The NY Day Concert from Vienna is always a brilliant way to start the year, - a musical celebration and professional in every way. For a pretty detailed overview of the technical aspects involved in broadcasting this year's concert (which is fascinating and worth sharing) this observation appeared from Martin Clarke on the Institute of broadcast sound forum:
"And the best bit of the New Year so far has been the inevitably excellent
New Year Concert, broadcast live from Vienna this morning. Did anyone else
In the course of the programme, there were two dance sequences from Palaces
outside Vienna, both featuring members of the Vienna State Ballet. One – the
last, danced to the Blue Danube – was live, but the other (in the middle of
the show) was pre-recorded. At least I assume it was, from the blue sky and
sunshine, the flowers in bloom, and the ability of the dance troupe to get
from one end of the gardens to the other in two bars of 3-4. So the due
process of technology comes into play, and we have a sequence that has been
shot to a guide track, with playback, foldback, click track, edit, dub – all
that stuff. And then the band plays along live.
But a vital part of the process is the need to give timing cues from the
pre-recorded item to the conductor and/or the band so that they can play
along and keep time and tempo. How many different ways are there? Foldback
bins, video monitor of the insert, in-ear monitor, cans for the ‘drummer’ –
any or all would do the job. But try has hard as I liked, I couldn’t spot
the magic ingredient. No-one seemed to be wearing cans, I didn’t see a
monitor (either sound or vision) in shot, and Zubin Mehta – the conductor –
wasn’t wearing an appliance, as far as I could tell.
The timing of the dance and the music seemed too good for it all to be
running wild, so what did they do? Is anyone in a position to tell, or
should I just carry on believing in the magic of television?"
And this is the reply from Florian Camerer of the Austrian Broadcasting Corporation (ORF) :
I have a broad smile on my face when reading Martin´s mail about the
ballet sequence during the New Year´s Concert ..... !
As I am quite involved in this production - especially from now on
as I do the audio edit for the DVD - I am happy to lift any "secret"
how synchronisation is done.
Martin, you spotted correctly that the second ballet to the Blue Danube Waltz was live - danced in Schoenbrunn - and sound was relayed to the site.
Any latency is not an issue, as the dancing sync is forgiving.
And yes, the other ballett ("Dynamiden") was pre-recorded in summer
and then played back druing the picture shoot at Schlosshof in September.
But ... there is NO sync signal whatsoever for the conductor or anyone else
at the actual concert.
This would be absolutely impossible and unacceptable for any conductor - these waltzes have so many accelerandi, so many rubato sequences, that this doesn´t work at all.
So - and if you watched the transition into the tape carefully you could spot it easily as there was a mistake on the picture editing side -
we have to take precautions to segue in and out of the tape seemlessly.
Ideally, the picture guys show a total shot of the orchestra or even a neutral picture of the ceiling or some flowers etc.
Right before the start of the music, the sound supervisor fades the live sound to a pre-recorded 5-channel atmosphere of an awaiting audience
(stitched together by me from several short bits and pieces from the last years concerts) so any earlier start of the music live in the hall
is not heard. This is where the trouble can start - if the director shows
a close up of the conductor (like today) you very evidently recognize that
something is not right: Mehta was already conducting and you heard no sound - they switched to the total shot too late because things went very quickly
and Mehta gave them a hard time today ...
Ideally the VTR guy has a monitor with a close up of the conductor and starts the tape in sync with the live thing - but we have a couple of seconds of "flesh" before the music starts (technical and esthetical reasons), so the start almost never is in sync. Therefore the prerequisite of a total shot or a neutral picture where it matters less.
Then both things go along in their individual tempo - and inevitably the live music is either shorter or longer; and this year it was a long tape,
more than 9 minutes, so things can be quite far apart ...
The measure we take is that I cut an applause (2 minutes) from last year at the end of the music on the tape which is what you hear when the music finishes.
And now there are 2 possibilities: when the live music was conducted quicker, the audience applaudes already and the fade from tape-applause
to real applause can be done immediately;
and if the tempo in the hall is slower (which it was today) then the engineer has to wait and play the tape applause until the live audience starts to clapp - and then he can again do the crossfade. In the second case, the picture guys have to cooperate - they have to stay on the tape or fade to a neutral picture in the hall (they did that today - there was a tilt from the ceiling to the orchestra) so that we don´t get in trouble.
Yesterday everything went perfect at the general rehearsal - but today
the transition into the tape was not good ... Back to the hall was ok.
You may ask about any audible differences between the different applauses.
Yes, there are some - but this year it was very similar, so almost not noticable.