Trying out Stouffer T2115 wedge on darkroom papers

[radiant]

I got my hands on Stouffer T2115 wedge. I bought T2115 to test my WS2812B LED conversion but then I decided to expose all kind of papers I had available.

I tested following papers:
- Harman Direct Positive paper
- Fomaspeed Variant 311 (natural gloss)
- Ilford Multigrade Deluxe V paper (pearl)

The image below is composition of one single scan of all the strips at once (to maintan same levels). I just cut-pasted the strips for easier examination. All strips are exposed at either full green or blue available at my setup. The rightmost strip is exposed at 50% which means both color channels were used for exposing.

What I found out for sure is that the direct paper has half stop difference in sensitivity for WS2812 green/blue leds. The most left strip is DP at full blue color. Could this be something to do with reality or is it caused by my setup - who knows.

When comparing Fomaspeed Variant 311 and MGRC V I cannot find large differences. Even the dMax seems pretty same on the scan. This is strange. Or is it just that the differences cannot be examined this way?

How does my LED head perform when looking at 100% blue exposures? How many stops do you see and how many steps the full blue should show?

Any other tips what did I mess up or anything where we could all learn are appreciated

 

[radiant]

.. and sorry about the dust :-/

Now looking at MGRC V at full blue it seems to be more dark than 311 paper (phew). But does MGRC V "deeper blacks" require exposing at higher grade too?

 

[relistan]

Very interesting! Thanks for sharing.

But does MGRC V "deeper blacks" require exposing at higher grade too?

I'm not an expert but that's sure what it looks like. Assuming that your green LEDs are in the right spectral range to allow the the lower contrast emulsion layer to reach its max density. (The blue clearly are for the other layer)

 

[albada]

For MGRC V with my Stouffer wedges, I count 14 steps from D-max to D-min for green (= grade 00), and 7 steps for blue (= grade 5). These counts include the steps at D-max and D-min.
On your scan, I count 17 and 9 steps, which is lower contrast for both.
Is your developer old? Or are you using a low-contrast developer? My developer is Liquidol, and I have not compared it to others, but I've never heard that Liquidol produces higher-than-normal contrast.
Mark Overton

 

[Bill Burk]

Look at the scale rating of the papers. The scale is density without decimal. Every other step is 0.3, in terms of paper scale, every other step is 30

So if the scale is 110 then you should have 7 or 8 visible steps on the print.

Obviously you are getting long scale for green and short scale for blue. So you’re doing something right

 

[Bill Burk]

Ilford shows its ISO numbers as ranging from 40 to 180, that's about 3 steps to about 12 steps. I'm just talking about casually counting the steps you can see, not really getting into the technical details of exactly what density range is expected from negative and the resulting density on print. The specs aren't complicated but I forget what they are right now (probably reflection density of 0.04 to 90% of the black the paper can achieve) but basically... what you see is what you get.

You're getting close to the 12 steps on the low contrast examples, but you probably could get more contrast out of the paper.

 

[radiant]

On your scan, I count 17 and 9 steps, which is lower contrast for both.
Is your developer old? Or are you using a low-contrast developer? My developer is Liquidol, and I have not compared it to others, but I've never heard that Liquidol produces higher-than-normal contrast.

I got 14 / 8 steps on my calculations .. The green strip is a bit difficult to interpret as the wedge wasn't in full contact with paper.

I was using freshly mixed Ilford multigrade developer but the bottle has been opened maybe 6 months ago, I cannot remember. It has a bit yellowish tint already. Mix from freshly opened bottle is totally clear in color.

I wouldn't want to open a new developer for only to test this out but maybe I should to find out how much aged stock developer makes difference.

BTW: isn't longer scale at low contrast better? The more steps the better (grade 0 / green)?

Ilford shows its ISO numbers as ranging from 40 to 180, that's about 3 steps to about 12 steps.

Bill, thanks for digging that out and calculating to steps.

On low contrast side I think my setup is fine, but on high contrast things are way off 3 steps vs 8 steps on mine. Damn.

One test I need to make is try find out dMax at green exposure to see if dMax is different on different grades.

 

[bernard_L]

IMO 0.15 density steps are fine for film sensitometry but maybe too coarse for paper; That is because a typical film gamma is 0.7 or less while for for #2 paper the defining points being typically at 0.1D and 1.8D, 1.0logE apart, the mean slope is 1.7 (rough estimate). Even with 0.1D steps, just counting steps results in coarse results esp. for hard grades (like, hmmm, are there three or four steps?). Measuring the densities and plotting the results with a proper interpolating curve can mitigate this issue to some extent.
This said, kudos and continue the good work.

 

[radiant]

IMO 0.15 density steps are fine for film sensitometry but maybe too coarse for paper; That is because a typical film gamma is 0.7 or less while for for #2 paper the defining points being typically at 0.1D and 1.8D, 1.0logE apart, the mean slope is 1.7 (rough estimate). Even with 0.1D steps, just counting steps results in coarse results esp. for hard grades (like, hmmm, are there three or four steps?). Measuring the densities and plotting the results with a proper interpolating curve can mitigate this issue to some extent.

Bernard you are absolutely right.

However at this accuracy level we can easily see that my blue isn't producing maximum contrast. If one wants to do detailed analysis, more fine wedges should be used. In this use case those are not needed.

I just made new strip and I couldn't produce any noticeable differences. Maybe I got a bit more dMax (because I did things more carefully) to my low contrast strips, but nothing radical. I even tested mixing new developer from unopened bottle. "Super fresh" developer didn't have any effect to anything. But hey, it was good to rule out - removed doubts when printing..

If I look at the strip exposed with 100% blue, I visually can see 7 different steps (including dMax and "dMin"). When scanned it is more easier to spot suddle differences but by eye and good heart I can separate only 7 tones. Should I feel sad?

Btw: back to reality. Let's think of frame shot in low SBR scene. It might have 2-3 stops of exposure range. If my high contrast range in stops is 3.5 I could almost print "stock" developed and exposed negative with good blacks and whites? I assume?

 

[bernard_L]

What I found out for sure is that the direct paper has half stop difference in sensitivity for WS2812 green/blue leds. The most left strip is DP at full blue color. Could this be something to do with reality or is it caused by my setup - who knows.

It is real and it is nothing to worry about. You cannot measure the exposure in physical units (erg/cm2); among film manufacturers, only Kodak does it, Ilford, Fuji, Foma give "spectrogram for 2850K/5400K" not specifying the dispersion law. So, what you see is a combination of the spectrum of your LEDs and the spectral sensitivity of the paper, and having this combined result within 1/2 stop is an excellent balance (except DP is not meant as a multigrade paper...).

On low contrast side I think my setup is fine, but on high contrast things are way off 3 steps vs 8 steps on mine. Damn.

I believe you are making it worse than it really is.

1. "counting steps" is a bit too vague; when you write that you count 8 steps for blue exposure, you are counting the extreme steps ---pure white and max black--; intervals is one less than steps;
2. as pointed out by Bill Burk, the ISO definition is from D=0.1 (not 0.04) to 90%Dmax, so, less than the total span from max white to max black

Btw: back to reality. Let's think of frame shot in low SBR scene. It might have 2-3 stops of exposure range. If my high contrast range in stops is 3.5 I could almost print "stock" developed and exposed negative with good blacks and whites? I assume?

Scene 3 stop range = 0.9 range in logE. Normal dev gamma 0.7 (remember gamma higher than G_bar or CI) so density range on neg is 0.63. On grade 4 according to ISO (R=0.60) this takes you from paper D=0.1 (light grey) to 90%Dmax (dark grey). On ISO grade 5, this would probably take you from pure white to pure black.
But... maybe a scene with such a low SBR should not be rendered on paper with total black and total white; an aesthetic, not technical statement.

And my impression is that grade 5 is partially a marketing concept. Look at grades 00-3, there is a clear increase in slope at each successive grade

Now look at grades 4-5

Hardly a difference.
And I did not make up these curves; they are from Ilford: https://www.ilfordphoto.com/amfile/file/download/file/1748/product/735/
The numbers in the table are the "official version", but the curves tell a different story.

So: shoot, develop, print. And beware of the modern fashion for thin negatives.

 

[ic-racer]

Yes, as mentioned, don't count the first and last gray step, just the ones in-between and multiply by 15, to get an estimate of ISO(R) contrast.

Typical figures might be Grade 5 = ISO(R) 40 to 45; Grade 4 = ISO(R) 60 to 70; Grade 3 = ISO(R) 80 to 90; Grade 2 = ISO(R) 100 to 110; Grade 1 = ISO(R) 120 to 130; Grade 0 = ISO(R) 140 to 150; Grade 00 = ISO(R) 160 to 180.

Also, as indicated in the above post, ISO(R) of 50 might be called grade 4.5 by some.

 

[radiant]

It is real and it is nothing to worry about. .... (except DP is not meant as a multigrade paper...).

Yes, this was side note. I was about to expose the DP paper and though, hmm, which color and decided to do both. I know it is single grade paper and not multigrade. I guess my test shows that DP paper has 3 stops of exposure range?

as pointed out by Bill Burk, the ISO definition is from D=0.1 (not 0.04) to 90%Dmax, so, less than the total span from max white to max black

Sorry Bill, I missed this. Is this top of not counting the extremes (your point 1.) or does this explain the standard?

But... maybe a scene with such a low SBR should not be rendered on paper with total black and total white; an aesthetic, not technical statement.

Yes, of course more imporant question; should it render from white to black. The example was just to analyze if I'm able to pull black+white on low contrast scene. I'm not saying I should.

continuing on next reply.. editor started failing.

 

[radiant]

Now look at grades 4-5

True. Pretty small difference.

So: shoot, develop, print. And beware of the modern fashion for thin negatives.

This is my main goal. I just happened to convert my enlarger to LED and as using WS2812 leds is probably quite new idea, I though I will analyse how my project succeeded - and in this case in terms of contrast control.

It seems that it is much easier to print dense negatives than thin negatives. I have always struggled with thin negatives as my highest contrast filter hasn't been enough for those.

Typical figures might be Grade 5 = ISO(R) 40 to 45; Grade 4 = ISO(R) 60 to 70; Grade 3 = ISO(R) 80 to 90; Grade 2 = ISO(R) 100 to 110; Grade 1 = ISO(R) 120 to 130; Grade 0 = ISO(R) 140 to 150; Grade 00 = ISO(R) 160 to 180.

Bernard's snip from Ilford datasheet says Grade 5 is 50 ?

If I try to calculate ISO(R) from my blue strip, I get ISO(R) of 75. That would be about grade 3.5 Pretty bad, I think. My low cotrast (green) would be ISO(R) 210 which is lower than the Grade scales lowest contrast (00) :O

 

[bernard_L]

Is this top of not counting the extremes (your point 1.) or does this explain the standard?

Both. (a) clarify the standard; (b), over that restricted range, and assuming one step wedge step each falls exactly on 0.1D and 90%Dmax, e.g. steps 3 and 8 respectively, there are a total of 6 steps (345678) but you should count 5 intervals only.

Bernard's snip from Ilford datasheet says Grade 5 is 50 ?

Never mind. In another Ilford document the range varies somewhat depending on paper. So ic-racer is just as "right" as regards documentation. As concerns reality, I would like see to see in real life an Ilford multigrade paper with a 40 ISO(R). The actual ISO spec I may have seen but cannot find a copy right now; paywall!
https://www.ilfordphoto.com/amfile/file/download/file/1952/product/1872/

If I try to calculate ISO(R) from my blue strip, I get ISO(R) of 75. That would be about grade 3.5 Pretty bad, I think.

By the way, and sorry if that is trivial or has already been mentioned: what are the specs, notably dominant wavelength, of your blue LEDs? Did you check them for absence of contamination at unwanted wavelengths, That is easy, just view the LED as a reflection off a DVD or CD acting as a diffraction grating, the various wavelengths are spread apart, look for a green dot.

 

[radiant]

By the way, and sorry if that is trivial or has already been mentioned: what are the specs, notably dominant wavelength, of your blue LEDs? Did you check them for absence of contamination at unwanted wavelengths, That is easy, just view the LED as a reflection off a DVD or CD acting as a diffraction grating, the various wavelengths are spread apart, look for a green dot.

Blue: 465-475nm and Green 515-530nm. Datasheet here: https://cdn-shop.adafruit.com/datasheets/WS2812.pdf By the way, does for example Ilford describe wavelength required for getting grade 5 results?

Here is some actual measurement of WS2812 wavelengths: https://www.luxalight.eu/sites/defa...60_leds_5050_protected_blue.jpg?itok=npd-FjK5


I have to try the CD test!

Maybe it is also good to say here; I'm totally satisfied with the LED conversion currently. Ignorance would be a bliss ..

 

[radiant]

I found out pretty interesting and matching to my foundings:

https://www.photrio.com/forum/threa...alternative-to-cold-light.75022/#post-1039007

Green and blue LEDs sound great in theory but don't work out so well in practice. I bought several discrete green and blue LEDs and tested them for VC work. Basically, what I found is green LEDs work great for low contrast and don't have any issues with VC paper. However, there are problems with blue LEDs. To the eye, they're an intense blue color--but the hitch is they're actually a combination of blue and green.

The slight green color hidden within the blue LED spectrum reacts with the low-contrast emulsion in VC paper and limits the VC papers I tried to about a grade 3 to 3 1/2. Hard contrast grades like 4 or 5 just can't be done with blue LEDs alone unless the green bias in the blue LEDs is filtered out somehow.

 

[ic-racer]

If I try to calculate ISO(R) from my blue strip, I get ISO(R) of 75. That would be about grade 3.5 Pretty bad, I think. My low cotrast (green) would be ISO(R) 210 which is lower than the Grade scales lowest contrast (00) :O

Discount the steps that are almost black and almost white to account for the points on the curve form which the ISO(R) is obtained. I'd say your test shows about 3 or 4 steps. I'd say that is pretty good contrast.

 

[radiant]

Discount the steps that are almost black and almost white to account for the points on the curve form which the ISO(R) is obtained. I'd say your test shows about 3 or 4 steps. I'd say that is pretty good contrast.

Nice to hear. Maybe my maximum contrast is pretty close to 4 which is totally fine. However my previous calculations were in par with that konakoa's post which would indicate grade 3.5 max. I'm pretty sure by the looks of the prints my max grade isn't close to 3 which is still quite mid-contrasty grade.

Maybe I will still continue my Tri-X + Rodinal development scheme just to stay on the contrasty side

 

[Bill Burk]

Sorry Bill, I missed this. Is this top of not counting the extremes (your point 1.) or does this explain the standard?.

I lazily counted steps. Others have clarified I was off by one.

In practice I would read and “graph” the resulting reflection densities and draw lines on the exposure axis bounding “practically white” to “practically black” to determine the scale for the paper with that filtration, exposure and development.

When I am lazy I just count the steps I can see.

 

[albada]

If I look at the strip exposed with 100% blue, I visually can see 7 different steps (including dMax and "dMin").

7 steps is the correct number. At least, that's the number of steps I got, as I mentioned in post #4 above. I think your set-up is working fine.
Mark Overton

 

[distributed]

Thanks for sharing your information! I am in the process of performing a very similar conversion. I am using an integrated RGB LED, but with custom drivers. Currently my focus is on establishing a relationship between blue/green expsoure ratio and ISO(R) or grade. I also bought a TSL 2591 breakout board (actually minutes before reading that you, vedostuu, are also using it) that I plan to use for intensity measurements on the baseboard. Using those and and ratio/grade/speed relationships I would like to have my system automatically propose exposure time and ratio for a print that fits the measured projected negative density range on the paper, i.e. a good starting point for experimentation towards the print I really want.

As the discussion around step counting has shown, it's sometimes not easy to know how you should look at your data. Looking at your test strips for MGRCV I would count between 5 and 6 strips (15 up to and including 9), which works out to 0.75 to 0.90 density range. Your ISO(R) is probably a bit better because the last step might have been foo far into the shoulder and the first step might be really in the toe. Still I would guess that you're not at the roughly ISO(R) 50 that the MGRCV datahseet mentions.

I have performed similar measurements and have to admit that it's quite easy to get unsatisfactory results. A number of times my developer was not fresh and I have gotten lower than expected DMax and lower than expected contrast. How fresh was your developer? I am using LEDs with a stated blue peak wavelength of 460-465 nm (https://mikroshop.ch/LED_LCD.html?gruppe=7&artikel=681) which is only 10 nm off the range stated for your LEDs. These are the two things I can think of off the top of my head and unfortunately not a conclusive answer.

As bernard_L suggests, if you manage to get your hands on a wedge with .10 density increments, you might get a better view of the steep curve resulting from blue exposure. I don't have a wedge yet, but I ordered the T3110 from Stouffer to be able to more quickly measure in the future. So far I printed my step wedges directly on the paper using exposure timing. I am driving the LEDs from a dedicated microcontroller, so millisecond precision is easily achieved. This is a way to make, say, .10 density increment test strips without a wedge. It works, but it takes a long time.

What I have observed with MGRCV Pearl is that with green only exposure I reach a lower DMax compared to blue only exposure, say roughly 1.75 vs 1.90. Do you have access to a densitometer? Or to a reflective step wedge?

@bernard_L @Bill Burk
I am confused about what Density to start measuring the ISO(R) at: 0.04 or 0.1? I have been unable to find free sources for the standard, but I did find this article in German from a B/W photo magazine: https://www.fotoespresso.de/SW-Magazin/swmag_sensito_5.pdf . They measure from 0.04 above B+F to 90% of DMax-(B+F).

 

[bernard_L]

@bernard_L @Bill Burk
I am confused about what Density to start measuring the ISO(R) at: 0.04 or 0.1? I have been unable to find free sources for the standard, but I did find this article in German from a B/W photo magazine: https://www.fotoespresso.de/SW-Magazin/swmag_sensito_5.pdf . They measure from 0.04 above B+F to 90% of DMax-(B+F).

My mistake. Sorry about that. Indeed 0.04 is the correct value for the low density point of the ISO(R) definition. Identical to American National Standard PH 22-1966. Arose when i went back to one of my plots, that was itself erroneous. Internal version of Internet error parroting.
See also: https://www.photrio.com/forum/threads/understanding-the-characteristic-print-curve.135144/

 

[radiant]

@distributed: Yes I'm pretty far off from ISO(R) 50 - that is pretty clear already. That would require 3.3 steps maximum. I guess the writing of "konakoa" is correct and "normal" RGB leds leak too much into green side and lowers the contrast.

My developer was 100% fresh. I opened new bottle of Ilford multigrade developer and mixed just before printing. I also proved myself that a bit yellowish developer works just as fine as completely fresh. So developer isn't lowering my contrast.

Actually I have never thought I could make this test without a Stouffer wedge. I can also control exposure times in 10 ms range so making really accurate exposures is possible. However do I want those since I'm not even close to grade 5? The tests I have performed are probably enough accurate for me at this point. It is what it is.

I also made a test strip comparing dMax on green/blue only exposures. I cannot see any difference, I exposed both next to each other but I cannot see any visible difference. Both are pretty black I found out that green exposure needs exposing 2 stops more than on blue to reach dMax.

 

[ic-racer]

Your blue shows about 4 steps to count for the ISO(R) determination. I think there is some confusion in this thread about how to determine the ISO(R).

 

[albada]

I am using an integrated RGB LED, but with custom drivers. Currently my focus is on establishing a relationship between blue/green expsoure ratio and ISO(R) or grade.

I've done much work on that as well, and my results are in the table below.

Grade	Green	Blue
00	0	OFF
0	.0.375	4.63
1	0.805	4.06
2	1.18	3.43
3	1.73	2.98
4	2.75	2.75
5	OFF	2.88

The Green and Blue columns are in units of stops below full power. So 0 is full power, 1 is half power, 2 is quarter power, etc. This is how I marked the knobs on my home-made controller.
When changing grades using this table, middle gray (Zone 5) will not change, and lighter and darker tones will shift equally as grades are changed.
The data in this table results in Stouffer prints that appear (to me) identical to those created by my tungsten-condenser head using Ilford filters on Ilford MGRC V Deluxe paper (their current paper), except that I adjusted the numbers to keep mid-gray the same among all grades.

The next table keeps Caucasion skin-tones the same as one changes grades:

Grade	Green	Blue
00	0	OFF
0	0	4.25
1	0.25	3.5
2	0.5	2.75
3	1.0	2.25
4	1.75	1.75
5	OFF	1.5

Note that the blue-green differences in this table are the same as in the first table. Blue-green difference determines grade.

Mark Overton