QuickCam use report

Alan Larson (EST)
Mon, 28 Nov 1994 01:09:33 -0500

This should have been sent out sooner, but I spent the holiday
playing with other things. I was still at work Wednesday when
the mail came announcing the new software, though. Since the
cameras had arrived a day or two before, I could not resist, and
stayed an extra hour or two playing with them.

The Quickcam does work with CU-SeeMe as those with the cameras
by now know. I wanted to compare it with a good quality Sony
camera and a Video spigot.

The first observation was that the QuickCam gave me sound input
capabilities. This was quite welcome, since the Mac IIcx did not
have any other form of sound input. Our 56K line to the internet
may continue to make sound impractical with the outside world,

The second observation came very quickly, and may have even been
the first, as I noticed it with the Connectix software. The input
from the QuickCam is somewhat slower than I had been used to with
the Video Spigot. Now, given the difference in technology, this
is not really a suprise, but it deserves some mention.

With CuSeeme, the extra load of processing the data from the
QuickCam on the serial port made the Mac quite noticably slower.
This was initially noticed when I tried playing with the sound
controls, but also was painfully obvious when dealing with other

There had been reports that the fixed light settings of the
QuickCam could be beneficial in keeping most of the light levels
constant when part of a scene changes brightness. This is a problem
with auto-exposure on camcorders as it causes many more blocks of the
image to need to be updated, while they really didn't change -- the
iris of the camera changed.

When looking at the image fromt he Connectix software, it became
obvious that this was going to be a potential problem with the
QuickCam. What was happening was that the image brightness was
changing from one scan to the next. This was apparently because my
office is lit by flourescent light, which was flickering with the power
line. (I didn't realize that it flickered that much.) The camera
samples came at different times, and with the exposure shortened by the
Connectix software, it was getting peaks or valleys of the brightness.

(My theory is that camcorders do not see this because their exposure
rate is close to the power line so the auto-iris can track the slow beat
between the two. Also, the 1/60 exposure commonly used will include two
brightness peaks, so variation is limited.)

It seems that some caution should be recommend with the possibility
of the QuickCam using unexpected amounts of bandwidth.

Another observation was that the (much more expensive) Sony camera
was giving better pictures. Not too suprising, given the difference
in price, and the fact that I had a *lot* more experience using the
Sony. Hopefully, experience with the QuickCam will improve the results.