I can't think of a time I ever disagreed with Matt and this is no exception. Exposure determines your shadow detail moreso than pushing your film. You can't get something from nothing. If you didn't expose for shadow detail, you're not going to get it by pushing your film -you'll get it by adjusting your exposure. Pushing your developing time simply adds contrast.
Originally Posted by MattKing
Now, I know this is a forum that promotes a vast array of knowledge from lots of people, but I'm always a little baffled at the lack of initiative in doing tests to get answers. Throwing out something like "what is the best developer to push Tri-x to retain shadow detail?" is going to bring an onslaught of subjective answers that are often untested against all the necessary variables. Rarely do I see anyone offer a well tested answer with examples like "This is a negative with tri-x in x-tol and then this with tri-x in hc-110 and then this with rodinal....etc. etc." to actually show the differences in the same shot with the same lighting conditions.
In short, your answer won't be on this thread. Spend a few bucks, get those developers you want to test, grab a roll of 35mm, a tripod, a camera, a scene and shoot a whole roll on the exact same settings in the same lighting conditions. Cut up the strip into as many sections as you want to test. Spend an hour or two developing and analyzing your results and you'll have YOUR answer, not someone elses. I'm not trying to be an a-hole and I'm sure if someone looked through my posts I have probably asked at one time or another, "what is the best........" so that I could have a quick and easy answer. But personal testing is the way to go. It takes more time, but it pays off in the end.
Now, all that being said, I've used x-tol, d-76, rodinal, pyrocat-hd and hc-110 extensively with Tri-x over the last 10 years and I've settled on hc-110 even though 4 years ago I swore it off. Often I push Tri-x to 800 and 1200 with great results, but I compensate and visualize my exposures and run tests based on my own conditions and workflow. I can't say it will work for you too. You'll have to test and see what works for you.