Trying to figure out exposure for Stouffer TP120-31

Bill Burk · Dec 13, 2013

qualsound said:
Bill Burk your opinion on my conclusion is very much appreciated! Thank you for your help!

Best regards, Christoph.

Looks good to me at first glance.

I'm graphing your HP5 to double-check.

qualsound · Dec 13, 2013

Bill Burk,

Thank you for your interest andd helpfulness!
I looked at the results closer now and made a few corrections concerning dev.time and filmspeed (see below) Why is it that at n+3 both, fp4 and hp5 go beyond their actual box speed of 125 and 400 (i printed the curves from the excel sheet on a dina4 paper and extended the curves by hand, considering they grdually went higher, until they intersected with the lines at +2 and +3)?
Bill you wrote i have to take the filmspeed results with caution...why is that...and how can i determine the correct film speeds and dev.times then...if my results should be too far off?

@tomasz: i will wait what bill suggests and see then what to do next...what about the results of your test?

FP4
N+3: iso 160 @ 23'
N+2: iso 125 @ 17'30"
N+1: iso 100 @ 11'35"
N: iso 80 @ 8'
N-1: iso 64 @ 5'30"
N-2: iso 50 @ min 4'55"
N-3: iso 40 @ min 4'30"

HP5
N+3: iso 640 @ 20'30"
N+2: iso 400 @ 16'30"
N+1: iso 320 @ 12'15"
N: iso 250 @ 8'50"
N-1: iso 200 @ 7'
N-2: iso 160 @ min 5'50"
N-3: iso 125 @ min 5'10"

Bill Burk · Dec 14, 2013

Hi Christoph,

This is your data as I would interpret it. Notice the ASA triangle. The tip of the triangle at the toe of the curve that fits that triangle, is where I marked 400, the ISO speed of that film. The other speeds marked are based on the point where their curves reach 0.10 above Base+Fog. The relative speeds can be read directly above that point off the chart pasted in the middle of the graph.

HP5
N+1 CI .72: ei 640-800 @ 16'
N CI .60: ei 400 @ 11'
N-2 CI .45: ei 320 @ 8'
N-3 CI .37: ei 250 @ min 5'30"
CI .25: ei 200 @ min 4'

wiltw · Dec 14, 2013

qualsound said:
what i still don't know is which bar (number) on the step wedge (either 21 or 31) represents medium gray...so that i could take a spotmeter reading off of that to determine expousre...

'Middle grey' by definition is the center of the range of densities, the midpoint between 'black' and 'white'. So on the 21 step wedge it would be #11 (plos and minus 10 steps to either side), and on the 31 step wedge it would be #16 (plus and minus 15 steps to either side), logically speaking!

So try spotmetering that step!

Bill Burk · Dec 14, 2013

But for the purpose of sensitometry, it's more important to select an exposure that gives you as much data as you need. In the graph I just posted, you can see my normal exposure "ruler" at the top of the sheet.

Christoph gave his film about four stops less exposure than I usually do. So you can see some wasted sample points where the toe is flat. If you gave four stops more exposure the next time you do a test, you will get a lot more information on the right-hand side of the graph.

You can see that I wasn't able to accurately measure Contrast Index (CI) on the 4 and 5.5 minute curves because one of the measurement points I needed (marked with a "C") is off the chart.

So Christoph, whatever you did before... Give four more f/stops of exposure next time.

qualsound · Dec 14, 2013

Bill thank you so much for spending your time on charting my data! it is very much appreciated!
What i don't get is why you gain so different results from the same data (the 30 density readings from the stouffer)?
what is it that makes the results of Ralph's excel sheets differ so much from your chart, considering filmspeed and dev. time...?

I think i will repeat the test giving both films 4 stops more exposure...

Best regards, christoph.

Bill Burk · Dec 14, 2013

Christoph,

That's a fair question. Ralph Lambrecht posts here regularly and often answers directly.

I've seen and tried to reconcile this difference between his programmed results and my paper graphs before.

First and most important. You can make excellent negatives following the results of his spreadsheet.

Second (and only less important because I want to stress the first point), Flare is distorting the results.

Third (this is where it becomes a journey), I think we define Quality differently.

I make negatives in the "West Coast" tradition of full shadow detail targeted for "Grade 2" for a "Diffusion" enlarger. I try to obtain the Best Possible Negative - For Printing.

I think Ralph prefers High Resolution, Minimal Grain, obtaining the Best Possible Print.

Notice, there is a subtle difference between what I want (pleasure in the darkroom) compared to what Ralph wants (pleasure at the Gallery Opening).

Bill Burk · Dec 14, 2013

I want to add more thoughts...

Ralph's spreadsheet times will not make negatives that are "hard" to print in the darkroom... It's just that ease in the darkroom wasn't necessarily his measure of "quality". The difference between a properly exposed negative developed for 8 minutes compared to 11 minutes is about one paper grade number. Well within what can be pleasantly printed.

A negative that is "hard" to print in the darkroom would be one that was underexposed and underdeveloped... And you wouldn't make that kind of mistake after going to this much testing. Obviously following Ralph or my guidelines you would never expose the film at EI 800 and develop for 4 minutes.

I stray from tradition by choosing a longer development time for N than Ansel Adams would have chosen. I keep forgetting to mention, I put N at the ASA triangle (that dashed-line triangle I drew on the graph). Ansel Adam's N is probably between 8 and 11 minutes.

wiedzmin · Dec 14, 2013

Bill,

Description to Ralph's test procedure says that it will print well on grade 2 paper. Do you think that difference can be caused by
Dmin = 0.17, Dmax=1.37 used in the spreadsheet (input data sheet)? Also normal gradient by default is 0.57 (summary sheet).

Tomasz

Bill Burk · Dec 14, 2013

I think the difference is due to trying to calculate Average Gradient (and for me, the Contrast Index) with less than enough information...

I aim for 0.62 Contrast Index, which I believe comes at a time slightly greater than 11 minutes. My aim fits 7 stops of subject brightness on Grade 2. Ralph's 0.57 would be 7 1/3 stops subject brightness on Grade 2, a trivial difference.

Ralph's aim for 0.57 Average Gradient, is calculated by the spreadsheet at 8 minutes.
But at 8 minutes I compute 0.45 Contrast Index !

The difference between 0.62 and 0.57 would be trivial. But the difference from 0.62 to 0.45 is significant. Now we have different development times advice that could impact your negative quality.

You really want a density difference from B+F to 1.05 above B+F for a negative that prints on Grade 2. So look at the graph! Which curve will get you 1.05 over a range of 7 to 7 1/3 stops? ... The 11 Minutes time. The 8 minutes curve gives you 0.90 above B+F in that range... Suitable for Grade 3.

Unless you are talking about a Condenser enlarger compared to a Diffusion light source...

wiedzmin · Dec 14, 2013

Bill,

thank you for the explanation. I'm not trying to question your method. I'm trying to learn and understand. I was reading regarding film test from WBM 2 and Phil Davis Beyond the Zone System (I use BTZS Plotter for Windows program) but I'm getting different results. That's the reason for those questions.

Do I understand it correctly that Ralph's spreadsheet aims for AG 0.57 and yours is based on the calculation?

in WBM 2 Ralph is referring to Diffusion light source

Bill Burk · Dec 14, 2013

Thanks wiedzmin for clarifying.

Even in my notes 0.57 average gradient is good for Grade 2 for Diffusing enlarger for 7 1/3 stops subject brightness range (which is a well-recognized standard for "normal").

0.62 average gradient is good for Grade 2 for Diffusing enlarger for 7 stops subject brightness range.

And this difference is insignificant - either way works for me.

So the main discrepancy is the question of... how many minutes of developing time gives 0.57 average gradient?

It just doesn't look like 8 minutes to me.

wiedzmin · Dec 14, 2013

so you think that short development time is because test was underexposed 4 stops? or spreadsheet is overoptimistic with using grade 2 paper?

Christoph, I hope that you do not mind that I'm using your thread.

Bill,

if you will have a moment could you take a look on attached files? xls it is Ralph's spreadsheet with my test of acros 100 in hc110, pdfs are from BTZS Plotter program one is without flare density entered , the other one with flare density of 0.02

spreadsheet calculates N development 6 min, BTZS plotter `5:15 with flare 0.02 , and just below 5 min without flare
ISOs are also different

Do charts from BTZS program look ok?

thank you
Tomasz

Bill Burk · Dec 15, 2013

I think curve-fitting is difficult when there isn't enough information, so yes having more data will help matters.

But flare really gets my goat. When you tape a Stouffer scale to a glass and photograph it, you just can't get any information for steps with density higher than 2.0. At that step, only 1% of the light should reach film. But due to scattering around in the optics and inside the camera, you can't reduce the light further, for example at 3.0 only 0.1% of the light should reach the film. But what I see looks like 1% of the light hits the film.

Is there any way you can contact print a step wedge?

qualsound · Dec 15, 2013

wiltw, thank you for your answer!
when i exposed the hp5 and fp4, i sticked the stouffer TP120-31 to a window masked it with black cardboard and took an average reading with the prism finder's internal light meter (relying that it would give me a more or less accurate exposure and compensating for the 2 stops for the bellows i used).
i wasnt able too take a reading off of bar nr. 16 with my spotmeter, because one bar on the TP120-31 is 3,3 mm wide, that is too narrow to take a reding with my pentax off of that single bar.
to take a reading of one single bar with a spotmeter, i think you have to use a Stouffer 4x5 inch with 31 (or 21) bars.
saying that, the spotmeter should have a close up lens, so that you can get close enough with the spotmeter to that one bar (16 or 11) to take a reflective reading...

FYI, i asked at Souffers for bar widths of different step tablet sizes,...so here is what i got as a reply:

The step size of the TP120-21 is 4.9mm and the TP120-31 is 3.32mm.

The step size of the TP4x5-21 is 9.525mm and the TP4x5-31 is 6.4mm.

qualsound · Dec 15, 2013

wiedzmin said:
Christoph, I hope that you do not mind that I'm using your thread.

@Tomasz, no i don't mind...it is rather a very good opportunity to learn and understand quite some things about film testing, etc......so thank you and bill for your input and time invested in this thread...

@Bill,...considering flare...isn't it that Ralph's test is taking that in account so that the results are more "real" when exposing the test films...like flare from the lens and camera you use...should be considered...because "out in the field" you are using that same equipment taking actual photographs...?!

Bill Burk · Dec 15, 2013

FIRST AND MOST IMPORTANT... I SEE WHERE THE DISCREPANCY CAME IN.

You manually overrode the Average Gradients that are properly calculated on "Input Data" page, when you got to each Curve page.

Cell L 48 - on every page "Curve 1" "Curve 2" "Curve 3" "Curve 4" "Curve 5" does not match the calculated cell K 48.

For specific example, Ralph calculated 0.59 for 11 minutes where I calculated 0.6 !!!!!

Were you aware of this adjustment? Or does it contain default values from the published spreadsheet?

If it contains default values from the published spreadsheet... Then my long-standing puzzle has been solved!!!

Back to the test design...

It's better science to eliminate variables from testing. Flare is a variable that can easily be eliminated from the experiment by placing the Stouffer scale in contact. But I didn't come here to harp on that.

The bigger problem flare causes IN CAMERA TESTS... Is that you just can't get anything useful from the steps denser than 2.0

I have another suggestion. Only take the results from steps with densities 1.60 and less. Make series' of exposures. One just as you did. Another shot with 4 stops more exposure. For the numbers you enter in the spreadsheet, only take the "top half results". Ignore any densitometer readings from the test film where the step wedge density was above 1.60 because you can't trust it. But because you have a high range and a low range, together you should have a full range test.

ACTUALLY, YOU ALREADY HAVE THE LOW RANGE, BECAUSE YOU UNDEREXPOSED FOUR STOPS!...
-Just put it together with another test where you expose for the High Range.

wiedzmin · Dec 15, 2013

@Christoph - thank you

@Bill, thank you for your help and your time. I really appreciate it.

Cell L48 yes I made those changes , it is according to manual Ralph provided with his spreadsheet (I was trying to do everything exactly like in WBM book and link to pdf). http://www.waybeyondmonochrome.com/WBM2/Library_files/FilmTestEvaluation.pdf page 5 right column in the middle. It looks like Christoph did the same.
quote from the pdf:
"Start by visualizing a smooth curve through the existing
data points in fig.5a. Then, imagine an extension of that
smooth curve all the way up to the right cross hair (Dmax).
Now, adjust the average gradient value in cell L48 slightly,
trying to improve its current position until it intersects with
your visualization of that smooth extension. This may need
a little practice, but as you will see, it does not have to be
perfect to significant improve the following calculations."

Is there any way you can contact print a step wedge?
I have only 35mm and 120 step tablets. For contact printing, I'm guessing I would need 4x5 and maybe cut it in half and tape it so it will fit on roll film length wise? What light source should I use for exposure?
For my tests I used something like that http://www.pentaconsix.com/C436_29.jpg, http://www.pentaconsix.com/C436_32.jpg but with Hasselblad 500cm. In addition I placed collapsible diffuser between camera and the flash.

Back to the test design...

It's better science to eliminate variables from testing. Flare is a variable that can easily be eliminated from the experiment by placing the Stouffer scale in contact. But I didn't come here to harp on that.

The bigger problem flare causes IN CAMERA TESTS... Is that you just can't get anything useful from the steps denser than 2.0

I have another suggestion. Only take the results from steps with densities 1.60 and less. Make series' of exposures. One just as you did. Another shot with 4 stops more exposure. For the numbers you enter in the spreadsheet, only take the "top half results". Ignore any densitometer readings from the test film where the step wedge density was above 1.60 because you can't trust it. But because you have a high range and a low range, together you should have a full range test.

ACTUALLY, YOU ALREADY HAVE THE LOW RANGE, BECAUSE YOU UNDEREXPOSED FOUR STOPS!...
-Just put it together with another test where you expose for the High Range.

sorry it is my fault by using Christoph's thread but above suggestion is for Christoph's or my test?

thank you
Tomasz

Bill Burk · Dec 15, 2013

OK so the adjustment caused the discrepancy.

I also had to do some visualization on paper to get CI. You can see my dashed lines reaching up to the letter "C" in the right margin.
But the "extensions" I drew were fairly short. You got the majority of the data needed to determine "CI".

For now, I recommend trusting the Average Gradient as-is and not adjusting.

Next tests when you have more data, you won't need to adjust Average Gradient.

Since you went to the trouble to setup a camera according to Ralph's instructions, go ahead and keep using that setup.

The one improvement from this thread I recommend going forward is to make two separate exposures, one with three stops more exposure than the other.

Bill Burk · Dec 15, 2013

Tomasz,

My comments regarding adjustment errors relate to Christoph's HP5 data that I graphed on paper.
His figures were significantly different than mine, while the calculated results are very close.

Your adjustments (HC 110 and Acros 100) are minor in comparison.

wiedzmin · Dec 15, 2013

Bill Burk said:
Tomasz,

My comments regarding adjustment errors relate to Christoph's HP5 data that I graphed on paper.
His figures were significantly different than mine, while the calculated results are very close.

Your adjustments (HC 110 and Acros 100) are minor in comparison.

thank you,

In my case do you recommend? :

the one improvement from this thread I recommend going forward is to make two separate exposures, one with three stops more exposure than the other.

Do you mean by that exposing 10 rolls of film?

Bill Burk · Dec 15, 2013

wiedzmin said:
thank you,

In my case do you recommend? :
Do you mean by that exposing 10 rolls of film?

Wow that's a lot of film for testing. I normally test on sheet film and fit three "tests" on a single sheet of film.

And for 35mm, I add the test at the end of any particular unfinished roll that might be in a camera.

But in a roll camera can't you take two shots and then cut it off and develop it for one of the tests?

wiedzmin · Dec 15, 2013

I never tried that but I guess it will work. What about taking 6 exposures (120 film) + 6 exposures (+3 stops )? I could cut roll in half before loading it into reels. This way I will use 5 rolls but have 10 tests. Will developing only half of the roll affect test accuracy, I would normally develop full roll in the same volume of developer?

I was under impression that it is better to perform film test using actual equipment which is going to be used to take photos based on film test. You pointed out flare problem but would not that flare exists in "real" photos taken with that equipment anyway?

If I understood your suggestion correctly:
I should for each developing time (4, 5.5, 8, 11, 16) read steps up to 1.6 density from normal exposure and above 1.6 from those with +3 stops of exposure? Enter values to the spreadsheet and do not modify AG for each curve.
Did I understand it correctly?

thank you

Bill Burk · Dec 16, 2013

I've reeled a 35mm roll with nicks at each test and then pulled and ripped off a strip when I reached pre-planned minutes. It'd be easier to do 2 reels with 2 strips of film reeled onto each reel, already cutoff. Use the normal volume of solution. Actually it does matter "where" in the reel a test is placed, but the difference is small enough to not be important. I have done tests where I graphed several strips from a single roll of 35 mm film. Each curve differed slightly from the others - despite being on the same reel and developed for the same time. At some point you have to accept these minor variations.

---

You can estimate flare and add it back in when you interpret the results.

---

I'm suggesting that you don't even bother reading any steps past step 16 on your 30-step test wedge because flare gives the entire sheet of film an exposure near what step 20 gets, which ruins the samples below step 20 and interferes with the accuracy of the steps from 16 to 20. But from steps 1 to 16 flare doesn't affect the exposure much.

qualsound · Dec 16, 2013

Bill Burk said:
Tomasz,

My comments regarding adjustment errors relate to Christoph's HP5 data that I graphed on paper.
His figures were significantly different than mine, while the calculated results are very close.

Your adjustments (HC 110 and Acros 100) are minor in comparison.

Bill thank you for clarifying.
I adjusted the average gradients of each curve in cells L48, like tomasz said, according to ralphs instructions in the pdf file accompanying the excel spreadsheet. It was my intention, that the curve would intersect the crosshair... i see now that it is better to leave these cells untouched and let the spreadsheet do the calculating.

Trying to figure out exposure for Stouffer TP120-31

Attachments

Privacy & Transparency

Privacy & Transparency