Millisecond Forums

+x

 

PepijnA_um - 5/21/2021

Hi all,
I’m trying to set up a Gradual Onset Continous Performance Task, which is a Go/No-Go Task in which participants see a continuous gradual blending of two types of images (mountains and cities, so city-city-city-city-mountain (press)-city-city etc.) so they have to press the space bar if they see a target (e.g. mountains) appear. The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255. I then want to make these images being played from transparent to non-transparent, over any previous images (so that it creates this blending effect). I have successfully done this in Python (where I can draw two stimuli on top of each other), but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time? I have created a script that presents the two images but it does not seem to work (they seem to be presented directly after each other but not at the same time).
Does anyone have suggestions on how to achieve this gradual blending effect?
I have appended a shortened version of my script, with only 1 stimulus read-in "mountain_9.png" that I want to make appear on top of another (saturated/non-transparent) image “city_1_g.jpg".

Thank you for your input!

<block myblock>
/ trials = [1-3=mytrial]
</block>

<trial mytrial>
/ontrialbegin = [trial.mytrial.insertstimulustime(list.masklist.nextvalue,0);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,13);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,26);
.
. Comment: In total 48x
.
trial.mytrial.insertstimulustime(list.masklist.nextvalue,767);]
/ontrialend = [list.masklist.reset(); trial.mytrial.resetstimulusframes();]
/stimulustimes = [0=mypicture; 13=mypicture; 26=mypicture;
.
. Comment: Again here 48x
.
767=mypicture]
/beginresponsetime = 0
/validresponse = (57,noresponse)
/timeout = 800
</trial>

Comment: Previous image as background (in this case just presented in its fully non-transparant form)
<picture mypicture>
/ items = ("city_1_g.jpg")
</picture>

<list masklist>
/items = (picture.p5 …. until, picture.p255)
/selectionmode = sequence
/selectionrate = always
</list>

Example stimulus read in (in real life 48 versions of multiple images, here just for mountain_9)
<picture p5>
/items = ("mountain_9_g_5.png")
</picture>
.
. 48x
.
<picture p255>
/items = ("mountain_9_g_255.png")
</picture>

+x

 

Dave - 5/21/2021

+x

 

PepijnA_um - 5/21/2021

Hi all,
I’m trying to set up a Gradual Onset Continous Performance Task, which is a Go/No-Go Task in which participants see a continuous gradual blending of two types of images (mountains and cities, so city-city-city-city-mountain (press)-city-city etc.) so they have to press the space bar if they see a target (e.g. mountains) appear. The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255. I then want to make these images being played from transparent to non-transparent, over any previous images (so that it creates this blending effect). I have successfully done this in Python (where I can draw two stimuli on top of each other), but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time? I have created a script that presents the two images but it does not seem to work (they seem to be presented directly after each other but not at the same time).
Does anyone have suggestions on how to achieve this gradual blending effect?
I have appended a shortened version of my script, with only 1 stimulus read-in "mountain_9.png" that I want to make appear on top of another (saturated/non-transparent) image “city_1_g.jpg".

Thank you for your input!

<block myblock>
/ trials = [1-3=mytrial]
</block>

<trial mytrial>
/ontrialbegin = [trial.mytrial.insertstimulustime(list.masklist.nextvalue,0);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,13);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,26);
.
. Comment: In total 48x
.
trial.mytrial.insertstimulustime(list.masklist.nextvalue,767);]
/ontrialend = [list.masklist.reset(); trial.mytrial.resetstimulusframes();]
/stimulustimes = [0=mypicture; 13=mypicture; 26=mypicture;
.
. Comment: Again here 48x
.
767=mypicture]
/beginresponsetime = 0
/validresponse = (57,noresponse)
/timeout = 800
</trial>

Comment: Previous image as background (in this case just presented in its fully non-transparant form)
<picture mypicture>
/ items = ("city_1_g.jpg")
</picture>

<list masklist>
/items = (picture.p5 …. until, picture.p255)
/selectionmode = sequence
/selectionrate = always
</list>

Example stimulus read in (in real life 48 versions of multiple images, here just for mountain_9)
<picture p5>
/items = ("mountain_9_g_5.png")
</picture>
.
. 48x
.
<picture p255>
/items = ("mountain_9_g_255.png")
</picture>

> but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time?

What makes you think that?

By default, stimuli presented during a trial will be erased from the screen at the end of that trial / before the start of the next trial. If you do not want that (and you don't, if I understand your description correctly), then set the various <picture> elements' /erase attributes to false.

+x

 

PepijnA_um - 5/22/2021

+x

 

Dave - 5/21/2021

+x

 

PepijnA_um - 5/21/2021

Hi all,
I’m trying to set up a Gradual Onset Continous Performance Task, which is a Go/No-Go Task in which participants see a continuous gradual blending of two types of images (mountains and cities, so city-city-city-city-mountain (press)-city-city etc.) so they have to press the space bar if they see a target (e.g. mountains) appear. The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255. I then want to make these images being played from transparent to non-transparent, over any previous images (so that it creates this blending effect). I have successfully done this in Python (where I can draw two stimuli on top of each other), but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time? I have created a script that presents the two images but it does not seem to work (they seem to be presented directly after each other but not at the same time).
Does anyone have suggestions on how to achieve this gradual blending effect?
I have appended a shortened version of my script, with only 1 stimulus read-in "mountain_9.png" that I want to make appear on top of another (saturated/non-transparent) image “city_1_g.jpg".

Thank you for your input!

<block myblock>
/ trials = [1-3=mytrial]
</block>

<trial mytrial>
/ontrialbegin = [trial.mytrial.insertstimulustime(list.masklist.nextvalue,0);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,13);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,26);
.
. Comment: In total 48x
.
trial.mytrial.insertstimulustime(list.masklist.nextvalue,767);]
/ontrialend = [list.masklist.reset(); trial.mytrial.resetstimulusframes();]
/stimulustimes = [0=mypicture; 13=mypicture; 26=mypicture;
.
. Comment: Again here 48x
.
767=mypicture]
/beginresponsetime = 0
/validresponse = (57,noresponse)
/timeout = 800
</trial>

Comment: Previous image as background (in this case just presented in its fully non-transparant form)
<picture mypicture>
/ items = ("city_1_g.jpg")
</picture>

<list masklist>
/items = (picture.p5 …. until, picture.p255)
/selectionmode = sequence
/selectionrate = always
</list>

Example stimulus read in (in real life 48 versions of multiple images, here just for mountain_9)
<picture p5>
/items = ("mountain_9_g_5.png")
</picture>
.
. 48x
.
<picture p255>
/items = ("mountain_9_g_255.png")
</picture>

> but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time?

What makes you think that?

By default, stimuli presented during a trial will be erased from the screen at the end of that trial / before the start of the next trial. If you do not want that (and you don't, if I understand your description correctly), then set the various <picture> elements' /erase attributes to false.

Thanks for your reply. Even with the /erase = False setting on the blend looks similarly discontinuous instead of a smooth blend, do you have any ideas why this is the case? Could it be that the requirement of drawing the image for each new frame is too resource intensive?

+x

 

Dave - 5/24/2021

+x

 

PepijnA_um - 5/22/2021

+x

 

Dave - 5/21/2021

+x

 

PepijnA_um - 5/21/2021

Hi all,
I’m trying to set up a Gradual Onset Continous Performance Task, which is a Go/No-Go Task in which participants see a continuous gradual blending of two types of images (mountains and cities, so city-city-city-city-mountain (press)-city-city etc.) so they have to press the space bar if they see a target (e.g. mountains) appear. The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255. I then want to make these images being played from transparent to non-transparent, over any previous images (so that it creates this blending effect). I have successfully done this in Python (where I can draw two stimuli on top of each other), but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time? I have created a script that presents the two images but it does not seem to work (they seem to be presented directly after each other but not at the same time).
Does anyone have suggestions on how to achieve this gradual blending effect?
I have appended a shortened version of my script, with only 1 stimulus read-in "mountain_9.png" that I want to make appear on top of another (saturated/non-transparent) image “city_1_g.jpg".

Thank you for your input!

<block myblock>
/ trials = [1-3=mytrial]
</block>

<trial mytrial>
/ontrialbegin = [trial.mytrial.insertstimulustime(list.masklist.nextvalue,0);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,13);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,26);
.
. Comment: In total 48x
.
trial.mytrial.insertstimulustime(list.masklist.nextvalue,767);]
/ontrialend = [list.masklist.reset(); trial.mytrial.resetstimulusframes();]
/stimulustimes = [0=mypicture; 13=mypicture; 26=mypicture;
.
. Comment: Again here 48x
.
767=mypicture]
/beginresponsetime = 0
/validresponse = (57,noresponse)
/timeout = 800
</trial>

Comment: Previous image as background (in this case just presented in its fully non-transparant form)
<picture mypicture>
/ items = ("city_1_g.jpg")
</picture>

<list masklist>
/items = (picture.p5 …. until, picture.p255)
/selectionmode = sequence
/selectionrate = always
</list>

Example stimulus read in (in real life 48 versions of multiple images, here just for mountain_9)
<picture p5>
/items = ("mountain_9_g_5.png")
</picture>
.
. 48x
.
<picture p255>
/items = ("mountain_9_g_255.png")
</picture>

> but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time?

What makes you think that?

By default, stimuli presented during a trial will be erased from the screen at the end of that trial / before the start of the next trial. If you do not want that (and you don't, if I understand your description correctly), then set the various <picture> elements' /erase attributes to false.

Thanks for your reply. Even with the /erase = False setting on the blend looks similarly discontinuous instead of a smooth blend, do you have any ideas why this is the case? Could it be that the requirement of drawing the image for each new frame is too resource intensive?

It shouldn't be. However, there's very little I can say based on a code snippet. Get me the complete code as well as the images the code needs to run, and I'll take a look.

+x

 

PepijnA_um - 5/24/2021

+x

 

Dave - 5/24/2021

+x

 

PepijnA_um - 5/22/2021

+x

 

Dave - 5/21/2021

+x

 

PepijnA_um - 5/21/2021

Hi all,
I’m trying to set up a Gradual Onset Continous Performance Task, which is a Go/No-Go Task in which participants see a continuous gradual blending of two types of images (mountains and cities, so city-city-city-city-mountain (press)-city-city etc.) so they have to press the space bar if they see a target (e.g. mountains) appear. The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255. I then want to make these images being played from transparent to non-transparent, over any previous images (so that it creates this blending effect). I have successfully done this in Python (where I can draw two stimuli on top of each other), but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time? I have created a script that presents the two images but it does not seem to work (they seem to be presented directly after each other but not at the same time).
Does anyone have suggestions on how to achieve this gradual blending effect?
I have appended a shortened version of my script, with only 1 stimulus read-in "mountain_9.png" that I want to make appear on top of another (saturated/non-transparent) image “city_1_g.jpg".

Thank you for your input!

<block myblock>
/ trials = [1-3=mytrial]
</block>

<trial mytrial>
/ontrialbegin = [trial.mytrial.insertstimulustime(list.masklist.nextvalue,0);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,13);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,26);
.
. Comment: In total 48x
.
trial.mytrial.insertstimulustime(list.masklist.nextvalue,767);]
/ontrialend = [list.masklist.reset(); trial.mytrial.resetstimulusframes();]
/stimulustimes = [0=mypicture; 13=mypicture; 26=mypicture;
.
. Comment: Again here 48x
.
767=mypicture]
/beginresponsetime = 0
/validresponse = (57,noresponse)
/timeout = 800
</trial>

Comment: Previous image as background (in this case just presented in its fully non-transparant form)
<picture mypicture>
/ items = ("city_1_g.jpg")
</picture>

<list masklist>
/items = (picture.p5 …. until, picture.p255)
/selectionmode = sequence
/selectionrate = always
</list>

Example stimulus read in (in real life 48 versions of multiple images, here just for mountain_9)
<picture p5>
/items = ("mountain_9_g_5.png")
</picture>
.
. 48x
.
<picture p255>
/items = ("mountain_9_g_255.png")
</picture>

> but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time?

What makes you think that?

By default, stimuli presented during a trial will be erased from the screen at the end of that trial / before the start of the next trial. If you do not want that (and you don't, if I understand your description correctly), then set the various <picture> elements' /erase attributes to false.

Thanks for your reply. Even with the /erase = False setting on the blend looks similarly discontinuous instead of a smooth blend, do you have any ideas why this is the case? Could it be that the requirement of drawing the image for each new frame is too resource intensive?

It shouldn't be. However, there's very little I can say based on a code snippet. Get me the complete code as well as the images the code needs to run, and I'll take a look.

Hi, I have attached the complete script and images.
GradCPT_ExampleScript

+x

 

Dave - 5/24/2021

+x

 

PepijnA_um - 5/24/2021

+x

 

Dave - 5/24/2021

+x

 

PepijnA_um - 5/22/2021

+x

 

Dave - 5/21/2021

+x

 

PepijnA_um - 5/21/2021

Hi all,
I’m trying to set up a Gradual Onset Continous Performance Task, which is a Go/No-Go Task in which participants see a continuous gradual blending of two types of images (mountains and cities, so city-city-city-city-mountain (press)-city-city etc.) so they have to press the space bar if they see a target (e.g. mountains) appear. The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255. I then want to make these images being played from transparent to non-transparent, over any previous images (so that it creates this blending effect). I have successfully done this in Python (where I can draw two stimuli on top of each other), but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time? I have created a script that presents the two images but it does not seem to work (they seem to be presented directly after each other but not at the same time).
Does anyone have suggestions on how to achieve this gradual blending effect?
I have appended a shortened version of my script, with only 1 stimulus read-in "mountain_9.png" that I want to make appear on top of another (saturated/non-transparent) image “city_1_g.jpg".

Thank you for your input!

<block myblock>
/ trials = [1-3=mytrial]
</block>

<trial mytrial>
/ontrialbegin = [trial.mytrial.insertstimulustime(list.masklist.nextvalue,0);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,13);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,26);
.
. Comment: In total 48x
.
trial.mytrial.insertstimulustime(list.masklist.nextvalue,767);]
/ontrialend = [list.masklist.reset(); trial.mytrial.resetstimulusframes();]
/stimulustimes = [0=mypicture; 13=mypicture; 26=mypicture;
.
. Comment: Again here 48x
.
767=mypicture]
/beginresponsetime = 0
/validresponse = (57,noresponse)
/timeout = 800
</trial>

Comment: Previous image as background (in this case just presented in its fully non-transparant form)
<picture mypicture>
/ items = ("city_1_g.jpg")
</picture>

<list masklist>
/items = (picture.p5 …. until, picture.p255)
/selectionmode = sequence
/selectionrate = always
</list>

Example stimulus read in (in real life 48 versions of multiple images, here just for mountain_9)
<picture p5>
/items = ("mountain_9_g_5.png")
</picture>
.
. 48x
.
<picture p255>
/items = ("mountain_9_g_255.png")
</picture>

> but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time?

What makes you think that?

By default, stimuli presented during a trial will be erased from the screen at the end of that trial / before the start of the next trial. If you do not want that (and you don't, if I understand your description correctly), then set the various <picture> elements' /erase attributes to false.

Thanks for your reply. Even with the /erase = False setting on the blend looks similarly discontinuous instead of a smooth blend, do you have any ideas why this is the case? Could it be that the requirement of drawing the image for each new frame is too resource intensive?

It shouldn't be. However, there's very little I can say based on a code snippet. Get me the complete code as well as the images the code needs to run, and I'll take a look.

Hi, I have attached the complete script and images.
GradCPT_ExampleScript

Thanks. I don't understand the setup of the script you attached.

You have 46 items in list.masklist (not 48) and /ontrialbegin you're sampling and are inserting stimuli from list.masklist 60 times. I'm having trouble reconciling this with the original description:

"The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255."

I'm also not sure what you mean by "[the images] seem to be presented directly after each other but not at the same time)."

+x

 

PepijnA_um - 5/24/2021

+x

 

Dave - 5/24/2021

+x

 

PepijnA_um - 5/24/2021

+x

 

Dave - 5/24/2021

+x

 

PepijnA_um - 5/22/2021

+x

 

Dave - 5/21/2021

+x

 

PepijnA_um - 5/21/2021

Hi all,
I’m trying to set up a Gradual Onset Continous Performance Task, which is a Go/No-Go Task in which participants see a continuous gradual blending of two types of images (mountains and cities, so city-city-city-city-mountain (press)-city-city etc.) so they have to press the space bar if they see a target (e.g. mountains) appear. The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255. I then want to make these images being played from transparent to non-transparent, over any previous images (so that it creates this blending effect). I have successfully done this in Python (where I can draw two stimuli on top of each other), but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time? I have created a script that presents the two images but it does not seem to work (they seem to be presented directly after each other but not at the same time).
Does anyone have suggestions on how to achieve this gradual blending effect?
I have appended a shortened version of my script, with only 1 stimulus read-in "mountain_9.png" that I want to make appear on top of another (saturated/non-transparent) image “city_1_g.jpg".

Thank you for your input!

<block myblock>
/ trials = [1-3=mytrial]
</block>

<trial mytrial>
/ontrialbegin = [trial.mytrial.insertstimulustime(list.masklist.nextvalue,0);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,13);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,26);
.
. Comment: In total 48x
.
trial.mytrial.insertstimulustime(list.masklist.nextvalue,767);]
/ontrialend = [list.masklist.reset(); trial.mytrial.resetstimulusframes();]
/stimulustimes = [0=mypicture; 13=mypicture; 26=mypicture;
.
. Comment: Again here 48x
.
767=mypicture]
/beginresponsetime = 0
/validresponse = (57,noresponse)
/timeout = 800
</trial>

Comment: Previous image as background (in this case just presented in its fully non-transparant form)
<picture mypicture>
/ items = ("city_1_g.jpg")
</picture>

<list masklist>
/items = (picture.p5 …. until, picture.p255)
/selectionmode = sequence
/selectionrate = always
</list>

Example stimulus read in (in real life 48 versions of multiple images, here just for mountain_9)
<picture p5>
/items = ("mountain_9_g_5.png")
</picture>
.
. 48x
.
<picture p255>
/items = ("mountain_9_g_255.png")
</picture>

> but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time?

What makes you think that?

By default, stimuli presented during a trial will be erased from the screen at the end of that trial / before the start of the next trial. If you do not want that (and you don't, if I understand your description correctly), then set the various <picture> elements' /erase attributes to false.

Thanks for your reply. Even with the /erase = False setting on the blend looks similarly discontinuous instead of a smooth blend, do you have any ideas why this is the case? Could it be that the requirement of drawing the image for each new frame is too resource intensive?

It shouldn't be. However, there's very little I can say based on a code snippet. Get me the complete code as well as the images the code needs to run, and I'll take a look.

Hi, I have attached the complete script and images.
GradCPT_ExampleScript

Thanks. I don't understand the setup of the script you attached.

You have 46 items in list.masklist (not 48) and /ontrialbegin you're sampling and are inserting stimuli from list.masklist 60 times. I'm having trouble reconciling this with the original description:

"The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255."

I'm also not sure what you mean by "[the images] seem to be presented directly after each other but not at the same time)."

I based this on an example my supervisor sent me in which a transparant mask appears in front of an image (attached below).
The difference in number of images is because there is a discreptancy between my previous ideal trial duration (48 images at a 60Hz screen resulting in 800ms) and the fact that Inquisit works with time in ms and not frame presentation. For your last line, to me it seems like there is more of an A-B-A-B-A-B sequence going on than a [B over A][B over A][B over A] which I am trying to set up, if that makes sense?



Example Script

+x

 

Dave - 5/24/2021

+x

 

PepijnA_um - 5/24/2021

+x

 

Dave - 5/24/2021

+x

 

PepijnA_um - 5/24/2021

+x

 

Dave - 5/24/2021

+x

 

PepijnA_um - 5/22/2021

+x

 

Dave - 5/21/2021

+x

 

PepijnA_um - 5/21/2021

Hi all,
I’m trying to set up a Gradual Onset Continous Performance Task, which is a Go/No-Go Task in which participants see a continuous gradual blending of two types of images (mountains and cities, so city-city-city-city-mountain (press)-city-city etc.) so they have to press the space bar if they see a target (e.g. mountains) appear. The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255. I then want to make these images being played from transparent to non-transparent, over any previous images (so that it creates this blending effect). I have successfully done this in Python (where I can draw two stimuli on top of each other), but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time? I have created a script that presents the two images but it does not seem to work (they seem to be presented directly after each other but not at the same time).
Does anyone have suggestions on how to achieve this gradual blending effect?
I have appended a shortened version of my script, with only 1 stimulus read-in "mountain_9.png" that I want to make appear on top of another (saturated/non-transparent) image “city_1_g.jpg".

Thank you for your input!

<block myblock>
/ trials = [1-3=mytrial]
</block>

<trial mytrial>
/ontrialbegin = [trial.mytrial.insertstimulustime(list.masklist.nextvalue,0);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,13);
trial.mytrial.insertstimulustime(list.masklist.nextvalue,26);
.
. Comment: In total 48x
.
trial.mytrial.insertstimulustime(list.masklist.nextvalue,767);]
/ontrialend = [list.masklist.reset(); trial.mytrial.resetstimulusframes();]
/stimulustimes = [0=mypicture; 13=mypicture; 26=mypicture;
.
. Comment: Again here 48x
.
767=mypicture]
/beginresponsetime = 0
/validresponse = (57,noresponse)
/timeout = 800
</trial>

Comment: Previous image as background (in this case just presented in its fully non-transparant form)
<picture mypicture>
/ items = ("city_1_g.jpg")
</picture>

<list masklist>
/items = (picture.p5 …. until, picture.p255)
/selectionmode = sequence
/selectionrate = always
</list>

Example stimulus read in (in real life 48 versions of multiple images, here just for mountain_9)
<picture p5>
/items = ("mountain_9_g_5.png")
</picture>
.
. 48x
.
<picture p255>
/items = ("mountain_9_g_255.png")
</picture>

> but Inquisit does not seem to have a straightforward approach of presenting two images (e.g. 1 saturated with an overlapping transparent .png) at the same time?

What makes you think that?

By default, stimuli presented during a trial will be erased from the screen at the end of that trial / before the start of the next trial. If you do not want that (and you don't, if I understand your description correctly), then set the various <picture> elements' /erase attributes to false.

Thanks for your reply. Even with the /erase = False setting on the blend looks similarly discontinuous instead of a smooth blend, do you have any ideas why this is the case? Could it be that the requirement of drawing the image for each new frame is too resource intensive?

It shouldn't be. However, there's very little I can say based on a code snippet. Get me the complete code as well as the images the code needs to run, and I'll take a look.

Hi, I have attached the complete script and images.
GradCPT_ExampleScript

Thanks. I don't understand the setup of the script you attached.

You have 46 items in list.masklist (not 48) and /ontrialbegin you're sampling and are inserting stimuli from list.masklist 60 times. I'm having trouble reconciling this with the original description:

"The individual images are supposed to be appearing over a span of 800 ms, where then the next image starts to appear on top of it (from fully transparent at 0ms to fully visible at 800ms, and subsequently being the ‘background image’ for the next image that then starts to appear on top of this over a span of 800ms). My set of stimuli consists of 48 .png versions (48*1/60Hz refresh rate = 800ms) for each stimulus image with alpha values (transparency) varying between 0 and 255."

I'm also not sure what you mean by "[the images] seem to be presented directly after each other but not at the same time)."

I based this on an example my supervisor sent me in which a transparant mask appears in front of an image (attached below).
The difference in number of images is because there is a discreptancy between my previous ideal trial duration (48 images at a 60Hz screen resulting in 800ms) and the fact that Inquisit works with time in ms and not frame presentation. For your last line, to me it seems like there is more of an A-B-A-B-A-B sequence going on than a [B over A][B over A][B over A] which I am trying to set up, if that makes sense?



Example Script

> The difference in number of images is because there is a discreptancy between my previous ideal trial duration (48 images at a 60Hz screen resulting in 800ms) and the fact that Inquisit works with time in ms and not frame presentation.

This is incorrect. Inquisit gives you the option to do either. If you wish to work in frames, use /stimulusframes and insertstimulusframe(). If you wish to work in ms timings, use /stimulustimes and insertstimulustime().

<values>
/ framenumber = 0
</values>

<block myblock>
/ trials = [1-3 = mytrial]
</block>

<trial mytrial>
/ ontrialbegin = [
    trial.mytrial.resetstimulusframes();
    list.masklist.reset;
    values.framenumber = 0;
    values.framenumber += 1;
    trial.mytrial.insertstimulusframe(picture.mypicture, values.framenumber);
    while (values.framenumber < 48) {
        values.framenumber += 1;
        trial.mytrial.insertstimulusframe(picture.mypicture, values.framenumber);
        trial.mytrial.insertstimulusframe(list.masklist.nextvalue, values.framenumber);
    }
]
/ beginresponsetime = 0
/ validresponse = (57, 0)
/ trialduration = 49*display.refreshinterval;
</trial>

<picture mypicture>
/ items = ("city_1_g.jpg")
/erase = False
</picture>

<list masklist>
/items = (picture.p5,picture.p11,picture.p16,picture.p22,picture.p27,picture.p33,picture.p38,picture.p44,picture.p49,picture.p55,picture.p60,picture.p66,picture.p72,picture.p77,picture.p83,
picture.p88,picture.p94,picture.p99,picture.p105,picture.p110,picture.p116,picture.p121,picture.p127,picture.p133,picture.p138,picture.p144,picture.p149,picture.p155,picture.p160,picture.p166,
picture.p171,picture.p177,picture.p182,picture.p188,picture.p194,picture.p199,picture.p205,picture.p210,picture.p216,picture.p221,picture.p227,picture.p232,picture.p238,picture.p243,picture.p249,
picture.p255)
/selectionmode = sequence
/selectionrate = always
</list>

<picture p5>
/items = ("mountain_9_g_5.png")
/erase = False
</picture>
<picture p11>
/items = ("mountain_9_g_11.png")
/erase = False
</picture>
<picture p16>
/items = ("mountain_9_g_16.png")
/erase = False
</picture>
<picture p22>
/items = ("mountain_9_g_22.png")
/erase = False
</picture>
<picture p27>
/items = ("mountain_9_g_27.png")
/erase = False
</picture>
<picture p33>
/items = ("mountain_9_g_33.png")
/erase = False
</picture>
<picture p38>
/items = ("mountain_9_g_38.png")
/erase = False
</picture>
<picture p44>
/items = ("mountain_9_g_44.png")
/erase = False
</picture>
<picture p49>
/items = ("mountain_9_g_49.png")
/erase = False
</picture>
<picture p55>
/items = ("mountain_9_g_55.png")
/erase = False
</picture>
<picture p60>
/items = ("mountain_9_g_60.png")
/erase = False
</picture>
<picture p66>
/items = ("mountain_9_g_66.png")
/erase = False
</picture>
<picture p72>
/items = ("mountain_9_g_72.png")
/erase = False
</picture>
<picture p77>
/items = ("mountain_9_g_77.png")
/erase = False
</picture>
<picture p83>
/items = ("mountain_9_g_83.png")
/erase = False
</picture>
<picture p88>
/items = ("mountain_9_g_88.png")
/erase = False
</picture>
<picture p94>
/items = ("mountain_9_g_94.png")
/erase = False
</picture>
<picture p99>
/items = ("mountain_9_g_99.png")
/erase = False
</picture>
<picture p105>
/items = ("mountain_9_g_105.png")
/erase = False
</picture>
<picture p110>
/items = ("mountain_9_g_110.png")
/erase = False
</picture>
<picture p116>
/items = ("mountain_9_g_116.png")
/erase = False
</picture>
<picture p121>
/items = ("mountain_9_g_121.png")
/erase = False
</picture>
<picture p127>
/items = ("mountain_9_g_127.png")
/erase = False
</picture>
<picture p133>
/items = ("mountain_9_g_133.png")
/erase = False
</picture>
<picture p138>
/items = ("mountain_9_g_138.png")
/erase = False
</picture>
<picture p144>
/items = ("mountain_9_g_144.png")
/erase = False
</picture>
<picture p149>
/items = ("mountain_9_g_149.png")
/erase = False
</picture>
<picture p155>
/items = ("mountain_9_g_155.png")
/erase = False
</picture>
<picture p160>
/items = ("mountain_9_g_160.png")
/erase = False
</picture>
<picture p166>
/items = ("mountain_9_g_166.png")
/erase = False
</picture>
<picture p171>
/items = ("mountain_9_g_171.png")
/erase = False
</picture>
<picture p177>
/items = ("mountain_9_g_177.png")
/erase = False
</picture>
<picture p182>
/items = ("mountain_9_g_182.png")
/erase = False
</picture>
<picture p188>
/items = ("mountain_9_g_188.png")
/erase = False
</picture>
<picture p194>
/items = ("mountain_9_g_194.png")
/erase = False
</picture>
<picture p199>
/items = ("mountain_9_g_199.png")
/erase = False
</picture>
<picture p205>
/items = ("mountain_9_g_205.png")
/erase = False
</picture>
<picture p210>
/items = ("mountain_9_g_210.png")
/erase = False
</picture>
<picture p216>
/items = ("mountain_9_g_216.png")
/erase = False
</picture>
<picture p221>
/items = ("mountain_9_g_221.png")
/erase = False
</picture>
<picture p227>
/items = ("mountain_9_g_227.png")
/erase = False
</picture>
<picture p232>
/items = ("mountain_9_g_232.png")
/erase = False
</picture>
<picture p238>
/items = ("mountain_9_g_238.png")
/erase = False
</picture>
<picture p243>
/items = ("mountain_9_g_243.png")
/erase = False
</picture>
<picture p249>
/items = ("mountain_9_g_249.png")
/erase = False
</picture>
<picture p255>
/items = ("mountain_9_g_255.png")
/erase = False
</picture>

> For your last line, to me it seems like there is more of an A-B-A-B-A-B sequence going on than a [B over A][B over A][B over A] which I am trying to set up, if that makes sense?

That is not my impression, no. The two images are displayed at the same time, B (at varying levels of transparency) superimposed over A.