Tchaikovsky PC1 arpeggio

[amr1980]

The second arpeggio (d dim7) in the opening cadenza is giving me trouble.  I was trying to watch Weissenberg to analyze his technique and see how he executes it.  However, after watching it many times in slow motion, I just can't seem to see what fingering he's using.  In fact, it almost looks like he doesn't play the a-flat (!). 

I'm wondering if anybody will have any observations about his fingering, and what he's doing with that a-flat.  Indeed, I hear the a-flat, but I wonder if he just catches the edge of the key, or I'm just not seeing something.  For example, at 00:13 it seems like he just doesn't play it.  It seems like the most practical fingering is to have the thumb on d, but I'm not positive that's what he's doing.  Below is the original and a slowed down version.

Thank you.

[birba]

Beats me.  It's probably just a play-back AFTER the recording, where you can screw up all you like, just as long as it looks REASONABLY correct...

[amr1980]

After rethinking it, I think the most practical fingering is the thumb on b, and to just have the extra thumb under at the end. 

I'm still perplexed by the missing a-flat. 

@birba: That sounds like something that might happen in certain circumstances, but we're talking about Weissenberg in his prime here.  I don't think he needed to do anything like that, especially in such a straght forward passage as this.  Of course, I could be wrong...

[mistermoe]

well i started with 2 on the a-flat, then 3 on the b, and then allways thumb on the d. This way the last d is played by the 5th finger. Works fine for me!

[birba]

After rethinking it, I think the most practical fingering is the thumb on b, and to just have the extra thumb under at the end. 

I'm still perplexed by the missing a-flat. 

@birba: That sounds like something that might happen in certain circumstances, but we're talking about Weissenberg in his prime here.  I don't think he needed to do anything like that, especially in such a straght forward passage as this.  Of course, I could be wrong...

That's exactly what they used to do when they first started making these video tapes in the 60's.  Nothing to be ashamed about.  And I bet you Karajan wouldn't have done it any other way.

[birba]

This IS the one with Karajan, right?

[amr1980]

So they would do extra takes - audio only - and use the existing takes of video to put them over?  That makes sense, especially with the price of film.  But that seems pretty risky since the audio and video might not line up.  Of course they can work wonders, those editors.  While your explanation makes more and more sense to me, I still have some doubt in that it's hard to imagine a pianist like Weissenberg playing an arpeggio and missing the fifth in every octave. 

Of course, at this point I could well be dwelling on minutiae, but even if he did miss those a-flats, it would imply something about his technique of playing the arpeggio - perhaps he only aims for the very edge of the black keys, and maybe the fourth finger is played very lightly. 

[birba]

Those arpeggios are really leggero and very fast - almost in one beat.  Could be you can't possibly play the video slow enough to catch all the keys-  At any rate, with the cadenzas, on videos, the audio is done first, perfectly,  and then played in playback while he "mimics".  E voilą! 

[scottmcc]

audio dubs onto video are certainly common...for instance, in the movie "Dr No," Ursula Andress had such a thick accent that they dubbed every single line of hers.  If you look closely (or even not that closely) it should be readily apparent.  Of course, if such a detail is enough to ruin your enjoyment of an otherwise perfect bit of cinema, then there are bigger issues. 

with modern video editing techniques and digital video and audio manipulation capabilities, such sloppiness is really relatively unheard of in something even semi-professionally edited.  Give me an hour with adobe premiere and I can take a verse out of a song and make the video line up so that nobody is the wiser, even on a careful listen, and I'm not even a pro. 

in general, the process is to make the audio track(s) and then to splice the video onto it.  When switching from camera to camera one simply has to make sure the timecodes line up, and magically it looks seamless.  Of course, if the tempo is slightly different between different takes, this creates complexities, but again, modern editing technology makes this easy--find a landmark, such as the beginning and end of a phrase, and then scale the video clip to match that.  hard to do on film, takes less than a minute on digital.