[Tig] FSR UDC Colorimeter

James DeLuca james_deluca
Mon Jun 2 22:18:28 BST 2003


MessageFred:

You wrote:
  Does anyone have any practical experience with the FSR "Universal Display
Calibrator"?
No I don't, but I can contribute...
   ... I have difficulty trusting the tool for setting up "critical viewing"
monitors because the pickup device (and serial port itself) is analog.
Sorry, but PC serial COMM. ports are DIGITAL devices not ANALOG. The
communication between the pickup and the PC is data sent as bytes, one bit
at a time, LSB first. Although there is a range of permitted voltages of the
Tx & Rx signal lines in the specification, the basic requirement is to be
able to distinguish between MARK and SPACE levels (high and low logic
levels) That is accomplished by a digital receiver IC operating as a
comparator, whose output is purely logic states, fed into a UART which is
basically a serial input shift-register with parallel output when all bits
of the byte are received (plus lots of control and timing logic). As long as
the serial input level is sufficient to unambiguously trigger the
comparators output states, levels beyond that (but less than the IC's max
ratings) are irrelevant. Hopefully, the baud rate generators of both units
are within tolerance, but unless the communications protocol has some form
of data validation (parity, checksum or CRC) you might not notice some data
errors.
   Won't different serial ports have slightly different voltage output and
resistance (termination) causing "slightly" different readings from the
sensor? I tested my theory using several different laptops on the same
monitor, and sure enough- slightly different readings from each laptop. The
discrepancy is subtle enough to use the system for a rough setup (I've
measured greater discrepancies from the factory probes on my Ikegami and
Sony monitors), and the readings are consistent with the same laptop, so it
is useful if all you need to do is match one display to another.
As long as the comm. port is accurately set up with regard to baud rate,
bits per word, and parity; and as long as no bytes are missed going in
either direction, the communication should be perfect. In the unlikely event
that the device's pickup head does not connect to a dedicated source of
power, then it is possible that, like a serial mouse, it relies on
scavenging power from the serial port's handshake pins. Variations in
current sourcing ability of those handshaking lines could affect the amount
of power available for the device in that case, but I doubt that is how the
unit is powered. But if the pickup is sensitive that way, it is a very poor
design and a good reason the company to drop the unit. There should be no
difference from PC to PC on the readings received!

By the way, you could prove this by placing a Y cable at the serial port on
the RECEIVE data only and monitor it simultaneously on a second PC using the
same software on both. If that works, then swap the PC ends of the cable and
try again. If as you say the PC affects the PICKUP readings then both PCs
should get the same readings as each other, but those readings may not match
the first set of readings. That's about all I can do from here.

Jim DeLuca, Senior Video Engr.
Crest National Digital Media Complex - Video Lab
1000 N Highland Ave; Hollywood, CA 90038-4207
james_deluca at crestnational.com (best option)
323.860.1300 ext 296           fax 323.461.8901
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://tig.colorist.org/pipermail/tig/attachments/20030602/33983f3e/attachment.htm 



More information about the Tig mailing list