Hi...First of all my apologies for leaving out the Snell/Kahuna booth from my previous explanation...they are well worth a look, and I should have suggested that Kurt see the "Big 4", not just 3.

Hosko, in answer to your question about Scenes - they can be used in a variety of ways.  The most basic way to look at the panel & equate it to what we all know - so you can think of the Stages on a DYVI/SVS as M/E's on a traditional switcher.  They are points of control for the Scenes & Layers that reside within the Stages.  So if you have one Stage, the Scenes within that Stage change the effect and layout of the Stage, much the same way as we now use E-MEMs.

A Scene can thus be thought of as a compliation of Layers (Keyers on a traditional switcher).  One Scene can activate one or more Layers, and tell those layers how to perform (a Chroma Key in one Scene vs a Linear Key in a different Scene for example).  Again, in this way Scenes are not really much different than E-Mems or Snapshots on traditional switchers.

But a Scene can also be an input to another scene.  For example, Scene 1 may be a Chroma Key of an anchor over a background with a DVE box over-the-shoulder.  However, the input to Scene 1 may be Scene A (a sub-scene), where Cam 1 is a green Chroma Key with color correction added to it AND it has a specific CG (or Still Store) keyed over it (maybe multiple graphics that are associated to the camera).  Scene B may be a different Chroma Key with a different CG (or Still Store) with different Color Correction.  Both Scene A & Scene B appear to the user as single inputs to the "Layer" of Scene 1 (think of them as upstream pre-keyers) - and Scene C can be one set of keyed inputs (remember that a full-screen keyer acts like a background) that feed the DVE OTS box, while Scene D, E, F, etc. can be a other sets.  You can "pre-composite" as many scenes as you need & use these as composited inputs into other scenes or stages - and because the actual Chroma Keys are done "upstream", the final keys can be mixed & matched - you can seemlesly dissolve from a Green Chroma Key on Cam 1 to a Blue Chroma Key on Cam 2 because they are both treated as Linear Keys by the time they get to the part of Scene 1 that is doing the final compositing.  Part of the beauty of this logic is that if you only need 3 possible (composited) inputs into your Chroma Key, you only map 3 buttons as CK inputs (you don't need to have every source mapped like on traditional switchers).  One of the demos that DYVI did was to treat different scenes as different blocks & additionally as different stories of a newscast...each scene remapped the whole switcher & included sub-scenes (such as all of the effects in each block of news), so that the panel emulated "station automation" software & thus let the Producer jump around as much as desired within the rundown.

So if you are only mapping what you need, you don't need the gihugous real estate of a traditional switcher (hence the smaller footprint).  You can control what buttons are needed for each scene, and so one of the things that scenes can do is either completly or partially remap the panel depending on what scene is called up.  For example - let's say you were doing a traditional football "coaches challenge" with a 3-box effect (1 for Coach A, 1 for Coach B, and 1 for the Referee); rather than use up 3 keyer buses or 3 DVE buses, you would have 3 buttons labelled "Coach A, Coach B, & Referee"...Pressing the Coach A button would light up only the cameras that you preprogrammed for use with that coach (probably only 3 choices [if that many] in real life...the StediCam isn't going to be involved in a shot for a challenge, so why waste real estate?).  You'd map the appropriate camera(s) for Coach B & map the possible Cams for the Referee.  You could set it up so that you 1st have to press the Coach A scene button & then make your choices before pressing the Coach B, etc. scene buttons (an example of traditional delegated assignments), or you could map it out so that when you press the "Coaches Challenge" Scene, you'd see all 3 sets of inputs at one time (more efficient).  The point is that you can use scenes to map out only what you need when you need it; and each button is dynamic so it can change roles as YOU see fit (not the manufacturer).  If you set up duplicate scenes for this effect, each scene can have its own source programming (like using multiple E-MEMs so that you're not having to call up the effect & THEN change the sources).

So the simple answer is that Stages = M/E's, Scenes = E-MEM's, and Layers = Keyers - if you wanted to treat the DYVI as a traditional switcher, that's how you'd set it up.  But if you go beyond the simple setup, it opens up a whole new world of resource management and operational possibilities.

Sorry to be so wordy - hope this all answers your question.

I would suggest just stopping by the manufacturer's booth & ask a sales rep.