View Single Post
  #17   Report Post  
Posted to rec.audio.pro
Scott Dorsey Scott Dorsey is offline
external usenet poster
 
Posts: 16,853
Default interested in audio fo video papers.

alex wrote:
In '93 i worked on a soundtrack for a short 35mm movie. i remember the
man at the film sound processing plant (international recording in
Rome), analyzing the level of my dat, correct the gain of the analogue
film magnetic recorder to reflect the very strict allowance of the
optical device used to impress the sound on the film.


Right. He's riding the gain on the fly, to keep perceived and recorded
levels correct. That's an important part of what audio engineering is
all about.

The man told me that there was a relationship with the tone, and other
than syncing purpuse, the 35mm magnetic film was used to "normalize" the
sound to avoid peaks dangerous for the optical device. They still used
an old western electric bulb which was out of production by decades.
The level of the tone on the magnetic film had an "exact" relationship
with the input level of the optical device. The "optical" operator just
played the tone and adjusted the level consequentially.


Well, here's the thing. The light valve, much like digital systems, cares
about peak levels rather than average levels, BUT the problem is that the
level that it fully opens or closes at changes with frequency somewhat.

So, you record the tone with a given average level, and since it's a known
waveform, you ALSO know the peak level for it.

Okay, stuff delivered to a film festival probably should be AC-3 encoded
and you probably SHOULD follow the Dolby recommendations because many other
people do. If you do it according to the Dolby recommendations, dialogue
will read 85 dBA in the center of the hall.

Unfortunately, lots of people are sloppy and don't, and the only solution
is for the projectionist to ride the gains by hand. Which is, after all,
his job.


No, most of them are documentary films, with stereo sound or, sometimes,
dolby prologic. The use of dolby E is very rare and we are not equipped
to decode it.


Dolby E basically exists to solve a lot of these problems. Part of the
issue is that many people are working with measurements designed in the
analogue world, but are living in the digital world. Part of the issue
is that some meters are designed to read peak levels or modulation levels,
while other meters (like VU and the EBU recommendation mentioned earlier)
are intended to estimate perceived loudness, and people use the wrong meter
for the job.

And so of course the end solution turns into having a man adjust a knob.

thanks for the advice, scott.
So what i have to do is valuate the level by hear with the help of a
standard VU meter or K20 (or 14) system where the zero reading will
equal to 83 dB in the control room.


Right, but THEN how do you match that to a given modulation level on your
tape or to a given dBFS value on the digital file?

In the case i use the k20 system, is ok to mix the dialogues around 0dB
on that scale and using the extra headroom for louder sound effects?
Doing this and putting a -20 tone (reading 0 on that scale), will tell
the operator where is the expected dialogs level. So he can manage to
set the theatre volume to obtain 83dB with the tone, and the dialogs
will follow. Correct? and if K14 is used, is better to set the tone to -14?


Yes, BUT you soon run into problems with running out of headroom, because
you can't go above 0dBFS at all. If your sound effects have spiky waveforms
you may have to limit them severely to get them loud enough.

Appear that, as opposite on what happened with 35mm film sound, the tone
will tell us something about the dialogs levels, instead being related
to the max VU program peak level? Is this correct?


Right, because everybody knows where the peak level is... it's a universal
scale that doesn't drift in the digital world. So there is no need for
another indication of maximum level.

BUT..... in the analogue world there is no abrupt peak saturation... even
with that light valve there's sort of a nonlinear range right before it
opens of closes completely. With analogue tape or magfilm you may have
a range of as much as 10db where the system sounds funny but hasn't clipped
completely yet. In the digital world, it is either fine or it is not,
there is no partway.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."