PDA

View Full Version : How Do I Extrapolate Reciprocity Times From Manufacturers Data Sheets

IanBarber
6-Apr-2017, 14:27
I need to make a chart for some new films I want to try

TMax 400 is the first film and the chart below is from Kodak's website but I want to make a chart in more easy to understand times like...

10 seconds = ?
12 seconds = ?
15 seconds = ?

and so on

From this chart, how can extrapolate the information

163536

Thalmees
6-Apr-2017, 16:26
Hello Ian,
Since Kodak provided only 3 points of time(1, 10 & 100 seconds), I think the relation is linear, at least the interval between 10" and 100".
Foe example, 100" needs 1.5 stops, and 10" needs 0.3 stop, so by simple calculation, a 55" needs 0.9 stop(1 stop practically), and so on.
Hope this helps.
Or, just draw axis and connect between 0.3 stop and 1.5 stop, and all times in between go to the corresponding stops on the other axis.

pjd
6-Apr-2017, 16:26
Make a graph with a line between the two points given (10 and 100 seconds) to get a rough idea, then keep detailed exposure notes and find out what works for you. It might be worth noting that for 100 second exposures Kodak data indicates that TMAX 100 is faster than 400 (1 stop extra, rather than 1 1/2 indicated).

Pere Casals
6-Apr-2017, 16:31
I need to make a chart for some new films I want to try

TMax 400 is the first film and the chart below is from Kodak's website but I want to make a chart in more easy to understand times like...

10 seconds = ?
12 seconds = ?
15 seconds = ?

and so on

From this chart, how can extrapolate the information

163536

If you are to deal with long exposures consider to do the easy thing: Use Fuji Acros, it has no remarkable failure.

Reciprocity failure given by manufacturer is orientative, an starting point. Shadow areas have more failure that lights, so calculation is not trivial, your exposure will depend on if you want to expose well shadows or highlights.

The information that gives Kodak is enough for a good guess, Ilford gives you that chart: http://www.ilfordphoto.com/Webfiles/20106281054152313.pdf

At the end the thing has another answer: do your calibration !!!

Take the Stouffer wedge and make calibration with different exposures of our interest. Then you can relate lux·second received during the long exposure with process and density achieved.

Another way is to take a 135 roll of same film (emulsion will be near the same) and make bracketings so you will know what happens with lights and shadows.

Also I suspect that reciprocity failure of same film may vary from batch/aging.

But I reiterate: Acros and no problem. For very long exposures you get rid of a lot of problems.

Regards.

Thalmees
6-Apr-2017, 17:19
Hello again Ian,
This is a fast graph shows Time in "X" axis, and stop compensation in "Y" axis.
It's just a rough guide.
163544
Still, it could be linear relationship between 10" and 100".
But this is the simplest approximation to find your result.
BTW, TMY2 sheets returned to its previous price at freestyles.
Hope this helps.

Neal Chaves
6-Apr-2017, 19:50
Do your own tests with the developer you intend to use. I found no loss in speed or significant change in the curve shape out to one minute.

Fr. Mark
6-Apr-2017, 20:47
I should ask for some acros for my birthday...it's also supposed to be super fine grain in xtol. More to the point there is an inexpensive app for the iPhone called reciprocity timer. It will help with bellows factors too. And filters and you can put in your own data too in some versions, I think.

Doremus Scudder
7-Apr-2017, 07:48
I need to make a chart for some new films I want to try

TMax 400 is the first film and the chart below is from Kodak's website but I want to make a chart in more easy to understand times like...

10 seconds = ?
12 seconds = ?
15 seconds = ?

and so on

From this chart, how can extrapolate the information

163536

Ian,

Everything you need to get started is here: http://phototechmag.com/black-and-white-reciprocity-departure-revisited-by-howard-bond/

The article is a bit old, but still applicable.

Best,

Doremus

IanBarber
7-Apr-2017, 09:06
Thanks everyone for the replies. I will study these links and as suggested, I have ordered some Acros 100 to try as well for long exposures on the coast

Jerry Bodine
7-Apr-2017, 13:43
Ian, I'm not sure where you got the chart you show, but it differs considerably from the Kodak chart presented on p.45 of my copy of The Negative. However, FWIW if you want to persist with using your chart, I've attached a screenshot of an Excel plot that allows you to read off all exposure corrections within the scope of the data. I've also attached a similar screenshot of the Kodak chart for comparison (I think if you right click on the attachments, you might be able to get a print). Note that Kodak indicates a need for reductions in dev time to counter the increased highlight densities that results from reciprocity corrections. The reason for this is explained by understanding that the exposure correction factor is only needed to preserve the negative's density level in the dimly lighted shadows, but the highlights also receive that same factor increase even though it's not needed and can get "blown out" unless development is reduced.

You may find this (http://www.unblinkingeye.com/Articles/LIRF/lirf.html) link interesting as well. I'm pretty sure the author has made use of Howard Bond's data points. Tmax400 is included and mentioned specifically at the very end. The article doesn't mention reduced development, but the relationship between exposure factor and f-stops (zones) is useful if you've done the Zone System tests for dev time. If you want to follow his system you'll need some log-log paper for your plots; this (http://customgraph.com/piart.php?art=446) link lets you print out whatever plot paper you want (4cycle x 4cycle base 10 is recommended).

IanBarber
7-Apr-2017, 14:29
Jerry, thanks for that, this looks extremely interesting to me. Out of interest, did you have to use many formulas for excel to draw that cure and plot the points

Jerry Bodine
7-Apr-2017, 16:34
Jerry, thanks for that, this looks extremely interesting to me. Out of interest, did you have to use many formulas for excel to draw that cure and plot the points

Not really. Excel offers a host of curve types to try and get a good fit to the data points. I find the polynomial curves most useful; one can select 2nd order through 6th order to visually see which works best. As you can see in the attachments, I've opted to have the equation for the curve placed on the plot and in both cases the 2nd order (e.g., first term in the equation indicates the x-value has an exponent of 2) was adequate. It takes a while to learn how to work with Excel but I think it's time well spent. I use it for many things. My engineering experience allows me to read curves of plotted data easily and I'm often able to spot errors in the data points when the curves have strange hiccups.

Doremus Scudder
8-Apr-2017, 02:38
Ian, I'm not sure where you got the chart you show, but it differs considerably from the Kodak chart presented on p.45 of my copy of The Negative. ...

Jerry,

The Kodak data sheet for TMY shows exactly what Ian posted. The info from The Negative is likely for older, conventional Kodak films; it seems to agree with the current published reciprocity characteristics for Tri-X.

The reciprocity characteristics of many Kodak films changed considerably when they were reformulated for the newer coating facility (sometime around 2002-3 if I'm not mistaken). Kodak, however, didn't change their recommended compensation, however, The point of Howard Bond's article was to test the new films and arrive at more reliable data. Kodak still hasn't revised their recommendations...

Charts like you posted above based on Howard Bond's data points for the films he tested would be really interesting. I'm pretty good at Excel, but don't know how to get the graphs going. Care to give me (us) a quick tutorial?

Best,

Doremus

Pere Casals
8-Apr-2017, 03:40
Jerry,

The Kodak data sheet for TMY shows exactly what Ian posted...

What I'd like to know is what Fuji do with Across to remove the failure...

I've been reading recently about hypersensitization, used in astronomy (lumicon etc), perhaps it's related to that, or by using the sensitizing dyes developed for color films...

koraks
8-Apr-2017, 04:18
Care to give me (us) a quick tutorial?

Make a scatter plot (be sure to pick the scatter plot type, not a regular line graph) of the desired data points
Right-click one of the points in the plot and pick the option 'add trendline'
Then right-click the trendline and choose options/properties/format trendline or whatever it's called (I don't have Excel running on this machine)
Expriment with the different types of trendlines to see what makes for the best match. You can even add an R-square value to the plot so you can see which line fits the data points the best.

The risk, of course, is that you'll find yourself picking a type of trendline that fits the data points perfectly, while in reality, you don't know what the exact relation between exposure time and reciprocity failure is. Could be logarithmic, polynomal, etc. It all remains guesswork, but obviously, the more data points you have, the more certainty regarding the shape of the curve.

Pere;

What I'd like to know is what Fuji do with Across to remove the failure...
I've wondered about this as well. Speculating a bit: the issue with LIRF seems to be that a grain of silver needs 2-4 photons in order to become activated (IIRC). However, the more time passes between the arrival of these photos, the larger the chance that the effect of the previous photon has already disappeared, eliminating the compounding effect of consecutive photons. Improving the situation would require that the effect of a single photon hitting the grain lasts as long as possible. No doubt this can be achieved through the controlled addition of certain doping agents to the emulsion and/or controlled grain growth during the emulsification process. Which of the two is the key factor and which dopants are particularly effective in this, I'm not sure. I've done some limited literature crawling, but haven't come up with anything conclusive.

Jim Becia
8-Apr-2017, 14:30
Ian,

Here are some times from a reciprocity table I have. Don't know if this helps as I do not use this film.

1 - 1.25
2 - 2.5
3 - 3.75
4 - 5
5 - 6
6 - 7
8 - 10
10 - 14
12 - 16
15 - 21
20 - 28
25 - 35
30 - 42
40 - 56
50 - 1'10"
1m - 1'24"
1'20" - 1'53"
1'40" - 4'42"
2' - 5'39"
2'40" - 7'32"
3'20" - 9'25"
4' - 11'18"
6' - 16'58"
8' - 22'37"
10' - 28'17"

Jerry Bodine
8-Apr-2017, 15:33
... I'm pretty good at Excel, but don't know how to get the graphs going. Care to give me (us) a quick tutorial?...

Glad to oblige, Doremus. It so happens that Ian pm'd me this a.m., asking if I'd send him my Excel files from the screenshots. I sent him a tutorial as well to help him out, using my own gmail acct, since I'm not certain emailing him through the forum's system would allow attaching the files. So here's an edited copy/paste of the tutorial I sent him. Coupled with the screenshots I think you can follow it OK:

...Putting the curves’ equations on the plots allows one to calculate precisely any point on the curve. Here’s a step-by-step tutorial for plotting the curve for your chart:

1. Open the file. You’re on the “Home” tab at the top of the page. Also notice the CCW arrow on top row; it can be used at any time to undo any change you just made if you’re not happy with the resulting change.

2. x-y data pairs from the chart are entered in columns A & B (column headings are optional but useful).

3. Select cells A1 to B4.

4. Select “Insert” tab at top of page. Hover pointer over the various chart types to see helpful tips on which type to choose. I usually choose “Scatter (X,Y)” type, so hover over that one, then click the arrow and select “More Scatter Charts” and select “XY (Scatter)” in the drop down window. It‘ll show an example of your plot points on a chart.

5. Beside the chart notice icons [+] [brush] and [funnel]. Click on [+] to see a list of “Chart Elements” you wish to include. I usually pick Axes, Axis Titles, Chart Title, Gridlines, Legend (sometimes), Trendline (this is the one that lets you choose which type of data-fitting curve you want - I frequently pick “Polynomial”). All choices can be edited at any time.

6. Click on the chart presented and see how it’s put in a box with o’s at each corner and mid-sides. These can be clicked and dragged to change the chart’s size as desired. Also seen is what I call a weathervane with N-E-S-W arrows. Click-drag that to relocate the chart on the worksheet as desired.

7. Right click on the Trendline shown to open window to see your curve-fitting options. Clicking on Polynomial (start with Order=2 and progress up to Order=6 and it’ll show you which best fits the data points). At the bottom of this window, click the box “Display Equation on chart”. Right-click the trendline and select “Format Trendline” to change as desired.

8. To edit any title, just click on it and it’ll be shown in a box. The contents of the box can be highlighted and edited (should be self-explanatory). Use the weathervane to move the box to where you want it on the chart.

9. The x or y scale can be adjusted by clicking on it; a window opens that you can play with to figure out. Same goes for editing grid lines (both major and minor).

10. Finally, the worksheet title/color (tab at bottom) can be changed by clicking it, then select “Rename.”

End of tutorial.

It seems to me that it'd be simple to know at which points on the curve a development reduction is called for. E.g., when the reciprocity correction FACTOR (adjusted time / indicated time) reaches 2, that would mean N-1 is needed; when it reaches 4, that would mean N-2, etc.

Hope this helps,
Jerry

Pere Casals
8-Apr-2017, 15:38
I've wondered about this as well. Speculating a bit: the issue with LIRF seems to be that a grain of silver needs 2-4 photons in order to become activated (IIRC). However, the more time passes between the arrival of these photos, the larger the chance that the effect of the previous photon has already disappeared, eliminating the compounding effect of consecutive photons. Improving the situation would require that the effect of a single photon hitting the grain lasts as long as possible. No doubt this can be achieved through the controlled addition of certain doping agents to the emulsion and/or controlled grain growth during the emulsification process. Which of the two is the key factor and which dopants are particularly effective in this, I'm not sure. I've done some limited literature crawling, but haven't come up with anything conclusive.

IMHO this wikipedia article https://en.wikipedia.org/wiki/Photographic_hypersensitization , section "Gas-phase hypersensitization" may give a clue of what kind of processing may lower LIRF.

So a possibility is removing oxygen and water from the gelatin matrix (degassing) and inserting hydrogen.

An ammonia bath (before exposure) also lowers LIRF, but if soon adds base fog.

It is clear that astrophotography was the field where there was great efforts to fight LIRF...

Doremus Scudder
9-Apr-2017, 02:10
Jeff and koraks,

Thanks very much for the help with Excel! I'll be making some graphs for my data soon.

Best,

Doremus

IanBarber
9-Apr-2017, 10:37
Thanks to Jerry for helping out with this, I have now created my first chart for Fomapan 100 based on his tutorial