Measuring luminance and color of a patch on the screen
If you just want to know lum and color of a patch rather than doing an entire screen calibration, follow steps 1-4 and 8 of the Spyder2Pro/Spyder3Elite instructions far below. Alternatively or in addition, follow the below instructions for getting a measurement from the SpectraScan PR670.
Using the Photo Research Spectroradiometer (SpectraScan PR670)
It should be noted that the USB cable that comes with this device is only for charging the battery, so it's unlikely you'll need to use it. All navigational terms used in this guide assume the spectroradiometer is facing you (I.e. the LCD display is not visible)
- Plug the power adapter into the small port on the right hand side of the device, next to the mini-USB port.
- Screw the standard objective lens (MS-75) into the front of the device next to the compartment that holds the battery. When screwing in the lens, hold it by the rubber grip strip at the base of the lens piece.
- Insert an SD card into the card slot, found beneath the View Finder (eye-piece on the back right of the device). The PR670 supports cards up to 1GB. You should be able to find a Kingston 512MB SD card in the spectroradiometer carry case.
- Open the View Finder shutter by twisting the small knob on the right hand side of the device.
- Switch the spectoradiometer on by pressing the power button found at the top of the button panel on the back left of the device.
- Wait for it to initialize. Once it's ready, the LCD screen at the back of the device will display a title screen with serial number and ownership information (property of THE UNIVERSITY OF SYDNEY).
- The LCD screen is touch-responsive. You may need to press "Ok" at the bottom left hand corner of the screen to exit the title screen.
- The first thing you want to do is calibrate the screen. Press the button in the top left hand corner of the screen labeled Menu. Under Utilities, you should find an option to Calibrate Touch Screen. Follow the on screen instructions. Note: if you don't want to use the touch screen there's a set of 5-way navigation keys on the back left of the device, under the power button. Use the arrows to move the red outline which represents the selection on the LCD screen to the desired on-screen button, press the square in the center of the set of 5-way navigation buttons to enter.
- Next, under File Settings in the main Menu, create a new file on the SD card within which you can store your data: Select New File and enter a file name. If you don't want to have to press save after every measurement, you can set it to autosave all data to the SD card immediately, however remember to turn this option off when you've finished with the device. Changes to settings are applied immediately, so once you're satisfied just press Back in the lower left hand corner of the screen to return to the main Menu.
- Under Setup -> Instrument Setup in the main Menu, you can change the number of measurements which are averaged to produce a result, the exposure time of each measurement, and whether you want the processor to automatically determine exposure time for each measurement in a reading in order to optimize accuracy and minimize time (if your desired measurement depends on controlling for exposure time, keep this disabled). If you enable this option, disable it before returning the device.
- Exit the menus.
- Before you can take a measurement you need to ensure the lens is focused. First, adjust the lens so the image through the View Finder appears blurred. Next, adjust the eye piece of the View Finder until the black circle in the center of the image appears sharp against the background. Note: this black circle represents the size of the aperture through which the light passes to the diffraction grating and the detector. You can change the size of this aperture in the Instrument Setup menu - see step 10 above. By default it's set to the largest size - 1 deg.
- Once the aperture appears sharp, adjust the objective lens to focus the background image around it. Note: even if you're just measuring luminance with this device, making sure the image is in sharp focus ensures the spectroradiometer is in a similar position for each reading, controlling for changes in background luminance between readings.
- To take a reading, just press the "Measure" button found on the top left of the device. Hold the device steady until you hear a tone.
- If you created a new file on the SD card and selected it in the File Settings menu, the data will be saved to that file, but in order to navigate the data on the LCD screen (called the 'Explorer') you need to select your file in the Explorer Settings menu under the Main Menu. You can navigate the data of the current reading as well as past readings, save data to the SD card or delete it from temporary memory (RAM) using the navigation buttons at the bottom of the LCD screen.
- A special note about the data files you create on the SD card. They cannot be read outside of the spectroradiometer on a mac machine (the software that comes with the device runs on Windows only), so you may want to manually extract the information from the device before you turn it off.
- To turn the spectroradiometer off, simply press the power button at the top of the button panel at the back left of the device. Don't forget to return all the settings back to how you found them. Note that the Spectroradiometer is set by default to ask you if you want to save unsaved data to the SD card at the end of a session (I.e. when you turn it off). However, it's best, if you forget to save certain data, to navigate back to the reading immediately and press save. If you want to change the save before quit option it can be found under File Settings in the main Menu.
- For any other information you might need, refer to the User Manual (PR-655 - 670 Manual.pdf) found on the CD that comes with the device.
- Ione Fine has a good explanation of monitor calibration
- OSX has built-in software that uses visual ramps to allow you to estimate gamma. It then creates an .icc monitor calibration file:
- Open System Preferences->Displays->Color. Click "Calibrate" on the display you need an estimate of the gamma for.
- Follow the Display Calibrator Assistant's instructions. Take note of the Native Gamma and Native White Point and don't set a target for either of these settings, simply keep the native settings.
- Psychopy also has a pair of demos to help estimate gamma. Just open and run "gammaMotionNull.py" from the "Demos" menu. The program requires you to press the up/down cursor keys depending on the apparent direction of motion of the bars. At the end of the program, it will allow you to save the data in two files. If you then run the "gammaMotionAnalysis.py" program on the data files (command-click to select multiple files at once), it should give you an estimate of screen gamma (the number labeled "threshold" above the right hand graph). Note. try not to move the mouse after the results have loaded in the "gammaMotionAnalysis.py" program as this may cause the program to crash.
Using the SpectraScan PR670 to do automatic calibration
- Follow the steps above through focusing the PR670 on the screen
- In Psychopy, go to Tools->Monitor Center
- Use the PR670's USB cable to connect it to the computer
- In the Monitor Center, select PR670 and click on "Get Photometer". Hopefully the result will be a message saying it has been found
- Click on "Chromatic Calibration" and follow the instructions so that the spectra of the three guns will be measured and the color space conversion matrices automatically filled in
- Click on "Gamma Calibration". For some reason when I did this, it did not allow me to change the number of values tested from 8 (FJ - a possible solution to this is updating to the latest version of psychopy and restarting the program).
a more-precise gamma correction function (Lum = a+(b+k*i)**g) coming in version 1.65
Using the SpectraScan PR670 to manually take readings and plot gamma curve
Instructions on using this device in general may be found above. In this section, instructions are given on using it to measure deviations in a monitor's gamma with higher accuracy than the above techniques for estimation.
- Download this Media:gamma.zip program and run it through psychopy.
- You'll be confronted with this dialogue box and required to fill in the details. Here are some hints and tips:
- Fill in all the fields, or the program won't work
- The greyscale option will allow you to plot luminances of solid greys of varying shades. The sine waves option is for plotting luminances across different parts of a grating. If all is well they should both be roughly equivalent, but if you're not using gratings in your experiment, stick to greyscale.
- Hot Points only apply to the sine waves option. It allows you to get more readings at particular parts of the grating cycle (e.g. if you want to focus on gamma around the trough) for this first dialogue box you only need to specify how many hot points you'll need. In the next one you can specify where on the sine wave you want it. Note that the maximum number of settable hot points is 3.
- The way the program is written, it's best to follow this guide when selecting the number of different luminances to test: with no hot points, select as many as you like provided it is more than 10. With one hot point stick to 15-20. Two hot points 20-30, and three hot points 25-40 (I.e. 5-10 more than the minimum of 10 for every extra hot point). This is not a hard set rule, you can choose to test more points if you like, but there is likely to be much overlap between points, and thus many discarded points - E.g. you'll ask for 40 but it'll only give you 20 because the other 20 were just repetitions of the same points. It's important to note that since hot points only apply to the sine waves option, if you're testing using greyscale, you can select as many readings as you like. Most importantly, you should know in advance that the number of readings you request is a maximum. You may get less readings than you asked for.
- The rest of the box is self explanatory, except for the Width of Monitor Image which refers to the size in cm of the image displayed on your screen. It is equivalent to the Screen Width (cm) measure that you input to the psychopy monitor center.
- When you press ok, you'll be presented with this aforementioned second dialogue. Actually, depending on how many hot points you asked for in the last dialogue box, your second dialogue box may look a little different. Either way, in each available input box, enter the location of the desired hot point as a proportion of the sine wave function (e.g. 0.0 and 1.0 would be the peaks (whitest points) of the function and 0.5 would be the trough (blackest point). This program will automatically allow you to test a range of extra values around the hot points.
- When you press ok, you'll see the first image. You can then use the SpectraScan PR670 to take a reading of the luminance value (see above), before repeating the process with the next image by pressing the right arrow key to progress. If you accidentally skip an image, you can backtrack by pressing the left arrow key.
- Once you've made your readings with the SpectraScan PR670, you can plot them against the values which are printed in a text file (named "Phases.txt") in the same folder as the psychopy program folder to get the gamma value for the screen (I.e. the spectroradiometer readings are luminance values (y axis) and the list of numbers in the text file are phases of pi (x axis) - the result should look like a cycle of a cosine curve - for technical reasons known only to Jon Peirce, the creator of Psychopy).
Using the Spyder2Pro or Spyder3Elite
After acquiring an estimate of the screen's gamma, the spyders can be used in conjunction with this psychopy program (Media:Test.zip) to test the screen luminance grating.
Instructions: The luminance of the first screen should equal the average of the luminances of the darker and lighter bars in the second and third screens. Press 'q' or 'escape' to quit, 'space' to progress to the next screen
Using Spyder2Pro/Spyder3Elite for spot measurements and gamma confirmation?
- It's best to measure screen luminance in a completely dark room.
- If you need the software code for the Spyder3Elite installation, it's on a post-it note on the front of the installation CD sleeve.
- The third screen in Test.py (with full contrast black and white bars) may give slightly inaccurate luminance measurements for the black bars, so take this into account when calculating the average.
- When removing the Spyder3Elite from a CRT, gently pull the small tab on the suction pad to release the suction.
- If calibrating a CRT and using Spyder2Pro, remove the plastic filter from the face of the device, making the suction pads accessible to the screen. If calibrating a LCD, replace the filter on the Spyder2Pro, remove the suction pad from the Spyder3Elite and use the counterbalance to position the Spyder on the screen.
- Plug the spyder into a USB port directly (not through a keyboard or USB hub).
- Start the Spyder2Pro (Spyder2Pro 2.2) or Spyder3Elite software as appropriate.
- Ignore the wizard which starts up and instead open Tools->Colorimeter (Shortcut: command-M). If using the Spyder2Pro, place the spyder on the screen and click initialise.
- Load the Test.py program code into the psychopy coder window and set the "screen=" parameter of the "win" object to the appropriate screen (0 for the Primary screen (the one with the OSX menu bar on it), 1 for the secondary screen etc.)
- Before you begin, create/modify the profile for the monitor being calibrated in the Monitor Center (Tools->Monitor Center). On the right hand side fill in the gamma values under linearisation with the gamma estimate obtained above. Fill in details about any modifications you made under "Notes", and save the profile before exiting monitor center.
- Enter the name of the appropriate monitor profile as it appears in Monitor Center between the quotes for the "monitor=" parameter of the "win" object.
- Follow the instructions in red (also found above), measuring the luminance by placing the spyder on the appropriate part of the screen and clicking "Take Reading". For luminance in cd per meter squared, use the Y of the CIE xyY
- The luminance of the first screen should equal the average of the luminances of the bars in the second and third screens.
General info about photopic,scotopic
The best reference for luminance values that are photopic (cone vision) versus scotopic (rod vision) is Donald Hood's chapter in Handbook of perception and performance. We have a good table of luminance, retinal illuminance, pupil size, etc. values from that chapter in the lab.
Alex White's experiment
In Alex White's buttonpress experiment we attempted to bring all stimuli down into the photopic range. To do this we got some neutral density filter sheets to put over the screen. We combined 3 filters: two 0.6ND (Lee Filters number 210) filters and one 0.9ND (Lee number 211) filter. This combined filter transmits 0.73% of light.
When using these filters on the Viewsonic G810 monitor, with contrast to max and brightness to min, the dimmest spinning bar (0.2 by 3 degrees visual angle) we could see well on a black background was 0.01 cd/m^2.
To find out if this luminance value is in the scotopic (rod-only) range, we tried to see how well we could perceive color. We did that by comparing that spinning bar with a green and a red spinning bar (painted with only the R and G guns), and adjusted the luminance of the red and green ones to see if those ever were indistinguishable from each other and from the gray bar. They weren't; Alex H, Alex W, and Dani could all still see color even in that range.
Actually for me (AH), as I take the red gun and slowly make it brighter, it seems that as soon as I can detect the bar I can see that it is red. This suggests that the cones are close to the most sensitive mechanism in this condition, probably because these stimuli are parafovea. When I look at them in the periphery just several degrees (after dark adaptation) the stimuli all look much brighter and it's harder to discriminate the colors. This tells us if we want to do a truly scotopic version of Alex's experiment, we'd have to use stimuli that extend farther into the periphery. This probably also would overcome the issue of the bars getting so fuzzy and occasionally invisible when very dim- they were too close to the rod scotoma of the fovea.--Alex O. Holcombe 05:41, 13 June 2008 (UTC)--Alex L. White 06:04, 13 June 2008 (UTC)