Beginner issue : gamma calibration with PR655 from Monitor Center

[Chencan QIAN]

Hi everyone, 

I've just borrowed a PR655 and would like to calibrate my monitor using Monitor Center.

However, I don't know how to establish connection between PsychoPy and the photometer.

When I plugged PR655 to my Windows 7 (64-bit) PC via USB, Windows complained that it couldn't install driver for the new device. (I didn't have a CD with the PR655...)

I tried "Get Photometer (Scan all ports)" immediately after power on PR655, but PsychoPy said "No photometers found".

Do I need to install a driver first? But I seemed to fail to find any driver download link in its website...

Thank you in advance!

Chencan

[Jonathan Peirce]

Yes, the PR655 should have come with a driver CD. I'll try mailing you
the files from the office on monday (unless someone else gets there first)

Jon

> --
> You received this message because you are subscribed to the Google
> Groups "psychopy-users" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to psychopy-user...@googlegroups.com.
> To post to this group, send email to psychop...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/psychopy-users/52eb05ad-df28-4f10-8512-26dcad239012%40googlegroups.com.
> For more options, visit https://groups.google.com/groups/opt_out.

--
Jonathan Peirce
Nottingham Visual Neuroscience

http://www.peirce.org.uk/

[Chencan QIAN]

Thanks, Jon.

I'll also contact the owner the PR655, check whether the CD is still available, and update here.

[Chencan QIAN]

I have my colleague copy the content of the CD to me, and now the PR655 and the Monitor Center is fully functional.

Thanks for the information!

He said that he had never used the CD because mac seemed to be able to recognize PR655 out of the box.

What's the difference between "Gamma Calibration" and "Gamma Test"?

I tried a 16 point auto Gamma Calibration. However, after that PsychoPy started to produce very bright image in experiment. (The photo looked over exposed.)

If I understand it correctly, after gamma calibration, PsychoPy will (approximately) map -1 to the minimum luminance the monitor can display, 1 to the maximum, and others linearly spaced in between, right?

E.g., 0 should be the min+(max-min)/2.

[Jonathan Peirce]

"Gamma test" is to make sure your calibration has worked. It collects a new set of measurements with the gamma value applied. You should always run that after performing a calibration (due to some annoying issues under the hood, you might have to close and reopen the monitor centre first though).

Yes, photographic images are going to look over-exposed on a linearised monitor. That's why your monitor is set to have a nonlinear gamma for most uses.

Jon

To view this discussion on the web visit https://groups.google.com/d/msgid/psychopy-users/4f928951-3418-46bc-b4a7-9b1917c3556b%40googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

[Chencan QIAN]

About the "close and reopen" thing, I've got a test result showing that the mapping was indeed linearized (even a little bit over) without close and reopen. Do you imply that this result could be unreliable? Or do you mean sometimes the test could fail to launch, but as long as it is started, the result is reliable even without a close and reopen?

My problem here is that the dark part of my photos looks too dark (and loses contrast for those pixels whose the gray-level is less than, e.g., 80) when they rendered in PsychoPy, although they look normal in Windows Photo Viewer. I guess the system has applied some (small) correction, while PsychoPy assumes gamma=1 by default. (BTW, after correction, the calculated gamma for L is about 2.7)

Is there a way to get a "natural" gamma (perhaps somewhere between 1 and 2.7), so that the photos will look normal/natural?

Thanks!

[Jonathan Peirce]

On 03/11/2013 12:39, Chencan QIAN wrote:

About the "close and reopen" thing, I've got a test result showing that the mapping was indeed linearized (even a little bit over) without close and reopen. Do you imply that this result could be unreliable? Or do you mean sometimes the test could fail to launch, but as long as it is started, the result is reliable even without a close and reopen?

Yes, if it fails it will fail completely.

My problem here is that the dark part of my photos looks too dark (and loses contrast for those pixels whose the gray-level is less than, e.g., 80) when they rendered in PsychoPy, although they look normal in Windows Photo Viewer. I guess the system has applied some (small) correction, while PsychoPy assumes gamma=1 by default. (BTW, after correction, the calculated gamma for L is about 2.7)

If the gamma table for your monitor has the default, uncalibrated, values then PsychoPy will simply use a pass-through set of gamma values (so images should look exactly like the system settings). If you've changed that gamme table in any way then the values you've inserted will be used and applied.

Is there a way to get a "natural" gamma (perhaps somewhere between 1 and 2.7), so that the photos will look normal/natural?

I think what you're wanting is an /uncalibrated/ system. Just create a new monitor (call it "natural"?) and set only the screen size values. Then you can use that monitor name for experiments with natrual images and use a different name for a linearised monitor for experiments that need it.

Joln

To view this discussion on the web visit https://groups.google.com/d/msgid/psychopy-users/529d2c12-f44e-48a3-be1d-6638849dc540%40googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.

[Chencan QIAN]

Thanks very much! Now my concepts become more clear.

Now I set the monitor in PsychoPy in /uncalibrated/ state as you suggested, and played around with Windows Color Manager / Display calibration to make those photos looks natural.

One final little question is, why there are four gamma value (LRGB) rather than three (RGB)? When display gray-scale image, am I using the L gamma or some combination of RGB gamma?

[Jonathan Peirce]

It uses the RGB is available
Jon

To view this discussion on the web visit https://groups.google.com/d/msgid/psychopy-users/2c4bf048-3397-4e5e-9a8a-14fef9b730d7%40googlegroups.com.

For more options, visit https://groups.google.com/groups/opt_out.