down times and emission cutoffs in the LED) -- as long as the total exposure ranges from a large fraction of a second to not much more than a minute, you should be fine with common multigrade papers.

Donald, duration of exposure does matter. Like I said, there's quite a bit of literature about it. One paper I found reported significant effects on sensitivity and curve shape for exposures as "long" as 1 ms and even 10ms. This was with very old film emulsions as it was a pretty old publication; 1950s iirc.

What I found no publications on, however, is the issue of repeated brief pulses within a fairly short period of time, as is the case with pwm.

@koraks Those exposures, however, were applying the entire exposure in a few milliseconds -- which, even for film in the 1950s, was pretty short (electronic flash was newish technology then, and common M type bulbs burned for about forty milliseconds). I'd expect HIRF to set in with enlarging exposures totaling under around half a second (at the least, I'd want to test for linearity if I were running exposures that short). Modern films typically don't show HIRF until your total exposure time goes well below a millisecond -- a hundred microseconds (= 1/10,000) is a figure that stuck in my mind, but again, few if any modern films exhibit LIRF with exposures shorter than a second, and in the 1950s that figure would have been in the 1/8 to 1/10 range (this is part of why Tech Pan was so popular in astronomy -- it could take very long exposures with no or very little LIRF effect).

There are enough workers now who've built their own PWM controlled multi-grade enlarging heads that we'd have heard about it if pulse width made a big difference -- it's been about fifteen years since the first time I read about one of those builds. For my own use, as a split-grade printer, I'd rather have the blue and green LEDs run sequentially and continuously (on separate timers), like exposing with grade 00 and grade 5 filtration -- but I can easily see a use case for PWM as equivalent to multigrade filters or use of a color head.

It would surely be easy enough to test -- pick an image, print it with PWM settings to give the desired contrast, then reprint with the same total times with the controller set to "solid blue" and "solid green" and use a reflection densitometer (or spotmeter with a diopter) to compare density of the same image area on the two prints.

Apparently the problem isn't severe enough to be noticed, but I suspect that very few people have actually tested it systematically. I haven't, but like I said ), no problems at around 1kHz. Duty cycle probably goes down to something like 1% on the blue channel for b&w; I didn't notice big bumps across grades that couldn't be explained by the characteristics of the paper itself.

Way Beyond Monochrome 2nd ed., by Lambrecht and Woodhouse, page 337, states that "The paper reciprocity test revealed a continuous reduction in print density when increasing exposure time from 8 to 256 seconds. The resulting overall speed loss is about 1/3 stop and can be averaged to about 1/16 stop per doubling of exposure time." The text mentions that all tested papers behave similarly. It adds that this reciprocity failure occurs in papers even at short times from 1 to 16 seconds, so the failure is not limited to long times.
I conclude that, with papers, this failure is not limited to high or low light-intensities.
This says nothing about the effect of PWM, but Donald Qualls makes a good point that if PWM were causing reciprocity failures, we would have seen postings about it.

Koraks, 1 KHz at a 1% duty cycle is an on-time of only 10 uSec! Can you double-check those numbers? If you have had no problems with 10 uSec on-times, then PWM is harmless.
Mark Overton

How are you measuring the light intensity pulses to test for reciprocity? Are you using some slow acting meter to average the pulses? Are you using a fast acting meter and then calculating the luminous flux? How are you testing the paper's response to the light?

My LED head controller offers both resistive and PWM dimming. It has an analog knob for resistive, and PWM is a simple switch for 25-vs-100%. I created a test-strip consisting of pairs of exposures: one done with 2 stops of resistive dimming, and the other with 25% PWM. I used blue because blue alone gives grade 5, which is most sensitive to exposure variation. Most steps in the strip were identical when viewed under bright light, but a few differed a little, which I attribute to my own inaccurate setting of the analog knob for resistive dimming or inaccuracy of my electro-mechanical Gralab 300 timer.
I calibrated both kinds of dimmers using an analog (needle) meter (Seconic Studio Deluxe), so it averaged the pulses for me. I tried calibrating PWM with the meter in my DSLR, but its lack of time-averaging made its measurements fluctuate wildly.
Mark Overton

I should explain "resistive dimming". This is a feature of the BuckBlock LED controller. It has two DIM wires (+ and ground), and you can dim your LEDs from 100% to 0% by putting a resistance across those DIM wires. They suggest using a 20K potentiometer. I did that. They also mention that the DIM wires can be PWM'ed, so I built an NE555 circuit for that. And you can do both in tandem. I found that the resistive dimming of those controllers becomes temperature-sensitive when dimming much below 25%. So I get 2.25 stops from resistive, and 2 stops from PWM, giving me 0 to 4.25 stops. And 4.25 stops is needed to hit grade 0.

@kevs suggested that I decrease the current for blue LEDs. That's a good idea. I'd like this LED head to work for color also, and I think RA-4 also requires little blue, so using a BuckBlock with lower current-output is feasible. I'll look into that.

Interesting factoid: Grade 2 for Ilford RC Deluxe paper requires that blue be 2.25 stops dimmer than green (assuming both have about the same luminance at full power). That will match the contrast of a condenser head using Ilford filters. I used test strips of a Stouffer step-wedge to determine that number. Foma requires 2 stops, so it's close to Ilford. Being able to set green and blue like this (by dimming blue) makes generating grade-2 test strips much faster because you only need to make one exposure per step.
Mark Overton

I must admit my LED head is rather a bodge-up job; it's no more sophisticated than some recycled silicon rectifier diodes wired in series and switched to drop the current, which comes from an AC-to-AC mains adapter.

LEDs output some light beyond their peak wavelength, which might be enough to affect contrast grade or colour balance. It might be a factor if you need Grade 0 or Grade 5, but modern LEDs may be better in that respect than my cheap, >10-year-old ones.

albada said:

Interesting factoid: Grade 2 for Ilford RC Deluxe paper requires that blue be 2.25 stops dimmer than green (assuming both have about the same luminance at full power).

Click to expand...

My WS2812 based LED "head" hasn't got such radical effect. I eyeballed it to be about 0.2 stops.

Edit: I assume 0.2 stops is needed for green too, but I will only know it after Stouffer wedges arrive.

Edit 2: BTW: why do you need adjustable dimming? Doesn't it add another variable to control and make everything more difficult?

@kevs : A silicon rectifier drops about 0.6 volts, so several will drop several volts. That's a clever way to drop the voltage down to what the LEDs need. Do those rectifiers get very hot?

@vedostuu : The data-sheet for your WS2812 LEDs is here: https://cdn-shop.adafruit.com/product-files/1138/SK6812+LED+datasheet+.pdf . About 2/3 of the way down is the section titled "RGB chip characteristic parameters", which shows that blue has a little less than 1/2 the luminous intensity as green (on average). According to my measurements, that green-blue ratio will be grade 3, which is probably why the results look fine.
Also, you are correct that adjustable dimming is not essential, because one can always use different times for the two LEDs. But dimming is convenient because, when using the same times, making test-strips at a given grade is easier (I use grade 2).
Finally, after your Stouffer wedges arrive, please try a blue-only test to determine whether your LEDS allow you to reach grade 5. You blue LEDs have a wavelength about 18nm longer than mine, so yours is a bit closer to green, making me wonder whether grade 5 is possible.

Edit: I just thought of another reason for dimming: Keeping exposure-times from being too short or long. Short times (for small prints) can be lengthened by dimming, giving enough time for dodging/burning. Long times (for large prints) can be shortened by boosting LED-power.
Mark Overton

albada said:

About 2/3 of the way down is the section titled "RGB chip characteristic parameters", which shows that blue has a little less than 1/2 the luminous intensity as green (on average).

Click to expand...

This is somewhat misleading, as the luminous efficiency is expressed in candela, and candela is in turn compensated for the spectral sensitivity of the human eye. You could use the numbers in the table together with the luminosity function to work out the absolute luminous flux to get a better estimation. Doing this will show that for instance the absolute blue output is likely much higher than green output, while for red it depends of photopic or scotopic sensitivity is used as a starting point (likely photopic). But red is of course not very relevant if this is about VC B&W materials.

kevs said:

silicon rectifier diodes wired in series and switched to drop the current, which comes from an AC-to-AC mains adapter.

Click to expand...

What do you mean by 'switched rectifier diodes'? Can you post a schematic of your solution?
On a sidenote, controlling leds by controlling voltage is generally not a good idea due to the steep I/V curve of leds. Small variations in voltage (due to variations in home AC voltage, aging of components, temperature of components etc.) will result in significant variations of led current and hence light output.

@albada, I still needed to get back to you on the question you asked about the minimum duty cycle of the leds in my current setup. Taking B&W as a starting point, I mix in a little bit of blue on the softest contrast grade in order to allow all papers to achieve full dmax. I found that only green light prohibits this (the result is an extremely long tonal scale and very low paper speed, but the tonal scale is so long that it doesn't have much practical value even for extremely contrasty negatives). To resolve this I add a little blue at the lowest contrast grade, which I do by switching the blue leds at a 6.25% duty cycle. Combined with a PWM frequency of 1.526kHz this gives a repeated 'on-state' of 41us.

This, however, doesn't help you much, because it says nothing about possible reciprocity failure at these short times. It only shows that the paper responds to these very brief (and repeated) pulses - but not of the response is linear if the duration is varied at the short end of the duty cycle range.

The most critical time variation is used for color work and that involves all 3 colors at a decent level. I.e. maximum duty cycles for color work in my setup are 100% for red, around 90% for green and 45% for blue, but this is at a Y/M filtration setting of 0, which of course isn't really used for color negative to RA4 paper - in practice, the green and blue duty cycles are probably closer to something like 60% and 20% or thereabouts, with variations (according to Y/M filtration setting) that are likely around +/- 20% for green and maybe +/-7% for blue for most negatives (I'd have to run the math on this as the variation is exponential and there's a bit of value mapping going on in the software). That means that for instance the blue channel tends to give pulses in the 100-150us range depending on filtration setting. Again, that doesn't give any insight into possible reciprocity failure. The only thing I can establish is that very small variations do indeed translate to color balance shifts and they do so predictably - which of course is the intended functioning of the setup in the first place.

So long story short: while the numbers above are nice for 'shits & giggles', they don't give any insight into reciprocity failure for repeated pulses in the tens-hundreds of microseconds range.

Here's an easy experiment that @koraks and @vedostuu (and anyone with PWM'd LEDs) can try. You will need one or more neutral density filters, and a light-meter that can measure PWM'd light (any analog meter will work).

1. Set the blue LED to full power (no PWM), hold the ND filters under the lens, and measure its light output.
2. Make a test strip with several times.
3. Remove the ND filters, and set the PWM for the blue LED such that its light output matches what you got in step 1.
4. Make a second test strip.

Because the light outputs were equal, the paper's densities should be equal, so any difference would be from reciprocity failure.

BTW, the spec for controller-chip in the WS2812 says its PWM frequency is 1.2 kHz, so it would be good for testing reciprocity.
Mark Overton

Yes, that would be a good test. Maybe I'll get round to it one of these days. I'll be sure to post the outcome when it happens.

albada said:

@kevs : A silicon rectifier drops about 0.6 volts, so several will drop several volts. That's a clever way to drop the voltage down to what the LEDs need. Do those rectifiers get very hot?

Click to expand...

They do get slightly warm after leaving the LEDs on for a while, but they haven't overheated. I'm not sure about the LEDS! I used an off-the-shelf adapter (1watt, 4.5 volts AC) rather than a self-built PSU.

koraks said:

What do you mean by 'switched rectifier diodes'? Can you post a schematic of your solution?
On a sidenote, controlling leds by controlling voltage is generally not a good idea due to the steep I/V curve of leds. Small variations in voltage (due to variations in home AC voltage, aging of components, temperature of components etc.) will result in significant variations of led current and hence light output.

Click to expand...

I used one rotary switch in each LED circuit to control the brightness; the diodes aren't themselves used as switches; one is used as a half-wave rectifiers and the others as voltage droppers. I've done a rough circuit diagram.

I agree it's not an ideal solution but I needed something simple; continuous mixing of blue and green channels would be nice. But I didn't want to construct complex designs involving microprocessors and lots of components. My setup works adequately for my needs (B&W only; no colour) and I can forgo the niceties, though I might improve it in the future. TBH, I don't do much in the darkroom these days!

Ah,thanks for clearing it up! Yes, that setup makes sense within its inherent limitations, but you prove it works well enough for you! In the end that's what matters. It is indeed nice and simple.