Noise Stim changing pattern and noise elements

Hello ! Can you help me with the noise stim function?

PsychoPy version (2022.2.5):

I want to create a dynamic noise pattern (like a blurred tv ->black and white). For now, I have a subrectangle pattern moving randomly, which is repeated several times on the screen. Hence my result is like a pavement of several times the same pattern. I would like to get rid of this pattern-like structure. I would like the whole screen to move randomly, without any repeating pattern.
Noise elements also are rectangle and I want squares.
I do not know if this is possible with this software though, any help would be warmly welcomed.

This is what I get :

This is what I want :

I changed this settings :

  • Noise Stim
  • Type of noise : uniform
  • Size : 10.5 DVA
  • Noise element size : 0.42 DVA
  • Timing : 1/60 NFrame

Thanks :slight_smile:

Hello

what does Size 10.5 DVA mean? Degrees of visual angle?. What about setting the spatial-unit to height? You should then get a square when using the same values for height and width.

What does Timing: 1/60 NFrame mean? When you set the Timing to NFrame and When to update noise sample to 1, the sample is updated on each frame.

Hi,

Yes DVA = degree of visual angle. I tried setting “height” but I still have rectangles… I’ve just tried changing the “Timing” and “when to update” but it do not work neither. Thanks a lot!

Hello Elise

I get square noise stimulus with these settings

Spatial units are set to height via Experiment settings.

What do you mean with?

What did not work?

Best wishes Jens

Ok, I understand what you meant with “height”. You’re right, the basic form (size) has to be a “square” to have squares noise element size! Thank tou very much!

Do you think you can help me with my other question?

Hello Elise

well, I don’t really understand your question regarding timíng. What does it mean, that it did not work?

This setting, for instance, results in updating the noise pattern on each frame, that is 1 sec/60 or approx. every 16.667 ms on my laptop.

Best wishes Jens

My second question is not really about de timing. “For now, I have a subrectangle pattern moving randomly, which is repeated several times on the screen. Hence my result is like a pavement of several times the same pattern. I would like to get rid of this pattern-like structure. I would like the whole screen to move randomly, without any repeating pattern”. I think the images on my original post are clearer. Let le know if I can give you more details. Thanks a lot for your help!

Hello Elise,

do plan to run the experiment online or offline? I am asking because online you don’t know the size of the screen/window in advance and you need some screen scaling procedure (see Units for the window and stimuli — PsychoPy v2023.1.0).

In case you run it offline, you know the size of your window and set the size of noise stimulus accordingly.

Best wishes Jens

Hi,

I’m running the experiment offline. What do you mean by “you know the size of your window and set the size of noise stimulus accordingly”?
I set my screen size in the settings and I chose 0.42 degree of visual angle but I don’t think it influences the “random” aspect of my pattern?

Thanks!

Hello Elise

didn’t you want to have the random pattern set to the size of your screen?

Best wishes Jens

No, I want a rectangle : 10;5 (W;H) degree of visual angle. I want the squares moving randomly in this rectangle. On this image, the pattern is the same

And here the rectangle has not any repeating items. This is what I want to set.

But the pattern has to be a rectangle on my screen. I don’t know if my explication is better! :slight_smile:

Hello Elise,

ok, I think I now understand what you mean and I think I can recreate the behaviour you describe.

When I set the size to 10:5 deg I get rectangle noise elements (size 1 or 2 deg) and the pattern seems to repeat over the trial. Also, I see a column of half a noise element on each side of the noise-stimulus.

This does not seem to happen when I set everything to height instead of deg. At least I do not see that the pattern repeats and the noise elements are squares.

The fact that the noise elements deviate from a square might be due the fact that degrees of visual angle span the same number of pixel at all parts of the screen, see here at the bottom of the page

https://www.psychopy.org/general/units.html

But I am not sure if my explanation is correct.

Why not try to set everything to height such that it spans 10 degree in width and 5 degrees in height with the appropriate noise element size?

Sorry, I hardly ever use the noise-component.

Best wishes Jens

Hi,

Thanks a lot for your reply, I just tried and I have exactly what I wanted! Thank you very much for your time and your reply!!

Best regards,

Elise