Timing between MAC and Windows

[jv]

Hi! I've been talking to a friend who has knowledge about computer engineering. He has told me that for what I am trying to accomplish (It has been a few years by now) I should switch to iOS,OSx. I just need to know if Windows really is that bad for accuracy and stimuli timing. Im using Windows 7 upgraded to 10.

Should I consider changing. I want the best set up I can get. I will use my project for a thesis and my professors are not familiar with the Reaction Timing and Cognitive Neuroscience paradigms. They are more inclined to Clinical practice not research.

Thanks!!

Juan

[David McFarlane]

Juan,

Generally speaking, *all* preemptive multitasking operating systems
suffer from similar problems with critical timing -- that includes
Windows starting with Windows 95 (prior to that Windows used cooperative
multitasking), Mac OS X, and any Unix-like OS (I don't know anything
about iOS or Android, but if you run experiments using a mobile OS on a
mobile device then you have a host of other challenges). The E-Prime
Users Guide speaks about this.

To get better control over timing you really need to go back to DOS, or
the Apple II, or a Commodore 64, etc. Or in modern times, use a true
real-time OS such as QNX. But good luck finding convenient
general-purpose experiment-generating software for any of those platforms.

Another question might be just how good of timing performance do you
really need? As also discussed in the E-Prime Users Guide (and in other
publications that you might find), in most cases the timing performance
of your response devices (keyboard, mouse, etc.) is already worse than
the timing performance of your OS, unless you go to the trouble of using
more reliable and expensive response devices. And in most cases, you
can make up for the sloppy timing in your systems just by running a few
more subjects -- think through the statistics. (Assuming that you
merely want aggregate timing measures -- if you really need to control
each individual event down to the millisecond (say, controlling the
spark timing of an engine), then you really will need real-time control.)

That said I don't know whether Windows and Mac OS X differ in any
particular way with regard to timing. But think about it -- for these
experiments, we essentially write real-time computer video games. And
ask yourself which platform is preferred by hard-core real-time gamers.
The answer is Microsoft Windows. So if Windows suits that demanding
use case, it must be at least as suitable as Mac OS X for psychology
experiments.

---------------
David McFarlane
E-Prime training online:
http://psychology.msu.edu/Workshops_Courses/eprime.aspx
Twitter: @EPrimeMaster (https://twitter.com/EPrimeMaster)

[jv]

Ok. I was thinking better gaming computer better timing and less overclocking. That is a good point.

This question arises because I followed some advice you gave me last year for sending trigger based on accuracy and a wait object. 

That worked, but I was receiving the marker for the stimulus after the marker for mouse press and accuracy. So these were the labels and the steps of the presentation:

0 = incorrect

1 = correct

49 = stimulus object presentation

so I was receiving when importing to another software for analyzing EEG:

0 or 1, then 49

That was confusing because the EEG acquisition software was presenting it in order.

But when importing to EEG analyzing software it was in the wrong order even though I labeled them correct.

One thing I was very motivated to do was to analyze the EEG on the same EEG analyzing software to do a P300 go/no go task.

This was through some code called LSL library. To do that I will need to communicate E-Prime with the LSL library at the same time EEG with the LSL library.

It is a lot I know but I seems to me that there is a lot of coding to do before actually synchronizing everything. This has proven to be an arduous task because of a lot of set backs like a robbery that have happened to me or my lack of knowledge. 

But,  I guess that sticking with what I have IS the best way to go for now. 

Juan