Sign up Calendar Latest Topics
 
 
 


Reply
  Author   Comment   Page 7 of 10     «   Prev   4   5   6   7   8   9   10   Next
Geenaianni

Junior Member
Registered:
Posts: 27
Reply with quote  #151 
excellent, thanks. 
0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #152 
* Changes in NIMH MonkeyLogic 2 (Jan 19, 2018)

 - A problem that EyeLink DLLs are not loaded when mlplayer is called before
   monkeylogic is fixed.

 + New adapters, FreeThenHold & SmoothPursuit, are added.

 + A new example task, task\runtime v2\pursuit eye movement, is added.

0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #153 
* Changes in NIMH MonkeyLogic 2 (Jan 23, 2018)

 - The data file recorded with SmoothPursuit is replayable now.

0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #154 
* Changes in NIMH MonkeyLogic 2 (Jan 25, 2018)

 - A memory leak that occurred when the SND objects were created from vectors,
   not from files, is fixed.

0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #155 
* Changes in NIMH MonkeyLogic 2 (Jan 29, 2018)

 + New adapters, TextGraphic and CurveTracer, are added.

 + A new example task, task\runtime v2\curve trace, is added.

0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #156 
* Changes in NIMH MonkeyLogic 2 (Feb 16, 2018)

 + When the 'task_aborted' hook is called in the alert function, the 'task_end'
   hook is called as well subsequently.

 - An error that NIMH DAQ Toolbox falls into an infinite loop in some systems
   and crashes MATLAB with a stack overflow is fixed.
 
 - In the 2-D Spatial Transformation method, only pre-selected fixation points
   (i.e., those that have a number behind) are clickable and can be presented
   to subjects now. This prevents some user errors.

 - minor fixes
0
MR_Riley

Junior Member
Registered:
Posts: 8
Reply with quote  #157 
Quote:
Originally Posted by 胡文娟
Hi Jaewon, I used to run the ML2 ok on the win7 .But ,now I upgrade my operating system to win10,and encounter the problem. If I want to use the " eyejoytrack('touchtarget', object_number, threshold, duration)" to detect the touchscreen(subjectscreen) touchpoint-position ,I must hold touching the touchscreen before the trial begining without releasing during the trial. Once I release the touchscreen the DAQ detect the mouse(controlscreen) position.And I touch the touchscreen again during the trial, it still detect the mouse position but not the touchscreen touchpoint-position.


Jaewon,

We are experiencing a similar issue. In order to detect any input from the touchscreen, we have to press and hold the touchscreen prior to the trial onset. If I let go early, no input is detected after that point. If I press after the trial onset starts, then there's no input detected either.

I've tried the "Run Message Loop" option, but nothing changes with that selected.

I can't testify if the mouse position is detected or not, but it seems very similar to what 胡文娟 describes.

Thank you for any assistance you can offer.
0
kms

Member
Registered:
Posts: 37
Reply with quote  #158 
Hi Jaewon, 

I just installed latest version of ML and loaded the older (20 Sept 17) version of condition file. I get an error: 

You cannot get the 'AdaptorName' property of analoginput.

The settings and channel assignements seem to be the same. What could be wrong?

Thanks.
0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #159 
I need more information. Please post which line of which function you got the error from.

I can test it, if you send me your task with the configuration file (*_cfg2.mat).
0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #160 
In Line 560 of your timing script, there is an asterisk inserted. It causes a syntax error. The task seems to work fine after I deleted it.
0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #161 
* Changes in NIMH MonkeyLogic 2 (Feb 23, 2018)

 + Previously a syntax error in the timing script produced just a warning so
   the task proceeded incorrectly with an empty runtime. Now it stops the task
   with an appropriate error message so that users can correct it.

 - When BlockChangeFunction and BlockSelectionFunction are identical, the value
   that the BlockChangeFunction returns indicates the next block. Previously ML
   tried to set the next block even when this number was 0 and caused an error.
   Now the problem is fixed. (Thanks to Peter Lauro)

0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #162 
* Changes in NIMH MonkeyLogic 2 (Feb 27, 2018)

 + Touchscreen support on Windows 10 is improved. Now touchscreens that use
   Windows 10's native driver should work fine. Since this requires to process
   Windows messages, the "Run Message Loop" must be activated always and
   therefore is removed from the menu. In addition, the "press and hold" is
   disabled programmatically, so users do not need to disable it manually.
   Reading existing configuration files with this version (or later) may
   generate a warning, due to the removal of the "Run Message Loop" menu, but
   it will disappear once the configuration is saved in the new format.

 - Fixed a bug that the SingleTarget adapter does not report the window-cross time
   correctly, when the subject is already fixating on the target, and therefore
   messes up the eye drift correction mechanism.

0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #163 
* Changes in NIMH MonkeyLogic 2 (Mar 8, 2018)

 - The update in Feb 23, 2018 made CurrentTrialWithinBlock start from 0. It is
   fixed.

 - Users can stop the eventmarker 13 from being stamped for skipped frames, if
   it is not necessary.

   TrialRecord.MarkSkippedFrames = false;  % true, by default.

0
crponce

Junior Member
Registered:
Posts: 22
Reply with quote  #164 
Hi Jaewon, quick question. For the past few days, upon exiting experiments from ML2, I lose control of the main Matlab desktop window: the ML2 Matlab GUI remains responsive (I can click on any button, type text in any window, run more experiments); however, if I click on the main Matlab window to work on the scripts, Windows doesn't highlight the title bar and I can't move it, resize it, or type in it. Closing the ML2 GUI doesn't return control - I have to force quit Matlab using Alt-Ctrl-Delete. Upon relaunch, everything is back to normal.

I'm not sure how to debug this one. Any thoughts? As part of my debugging, I re-installed Matlab, and the problem returns upon multiple computer restarts.
0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #165 
Is this a new problem that occurs after something changed? Is there anything else you found unusual? Did the main window look blurry when it happened?

The main window being unclickable and immobilized is typical when there is a modal dialog opened, like a file-open dialog or a color-pickup dialog.  For example, the ML main menu is a modeless dialog, so you can move around the MATLAB command window while the ML main menu is open. However, if you open a color-pickup dialog from the ML main menu, like the one in the "Subject screen background" option, the MATLAB command window will become unresponsive until you close the color-pickup dialog.

I wonder if it is possible that your task opens such a dialog during the experiment and leaves it open somehow. To check this, press ALT + TAB (type TAB while holding down ALT). It allows you to move the focus from window to window. If there is any dialog unclosed, close them (move the focus to it and type ALT + F4) and see if the control comes back to the MATLAB.
0
pschade

Junior Member
Registered:
Posts: 4
Reply with quote  #166 
Hi Jaewon, I am attempting to read in images that are in a different folder than my userloop script by defining the whole path to the images. However, I am getting an error in the command window pointing to mlconditions.m at line 123 but the error is due to the 'find_ext' function at line 199. It searches if  [obj.MLPath.ExperimentDirectory 'path to filename'] exists, but since I've tried to take images from another location, it creates a nonsense path, which does not exist. 

I added a line of code that seems to fix the above issue (it runs); though it may cause others that I am not aware of. 

if ~isempty(e) && 2==exist(n,’file’);  p = n; return ;end

after the two lines that start with ‘p =’
in my stock mlconditions.m file, this turned out to be line 201

Not sure if this is the correct solution but it would be a useful feature to be able to pull images from folders outside of experiment or base directory. 

0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #167 
Hi pschade,

I like your solution and will add it to the next version. Thanks!
0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #168 
* Changes in NIMH MonkeyLogic 2 (Mar 15, 2018)

 - The "I/O Test" couldn't be run, if ML was installed with the zip file and
   the "I/O test" was the first test you ran after starting ML. It works now.

 - In the mlplayer, unused non-visual objects were displayed from Time 0. It is
   fixed.

 + You can use the full path in the conditions file, to indicate the location
   of TaskObject items. (Thanks to pschade from ML forum)

0
W F Asaad

Administrator
Registered:
Posts: 9
Reply with quote  #169 
I recently found out from my lab members that ML2 doesn't save the actual images used as visual stimuli in the data file.  While I understand that this saves space, especially if images are very large, is there some way to have an option to save the images in the file itself, thereby reducing the need for users to manually save the appropriate images for each data file in a particular place (which can be cumbersome and allows for user error)?  The initial philosophy of ML was that everything needed to reconstruct exactly what happened during an experiment should be in the data file.

If there is a way to do this now in ML2, and we're simply unaware of it, would be happy to learn about it.

Thanks, Jaewon!

-Wael
0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #170 
* Changes in NIMH MonkeyLogic 2 (Mar 19, 2018)

 + Now stimulus files used in the task can be saved within the data file. NIMH
   ML does not save all the files defined in the conditions file but the ones
   actually used in the task.
   mlexportstim, a function that extracts saved stimuli from a data file, is
   added to the package.

save_stimuli.png

0
aboharbf

Member
Registered:
Posts: 55
Reply with quote  #171 
Does this new stimuli saving extend to avi stimuli?
0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #172 
Yes, whatever stimuli that are created from file sources, including WAV and MAT, will be saved.

I thought keeping all stimulus files in one directory was not a big deal, but some people rename their stimuli daily not to edit the conditions file, so this feature may be useful to those.
0
W F Asaad

Administrator
Registered:
Posts: 9
Reply with quote  #173 
Thanks, Jaewon!  Saving the stimuli automatically is a huge help.
0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #174 
* Changes in NIMH MonkeyLogic 2 (Mar 22, 2018)

 + A new tracking option, '~touchtarget' is added to eyejoytrack(). This option
   allows you to monitor whether touch is placed outside the target window.

   ontarget = eyejoytrack('touchtarget',1,3,'~touchtarget',1,3,5000);

   % ontarget will be [1 0], if the subject touches the target, or [0 1], if
   % the subject touches outside the target.

 - An error occurred during stimuli saving, when there was no TaskObject or
   only one created from file sources. Plus, mlexportstim was not included in
   the app package. All are fixed.
0
Jaewon

Administrator
Registered:
Posts: 708
Reply with quote  #175 
* Changes in NIMH MonkeyLogic 2 (Mar 23, 2018)

 - When eyejoytrack was called with the 'releasetarget' option (touch input),
   tracking didn't stop upon an early release of the touch. It is fixed.
   (Thanks to Simon Nougaret)

0
Previous Topic | Next Topic
Print
Reply

Quick Navigation:

Easily create a Forum Website with Website Toolbox.