These drawings are only speculation, based upon fragments of published data and the photographs of the Kastonome piano owned by Sr. Perez.  It is clear that the basis of his piano is the American-made Autopiano which was modified by Kastner.

An old catalog by Player Piano Company (PPCo), of Wichita KS, included photos of various player types for identification purposes.  The picture of the Autopiano, and the accompanying text, made it clear that the Autopiano is a double-valve system with horizontal main valves.

Two unique features of the Autopiano are of interest.  The Autopiano roll-tracking system is a four-hole Standard Pneumatic type, except that the two tracker pneumatics are mounted vertically on the ends of the tracker logic block, instead of horizontally.  This can be seen in the middle photo, and is exactly the same as the one pictured in the PPCo guide.  The straight parts of the Autopiano lead tubes from the stack to the tracker bar area are enclosed in pieces of stiff loom tubing (something like old fashioned electronics spaghetti).  This is the purple stuff seen on the front layer of tubes in the bottom photo.  So indeed Kastner did import ordinary Autopianos to use as the foundation for the Kastonome.

Fig. 1 below shows an approximate cross section of an ordinary Autopiano, based on my knowledge of the similar player action by Standard Pneumatic Player Company.

But there are differences.  Another deck was added in the Kastner piano under the normal primary deck, presumably for the Kastonome primary pouches.  Also look at the lower right side of the lower picture, where the opening for the vacuum to the main valve box is shown.  This is about an inch or so more forward than it ought to be, based on extrapolation of the position of the hinge ends of the pneumatics.  Something was inserted or replaced between the back of the valve box and the normal shelves on which the pneumatics are glued.  This must be the 'guts' of the Kastonome, where the switching the between normal or restricted playing flow path was located.

Fig. 2 is the way I imagine it could be done.  The proportions may not be quite right, but the concept is there.  The adaptation was not done too neatly, evidenced by the crude sawing around the player lever rails.  Note the crudely made spool box deck support blocks, shorter than the original black ones, attached with new screw holes in different places than the original, unused screw holes.

Figures 3 and 4 show how the Autopiano valve was possibly modified.

Richard Vance