[78-L] Speaking About Really Early Film Sound Technology -- Read This. Whoa!

[Michael Biel]

From: Michael Shoshani <mshoshani at sbcglobal.net>

On Mon, 2010-06-21 at 21:58 -0700, Michael Biel wrote:

>> I uploaded the videotape of the demonstration of the playback machine
>> that Leah shot at the Schenectady Museum. You can hear the entire track
>> he played, not just excerpts. This is track two of the film recorded
>> April 26, 1930. Hearing extended music you can hear the extreme wow
>> problem. You can see the capstan which contacts the film only at the
>> edges, especially when she zooms in. 

> Yeah. That contraption does more harm than good, it seems. I'm glad
> they have the basic idea, but....along with other faults that have
> been discussed, the flanges of the film reels are warped, and you
> can hear them scrape the edges of the film! 

Somebody commented about that on YouTube.  Remember, he bought those
reels on ebay.  

> And his talking about how the film needs to be flipped for alternate
> tracks? ...No. Reversible mechanism. If a 16MM motion picture projector,
> the kind that once graced every school classroom, can have a "reverse"
> switch, so can this. 

There is a "capstan direction" switch in my still shots under the deck. 
But that would be "pushing" the film thru the machine, just like that
asinine Ampex 400 which had the capstan to the left of the head block.  

But also, don't you remember that the classroom projectors always lost a
loop whenever you put it in reverse?!  Especially if you did it too
fast!  

Most of the films had the tracks going in the same direction, and these
guys didn't understand the concept of a continuous loop at first.  It
turns out that the machine DID record and play with a continuous loop. 
But they also had experimented with using reels and a reverse mechanism
on the recorder.

> Handle that film as little as possible. Have as few adjustable parts as possible. 

He IS using white gloves!

> Would it be so wrong (or impossible) to mount a dozen small photocells,
> each with its own independent output, and pass the tracks over that
> assembly *once*? Take the reading from every track at once straight into
> digital, and digitally reverse the tracks that need it. Or, record two
> passes, one in each direction. 

There is a lot of cross-talk if the play cell is not adjusted just
right.  This would complicate things and lead to MORE handling of the
films!  

On a similar note, we had a presentation at ARSC warning us that you
CANNOT record an analogue track in reverse onto a digital medium and
expect to be able to digitally reverse it.  You wouldn't believe what
the waveforms looked like when he shows them to us.  Reversed square
waves had loop-de-loops!!!  Analog reversal is generally OK -- there
were some occasional problems that tape duplicators had discussed in the
1960s-- but digital is absolutely OUT.

Mike Biel  mbiel at mbiel.com