Some years ago I ran a (somewhat) systematic test series of Tri-X with different developers. D-76, Rodinal and XTOL.
I started using Tri-X and D-76 in 1963. Loved the tonality, but quite often wished for more sharpness. The perceived sharpness of Rodinal but without the sandpaper-like grains of Rodinal, even when diluted.
Seems that XTOL gives smaller grain clusters than D-76 and thus the extra sharpness while preserving the tonality. Ended up using 1+1 dilution.
Had to standardize tank agitation to get consistent results.
Once you find a favorite agitation pattern you should stick to it.
Ended up with the "textbook standard" of 30 seconds slow agitation from the start and then 10 seconds of slow turns every minute.
The other objective was to test the emulsion speed of Tri-X using a densitometer.
If you only observe mid gray, the 400 ISO box speed may seem correct.
Found though (like may others have before me) that if you want rich detail in the lower grey tones, exposing for 160 ISO gave the best results. This gives quite heavy highlights and you may wish to reduce the development time to make the film easier to print.
Textbook recommendation is 20%. To me that gave prints that were a little "dull and flat". Ended up with 320 ISO and a 10% cut as a compromise between shadow detail and preserving the nice original Tri-X tonality.
You may push the Tri-X to 800 or 1600, but it seems always to be at the expense of shadow detail. Lower and deep dark grays are not there anymore, just the blank film base with no information in it. That may well be an artistic choice, but one should be aware of the trade-offs.
As so often in life this information was readily available beforehand in textbooks, but one often needs to try it out for one self to believe the author 🙂 🙂