Hi Kristina,
no practical (test) experience with these, so all comments
just based on the official specs and under the assumption
that the drivers don't contain any unknown bugs specific
to that chip.
The 9400M is feature-wise state of the art and provides
support for all functionality of PTB. The timing of
stimuli should be correct, i.e., stimulus onset frame-
accurate and perfectly synced to display retrace,
timestamps reliable and sub-millisecond exact,
of course always assuming you use a CRT display
for stimulus presentation, as exact timing on
any current LCD flat panel is impossible.
Performance will be lower than with the 9600M GT in
MacbookPro for some types of stimuli, as the 9400M
doesn't have dedicated VRAM memory and a lower number
of stream processors, but that will only affect you if you
try to do fast animation with complex stimuli or complex
post processing. As always, it depends on the complexity
of your stimuli.
Wrt. to display configuration. Single display is always
lower load than dual-display. Mirror mode should be
lower load than "true" dual display with separate content
and dual display under control of PTB (e.g., for binocular
stimulus presentation on two displays) should always be
lower load than having 1 display for PTB and 1 display
for the regular Matlab GUI etc. These are rules of thumb,
they are always true for technical reasons, but their relative
impact is highly dependent on what you're doing, so the
only way to find out if it's a problem is to find out if it's
a problem for your specific setup.
What do you mean with "suitable for stimulus generation"?
One reason to go for a MacBookPro in the past was that
Apple used Intel graphics chips in their previous MacBooks,
and Intel is not exactly famous for high performance graphics
or bug free graphics drivers, so the MacBookPro with an
ATI or NVidia GPU was a safer bet.
That said, there are still two serious unfixed bugs
in Leopard 10.5.6 at least on the NVidia 8800 cards:
a) Trying to open two fullscreen windows on a dual-display
setup causes serious system malfunctions up to the point
of a system crash, so real dual-display mode with PTB is
a no-go until f%$#%#$Q#$ Apple fixes their operating system,
instead of completely ignoring the open bug reports.
b) Frame sequential stereo for shutter glasses is also
broken on Geforce 8800 with 10.5.6. See a) for the
solution.
However if you don't need stereo display via a) or b),
it should be fine. As far as i know, one can turn the
internal display of MacBooks off by connecting an
external keyboard, putting the machines to sleep by
closing the lid, then waking the machines up via the
keyboard, so only the external display turns on. Or
you use mirror-mode.
-mario
--- In psychtoolbox@yahoogroups.com, Kristina Visscher <visscher@...> wrote:
>
> Hi, Psychtoolbox users,
> I am going to be using psychtoolbox for display of stimuli for EEG.
> I'm particularly interested in knowing that the timing of stimuli is
> correct. For various reasons, I'd like to use a laptop to present
> stimuli. What do you think of using the new macbook laptops for this
> purpose?
> These computers have a pretty good graphics card (NVIDIA GeForce
> 9400M). Has anyone extensively tested timing for the new batch of mac
> laptops with psychtoolbox? Do you anticipate any timing problems due
> to having either a dual display or mirrored display configuration?
>
> I had a look at the page http://psychtoolbox.org/wikka.php?wakka=GraphicsHardwareRequirements
> and didn't find my answer, but I may have missed another discussion
> of the topic.
>
> In addition, I'm interested in whether anyone has used the new macbook
> minis for stimulus generation. Are these computers (graphics card is
> NVIDIA GeForce 9400M) up to the challenge of having a single display?
> I notice from the message board that many people are using macbook pro
> -- is there a reason to upgrade to that if the computer is used almost
> exclusively for stimulus generation?
>
> Thanks for any help you may have,
>
> Kristina
>