RT Correction Keyword
R switch alternative.
This is largely redundant these days as it was implemented in the days of DM where one machine controlled four subjects and with each subject's display being on a different portion of the screen there were four different constant errors that could be in one experiment's data. With this removed and given the fact that we can no longer say where a given stimulus will be presented on the screen use of RT correction has fallen by the wayside. The only time someone would need to use this would be if they presented a probe on the screen at widely different Y coordinates -- and having just looked closely at the code it couldn't be used that way as the
<AcousticControl>had to be set before the <rtc> value was used, this is now fixed in version 2 of DMDX.
If <AcousticControl>is turned on then this value will need to be specified for each RT as it is cleared to 0 at the beginning of the item, in version 2 using <rtc> will set the display correction factor and once specified all following items will use this. Note that this value is a floating point value if used in the keywords, fractions of a millisecond are able to be specified, when used with the R switch it is still an integer. If you wish to determine the delay for a raster travelling from the top of the screen to a given location you will have to use a photosensitive device and a logic analyzer and a signal from the PIO, bearing in mind that timing changes between video modes. A crude estimation can be obtained by dividing the refresh interval by the number of lines in the display and multiplying by the line that the target is presented on, however not all the retrace interval is spent drawing the display, there is a good deal of overdraw and the time for retrace itself, usually totaling more than a couple of milliseconds.