random noise frametiming

1
Hi Mario (or anyone that has tried this),

i am trying to deliver a random noise to a retina in a multielectrode
array set up. It appears that going at the refresh rate of the screen
(75 HZ in my case), may be too fast for the RGCs to "get it" and
respond. Therefore i want to prolong the presentation of each frame.
Is Waitsecs (delay), before flip the way to go ?


Thanks for your help, you guys have enabled a molecular biologist to
do physiology.

Tudor
2
Hi Tudor,

The third (optional) argument to the flip function is "when", see the
help file for the function by doing Screen('Flip?')

If you're using a CRT, the RCGs will still be entrained to 75 Hz,
because that flicker frequency is still present even if multiple
sequential frames are identical.

keith

On Sep 22, 2008, at 11:29 AM, tudorcb wrote:

> Hi Mario (or anyone that has tried this),
>
> i am trying to deliver a random noise to a retina in a multielectrode
> array set up. It appears that going at the refresh rate of the screen
> (75 HZ in my case), may be too fast for the RGCs to "get it" and
> respond. Therefore i want to prolong the presentation of each frame.
> Is Waitsecs (delay), before flip the way to go ?
>
>
> Thanks for your help, you guys have enabled a molecular biologist to
> do physiology.
>
> Tudor
3
Hi Keith,

if i get this right, then, perhaps it is easier to tell the program
to repeat flip the needed amount of times.
Say my frame rate is 75 Hz, which amounts to roughly 13 msec/frame.
If i ask for a fixed time, like 60 msec, it will come up with the
closest estimate which is either 4 or 5 screen redraws, and it may
completely screw up the overall timing. Might as well settle for 65
msec presentations and tell it to redraw the screen 5 times or 52 msec
for 4 redraws.

Thanks for pointing that out to me.

Tudor


--- In psychtoolbox@yahoogroups.com, Keith Schneider <ks@...> wrote:
>
> Hi Tudor,
>
> The third (optional) argument to the flip function is "when", see the
> help file for the function by doing Screen('Flip?')
>
> If you're using a CRT, the RCGs will still be entrained to 75 Hz,
> because that flicker frequency is still present even if multiple
> sequential frames are identical.
>
> keith
>
> On Sep 22, 2008, at 11:29 AM, tudorcb wrote:
>
> > Hi Mario (or anyone that has tried this),
> >
> > i am trying to deliver a random noise to a retina in a multielectrode
> > array set up. It appears that going at the refresh rate of the screen
> > (75 HZ in my case), may be too fast for the RGCs to "get it" and
> > respond. Therefore i want to prolong the presentation of each frame.
> > Is Waitsecs (delay), before flip the way to go ?
> >
> >
> > Thanks for your help, you guys have enabled a molecular biologist to
> > do physiology.
> >
> > Tudor
>
4
That's right, you have to determine which is more important, either
having the exact same number of screen refreshes, or keeping the
overall timing precise. For example, if you want a stimulus that
alternates every 60 ms, is it more important that every alternation
period includes 4 screen refreshes, or do you want to keep the time as
close as possible to 60 ms, in which case subsequent periods might
include either 4 or 5 refreshes? If you include the "when" argument,
the screen will flip on the next refresh that is later than the time
you specify. Keep in mind that the refresh rate isn't *precisely* 75
Hz, it's just some number close to that, which the PTB can estimate
accurately.

By the way, you can only give the flip function a timing arguments,
not an integer number of frames. That is, you can't tell it to wait 4
frames, you would have to (knowing your exact refresh rate) calculate
the time this would take.

keith

On Sep 22, 2008, at 4:37 PM, tudorcb wrote:

> Hi Keith,
>
> if i get this right, then, perhaps it is easier to tell the program
> to repeat flip the needed amount of times.
> Say my frame rate is 75 Hz, which amounts to roughly 13 msec/frame.
> If i ask for a fixed time, like 60 msec, it will come up with the
> closest estimate which is either 4 or 5 screen redraws, and it may
> completely screw up the overall timing. Might as well settle for 65
> msec presentations and tell it to redraw the screen 5 times or 52 msec
> for 4 redraws.
>
> Thanks for pointing that out to me.
>
> Tudor
>
>
> --- In psychtoolbox@yahoogroups.com, Keith Schneider <ks@...> wrote:
>>
>> Hi Tudor,
>>
>> The third (optional) argument to the flip function is "when", see the
>> help file for the function by doing Screen('Flip?')
>>
>> If you're using a CRT, the RCGs will still be entrained to 75 Hz,
>> because that flicker frequency is still present even if multiple
>> sequential frames are identical.
>>
>> keith
>>
>> On Sep 22, 2008, at 11:29 AM, tudorcb wrote:
>>
>>> Hi Mario (or anyone that has tried this),
>>>
>>> i am trying to deliver a random noise to a retina in a
>>> multielectrode
>>> array set up. It appears that going at the refresh rate of the
>>> screen
>>> (75 HZ in my case), may be too fast for the RGCs to "get it" and
>>> respond. Therefore i want to prolong the presentation of each
>>> frame.
>>> Is Waitsecs (delay), before flip the way to go ?
>>>
>>>
>>> Thanks for your help, you guys have enabled a molecular biologist to
>>> do physiology.
>>>
>>> Tudor
>>
>
>
>
> ------------------------------------
>
> Post your message to: psychtoolbox@yahoogroups.com
> Please indicate OS9, OSX, or WIN version, and include your full name.
> Denis Pelli, David Brainard, and Allen Ingling.
> http://psychtoolbox.org
> Yahoo! Groups Links
>
>
>
>
5
DriftWaitDemo and VBLSyncTest show nicely how
to achieve this "flip every n frames".

The PTB overview PDF in Psychtoolbox/PsychDocumentation
subfolder contains useful info and examples as well.

-mario

--- In psychtoolbox@yahoogroups.com, Keith Schneider <ks@...> wrote:
>
> That's right, you have to determine which is more important, either
> having the exact same number of screen refreshes, or keeping the
> overall timing precise. For example, if you want a stimulus that
> alternates every 60 ms, is it more important that every alternation
> period includes 4 screen refreshes, or do you want to keep the time as
> close as possible to 60 ms, in which case subsequent periods might
> include either 4 or 5 refreshes? If you include the "when" argument,
> the screen will flip on the next refresh that is later than the time
> you specify. Keep in mind that the refresh rate isn't *precisely* 75
> Hz, it's just some number close to that, which the PTB can estimate
> accurately.
>
> By the way, you can only give the flip function a timing arguments,
> not an integer number of frames. That is, you can't tell it to wait 4
> frames, you would have to (knowing your exact refresh rate) calculate
> the time this would take.
>
> keith
>
> On Sep 22, 2008, at 4:37 PM, tudorcb wrote:
>
> > Hi Keith,
> >
> > if i get this right, then, perhaps it is easier to tell the program
> > to repeat flip the needed amount of times.
> > Say my frame rate is 75 Hz, which amounts to roughly 13 msec/frame.
> > If i ask for a fixed time, like 60 msec, it will come up with the
> > closest estimate which is either 4 or 5 screen redraws, and it may
> > completely screw up the overall timing. Might as well settle for 65
> > msec presentations and tell it to redraw the screen 5 times or 52 msec
> > for 4 redraws.
> >
> > Thanks for pointing that out to me.
> >
> > Tudor
> >
> >
> > --- In psychtoolbox@yahoogroups.com, Keith Schneider <ks@> wrote:
> >>
> >> Hi Tudor,
> >>
> >> The third (optional) argument to the flip function is "when", see the
> >> help file for the function by doing Screen('Flip?')
> >>
> >> If you're using a CRT, the RCGs will still be entrained to 75 Hz,
> >> because that flicker frequency is still present even if multiple
> >> sequential frames are identical.
> >>
> >> keith
> >>
> >> On Sep 22, 2008, at 11:29 AM, tudorcb wrote:
> >>
> >>> Hi Mario (or anyone that has tried this),
> >>>
> >>> i am trying to deliver a random noise to a retina in a
> >>> multielectrode
> >>> array set up. It appears that going at the refresh rate of the
> >>> screen
> >>> (75 HZ in my case), may be too fast for the RGCs to "get it" and
> >>> respond. Therefore i want to prolong the presentation of each
> >>> frame.
> >>> Is Waitsecs (delay), before flip the way to go ?
> >>>
> >>>
> >>> Thanks for your help, you guys have enabled a molecular biologist to
> >>> do physiology.
> >>>
> >>> Tudor
> >>
> >
> >
> >
> > ------------------------------------
> >
> > Post your message to: psychtoolbox@yahoogroups.com
> > Please indicate OS9, OSX, or WIN version, and include your full name.
> > Denis Pelli, David Brainard, and Allen Ingling.
> > http://psychtoolbox.org
> > Yahoo! Groups Links
> >
> >
> >
> >
>
6
Thanks Keith and Mario,

i have read through the DriftWaitDemo and it does seem the best you
can get this to work if you have a given precise timing in mind.

In my case though, the delay is somewhat arbitrary (or my knowledge of
it, anyways) so i think i will just loop the presentation x times to
get something closer to my timing. As i am trying to reverse
correlate spikes from the random noise, and, obviously have to run
many iterations, even after the calibration, the error may be
important enough so that, after say 80,000 frames the timing will be
off. I have enough trouble with delays induced by filters and
recording software. So yeah, i will go with 4 screen refreshes.
Anyway, each trouble i run into is a learning experience.
Thanks a lot for the pointers and the explanations.

tudor



--- In psychtoolbox@yahoogroups.com, "Mario Kleiner"
<mario.kleiner@...> wrote:
>
> DriftWaitDemo and VBLSyncTest show nicely how
> to achieve this "flip every n frames".
>
> The PTB overview PDF in Psychtoolbox/PsychDocumentation
> subfolder contains useful info and examples as well.
>
> -mario
>
> --- In psychtoolbox@yahoogroups.com, Keith Schneider <ks@> wrote:
> >
> > That's right, you have to determine which is more important, either
> > having the exact same number of screen refreshes, or keeping the
> > overall timing precise. For example, if you want a stimulus that
> > alternates every 60 ms, is it more important that every alternation
> > period includes 4 screen refreshes, or do you want to keep the
time as
> > close as possible to 60 ms, in which case subsequent periods might
> > include either 4 or 5 refreshes? If you include the "when"
argument,
> > the screen will flip on the next refresh that is later than the time
> > you specify. Keep in mind that the refresh rate isn't *precisely*
75
> > Hz, it's just some number close to that, which the PTB can estimate
> > accurately.
> >
> > By the way, you can only give the flip function a timing arguments,
> > not an integer number of frames. That is, you can't tell it to
wait 4
> > frames, you would have to (knowing your exact refresh rate)
calculate
> > the time this would take.
> >
> > keith
> >
> > On Sep 22, 2008, at 4:37 PM, tudorcb wrote:
> >
> > > Hi Keith,
> > >
> > > if i get this right, then, perhaps it is easier to tell the program
> > > to repeat flip the needed amount of times.
> > > Say my frame rate is 75 Hz, which amounts to roughly 13 msec/frame.
> > > If i ask for a fixed time, like 60 msec, it will come up with the
> > > closest estimate which is either 4 or 5 screen redraws, and it may
> > > completely screw up the overall timing. Might as well settle for 65
> > > msec presentations and tell it to redraw the screen 5 times or
52 msec
> > > for 4 redraws.
> > >
> > > Thanks for pointing that out to me.
> > >
> > > Tudor
> > >
> > >
> > > --- In psychtoolbox@yahoogroups.com, Keith Schneider <ks@> wrote:
> > >>
> > >> Hi Tudor,
> > >>
> > >> The third (optional) argument to the flip function is "when",
see the
> > >> help file for the function by doing Screen('Flip?')
> > >>
> > >> If you're using a CRT, the RCGs will still be entrained to 75 Hz,
> > >> because that flicker frequency is still present even if multiple
> > >> sequential frames are identical.
> > >>
> > >> keith
> > >>
> > >> On Sep 22, 2008, at 11:29 AM, tudorcb wrote:
> > >>
> > >>> Hi Mario (or anyone that has tried this),
> > >>>
> > >>> i am trying to deliver a random noise to a retina in a
> > >>> multielectrode
> > >>> array set up. It appears that going at the refresh rate of the
> > >>> screen
> > >>> (75 HZ in my case), may be too fast for the RGCs to "get it" and
> > >>> respond. Therefore i want to prolong the presentation of each
> > >>> frame.
> > >>> Is Waitsecs (delay), before flip the way to go ?
> > >>>
> > >>>
> > >>> Thanks for your help, you guys have enabled a molecular
biologist to
> > >>> do physiology.
> > >>>
> > >>> Tudor
> > >>
> > >
> > >
> > >
> > > ------------------------------------
> > >
> > > Post your message to: psychtoolbox@yahoogroups.com
> > > Please indicate OS9, OSX, or WIN version, and include your full
name.
> > > Denis Pelli, David Brainard, and Allen Ingling.
> > > http://psychtoolbox.org
> > > Yahoo! Groups Links
> > >
> > >
> > >
> > >
> >
>