Thank you, Rainer!
My Granados second version has a timing of 2:44 (only 7 seconds faster than the first version). The second version plays for me when I click on the link. But I noticed that my first version (timing 2:51) is still on the "new recordings" list and it shouldn't be. I may be stuck with it on the server because perhaps I uploaded the files with two separate names...I can't remember. Or maybe the list just needs to refresh over night. Can you please try clearing your cache and then click on the Granados link here in the thread again and see if the file that is 2:44 plays for you now?
I was getting the earlier 2:51 (4128810 byte) file, but clearing my browser cache fixed it for me and I got the newer 2:44 (3950760 byte) one. This surprises me a bit because I thought these files would be too big to cache.
The newer one flows a bit more, but I actually prefer the dreamy quality of the earlier one. This is interpretive, I guess, but I still don't like those inter-section breaks you put in, they seem to artificially disconnect the sections from each other in a way I just can't relate to, and almost make it sound as though the sections had been separately recorded and then spliced with scant regard for continuity.
The first of these hiatuses, before the triplets 3 bars before the Gallardo
section, is all the more noticeable because there is no rit
in the previous bar, which makes the gap appear out of nowhere.
The second one is at the start of the Gallardo
itself. But this one is OK because here you do rit
in the bar before. Although this rit
isn't in the score, I think it works well.
The third is at the beginning of the 4-bar "meno" section that links Gallardo
to reprise. It seems to me that the last pair of LH sixteenths D-E should lead without interruption to the F in the "meno" bar. Why else continue this timed trill right to the end of the bar line?
At the end of the "meno" section, you are cutting short the dotted half-note chord, jumping into the reprise a bit early. But this is only on the re-recording, the earlier one was fine in this respect.
Finally, in the penultimate bar of the piece, and bearing in mind that the rall
marked in the previous bar should still be in progress, you are playing the 8th-note B-A too quickly (the A comes too soon), almost as quickly as a rall
ed version of the 16th-note B-A an octave higher in the previous beat.