Although this thread is a bit old I discovered it while working on a somewhat related project, and thought I'd offer a few observations based upon my nearly fifty years of work with both photographic filters and theatrical gels.
One note - I posted the original version of these comments earlier this year but discovered an error in my methodology. Late spring to early fall is my busy season at work so it has taken this long to get my comments and illos reworked.
First, there is no such beast as a "Wratten 25A" filter. I believe that this error arises from a confusion between the older and newer designations for Wratten filters.
The standard reference work concerning Wratten filters has, for nearly a century, been "Wratten Light Filters." The first two editions of "Wratten Light Filters" were published in Croydon, England, by Wratten & Wainwright, Ltd. in 1900 and 1913. I do not have access to those but do have access to the third and later editions. Frederick Wratten sold his company to Eastman Kodak and they published the third edition at Rochester NY in 1916. Over the decades the title was changed successively to "Kodak Wratten Filters," "Handbook of Kodak Photographic Filters," and "Kodak Photographic Filters Handbook." I own the 1990 edition of "Handbook of Kodak Photographic Filters" and it is my source for filter data.
From the 1916 edition until WWII Wratten filters had both a number and a name. Filter No. 25 was named "A (Red Tricolor)." The portion of the name in parenthesis indicated the intended use of the No. 25 filter in preparation of tricolor separations (the corresponding green and blue filters were named "B" and "C" and were numbers 58 and 49). Beginning after WWII Kodak abandoned the letter names and used only filter numbers. The use of letter names continues among some manufacturers of photographic filters (notably Hoya), and Kodak for many years after WWII stamped Wratten envelopes with both the number and the letter name, like this:
No. 25 A
The confusion arises when someone reads such an envelope as meaning "25A" rather than "25 or A." The inaccurate designation of this filter as "25A" is found in many places online.
Now to the gist of this thread - finding the Rosco theatrical gel which most closely mimics the spectral characteristics of any particular Wratten filter. I own a Pentax DSLR with a number of Tiffen filters (which use Wratten numbering) to alter the contrast in black-and-white images. I embarked on the quest because changing my threaded filters to evaluate which of, say, my three red filters (23A, 25 and 29) produces the black-and-white contrast effect I want is cumbersome. I resolved to use my older Rosco swatchbook of 3" x 5" swatches (current books have 1-3/8" x 4" swatches) to build a flipbook of the swatches that best mimic my filters. That way I can try a series of gels in front of the lens to choose which filter I should mount.
I began by preparing graphs for Wratten filters mimicing those provided by Rosco for their gels. Here is an example:
I initially chose gels based upon graph comparisons, much as Mathew did upstream. Here is my Wratten 25 graph compared with the Rosco R19, R25, R26 and R324 graphs:
Test shots showed that this approach was not entirely satisfactory. I decided to attempt a more objective comparison using test shots with my Pentax.
I own and use an X-Rite ColorChecker Passport, and used it as my target:
All of the test shots were taken in my tabletop studio using a pair of 5600K LED lights with softboxes for illumination and with all exposures made in black-and-white mode (because I wanted to compare the effects on filters on colors as rendered in black-and-white). I photographed the Passport first with no filter, then through my Tiffen photographic filters, and finally through a series of candidate Rosco gel swatches held over the lens. The unfiltered image provided a standard for calibrating the grayscales of the filter and gel images since all of those had a shifted grayscale. I could then measure how each filter or gel affected the the three additive primary colors (red, green and blue) and the three subtractive primary colors (cyan, yellow and magenta). I graphed the results on polar graphs with the white target at 100% and the black target at 0%. White circles represent the six colors in the unfiltered black-and-white image, black circles those colors through a filter, and red circles those colors through a gel. Here are the polar graphs comparing the Rosco R19, R25, R26 and R324 filters with my Tiffen 25 filter (equivalent to a Wratten 25):
As noted by Chris upstream any of the first three would probably be a good substitute for the 25, and I added the R324 as a fourth candidate. Although no theatrical gel can be expected to duplicate a photographic filter I have chosen the Rosco R25, based upon my polar graphs, as my gel to represent the 25 filter. The object of my polar graphs is to find the Rosco gel whose red dots most closely match the black dots of the reference filter. In the case the Rosco R25 is the best match.
The Wratten 25 filter and the Rosco candidate gels, though, are visible filters. It seems to me that the goal of providing a false-color image to indicate the health of vegetation would be better served by shooting monochrome images first through a cyan filter such as a Wratten 44 or 44A which would maximize the blue and green channels while minimizing the red channel, and then through a near-infrared filter that would pass primarily wavelengths in the 700 to 900 nanometer range while blocking most visible light, provided the camera has usable infrared sensitivity in the NIR range. The two could then be combined into a single false-color image in post-production. Likely candidate near-infrared filters include the members of the Wratten 87, 88 and 89 series, none of which I own at this time. I hope to soon have a chance to evaluate the NIR sensitivity of my Pentax with one or more borrowed IR filters.
I have a few additional notes. The first is that Mathew had some difficulty comparing the Wratten 25 graph with the Rosco graphs because the former uses a logarithmic Y axis and the latter use a linear Y axis. The Wratten graphs that I have created for my Tiffen filters (which use Wratten numbers as Tiffen now makes the Wratten photographic filters) use the same scales as the Rosco filters. The "Handbook of Kodak Photographic Filters" includes tabulated data for each filter as well as a set of scale conversion tables that made it possible to create Wratten graphs with a linear Y axis.
The second note is that the Wratten graph that Mathew displayed was taken from the Wratten 2 series web site, and differs from the graphs published for seventy years by Kodak for Wratten filters.
I am still working on overhauling a pair of web pages detailing the Rosco gel equivalents in my filter flipbook. I have had to correct the graphs for the methodological error mentioned earlier but it has not affected most of the choices. My Rosco gel for my 58 Green filter appears to be the one most affected, so that assignment is still tentative, but the pages are reasonably close to the final version so I am posting a link to them:
Dale H. Cook, many years as 35mm SLR photographer,
now Pentax K-70 w/ Pentax-DA 18-270mm walking-
around lens or SMC Pentax-A 50mm/f2 lens