Does failure to do the VBL synch affect the stimulus duration?

1
Hi everyone,

I know that VBL sync failure will greatly affect the timestamp, but does this also affect how long the stimulus appears on the screen if I use waitSecs?

I have tested with waitsecs = 3s, and time it as soon as the stimulus appears on the screen, it looks like it's perfectly 3s. But for the real experiment, i will be using 250ms (which is too short for me to be able to time it), so not sure if this will be affected. Can anyone advice me?

Thank you.

Sabrina
2
XX---In PSYCHTOOLBOX@yahoogroups.com, <ultra_violet212@...> wrote :

Hi everyone,

-> Hi Sabrina

I know that VBL sync failure will greatly affect the timestamp, but does this also affect how long the stimulus appears on the screen if I use waitSecs?

-> VBL sync failure means that something is wrong with the way your system does visual stimulus onset, so the general answer is yes. It could be only timestamping affected, however if you don't use the timestamps and mechanisms of 'Flip', but WaitSecs then i can't guarantee anything wrt. timing. It could also/or mean the stimulus onset timing is affected. It could mean that vsync is broken and your stimulus is tearing at onset, which would make the stimulus itself undefined, especially on short presentations or animations. It could mean certain stimulus frames don't get shown at all with fast animations.

-> The failures and other problems affect timescales in the range of a few video refresh cycles, think of errors in stimulus onset/presentation duration/timestamps in the 50 msecs range on a 60 Hz display, so none of this would probably a big deal at stimulus durations of 3 seconds. At 250 msecs i personally would rather try to fix my system than hope it doesn't do anything bad to my experiment. "help SyncTrouble" as a starter.

-mario

I have tested with waitsecs = 3s, and time it as soon as the stimulus appears on the screen, it looks like it's perfectly 3s. But for the real experiment, i will be using 250ms (which is too short for me to be able to time it), so not sure if this will be affected. Can anyone advice me?

Thank you.

Sabrina
3
Thanks Mario.

On 5 Dec 2016, at 05:16, mario.kleiner@... [PSYCHTOOLBOX] <PSYCHTOOLBOX@yahoogroups.com> wrote:

 

XX---In PSYCHTOOLBOX@yahoogroups.com, <ultra_violet212@...> wrote :


Hi everyone,

-> Hi Sabrina

I know that VBL sync failure will greatly affect the timestamp, but does this also affect how long the stimulus appears on the screen if I use waitSecs?

-> VBL sync failure means that something is wrong with the way your system does visual stimulus onset, so the general answer is yes. It could be only timestamping affected, however if you don't use the timestamps and mechanisms of 'Flip', but WaitSecs then i can't guarantee anything wrt. timing. It could also/or mean the stimulus onset timing is affected. It could mean that vsync is broken and your stimulus is tearing at onset, which would make the stimulus itself undefined, especially on short presentations or animations. It could mean certain stimulus frames don't get shown at all with fast animations.

-> The failures and other problems affect timescales in the range of a few video refresh cycles, think of errors in stimulus onset/presentation duration/timestamps in the 50 msecs range on a 60 Hz display, so none of this would probably a big deal at stimulus durations of 3 seconds. At 250 msecs i personally would rather try to fix my system than hope it doesn't do anything bad to my experiment. "help SyncTrouble" as a starter.

-mario

I have tested with waitsecs = 3s, and time it as soon as the stimulus appears on the screen, it looks like it's perfectly 3s. But for the real experiment, i will be using 250ms (which is too short for me to be able to time it), so not sure if this will be affected. Can anyone advice me?

Thank you.

Sabrina