Background
A single Waveform can or could contain as many as 256 samples. By definition, a Waveform, in Yamaha-speak, is a collection of samples. A maximum of two samples can share a KEYBANK. Samples can be mono (1 channel) or stereo (2 channels). That is why a maximum of two samples can share a Keybank. A Keybank is defined by a Note Range and Velocity Range. Once a stereo sample occupies a Keybank that Keybank is filled.

The Theoretical Maximums
So you could, in theory, assign 128 stereo samples to the same Note. Say you assigned them all to C3. Only one of the stereo samples could sound at a time. To make it work you'd need to assign each it's own Velocity domain. That is how they all could exist (nicely) on one key. Each velocity unit would trigger a different stereo sample.

You could (again in theory) map 128 stereo samples across the keyboard, C-2 through G8. This is easy to understand - each key would sound a different stereo sample as you triggered it's particular assigned note.

Neither of the above conditions would ever be done. The mapping is always less that the theoretical maximum, because the goal is make something playable.

Secret Code
I'm not completely sure why the naming abbreviations are kept such a mystery - it's like a "secret code" only the members of the programming cult seems to know... Perhaps they are suggesting that you select the Waveforms by sound, not by the name. (And at the end of the day, that IS probably the best way to begin programming).

SW signifies a velocity switch or swap... So that as you increase velocity you are transitioning through different samples. For example, the Waveform 0001 is "CF3 Stretch Sw St".., I'm quite sure that translates: It's a sample of a "stretched tuned" Yamaha CFIIIS Concert Grand with multiple velocity switches in stereo. They don't tell you (in this case) how many velocity switches but there are several. (Probably 3... My guess)

Velocity Switching
Velocity switches are simply sampling the piano at different velocity striking strengths to capture the distinctively different qualities of the piano at different velocities. The pianoforte is a percussion instrument - and like all percussion instruments the behavior changes dramatically as you strike it harder. It is easier to explain this behavior using a crash cymbal... Where the change is extreme. If you set up a microphone and hit a crash cymbal as hard as you can, there is an explosion of harmonic content.

Now say you make a sample out of this recording - it doesn't take much to imagine assigning velocity sensitivity to this sample so that when you play the key soft, the volume output is soft, and when you hit the key hard the volume is loud.

But that wouldn't sound right. The soft strike will not sound right to your ear. It will sound like a crash cymbal that was struck very HARD, but played back at a low volume. Your ear/brain will not be fooled. Because your ear/brain expects a different harmonic array when the cymbal is actually played softly, so this will cause it to sound unnaturally. It's like if you yelled into the mic, playing it back softly, you still hear it / understand it as a person yelling... You'd know its not behaving any different... It's still a yell... That you just hear less well.

Not to put the competition down, but for many years pianos where sampled with only the Hard Strike sample (which is an explosion of harmonic content) and the theory was to use a velocity sensitive filter to dampen the harmonics so it sounded more natural as you played with increased velocity. But just like a hard struck cymbal played soft, or someone shouting, you cannot hide the fact that it was struck hard or the fact that the person was yelling. A softly struck cymbal, or a softly struck piano note is vastly different from a hard strike. And your ear/brain knows the difference. This is why velocity switching sounds more natural. There is less chaotic motion in a soft strike piano note.

Harmonic content of acoustic instruments increases the harder you hammer, pluck, strike, blow or bow them. So a "Sw" or switching Waveform, represent multiple samples stacked such that it switches from one to another as you increase velocity. On percussive family instruments this is a workable method.

The R is Right, the L is Left. The RL might have something to do with the mic placements used to create the sample. (It is clearly different from "St". There are lots of things going on in the miking of drums... You might be listening to a combination of mics in different placements. Or it simply might be the Right mic position on one (mono) assigned with the Left mic of another (mono) sample. To be true stereo the data in the left and right channel are acoustically "related". When combining two mono recordings they do not have to have this relationship. They are just a combination played together. Notice you are often given the L sample, and the R sample separately. Assigned to the same Keybank with one panned hard left, the othe hard right would result in a stereo result.

Range Examples
Something like a number range, like "Sd PowerD 1-5St", indicates there are 5 velocity switches when playing the Waveform. This Waveform is the Snare Drum assigned to the key E1 in the Power Stanard Kit 1. As you play the key E1 and increase the velocity, you will hear 5 different snare hits. St is stereo. The letter "D" here signifies there are several power snare drums in the Wave Rom and this is the fourth (alphabetically).

Try it! Play that key increasing your velocity as you repeat hitting it.

Contrast this to "Sd PowerD 5St" assigned to A#0 of this same kit ...it is clearly just the 5th or hardest strike of the previous snare I just described... No velocity switching... Just the single sample mapped to all velocities 1-127. It's a hard strike... And even at a velocity of 50 it's a hard strike.

I do not know all of the codes... Perhaps we can ask the programmers and post it here in an article... but at the end of the day, you choose by ear... What sounds best... still, we agree, knowing can be interesting,

Thanks! That helps a lot, especially on the velocity side of things. As for what you said about asking the programmers to post it in an article, I'm 110 percent in favor. Since there are many others who wish to know, having a result come up in an online search (or even just a page to refer to) would be super helpful.

We'll see what we can do. Thanks for the question.