PDA

View Full Version : Measuring Film Density



ageorge
6-Dec-2006, 12:26
In order to calibrate film development times one needs to measure the resulting densities on the negative. How do you guys do this? I was wondering if there is an economical path. Is a densitometer the only way to go? What do most people use? They seem expensive, can these be rented?

Thanks in advance,
-alan

Leonard Evens
6-Dec-2006, 12:51
If you are talking about using the "Zone System", it is actually possible to do what needs to be done without a densitometer. You just establish a standard way to make prints and then you draw conclusions from visual examination of such prints. I'll leave it to others to explain how to do that or to give references which explain it. But, it is easier if you measure densities. A densitometer would certainly be helpful in doing that, and you may be able to find one for a reasonable price at ebay. You can also, in principle, use a negative scanner. The RGB values you get from such a scan are not very helpful because they are the end of a process of fitting the dynamic range in the negative to the values in the range 0..255. The same density values in different negatives will generally give you different values in that range. But some scanning software allows you to read the the actual densities. I use Vuescan which has such an option, and I check densities regularly. I haven't ever compared the readings with those from a densitometer, but they seem to be plausible. Also, it doesn't really matter since you just need some consistent way to produce readings which are coherently related to the actual densities. You have to relate everything you do to visual inspection of prints or scans in any event.

Ron Marshall
6-Dec-2006, 14:08
I use a light table and my spotmeter. Definately not a rigorous method, but the results are repeatable and it gets me in the ballpark for development times, which I then fine tune visually.

John Cahill
6-Dec-2006, 15:56
Alan,
If you have a good light meter, as Ron says, you can get yourself in the ball park.
IIRC, the difference between film base plus fog level of developed film and the zone I exposure of .1 above film base plus fog level, is 1/5 of an f/stop.

Kirk Gittings
6-Dec-2006, 17:00
Try this, it is a simplification of Fred Pickers approach that Steve and I happened upon independently some 25 years ago. It has worked for me.

http://www.viewcamera.com/pdf/2006/VC_Getting%20Started.pdf (http://www.viewcamera.com/pdf/2006/VC_Getting%20Started.pdf)

Brian Ellis
6-Dec-2006, 18:17
It's possible to pick up a respectable older but still good used transmission densitometr on ebay for $50 - $100 with a little patience. If you're looking make sure to look for "transmission" densitometer, they're the ones that will be relatively inexpensive because they're tied to film. Reflection densitometers are useful in digital printing so they've tended to hold their value better.

ageorge
6-Dec-2006, 18:33
Maybe I should clarify my goal, which is to scan the negatives. So I guess fundamentally I want to tune my development times for the best scanning potential and not a specific scanner per se as I have a 4990 and would like to have negatives that are applicable for other scanners in the future. So I am guessing that for example a high contrast scene would benefit from a N-1 development to draw down the highlights in a range manageable. Please note that I am new to the arena and maybe way off base. So given this, how do I determine what development time will achieve this. How do you define what N-1 is in terms of scanning and that being a generic scanner?

Thanks again for your input. Hopefully I will get to the point where I can contribute at some point.

-alan

steve simmons
6-Dec-2006, 19:12
I develop my film the same way for scanning and traditional printing. That way I am covered. I use the test Kirk referred to above. The goal for both uses is to make the best use of the tones availabable with that film and developer.

steve simmons
www.viewcamera.com

Kirk Gittings
6-Dec-2006, 19:21
As Steve said.....Since I print both digitally and traditionally, I want my negatives to service both possible mediums. Generally what services for a silver print will scan well too. I.E. the above suggestion works well for both.

next question....are you using Silverfast with the 4990? i am working on a method that looks promising for determing proper dev via a scanner, but it requires SF. It looks promising.

Leonard Evens
7-Dec-2006, 08:37
Maybe I should clarify my goal, which is to scan the negatives. So I guess fundamentally I want to tune my development times for the best scanning potential and not a specific scanner per se as I have a 4990 and would like to have negatives that are applicable for other scanners in the future. So I am guessing that for example a high contrast scene would benefit from a N-1 development to draw down the highlights in a range manageable. Please note that I am new to the arena and maybe way off base. So given this, how do I determine what development time will achieve this. How do you define what N-1 is in terms of scanning and that being a generic scanner?

Thanks again for your input. Hopefully I will get to the point where I can contribute at some point.

-alan

I scan using an Epson scanner. For b/w negative film, you don't really have to worry too much about which scanner you use since the maximum density any modern scanner can handle is high enough to handle what you are likely to encounter with a b/w negative developed in some reasonable way.

But, as I indicated before, you want to measure the actual desnities with your scanner/scanning software combination rather than the RGB values. As I said, Vuescan can do that, and other software may also be able to do it. I like Vuescan a lot and would recommend it anyway. (You do have to enable that capability; it is not there by default. but then using the Ctrl key will show the densities.)

So let's assume you have figured out how to measure desnities with your scanning software. You first task is to determine what you actual film speed is using your equipment and with the film devloped in what you expect your normal way to be, including your development method, developer and time-temperature combination. There are a variety of ways to do that. If you have a spot meter, take a picture of scene with a variety of light intensities ranging over a fiarly wide range, and make sure you measure the EV values carefully. That may require getting fairly close so that the field measured by the meter is uniform in each case. Take the picture so some of the regions should be dark enough not to show any detail. Develop the film normally, scan it, and read the densities, starting with the density of the clear film base. That way you can determine the EV at which you begin to see somethig above base. That should be your Level 0 (or perhaps I).

if you scan, it is not quite as important to worry about N+ or N- developement. As long as you make sure there is sufficient detail in the shadows, the film latitidue and the dynamic range of the scanner should be able to handle you maximum density. But, if you find by experience that such is not the case, you can adjust development approriately. The point is that since you still have considerable freedom to adjust the range in the scan by choice of white and black points and "gamma" (in Vuescan called birghtness), the degree of over or under-develpment won't be critical. So the kind of testing needed for conventional processing using the Zone System is not necessary.

If you want to use densities for more precise control, photograph a scene containing a variety of EV values which would take you from Zone I to Zone VIII. Again make sure you read these values carefully using uniform fields to do so. Then measure the desnity values of these different fields in the developed, scanned negative. Adjust your development until these values are what is typically recommnded. But remember those recommendations are usually for some specific method such as using a condenser enlarger and printing on normal paper. You may find that other values work beter for you with your method.

In all of this, the RGB values in the final scan are what will determine what happens in the print. No matter what the density values are, you can adjust the RGB values quite a lot when scanning and then afterwards in photoediting software such as Photoshop, useing the curves tool. (I use the Gimp under Linux, myself.) So you need to do some testing to determine which RGB values correspond for your equipment to which Zones. Here is a rough outline which works for me:
O, 0; I, 25; II, 50; III, 75; IV, 100; V, 130; VI, 165; VII, 195; VIII, 225;, IX, 245; X, 255.

steve simmons
7-Dec-2006, 09:03
Leonard,

your long answer above just too typically promotes the attitude that you can fix/adjust the image in the computer so don't worry about being to careful with the negative densities. It is the same as not really caring about the densities in your neg and then taking more time in the darlroom to make a mediocre print when a little more time and care would get you a better negative and you would have a much easier time making a much better print

I suggest taking the time to test for your personal EI and developing time as suggested in earlier posts. Then you will knowyou are getting the best negative for scanning or traditional printing and that you are getting a good set of tones in the film to work with regardless of the 'output'.


steve simmons

Ted Harris
7-Dec-2006, 09:48
I spend a lot of time working with all sorts of negatives and chromes digitally; scanning them and working with them in PS CS2 and LightZone. While I agree that you can correct a lot, not all but a lot of exposure/development errors with software it is neither the best nor the most efficient way to accomplish this. One of the very first thigns we tell participants in our scanning workshops is that scanning and then editing a properly exposed negative (or working with it in a traditional darkroom for that matter) is rather simple and straightforward, little or no heroic adjustment is required and, in many cases, the auto adjustment in Silverfast works fine (we don't recommend using the auto settings in any software but want to point out that they are set for the 'normal' negtive or chrome). Nor do we recommend exposing/developing differently for scanning than for traditional post exposure processing. A well exposed negative is just that.

The next step, assuming you don't have a perfectly exposed negative/ideal lighting conditions/the best color rendition, etc. is to determine where in the process and how you want to make th ecorrections. As for the where, you must set your white and black points correctly before you scan or you will lose information that cannot be recaptured once you have scanned (e.g. shadow detail that isn't in the scan) and if it can't be recaptured it can't be adjusted in PS or other editing software. Next, color masking, gross color shifts, etc. can be handled in either the scan or later but hanbdling them with Silverfast Ai is much easier than doing it with Photoshop. The same may be true with other scanning software too (it is true with the software available for the high end scanners such as Color Genius for the Screens). All this is by way of saying that while the availability of computer software may be slightly more forgiving than traditional darkroom processes it is not a good reason to be sloppy in your exposure and development ... not to say that there aren't lots of situations where you may not be able to get the perfect exposure/development combination.

One quibble with Leonard, while I know he and many others get good results with VueScan, we have found that it is less flexible and less adept at adjusting film than Silverfast Ai. When used in an 'auto' mode VueScan, OTOH, works quite well. YMMV but we just don't find it the software of choice for scanning. It is a great tool if your budget is limited but probably no better than what comes with your scanner out of the box if we are talking a current generation scanner.

Finally, any of the scanning software packages discussed will allow you to do far more manipulation than is usually required or wise. Less is usually better.

Leonard Evens
7-Dec-2006, 20:00
Leonard,

your long answer above just too typically promotes the attitude that you can fix/adjust the image in the computer so don't worry about being to careful with the negative densities. It is the same as not really caring about the densities in your neg and then taking more time in the darlroom to make a mediocre print when a little more time and care would get you a better negative and you would have a much easier time making a much better print

I suggest taking the time to test for your personal EI and developing time as suggested in earlier posts. Then you will knowyou are getting the best negative for scanning or traditional printing and that you are getting a good set of tones in the film to work with regardless of the 'output'.


steve simmons

Perhaps I didn't express myself clearly. I certainly endorse determining the exposure index or speed by testing. Indeed, if you look at what I said, I suggested a method for doing that given a choice of development. Needless to say, one might also need to modify the exposure index for different choices of development, but, for me that would be a refinement to be added later.

The original Zone System assumed one was aiming at printing with a specific enlarger on normal paper. Hence, in addition to being sure there was adequate detail in the shadows, one needed to make sure the density at other zones was appropriate for producing the right shade of gray for those zones in the print. If, say, one wanted what fell on Zone V to end up at Zone VII, one might need to choose N++ development. If you got it wrong, you would not produce the desired result in your print. But even with conventional printing, you could choose instead to develop normally and use a higher grade of paper for the same result. (Or you could do it with variable contrast paper and filters.) When you scan, you have so much control of the distribution of values, that it seems to me you needn't worry quite as much about determining just the right development to get a specific density in the negative. You might still want to overdevelop some, but the exact amount will not be critical. In effect, it is as if you had a continuous choice of paper grades and need not restrict yourself to a small number of specific contrasts.

Let me emphasize again that it is the transition from densities to RGB values which is the tricky part when scanning.

Leonard Evens
7-Dec-2006, 20:19
One question.

Do other scanning programs, e.g., Silverfast Ai, let you measure the raw density values directly? The last time I looked, the default Epson scanning program did not have that capability.

Ron Marshall
7-Dec-2006, 21:16
Silverfast-Ai just gives values from 0 to 255.

Leonard Evens
8-Dec-2006, 07:54
Ron,

Does Silverfast Ai give you a "raw" option which just gives you values without attempting to adjust white and black points, gamma, or type of film? If so, those values could be used as substitutes for densities, but on a different scale, of course. That seems to be what Vuescan's "None" option for color balance does except that you can still vary the gamma (labelled 'brightness') and it may also depend on the film type. Density readings give you the actual raw densities above base.

Personally, I find that without knowing precisely what the scanning software is doing, I can't determine just what the RGB values mean and how they relate to densities. And the only way to be sure about such matters is to look at the code, which no vendor (except open source such as Sane) is going to let you do.

Without being able to measure the raw densities as seen by the scanner, I find it hard to relate any of the usual methods for working with the Zone System to what you get in a scan.

It might also be noted that a similar problem exists with conventional processing, but it is not as serious. Films and papers, as well as enlargers, will differ in how they respond to light, i.e., they will have different characteristic curves. That means the net result of exposing the negative, developing it, and then printing the result, won't yield a linear relation between EVs in the scene and reflection densities in the print. That means that any standardization you do will generally only work for specific choices of film, development, enlarger and paper. If you change any of those, you will generally have to recalibrate. Changing from enlarging/printing to scanning (and subsequently printing) is a much larger change, so, for me at least, it makes sense to do one's calibration from scratch for your choice of film/development and your scanner. If your scanner doesn't show you raw values, it seems to me that is going to be very difficult, since you certainly don't want the same EVs in different scenes to result in different RGB values in the scan when you are trying to calibrate.

Ron Marshall
8-Dec-2006, 08:22
Leonard, it appears to be an uncorrected value, but as you said it is unceartain if something is going on in the software. That is why I judge densities with my spotmeter on a light-table. This is not optimal, and only to get me in the ballpark, but at least I know where the numbers come from.