Dear Colleagues:
Those of you using the ³Freiburg Visual Acuity and Contrast Test²
(FACT), please read on:
It took only another 3 months to find the "second real bug":
If and only if the monitor was set to "Millions of colo(u)rs", the
stimuli for contrast sensitivity were not correctly dithered, thus
the stimulus contrast never went below about 0.4-0.7%. Consequently,
spuriously high results (CS>>300) could erroneously obtain. Subnormal
contrast values (due to pathologic conditions or experimental
degradation of optics) below about 100 would still be fine. In other
words: Any spurious high CS was at least as good as 100.
Special thanks to Adele Diamond, Dima Amso & Ping Situ for spotting
this sizable problem and prodding me to unearth the cause.
Version 5.5 of the FAT corrects this problem. It further adds a demo
of liminal (and subliminal) contrast stimuli, and a logarithmised
contrast sensitivity readout.
The revised version is available for free download, as usual, at
<http://www.michaelbach.de/fat.html>. Please note that it only works
on PowerMacs and requires MacOS 8.6 or later.
I apologize for any inconvenience my error may have caused you and
look forward ;-( to further feedback.
Best, Michael.
PS: What lesson do I learn from that?
1. User feedback should be taken very seriously (well, I've always
done that ;-)
2. Programming (sub)liminal stimuli is very difficult...
3. Improved technology (e.g. 5 years ago no one set her Mac to
"Millions of colours") must be anticipated
4. My own CS seems low, otherwise I would have recognized this
problem much earlier
PPS: May I take this opportunity to enquire who has used and cited
this test? Someone recently asked me about this and I had
insufficient information.
-- Prof. Michael Bach PhD, Dept. of Ophthalmology, University of Freiburg, Killianstr. 5, 79106 Freiburg, Germany. fon +49(761)270-4060, fax +49(761)270-4052. bach@uni-freiburg.de, <http://www.michaelbach.de>--------- About this list: http://www.visionscience.com/vsList.html
This archive was generated by hypermail 2b29 : Tue Jul 31 2001 - 13:58:37 PDT