Timing issues

From Center for Cognitive Neuroscience
Jump to: navigation, search

Samantha O'Connell <samroconnell@gmail.com> 10/27/14

to supportteam, Austin Hi EGI support,

We're in the process of transitioning Net Station 5 into our EEG testing and have been running into some issues:

1) We deliver our stimuli and manage our events from MATLAB. With that said, when we send simultaneous events from MATLAB to Net Station, it says in the event log that "events that exist at the same time as existing events will not be added". Is there a way to fix this issue? There are some experiments in our lab that depend in sending simultaneous events and this would present a big problem. This was not an issue that we had with earlier versions of Net Station.

2) The events that we send to Net Station from MATLAB are recorded properly as they occur in time in the event log; however, when opening the recorded file in Net Station Review, all of the events have the same onset time and occur at the end of the recording, though durations of all events are preserved. We haven't had timing issues in the past and are wondering if there is a set up feature involving timing that we overlooked.

If you could help us solve these issues, this would be great.

Thanks, Sam


Mark Moran <mmoran@egi.com> 10/28/14

to me, supportteam, Austin Dear Samantha,

Are you using psychtoolbox? We are investigating a timing issue with the synch method used by Psychtoolbox. Do you have a script known to work with an older version of NS I could test with

I'll need to investigate a bit more on how you were doing simultaneous events before. That was during acquisition I assume?

Can you remind me which lab you are at?

Thanks,

Mark


Samantha O'Connell <samroconnell@gmail.com> 10/28/14

to Mark, supportteam, Austin Hi Mark,

Thanks for your reply!

Yes - we are using Psychtoolbox. I'll see if Austin (the one having trouble with the simultaneous events) can send you some code. And yes - we were delivering simultaneous events during acquisition.

We're in Mark Cohen's lab at UCLA.

Send you some code in a bit.

Thanks, Sam


Mark Moran <mmoran@egi.com> 10/28/14

to me, supportteam, Austin Dear Samantha,

I'll need to investigate the issues separately, so if you have an experiment known to run in 4.5 that does not have simultaneous events, that would be perfect.

Thanks,

Mark


Samantha O'Connell <samroconnell@gmail.com> 10/28/14

to Mark Oh ok - great. Sounds good.


Austin Head <alheadbme@gmail.com> Attachments10/28/14

to me Here's the script. Hold off on sending it; I want to confirm that it works in 4.5 first.

Attachments area Preview attachment testNS5.m

Binary File testNS5.m


Austin Head <alheadbme@gmail.com> 10/28/14

to me Use this version. It works in NS 4.5 just fine.


Samantha O'Connell <samroconnell@gmail.com> 10/28/14

to Austin Perfect! Thanks Austin!


Samantha O'Connell <samroconnell@gmail.com> 10/28/14

to Mark, supportteam, Austin Hi Mark,

Here's a code that works fine in NS 4.5 that's been having timing issues in NS 5. Let us know how it goes.

Thanks, Samantha​

testNS5.m

Attachments area Preview attachment testNS5.m

Binary File testNS5.m


Austin Head <alheadbme@gmail.com> 10/28/14

to me I'd like to point out another issue. Whenever the recording stops, NS forces you to enter a new patient ID before you can interact with the program again. This is very disruptive, particularly when a user does not intend to save new data. If I have finished my session, I shouldn't have to enter a new dummy patient ID (e.g. "asdf") just to look scroll down.

This behavior may be a design choice to prevent users from accidentally running a session without entering the patient ID, but there is a bigger problem. The prompt for the patient ID causes Net Station to remain as the topmost window even if users click to other programs (including Finder, Safari, System Preferences, etc.). The program should never capture focus in that manner. I discovered this bug because I was trying to open Finder to determine the next unused subject number, but the window was not visible. I was forced to enter a placeholder ID so that I could use Finder before re-entering the correct ID.



Austin Head <alheadbme@gmail.com> 10/28/14

to me Do you want me to email him directly?



Samantha O'Connell Yeah, that would be best. Mind sending your email to him? 10/28/14

Austin Head <alheadbme@gmail.com> 10/28/14

to me Sure. I'll email him.


Mark Moran <mmoran@egi.com> 10/28/14

to me, supportteam, Austin Dear Samantha,

Thank you, That's just what I needed.

Thanks,

Mark


Mark Moran <mmoran@egi.com> 10/29/14

to me, supportteam, Austin Dear Sam,

I have verified your code works in 4.5 but not in 5.0, I've turned the issue over to the programmers and will let you know what a fix is available.

Thanks,

Mark


Sam O'Connell <samroconnell@gmail.com> 10/30/14

to Agatha I guess I can try your code and see if we're running into similar issues, but I'm guessing so because of PTB

Sam O'Connell samroconnell@gmail.com (847) 915-5355

Staff Research Associate Staglin Center for Cognitive Neuroscience, UCLA www.brainmapping.org Think Now, Inc. www.think-now.com

Begin forwarded message:

From: Mark Moran <mmoran@egi.com> Date: October 29, 2014 at 3:49:46 PM PDT To: Samantha O'Connell <samroconnell@gmail.com> Cc: "supportteam@egi.com" <supportteam@egi.com>, Austin Head <alheadbme@gmail.com> Subject: Re: NS 5 events


Agatha Lenartowicz <alenarto@ucla.edu> 10/30/14

to me Yes pls do. You ought to get Andrew on this as well.


Austin Head <alheadbme@gmail.com> 11/3/14

to Mark, me, supportteam Dear Mark,

I'm a doctoral student working with Mark Cohen. Thank you for looking into our event problem. I'd like to share a few general comments on the new software as well as an interface issue that I've encountered.

The interface feels much more cohesive and organized now that the various controls have moved from separate windows to a single sidebar. I'm glad that I no longer need to drag things around to search for a lost dialog box. This is a great improvement.

I'm a big fan of the separate icons for Acquisition/Review/Tools. I also like the "Not Recording" text that shows up over the EEG waveforms; that indicator makes it so much easier to quickly check that things are working. I've noticed small touches like these that give the program a greater sense of polish.

I do have one issue with the new interface. When the program first starts and whenever the recording stops, Net Station forces you to enter a new patient ID before you can further interact with the software. I assume that this behavior is intended to prevent users from accidentally recording a session without a patient ID. However, this can be rather disruptive, particularly when a user does not intend to save new data. For example, if I want to check the TCP/IP event settings that were in place for a recording that just ended, I first have to enter a patient ID before I can scroll through the control interface, even if I will not record another session. Personally, I would prefer a warning like the "Not Recording" text instead of a required dialog box.

I recognize that the mandatory patient ID dialog box may be a design choice, but it also causes a bigger problem impacting the operating system as a whole. The prompt for the patient ID causes Net Station to remain as the topmost window even if users click to other programs (including Finder, Safari, System Preferences, etc.). I discovered this bug because I needed to check a file to determine the correct patient ID. However, Finder opened in a window hidden behind Net Station. The top panel showed that Finder was the open application, but there was no way to bring the Finder window into view. My only option was to enter a placeholder patient ID so that I could use Finder before re-entering the correct ID. Perhaps you are aware of this issue, but I wanted to point it out just in case.

Let me know if you would like any more feedback on these or any other issues.

Best, Austin Head


Mark Moran <mmoran@egi.com> 11/4/14

to Austin, me, supportteam Dear Austin,

Thank you for the feedback.

Yes what to do about the patient ID is an open issue, you're not the only person who does not like the current implimentation. I will log your comment into our feature request tracking system

I'll also log the detail about NS blocking the finder.

Thanks,

Mark


Samantha O'Connell <samroconnell@gmail.com> 11/5/14

to Andrew This is to keep you in the loop. We'll need your help


Samantha O'Connell <samroconnell@gmail.com> 11/12/14

to Mark, Austin, supportteam, Agatha, Andrew Hi Mark,

Thank you for your help with the sync timing we've been experiencing with NS5.

I wanted to check in on the progress of testing out our code with NS5. We're keen to keep working with NS5, but won't be able to make any progress without figuring out this road block.

Please let us know at your earliest convenience.

Thanks, Sam


Samantha O'Connell <samroconnell@gmail.com> 11/17/14

to Mark, Austin, supportteam, Agatha, Andrew Hi Mark,

Before we get hit with the holiday season, wanted to check in again and see if there's any updates on the sync issues we were having with NS5 and Psychtoolbox. Please let us know - we're eager to continue testing out the new program

Thanks, Sam


Gaynor Foster <gfoster@egi.com> 11/17/14

to me, Mark, Austin, supportteam, Agatha, Andrew Dear Sam,

I am sorry but we do not have an update for you at the moment. We understand the need for this fix, and will provide you with an update as soon as it is made available to us.

Best Wishes, Gaynor


Return to EGI main page