Hitting ASA Triangle Does Not Mean You Got Full Film Speed

I'm consistently getting EI 250 from TMY-2. Have been for two years. The curves look great. The negatives look great (except for scratches that I complain bitterly about).

All along I thought I was getting full rated speed of 400 because I hit the ASA triangle. I made 400 my benchmark.

(Real quick: ASA triangle is 0.8 rise over 1.3 run from 0.1 over base plus fog).

I didn't question my benchmark until a roll of Panatomic-X rated extremely fast.

Reference post 146 of this thread...

(there was a url link here which no longer exists)

I know tray processing in stacks of six sheets gives less agitation than small tanks with plenty of developer circulating around the film. But I thought hitting the triangle made up for it.

In that thread at least two estimates discredited my 400 benchmark. Now I believe my benchmark is close to 250 (may move again if new trends emerge).

Ironically, I felt clever using half box speed from the beginning. And even more ironic... after a while I felt like bumping it up to 250 because of consistently "more than enough" shadow density.

So even though I thought I was getting 400, and thought I was overexposing... I was actually using the right speed all along. That is just too scary. I could just as easily have used 400 and botched a lot of film.

I know I have insufficient agitation to meet ASA specification. I may continue to develop the way I do because the negatives meet my quality standards. I currently have no "need for speed".

But at least now there is little risk that I will overrate my film. It also emphasizes the need for photographers to determine their own effective EI. From now on, I'll be using 250 for TMY-2 and it's not just because it's "half box speed"...

Bill Burk said:

I'm consistently getting EI 250 from TMY-2. Have been for two years. The curves look great. The negatives look great (except for scratches that I complain bitterly about)

Click to expand...

You are "there."

I'll mention that my in-camera testing (where I can associated an exposure to a number on the camera or meter) produces similar results to what you have found (EI 200 to 250).

I'd not worry about the number. Maybe we should call it the "Exposure Index Triangle" because the results usually are used as an Exposure Index and the conditions under which the test is performed frequently do not follow all ASA or ISO 6-1993 guidelines.

Stupid question, way out of my league here:

Is this the same as the ASA (angle, side, angle) triangle?

http://www.mathsisfun.com/algebra/trig-solving-asa-triangles.html

This is what Bill is referencing.

Bill,
I'm graphing your data to double check it. Maybe Steve can do the same.

I started with the 6min graph, just because it was first in line, though, I could see the gamma was going to be low.

My first graph has some oddball things, so maybe there are some issues that need to be addressed.

1) You indicate "calibrated grayscale" but are the values you gave your actual densitometer readings from the grayscale? It would be hard to believe your densitometer matches that gray scale exactly. I use readings from the same densitometer on both X and Y axis. If you densitometer is far from the calibrated step wedge on a lot of the range, the curve might look better if you add a correction factor to each film step checked. (unless your densitometer agrees with every calibrated step).

2) I process TMY 2 for 6 minutes and get a gamma around 0.7. In your case the gamma was only 0.3. I know you used the steeper curves to fit the ASA triangle (per protocol) and Steve posted that ISO criteria for specific developers was relaxed, I think you may be using a developer combination or process that is not giving you full speed available from the film. Not all developers give the same 0.1 point when processed to match the ASA triangle.

Which dataset did you use to match the ASA triangle? I'll graph that one next.

3) The curve just 'looks odd' I have double checked all the points and I have graphed exactly the dataset below. Perhaps there is an error in data acquisition?

We can look at these speed numbers 'for educational purposes' but since this particular sample is not processed to a gamma you are likely to ever use in practice, I'll ignore those values.
The "W" point is a mathematical estimate of 0.3G point (likely similar to Delta-X point, maybe Steve can calculate the Delta-X on this dataset)
The ISO estimate is one stop to the right of the "W Speed" point
The inertia point is where the gamma line hits the X-axis
The ASA point is not indicated because the gamma is too low to fit the ASA triangle in this sample
The Delta-X point is not indicated but I may add that functionality to my spreadsheet later.




Step,Density
01,0.05
02,0.22
03,0.34
04,0.48
05,0.62
06,0.76
07,0.93
08,1.07
09,1.20
10,1.35
11,1.50
12,1.65
13,1.80
14,1.94
15,2.08
16,2.22
17,2.37
18,2.52
19,2.65
20,2.78
21,2.92
22,3.07
23,3.20
24,3.32
25,3.47


9/9/2011 TMY-2, 6 min, B+F 0.03
---
Step,Density
01,1.22
02,1.05
03,0.97
04,0.90
05,0.84
06,0.76
07,0.69
08,0.63
09,0.55
10,0.49
11,0.44
12,0.38
13,0.33
14,0.29
15,0.25
16,0.21
17,0.17
18,0.13
19,0.10
20,0.07
21,0.05
22,0.04
23,0.01
24,0.00

There's something problematic going on in the toe. In my opinion, this questions the legitimacy of the test. I don't think you can confidently conclude anything from this. Sorry.

Thanks, Stephen. I am a bit wiser now that I see the graphic.

The toe data had a transcription error: I corrected the numbers to agree with what I used for graphing and replaced the original file on web server.

---
9/9/2011 TMY-2, 6 min, B+F 0.03 *Corrected toe*
---
Step,Density
01,1.22
02,1.05
03,0.97
04,0.90
05,0.84
06,0.76
07,0.69
08,0.63
09,0.55
10,0.49
11,0.44
12,0.38
13,0.33
14,0.29
15,0.25
16,0.21
17,0.17
18,0.13
19,0.09
20,0.07
21,0.04
22,0.01
23,0.00

---
I have not verified the calibrated step tablet. Can't remove it from glass to clean without possibly destroying it. (First on list of things to replace). It is a 21 step tablet, and I extend its range to 25 steps by using a gelatin 0.6 ND filter. (Steps 22-25 are not really calibrated). I put two or three step wedge exposures on a sheet of film, and average the readings.

My ASA triangle typically is 13 minutes. (Jerevan, it's just a nickname I gave the ASA/ISO standard that defines two sides of a right triangle). I use D-76 1:1 and process in trays with 6 sheets at a time, tightly stacked.

I used to "gently" and "loosely" stack the film which gives these curves upswept high values. That is not a film characteristic, it is a common development anomaly in all my curves. I ignore it.

I assume upswept curves happen because edges with more access to fresh developer - develop more - and that is where the high steps are positioned, near the film sheet edges. Lately I have been "jogging" the sheets. This reduces the upswept curves... At the same time restricts the entire sheet's access to fresh chemistry.

The high end of the step wedge is where it is important to know the correct values. That is where the speed point will fall. My densitometers calibrates to a standard around 2.0, so when I measure my step wedge in the 2.0 range I know pretty well that the densities I'm using are correct.

So, with unknown high densities on wedge and the addition of extra density, you could easily be 1/3 stop off in the calculations of the light reaching the film at the 0.1 point, thus giving you the wrong speed.

That is interesting that the step wedge is between glass. On my EG&G, and the description in the manual, the plastic chamber ("gray scale box" per the manual) has a clear glass on top.

On my Wejex, the step wedge is also between glass, but the top of the platform, on which the glass resides, actually fits under the arm of my Tobias densitometer. Maybe they planned it that way because both devices are made by the same company.

Extra agitation or development should have minimum effect on the speed point, so I am suspecting errors in measurement of the step wedge density. But just speculation at this point.

With the revised numbers it still looks pretty odd at the toe. But this is not the dataset you fit to the ASA triangle, I suspect you used on of the others (which one?).

13min Graph looks ok.

ic-racer said:

With the revised numbers it still looks pretty odd at the toe. But this is not the dataset you fit to the ASA triangle, I suspect you used on of the others (which one?).

Click to expand...

Dataset error again. Sorry. The numbers are scrawled on unlined paper and I averaged them in my head when I graphed the first time. I did it right the first time, the graphs are good, but yesterday and today it wasn't easy lining up the columns...

The test wedge is just "taped" to top of glass, but I just am leery of tampering with it. The higher densities are "calibrated by me". I read the 2 stop ND filter on the densitometer, and added the result to the calibrated step it overlaid to get the simulated steps above 21. Next time I am getting a 31 step wedge.

The 13 minute fits the ASA triangle.

Bill Burk said:

The 13 minute fits the ASA triangle.

Click to expand...

I took the opportunity of adding some functionality to my spreadsheet. Now it also calculates the "0.1" point and automatically shows the correct placement of the ASA triangle on the grid. The user can easily determine if the data passes through the red circle or not. In this case you can see that it does pass through the red circle (1.2 log out from the speed point and 0.8 log up from the speed point).
So the spreadsheet automatically calculated the 0.1 point as 3.11. How does that compare with what you got by hand?



The 0.1 point is the GREEN CIRCLE and the ASA/ISO speed point is the GREEN TRIANGLE. To save weeks of programming I did NOT use a spline. I used linear interpolation. As you can see it is as good as one could eyeball it on graph paper. I'm happy with it.
The WHITE TRIANGLE is the W-speed point with the safety factor added (surrogate for 0.3G point with one stop safety factor). Delta-X plus safety factor would likely be similar. For educational purposes, I guess the next step is to program the Delta-X into the spread sheet....

At this point I'm reluctant to pass out the spreadsheet as the calculations are hidden all over the place and the spreadsheet is essentially held together with bailing wire and duct tape. One click in the wrong place and it could start giving bogus numbers or stop working all together.

For example there are over eighty cells with the following equations to systematically search for the W point and the 0.1 point:

=IF(H23=0, 0,(C23+(H23*($E$35-B23))))
=IF(AND(B23>=$E$35,B22<$E$35),(((C22-C23)/(B22-B23))),0)
=IF(AND(C25>=0.1,C26<0.1),(((C25-C26)/(B25-B26))),0)
=IF(K25=0,0,(B25-(C25-0.1)/(K25)))

3.11 is what I got by linear interpolation as well. And right through the circle too.

Not Hitting Full Film Speed In Tests Does Not Mean You Didn't Get Full Film Speed

I did quite a few development tests the past few days, where I developed one sheet at a time versus my usual six sheets at a time.

The only thing that seemed to change was decreased time to develop to similar contrast.

I didn't see any speed increase.

The EG&G results remain around 200 to 250. I am used to that.

I may make other attempts to increase speed. But I am beginning to think maybe I "really" get full speed. Maybe the tests just aren't confirming. There could be a good reason. Maybe the "discontinuous spectrum" of the xenon flash doesn't match daylight... Maybe excessive "blue" absorbtion by the No. 96 ND filter makes it a poor choice to include in sensitometry tests...

I believe now that stacks of six in tray provides "pretty good" agitation. Earlier I posted a significant difference but that was a mistake because I accidentally developed in Dektol 1:1 instead of D-76 1:1 as I planned.

Bill Burk said:

But I am beginning to think maybe I "really" get full speed. .

Click to expand...

I'd re-do the estimated EG&G light output based on your tests. With all the estimated densities in your calculation of ISO 200/250 it could easily be 1/2 stop off. That is, work backwards, saying the film is ISO 400 by definition and your light out put is 2 millilux-seconds at the 0.1 point after you subtract all the density you have between the lamp and the film. That is the way I calibrated mine.

ic-racer said:

I'd re-do the estimated EG&G light output based on your tests. With all the estimated densities in your calculation of ISO 200/250 it could easily be 1/2 stop off. That is, work backwards, saying the film is ISO 400 by definition and your light out put is 2 millilux-seconds at the 0.1 point after you subtract all the density you have between the lamp and the film. That is the way I calibrated mine.

Click to expand...

Big issue is the contradiction that I got from Panatomic-X. I am using that placement as the position for the film speed scale.

Differences I think play into it are:

-> Panatomic-X is a traditional emulsion, TMY-2 is T-grain.
-Different spectral sensitivities, Panatomic-X vs TMY-2... Maybe the EG&G is "bang on" for traditional emulsions so it was deemed a good sensitometric light source on that basis... But maybe it is not as great a match for TMY-2.

-> No. 96 ND 0.6 has its limitations in sensitometry.
-I do not use it at all to catch the toe of Panatomic-X.

Bill, why not do a TMY test with the ND 0.6 on top of the flash housing? Forget about trying to extend the range of the step tablet and limit the variable.

Just to record a thought, I had not been consistently applying "hold time", as Stephen pointed out in a similar thread the temptation is there, the lab is setup and ready, to expose and process immediately.

And so. It is quite possible the Panatomic-X reads higher speed than its rated ASA 32 in my testing - because I developed immediately after exposing.

(there was a url link here which no longer exists)

Bill Burk said:

I'm consistently getting EI 250 from TMY-2. Have been for two years. The curves look great. The negatives look great (except for scratches that I complain bitterly about).

Click to expand...

Bill, what is causing the scratches?

cliveh said:

Bill, what is causing the scratches?

Click to expand...

Me. Sheet film in trays, emulsion up. I cannot resist moving the sheets constantly. I work diligently too. Wear gloves, attend to every moment the film is wet, including wash.

The scratches are very fine, and I only get one or two every dozen sheets. But, they are there.

Bill Burk said:

Me. Sheet film in trays, emulsion up. I cannot resist moving the sheets constantly. I work diligently too. Wear gloves, attend to every moment the film is wet, including wash.

The scratches are very fine, and I only get one or two every dozen sheets. But, they are there.

Click to expand...

Well perhaps you should not move the film constantly and when you do, just touch them at the edge.

Reviving an old thread because I just developed a roll of TMAX 100 35mm in D-76 that was mixed 41 days before developing...

And I believe I lost 1/3 stop of speed compared to the same batch of TMAX 100 where I ran a family of tests on 1 day old D-76.

Development time was the same for the family curve that met ASA Triangle and my recent roll, which also met the ASA Triangle.

But the LogE displacement caught me off-guard. The new curve is 1/3 stop to the right. Speed is still above 80 by my calculation (but only 0.02 density units above).

So where I formerly held the position that D-76 can be used "practically forever" I must now take the position that "Yes, D-76 can be used practically forever, but you might lose a little speed - therefore it falls in the category I would call alternative processing." The alternative I am referring to in this case is "instead of using known good developer, experimenting with old developer." And that can be fine for thrift or occasional hobby developing.

p.s. I work for Kodak but this position I'm taking is my own and not necessarily that of EKC.

That thread shouldn't be titled "TMY is 250 speed", but "TMY in D-76 full strength is 250 speed". AFAIK box speed is no longer tied to a specific developer. There is a good chance that TMY will reach true ISO 400 speed in a more modern developer, or even D-76 1+3.