So an 89 filter is the best for SFX, if you do not have the specific SFX filter?
SFX has extended spectral sensitivity (i.e. "extended" compared to your typical panchromatic b/w film, which stops being sensitive in the mid 600 nM area) out to around 720 nM, if my memory is correct. So, theoretically, the use of any "ideal" filter with a cutoff of 720 nM or higher will result in no exposure.
However, photography filters are not "ideal." Even with filters with cutoffs at or above 720 nM, there is some passage of shorter wavelengths. So, even the heavier filters than 89a/b can be used with SFX, and exposure will result. Bring your tripod and experiment to see just how much, and what effects you get. From my personal experience, I can say that Hoya R72 (720 nM), works with SFX. You just have to let the shutter hang open long enough.
With any infrared/extended red film, the filter to be used should be decided upon based on much experimentation and experience in real-world scenarios. These films are always exposed by
all the kinds of energy to which they are sensitive, with or without filters. The selection of a filter just determines how much of the visible energy you remove from the equation.
If you only want to make use of the extended range of SFX, a filter that cuts off at about 665 nM should do the trick. That will theoretically trim out all the wavelengths to which most pan films are sensitive, and only use the extended red sensitivity of the SFX.
If you only want to expose your film using only the deepest of visible reds, a heavier filter would be a better choice, like one with a 695 nM or 715 nM cutoff (and, as mentioned, Hoya R72 will work).
The more of the exposing wavelengths you block from being able to expose the film, the longer your exposures will need to be. This is pretty simple to visualize. Imagine a row of 20 guys pissing in the communal "trough-style" urinals at Dodger Stadium. This is a metaphor for shooting your IR film unfiltered. How quickly does the level of piss in the bottom of the trough reach a certain marked point (the point at which your film has received enough exposure, metaphorically)? Now, imagine that 3/4 of those guys finish, and only five remain pissing into the trough. This is metaphorical for putting on a filter that blocks three-quarters of the wavelengths that are exposing your film. How quickly does the level of piss in the trough reach that same mark now? ADDITIONALLY, assuming equal amounts of piss in each of the 20 guys' bladders, why are these five guys the last ones pissing out of the 20?
Because they piss slower than the ones who are already zipping up and not washing their hands before heading back to their seats. Therefore, the rate at which the remaining five can fill the trough to reach that mark is actually lower than the rate at which any five out of the other 15 guys who already finished pissing could do it. Metaphorically, this difference in pissing speed between the first 15 guys and the last five guys is the difference in intensity between visible light and IR/deep red light that is illuminating a scene that is being photographer. So, not only does filtering cut the percentage of the spectrum that is exposing your film, but it cuts out the parts of the spectrum that are exposing the film
the most. This is why your exposure times go up dramatically when you filter.