Forum
Register Calendar Latest Topics
 
 
 


Reply
  Author   Comment   Page 3 of 4      Prev   1   2   3   4   Next
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #51 
* Changes in NIMH MonkeyLogic 2 (May 25, 2017)

 + mlbhv2.m caches the positions of the variables in the file and improves the
   reading speed after the first access.

 ~ The problem that structure arrays are read backward from the data file
  (bhv2, h5) is fixed.
0
aboharbf

Junior Member
Registered:
Posts: 16
Reply with quote  #52 
Quote:
Originally Posted by Jaewon
There was a network problem in NIH. It seems that anonymous FTP still doesn't work.


So How do you recommend acquiring the latest stable build of NIMH ML2?
0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #53 
I just emailed the system administrator. Hopefully the issue will be fixed soon. They said the main issue was resolved, but might forget to turn on the anonymous FTP.
0
aboharbf

Junior Member
Registered:
Posts: 16
Reply with quote  #54 
Alright, Thanks a lot for the files!
0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #55 
* Changes in NIMH MonkeyLogic 2 (Jun 14, 2017)

 + The trial replay tool, mlplayer, is added.

 ~ The problem that the conditions are not refreshed when the block is switched
   is fixed (Thanks to Sheena from the ML forum).

mlplayer.png   

0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #56 

* Changes in NIMH MonkeyLogic 2 (Jun 27, 2017)

 ~ Movie-related codes are refactored. Decoding is twice faster now.
 
 ~ mlplayer can handle repositioning of visual objects, frame manipulation of
   movies and background color changes. It is required to record data with this
   version of ML (or later), to get those features replayed correctly.

0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #57 

* Changes in NIMH MonkeyLogic 2 (Jun 30, 2017)

 ~ When stimuli are pre-loaded with the userloop and not turned off by the end
   of a trial, they can show up in the next trial. To prevent this, pre-loaded
   stimuli are deactivated at the end of each trial.
   
 ~ The problem that SND created from wav data, not from the file, does not play
   is fixed.

0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #58 
* Changes in NIMH MonkeyLogic 2 (Jul 5, 2017)

 + The I/O test panel is added. This panel is accessible from the main menu and
   users can test analoginput (eye, joystick, touch, etc.), stimulation and
   TTL with it.

 ~ Minor fixes

iotestmenu.png  iotest.png 

0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #59 
* Changes in NIMH MonkeyLogic 2 (Jul 7, 2017)

 + The main menu GUI can be collapsed so that ML2 can be run on a smaller
   screen (1024 x 768).

 ~ Minor fixes

collapse.png 

0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #60 
* Changes in NIMH MonkeyLogic 2 (Jul 12, 2017)

 ~ The imresize() function used in the recent releases is replaced with
   mglimresize(), which is not dependent on Image Processing Toolbox.
 
 ~ NIMH daqtoolbox prints out more detailed information for NI-DAQmx errors.
 
 ~ The default input range of analoginput is changed to [-10 10] for better
   compatibility with MATLAB DAQ.
0
crponce

Junior Member
Registered:
Posts: 16
Reply with quote  #61 
Hi Jaewon, loving these latest releases. Data is flowing, so thanks for your work.

Quick question: when loading a previous configuration file (cfg2), is there any way to reset the previous eye calibration settings, specifically the 2-D spatial transformation values?
0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #62 
Hi crponce,

How would you like to reset? Do you want to go back to the very initial settings I put in (9 calib points, 5 deg apart)? Or do you want to keep your own calibration points but just clear up the associated voltage values?

You can overwrite the previous settings by simply performing the calibration again. By the way, the recent releases have a problem that the origin of the transformation was not consistent. I will upload a new version next week after some testing.
0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #63 
* Changes in NIMH MonkeyLogic 2 (Jul 17, 2017)

 + Image Processing Toolbox is not requied any more, even when the 2-D Spatial
   Transformation is used for the calibration.

 ~ The origin-gain calibration supports the rotation manipulation, to
   compensate the distorsion in the input signal due to, for example, the
   tilted eye camera.
 
 ~ All calibration methods, including 'Raw Signal', allow the temporary
   rotation of the space with the following command. This change is reset in
   the next trial and, for the origin-gain calibration, its effect is
   additional to the rotation change during the calibration.

   EyeCal.rotate(degree);
   JoyCal.rotate(degree);
0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #64 
* Changes in NIMH MonkeyLogic 2 (Jul 19, 2017)

 + The pause menu items are mouse-clickable now.

 + A button to reset eye/joystick calibration is added to the main menu.

 + When the calibration tools are closed without saving the changes, a question
   dialog will be pop up to confirm.
 
 ~ A testing code accidentally inserted in the sound module is removed.
0
crponce

Junior Member
Registered:
Posts: 16
Reply with quote  #65 
Hi Jaewon, just noticed something new with build 85: pressing 'C' to correct an eye position offset during runtime introduces a large offset in subsequent trials (up and to the left, if that helps). Pressing 'C' again corrects the offset during the new trial, but it returns on the next trial again

I was only able to correct the offset by re-starting eye calibration...had to do it several times today.

Otherwise it runs pretty well
0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #66 
Hi crponce,

Do you happen to have a data file saved when it occurred? Can you send it to me?
0
crponce

Junior Member
Registered:
Posts: 16
Reply with quote  #67 
yep, any email in particular?

edit: just sent it to your nih account
0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #68 
* Changes in NIMH MonkeyLogic 2 (Jul 24, 2017)

 ~ The problem of the space origin floating when the slip correction is made by
   the 'C' key during the 2-D spatial transformation calibration is fixed.
   (Thanks to Carlos Ponce)
0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #69 
* Changes in NIMH MonkeyLogic 2 (Aug 25, 2017)

 + reward_function.m supports multi-channel reward devices via multiple digital
   lines by default. To trigger a different reward channel, use the 'juiceline'
   option as below.
   
   goodmonkey(100,'juiceline',2);      % trigger Reward 2 for 100 msec
   goodmonkey(100,'juiceline',[1 3]);  % trigger Reward 1 & 3 simultaneously

 + Now gray-scale bitmap files can be used. (Thanks to Xiaomin Yue)
 
 + Now the ML main menu shows up on the screen where the command window is
   located (Thanks to Sebastien Ballesta)
   
 + Quotation marks (double or single) added to each field in the conditions
   file are processed properly now. (Thanks to hoptman from ML forum)
   
 ~ Minor fixes
0
kms

Junior Member
Registered:
Posts: 27
Reply with quote  #70 
Thank you, Jaewon! 

A quick question on using multiple channels for reward in the Aug 25 version. I have selected 4 lines from the main menu for reward (port 2, channel 4:7). Does that automatically correspond to Juiceline 1:4 such that specifying "juiceline, 1" in the goodmonkey function corresponds to sequentially the first channel assigned to reward (port 2, channel 4 in this case) & so on?

I would like to assign "juiceline,1" to (port 1, channel 0) and the other three juicelines to (port 2, channel 5:7). Is it feasible?

Thanks.
0
crponce

Junior Member
Registered:
Posts: 16
Reply with quote  #71 
Hi Jaewon! We're expanding our use of ML2 with new monkeys and computers, but suddenly we've run into a new problem.

- After we enter a new value in the "Subject name" box for a new monkey, preserving all other field entries in the GUI, we go to "Eye Calibration" and press "Reset." We then press "Calibrate Eye" to bring up the calibration window. Instead, we get an error message:  

<<< MonkeyLogic >>> Invalid input argument of type 'double'. Input must be a structure or a Java or COM object. (mlcalibrate_spatial_transform/copyfield, Line 677)

This also happened when we installed ML2 in a new computer and populated all GUI fields by hand. The only way around it is to load another subject's default preferences and to modify that subject's eye calibration values.

Any input would be greatly appreciated!

NINJA EDIT: just tried the new version (Sep 20, 2017), same problem...

0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #72 
* Changes in NIMH MonkeyLogic 2 (Sep 20, 2017)

 - The problem that the mlplayer screen doesn't come back when "Export to AVI"
   is cancelled is fixed.

 - In previous versions, movies are rewound automatically when they are toggled
   off so that they can be reused in the same trial. This introduces a
   unnecessary delay when the movie is big. Now users have to explicitly rewind
   them with rewind_movie().
   
 - When a new configuration profile is created for a new subject in previous
   versions, the menu settings are copied from the last subject's profile, but
   the editable variables of the new profile are reset to the values defined in
   the timing file. Now the editable variables are also copied from the last
   subject's profile.
   
 - Minor fixes
0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #73 
Hi kms,

The answers to your questions are both yes. The lowest line of the lowest port is assigned to juiceline 1. Then the next line in the same port, if there is any, or the lowest line in the next port gets juiceline 2.
0
Jaewon

Administrator
Registered:
Posts: 380
Reply with quote  #74 
Hi crponce,

Please try the package I just uploaded. If I am right, the problem should be fixed.
0
kms

Junior Member
Registered:
Posts: 27
Reply with quote  #75 
Thanks, Jaewon! Unfortunately, I am unable to assign channels from different ports to 'Reward' - for e.g. juiceline 1 to port 1, channel 0 and juiceline 2:4 to port 2, channel 5:7 . It seems only one port can be selected at a time while assigning. Is that easily fixable?

Thanks.


0
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.