MIMF

I've done a few, but all of them pretty subjective. On my main guitar, which has a split saddle, I ended up with a piece of agate on the wound strings and bone on the unwound strings. Based solely on subjective A/B testing. I had also included brass, aluminum, and tusq in the tested materials.

Anyone here experiment with different materials, or know of more objective tests?

That's on my 'to do' list, but it's not going to be easy. I ran a fairly simple experiment some years ago, using bone and HDPE for saddles on a steel string and a Classical. They certainly did seem to sound different subjectively, but I was unable to measure any changes.

A/B tests of this are really hard to do, for the simple reason that we don't really remember sounds all that well. I've been told that if there's more than about fifteen seconds between two sounds it's hard to make an accurate comparison, and if there's any sound in between it's much harder still. Since you can't exactly swap out saddles silently in ten seconds it's difficult to be sure of what you're hearing. It would really help if we could make 'identical' guitars, and that's the project I'm working on now. With two guitars that sound the same you could just set them up with different saddles and run an A/B very quickly. You would, of course, have to reverse the saddles and do the test over another time to be sure that there was no confounding variable, but that's all SOP in this sort of thing.

There was this thread at the other lutherie forum recently. Some criticism of the methodology

http://luthiersforum.com/forum/viewtopi ... 01&t=48334

Maybe two half saddles, and two half sets of strings?
I've only used Corian and Aluminum. The Aluminum is a little brighter, but I'm pretty sure I could not tell the difference blind folded.