DDPI MK1

DDPI MK1 refers to the experimental rig that currently sits in room 266. This system runs EyeRISM, which is a version of "old" EyeRIS, the lab's stimulus display software, "hacked" to also run an image-based DPI tracking algorithm developed by Ruei-Jr Wu. The camera, and thus the tracking algorithm, runs at 340Hz (changed on January 2021, previously was at 330 Hz).

New eyeris (linux) & mk1 DDPI_MK1_ChangeLog

Link to filling out notes/problems form https://docs.google.com/forms/d/e/1FAIpQLScdX1puBFLEQgGfEkHb_AOCTXKjxN88rh0lHg_8BgaaPxDrLg/viewform?usp=sf_link

Spread sheet of notes https://docs.google.com/spreadsheets/d/13FcWE_Q2GnNrF5maUB7KfO_pDLjLaxHwEiuJoWGErRM/edit?usp=sharing

Record of Recordings

Approved Users

Expert: If it’s broken, you can figure out why and fix it. You can align the optics when out of alignment.

Advanced: You can collect high-quality data with the DDPI in a variety of circumstances. You can typically determine the cause of issues that may arise during data collection. You can train novice users and supervise beginners.

Normal: You can reliably use the DDPI without damaging it or causing issues for other users. The majority of your data is high-quality.

Beginner: You MUST have an Advanced or Expert user with you when setting up the subject and beginning tracking, i.e. the 1st 10 min:

Novice: You should be shadowing to learn the DDPI.

Experimenter Instructions for DDPI

Look here for dDPI Protocol: https://docs.google.com/document/d/1g-ghOYonxl1CF9SxTAMJS5ZEkB1XDgg2ZsW_I8mPVr8/edit?usp=sharing

Powering Up

  1. Turn on the power strip underneath the table (Supp Figure 1).

  2. Toggle the joypad (X -> D -> X). (MAKE SURE TO DO THIS IF YOU HAVE JOYPAD ISSUES)

  3. Open your experiment project in Visual Studio.

Setting Up dDPI

  1. Run experiment project with ExpDDPICalibration
  2. Two windows will pop up – press Enter both times
  3. You will now see the experiment console and 5pt grid on subject's screen
  4. Use experiment console to move stage to center of dynamic range on all axis - Artificial Eye Set Up: This step is optional for advanced and expert users.
    1. Get artificial eye from top left compartment in toolbox. b. Get short (150mm) rail with screw from the bottom of tool box. c. Screw this rail into the top platform of the stage through marked hole. d. Slide artificial facing monitor onto the rail. e. Adjust X Z until the 1st and 4th image are focused in camera. If image is too low/too high, change tightness of rail screw. f. Slide on 2-connected pinholes and laser onto far rail point towards the artificial eye. g. Move stage using experiment console so that the artificial eye is inline with the laser. h. Now you can remove the artificial eye.
  5. Re order pinholes and laser so that the laser is facing the monitor and projecting through the pinholes.
  6. Align monitor using the center point displayed from the DDPICalibration Experiment.
  7. [optional] Check view distance with laser, screen size, and resolution
  8. Remove laser and keep pinholes on rail for subject.
  9. Make sure subject baby camera is aligned to face the dichro and subject
  10. Roll out experimenter monitor and table, along with baby camera monitor.
  11. Set up computer camera to face subject monitor.

Setting Up the Subject

Finding the Subject's Eye

Assuming that experiment console is still open and ExpDDPICalibration's 5pt grid is visible. If not, see steps 1-3 in "Setting Up dDPI"

  1. Open camera app and make sure that you can see subject monitor.
  2. Click "Monitor" to turn on camera view. NEVER close with window X, always click "Monitor"

  3. Use the Up/Down, Far/Near, and then Left/Right buttons to position the stage so that you can see the 1st and 4th (gross adjustment at this point)

  4. Have the subject sit in and gaze at center point, again
  5. Use Far/Near buttons to focus image first. This ensures that you won't hit the limit of the stage.
    • - If at this point you can not find the eye, use the artificial eye again.
  6. After focusing, use the Up/Down, and Left/Right buttons to position the stage, focusing on getting a clear 4th reflection .
  7. Verbally instruct subject to look at grid points and confirm 4th reflection is equally focused at each point and stray reflections (eye lash, scelara, etc.) are outside the field of view for all eye positions. If not, continue to adjust. ***The center point will have the most focused reflection.

  8. Turn on tracking, adjust Intensity and Radius parameters until the 4th reflection is correctly identified and the score in the top left corner is as small as possible (0.4 or below, ideally 0.2/0.3).

  9. Make sure IR light source is turned on to max.

  10. Click on experiment window. This is critical, or else you will experience intense frame skipping.

  11. Press space bar for dDPI calibration, subject should follow dot and try not to blink

  12. Click "q" or "Quit Experiment" to proceed to auto calibration. During the experiment DO NOT click on any other window – the PC must stay focused on the experiment to maintain the correct refresh rate

  13. If you re adjust focus during the experiment:
    • - Have subject maintain fixation during a re calibration trial (this will prevent corrupting data by moving the eye tracker - Adjust accordingly - Click back on monitor window - Tell subject to continue

Shutting Down

  1. Turn off camera using the power strip after experiment
  2. Move data to opus
  3. Shut down computer if you are the last one within 4hrs of using the machine.
  4. Cover machine with plastic tarp to prevent dust.
  5. If you encountered any issues during your session, log them on gitlab and post them to slack.

Notes and Tips

Cleaning the Dichromatic Mirror

This is one of the most expensive parts of the system, so be careful! Wear gloves to avoid getting oils from your skin on the mirror. Use compressed air to blow off dust and dirt. Inspect for dirt, grit, etc. You do not want to drag any of these across the mirror, so spot clean these first if needed. Apply alcohol or acetone to mirror. Wipe off gently with a lens wipe (found with the optics supplies) in a single left/right or up/down direction (not in a circular pattern). DO NOT USE ALCOHOL WIPES, PAPER TOWELS, OR ANYTHING OTHER THAN LENS WIPES.

DDPI Issues & Development

Issue Board

Bugs are posted and resolved on the EyeRISM Issue Page. Any discussion etc should be on the slack channel #ddpi-mk1

Software

See EyeRISM repo on Git Lab.

Where to Save Data

As discussed, all experimenters will save a copy of their data for debugging and development purposes on BOX. Specifically, use the folder dDPI_Testing. Put your data in a unique folder and include a note about what you ran, online or offline observation, etc.

Online Detection

As of 5/6/2019

Patch for 2-Tag EyeRIS on DDPI:

P4 - Present

P4 - Absent

P1 - Present

No Track

P1 - Absent

No Track

Blink

If using custom Debug mode:

P4 - Present

P4 - Absent

P1 - Present

No Track

P1 - Absent

Out of Range

Blink

Additional Info

Experimenter manual

instructions to port code from DPI

Video instructions on running dDPI

Contents of the D: drive of the current DDPI PC (excluding .avi, .dat, .eis, .fig, .mat, .mov, .pdf, and .zip files) have been copied to the Windows side of the "ARGUS" PC, which will be the new dual-boot DDPI PC.

DDPI MK1 as of Dec 2020 (and before):

Jumps Investigation

Rotation stage for DDPI artificial eye

A new rotation stage from PI was purchased in late 2020 to run artificial eye tests with the DDPI artificial eye. The DDPI artificial eye is much heavier than the DPI artificial eye, which makes the old DPI galvo unstable and not able to reliably run the dynamic artificial eye tests. The part number for this new stage is M-116.DGH.

Development of the protocols and procedures for this stage are currently underway (2/11/21, BM). Details of the development are here.

Equipment-DDPIEyeTracker (last edited 2022-07-29 15:00:10 by SamJenks)

MoinMoin Appliance - Powered by TurnKey Linux