Typically when I do 1095 (which is close to W2) for thicknesses of 1/8" to 1/4" I'll dip the 1425 to 1450degF blade into 100degF pre-heated brine for 3 secs, pull for 4 or 5sec, then back into the brine until touch-safe (140 or less). Once cool I'll run a file over it to check that it skates off like glass (a check to make sure it hardened the steel), then I'll immediately put it on my oven at 400degF for 2hrs, air cool for 3mins, then another 2hr temper, and finally cool. Any bare metal parts should have a slight straw or light yellow look to them. Once tempered I would remove the scale and sprinkle some ground up metal on top of the blade (gathered from around my grinders) and then run a strong, rare-earth, magnet along the bottom side of the blade. This moves the metal powder on top of blade aound a bit and makes it gather a little by any stress crack that might have occured. This is an ad-hoc magnetic particle test and will show cracks pretty readily (they'll show up as lines as you move the magnet around). Cracks in the metal alter the magnetic field passing through the metal and causes the metal powder to concentrate along crack lines.
A preliminary test that I'll do before and after quenching (not after tempering) is just barely hold the edge of the handle side of the blade and lightly tap it with a file - just enough to make the metal ring like a bell. After quenching I'll repeat the same test and the metal should have the same tone of ring as well as maintain the ring for the same lenth of time as before the quench. Cracks in the blade will dull the ring and shorten the ring time. Similar to putting a crack in a bell and trying to get it to ring - it'll sound dull. The best way to describe this is that, before heat treat when you tap the blade with something else metal, it sounds like "tiiiiiiiiiiiiinnnnnnnnnngggggg". If there are any cracks after quench it'll sound like "tunk" if held in the same spot as before. I don't know if this a tried and true method of discovering cracks as it's only made on a couple of my observations. Even if it sounds dull I would still temper it and see just how large the cracks are. Sometimes the cracks can be removed by grinding away the cracked area and making a shorter blade (worked for one of my blades), but for the most part the cracks might occur in more than one place and usually prove to be fatal.
If you do have a cracked blade don't discard it as it's a valuable learning example. I cracked a couple because I ground the blade edge thinner than 0.040". I keep these mulligans mounted on the wall above my workbench to remind me not to do that again. Also keep notes on what you did and the process you used when heat treating so you can go back and isolate the variables that might have caused issues (blade thickness, edge metal too thin, different interrupted quench timings, etc). Makes it easier to not do the same thing twice as you can read back over your notes again before heat treating again.