Errors running two movie objects back to back (white screen and audio problems for 2nd video)

[Katie Jankowski]

I created an EPRIME task for an fMRI scan (timing is sensitive) but I am having troubles running my video stimuli. In each trial, participants view a fixation cross (text slide, 12000ms), a block introduction (text slide, 2400ms) , video1 (movie object, ~12000ms), and video2 (movie object, ~12000ms). Since I am using a block-design fMRI task, not event-design, I want to treat Video1 and Video2 as 1 block. Thus, they are displayed back to back.

Each video stimuli includes both audio and visual information. The durations of each video are variable (approximately 11000ms-14000ms, some 11500ms, others 13500ms, others 12000ms, so I included a duration attribute [VidDur1] and [VidDur2] for video1 and video2. respectively). The the final "block" (video1 and video2 combined) is always 24000ms.

I am having several issues running the movie files. Video1 runs well, but video2 has both visual problems (blank white screen) and audio problems (starts about 2s into the clip, then after about 2s starts from the beginning at 0s, then runs through the audio but ends early). I think my problems might be related to preloading the stimuli, but I don't know what to do. I have been emailing support for about 2 weeks without coming to a solution, so I thought I would ask the google group. I have also looked on the knowledge base, but I have not found a solution.

Currently, pre-release for video1 is set to 5000ms and pre-release for video2 is set to 0ms (but I have also tried 5000ms for video2 with similar results). Start and stop position are 0, stop after is set to "yes", stop after mode is set to "nextonsettime", end movie action is "none", duration is "[VidDur1]" and "[VidDur2]", and timing is "event". Under experiment properties, "flip enabled" is selected and both "display" and "sound" under "devices" are selected.

I would greatly appreciate any advice!

(As an aside, if I set "duration" to 12000ms for video1 and "duration" to 12000ms for video2 (instead of relying on an attribute to represent the duration, I have the same problems. If I set the duration to "infinite" for video1 and "infinite" for video2, only video1 runs, regardless if "stop after mode" is set to "offsettime" or "nextonsettime".)

[David McFarlane]

Movies make life difficult for any experiment software I know. Often
the problem comes from the movie files, not necessarily the
experiment software. Remember that how a movie file works in Windows
Media Player or any other software tells you nothing about how that
file will work in E-Prime. So first you need to find out whether the
problem pertains to your movie files, or to E-Prime.

I would start by making a simple E-Prime program that does nothing
but play my movie files, in Sequential order. Then I would try
running in Random order. If all my movie files succeeded under those
conditions, then I would incrementally change my test program (or
actually, a copy of my test program) to make it more and more like my
experiment program, hoping to sneak up on the culprit that breaks my
experiment program.

Oh, I would also use the Codec Config utility that comes with EP2.

Just some general troubleshooting advice there.

-----
David McFarlane
E-Prime training
online: http://psychology.msu.edu/Workshops_Courses/eprime.aspx
Twitter: @EPrimeMaster (https://twitter.com/EPrimeMaster )

/----
Stock reminder: 1) I do not work for PST. 2) PST's trained staff
take any and all questions at https://support.pstnet.com , and they
strive to respond to all requests in 24-48 hours, so make full use of
it. 3) In addition, PST offers several instructional videos on their
YouTube channel (http://www.youtube.com/user/PSTNET ). 4) If you do
get an answer from PST staff, please extend the courtesy of posting
their reply back here for the sake of others.
\----

[Katie Jankowski]

Thank you very much for these suggestions. I had downloaded the latest version of DivX (http://www.divx.com), selected create default mappings, and rendered the first two files (video1 and videos2 of the first trial) and received no errors. This lead me to believe that my problem was not related to the codecs.

To confirm that the problem originated from my task, and not the video files, I made a simplified version of the task that included just 1 trial (fixation, intro slide, video1, and video2). When I ran this simplified version with just 1 trial, video1 ran fine, but video2 had the same errors (no visual, audio problematic). Next, I switched the name of the specific files listed in the trial proc so that the videos that originally played for video1 now played for video2, and visa versa. Again, only video1 worked and video2 had problems. This lead me to believe that the video files were not the problem, and the problem was the script (or script properties) which effected how video2 played.

I took your advice and made a simple script (tester 1) which had 1 object (movie object video1) that played videos one after the other with a set duration of 12000ms (keeping the prerelease 5000ms). Each video played, starting at the beginning of the video (0s), but each video cut off after ~6000ms. Since the duration was set to 12000ms and the prerelease was set to 5000ms, could the pre-release, or how the video is loaded, be the problem?

I made another simple script (tester 2) which only played videos one after the other (same as above), but this time I kept the attribute VidDur1 to specify the duration of each video (keeping the prerelease 5000ms). Similar to the above, each video played, starting at the beginning of the video (0s), but each video cut off after ~6000ms. The only difference between this simplified script with a duration attribute and the above simplified script with a set duration of 12000ms was that instead of immediately moving onto the next video, the preceding video cut off and froze for ~5000ms before moving onto the next video.

Next, I investigated if the problem was caused by the prerelease setting. I used the tester 2 script (which included the duration attribute) and changed the prerelease for the movie object to "0". This resolved the problem- all videos played back to back correctly (good visual, good audio). However, when I changed my original script (2 movie objects back to back in the same trial) to have the prerelease set to 0 for video1 and video2, I continued to have problems playing video2. So, it seems like the problem is related to how video2 is loading/playing after video1 plays, when 2 movie objects are presented consecutively in the same trial.

Would it be helpful to attach my script?

Thank you for your assistance

--
You received this message because you are subscribed to a topic in the Google Groups "E-Prime" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/e-prime/LHq3Niv1zfk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to e-prime+unsubscribe@googlegroups.com.
To post to this group, send email to e-p...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/e-prime/5339c347.8770320a.7697.55a1SMTPIN_ADDED_MISSING%40gmr-mx.google.com.

For more options, visit https://groups.google.com/d/optout.

[Katie Jankowski]

One other question and update: Does EPRIME have problems presenting 2 movie objects in the same trial?

The reason I ask is because I thought the problem was that the second video didnt have enough time to load immediately following video1. So, I made a simplified script where each trial included just a fixation, video1, and video2; however, I rearranged the order to be: video1 (~12000ms), fixation (~12000ms), video2 (~12000ms). Here, I made sure that all objects had prerelease set to 0.. However, video2 still had audio and visual problems.

Finally, I thought perhaps that I had set the properties incorrectly for the video2 movie object. So, I copied the video1 object (which runs correctly) and renamed it video2. However, when I ran this new script,  video2 still had audio and visual problems.

In summary:

In a simplified script, where each trial includes only 1 movie object, so that all movies are presented back to back, the videos run correctly if the pre-release is set to 0. (Setting the duration of each video using an attribute "viddur1" does not affect the script).

However, in another simplified script, where each trial includes only 2 movie objects (whether these movie objects are either presented back to back or separated with a fixation cross slide), the 2nd video always has visual and audio problems. (Changing the prerelease from 5000ms to 0s does not affect the script. Setting the duration of each video using an attribute "viddur1" vs including a set duration for all videos does not affect the script.)

Thank you,

Katie

[David McFarlane]

To answer your first question, I am pretty sure that E-Prime can run
2 or more movies per trial, but I do not recall exactly if I have
ever tested that. Perhaps I should.

Next, please see http://www.pstnet.com/support/kb.asp?TopicID=3299
for "gotchas" that may happen with the new GeneratePreRun
feature. (You may also look at the following
threads: https://groups.google.com/d/topic/e-prime/IV1KWxm3Q-U ,
https://groups.google.com/d/topic/e-prime/Azi36c7sXSI ,
https://groups.google.com/d/topic/e-prime/0Yqv3FHXm6k , and
https://groups.google.com/d/topic/e-prime/Q6si9tksg4Q .)

Since you already set PreRelease to 0 in your demos (nicely done),
those gotchas should not hit you. Still, just for fun, open the
Properties Pages for your Procedure, and set Generate PreRun to
"BeforeObjectRun" (I assume that objects in your Procedure have their
Generate PreRun set to the default of "Inherit"). Then try your
demos and see what happens.

Beyond that, I have no ideas, other than trying to encode your movies
in another codec (see links above). We had a case here of a very
simple E-Prime program that kept intermittently balking with movies
(running merely *one* movie per trial), with no apparent pattern. We
gave up and redid the program in PsychoPy, and still had
problems. The student finally recoded the movies using another codec
(sorry, don't remember which one), and then everything worked in
PsychoPy (and I bet would also work in E-Prime now). So as I said
earlier, when you use movies with high-performance software, you are
asking for trouble. Sorry,

-- David McFarlane

>latest version of DivX (<http://www.divx.com/>http://www.divx.com),

><http://psychology.msu.edu/Workshops_Courses/eprime.aspx>http://psychology.msu.edu/Workshops_Courses/eprime.aspx
>Twitter: @EPrimeMaster
>(<https://twitter.com/EPrimeMaster>https://twitter.com/EPrimeMaster )

>
>/----
>Stock reminder: 1) I do not work for PST. 2) PST's trained staff
>take any and all questions at

><https://support.pstnet.com>https://support.pstnet.com , and they

>strive to respond to all requests in 24-48 hours, so make full use
>of it. 3) In addition, PST offers several instructional videos on
>their YouTube channel

>(<http://www.youtube.com/user/PSTNET>http://www.youtube.com/user/PSTNET

[Katie Jankowski]

Hi David,

I apologize for the delay.

Thank you for your helpful advice!

Changing the trial proc properties from "topofprocedure" to "beforeobjectrun" solved my problem! (Note, I changed the trial proc properties, not each objects' properties, which seemed to work better for me). In fact, it sounds like changing the procedure properties affects all the objects properties within the procedure, when set to inherit (new features guide on pg. 18), although I am not positive. In my full script, all movies ran without errors and both the auditory and visual output were presented.

Also, I just found this post on the EPRIME knowledge base, which sounded very similar to what I was experiencing: "NOTE: Use of OffsetTime StopAfterMode option with any PreRelease value including the default (same as duration) will likely cause no sound or only a brief portion of the sound to occur. " Setting the prerelease to 0 sounds like a good fix.

My only remaining concern is that the last frame in the 1st movie of each trial freezes, while the 2nd movie (last trial object) does not. If you have any suggestions for how to avoid the last frame of the movie from freezing and achieving a smoother transition, please let me know. If I figure it out, I will post it here. One thought is perhaps it is related to the "stopaftermode" set to "nextonsettime". I dont know if this is ok, if it should be set to "stopoffsettime", or if an inline script is necessary, as mentioned here.

Again, thank you for your help.

In summary, I changed the prerelease settings to "0", changed generate prerun to "beforeobjectrun", and made sure my script was saved as an EPRIME Professional file (I have Professional but I was saving the script as EPRIME Standard.) Some changes may or may not have been necessary, but I wanted to list them all here, in case some people have similar problems in the future.

--
You received this message because you are subscribed to a topic in the Google Groups "E-Prime" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/e-prime/LHq3Niv1zfk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to e-prime+unsubscribe@googlegroups.com.
To post to this group, send email to e-p...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/e-prime/533ad005.2327320a.4496.69ebSMTPIN_ADDED_MISSING%40gmr-mx.google.com.

[David McFarlane]

Glad you got it (mostly) sorted, and thanks for writing back.

Perhaps I can explain a few things here...

First, "End Movie Action" specifies what action the MovieDisplay
object will take in the event that the movie file reaches its end
before the MovieDisplay object does. "(none)" means, well, just
continue as if nothing happened; "Terminate" means to terminate the
MovieDisplay object (not clear to me what happens when this happens
during the PreRelease period, perhaps someone could test that); as
for "Jump", well, let's just save that for another discussion (it gets tricky).

"Stop After", by contrast, indicates whether or not movie playback
will be stopped when the event indicated by "Stop After Mode"
occurs. With "Stop After Mode" set to NextOnsetTime, movie playback
would be stopped at the OnsetTime of the next object; set to
OffsetTime, it would be stopped at the OffsetTime of the MovieDisplay
object. Using a PreRelease with "Stop After Mode" set to OffsetTime
means that the movie playback might be stopped while the next object
is still waiting to start, so usually you would keep this set to
NextOnsetTime unless you want to create some special effect and know
what you are doing.

Now, as I recall (someone please run the tests for me to confirm or
disconfirm), if movie playback reaches its end before anything else
makes it stop, then the last frame of the movie will linger until
something else replaces it. If you do not want that, then you might
set "End Movie Action" to "Terminate", and follow the MovieDisplay
with a visual object that will overwrite the last frame. Of course,
then your trial timing will depend on the duration of your movie
file. If you want to make the trial timing constant for all movies,
then you could use a little bit of inline code to adjust the duration
between the end of the movie and the continuation of your
trial. There are several ways to do that (my own favorite technique,
using SetNextTargetOnsetTime, is a bit esoteric), so if that
interests you please write back.

Anyway, please give some of this a try and let me know whether I know
what I am talking about :).

Thanks,

-----
David McFarlane
E-Prime training
online: http://psychology.msu.edu/Workshops_Courses/eprime.aspx

Twitter: @EPrimeMaster (https://twitter.com/EPrimeMaster )

/----
Stock reminder: 1) I do not work for PST. 2) PST's trained staff

take any and all questions at https://support.pstnet.com , and they

strive to respond to all requests in 24-48 hours, so make full use of
it. 3) In addition, PST offers several instructional videos on their

YouTube channel (http://www.youtube.com/user/PSTNET ). 4) If you do

get an answer from PST staff, please extend the courtesy of posting
their reply back here for the sake of others.
\----

At 4/4/2014 07:33 AM Friday, Katie Jankowski wrote:
>Hi David,
>I apologize for the delay.
>
>Thank you for your helpful advice!
>
>Changing the trial proc properties from "topofprocedure" to
>"beforeobjectrun" solved my problem! (Note, I changed the trial proc
>properties, not each objects' properties, which seemed to work
>better for me). In fact, it sounds like changing the procedure
>properties affects all the objects properties within the procedure,
>when set to inherit

>(<http://www.kent.ac.uk/psychology/downloads/EPrimeNewFeaturesGuide.pdf>new

>features guide on pg. 18), although I am not positive. In my full
>script, all movies ran without errors and both the auditory and
>visual output were presented.
>
>Also, I just

>found<http://www.pstnet.com/support/kb.asp?TopicID=4565> this post

>on the EPRIME knowledge base, which sounded very similar to what I
>was experiencing: "NOTE: Use of OffsetTime StopAfterMode option with
>any PreRelease value including the default (same as duration) will
>likely cause no sound or only a brief portion of the sound to occur.
>" Setting the prerelease to 0 sounds like a good fix.
>
>My only remaining concern is that the last frame in the 1st movie of
>each trial freezes, while the 2nd movie (last trial object) does
>not. If you have any suggestions for how to avoid the last frame of
>the movie from freezing and achieving a smoother transition, please
>let me know. If I figure it out, I will post it here. One thought is
>perhaps it is related to the "stopaftermode" set to "nextonsettime".
>I dont know if this is ok, if it should be set to "stopoffsettime",
>or if an inline script is necessary, as mentioned

><http://www.pstnet.com/support/kb.asp?TopicID=2720>here.

>
>Again, thank you for your help.
>
>In summary, I changed the prerelease settings to "0", changed
>generate prerun to "beforeobjectrun", and made sure my script was
>saved as an EPRIME Professional file (I have Professional but I was
>saving the script as EPRIME Standard.) Some changes may or may not
>have been necessary, but I wanted to list them all here, in case
>some people have similar problems in the future.
>
>
>On Tue, Apr 1, 2014 at 11:40 PM, David McFarlane
><<mailto:mcfa...@msu.edu>mcfa...@msu.edu> wrote:
>To answer your first question, I am pretty sure that E-Prime can run
>2 or more movies per trial, but I do not recall exactly if I have
>ever tested that. Perhaps I should.
>
>Next, please see

><http://www.pstnet.com/support/kb.asp?TopicID=3299>http://www.pstnet.com/support/kb.asp?TopicID=3299

>for "gotchas" that may happen with the new GeneratePreRun
>feature. (You may also look at the following
>threads:

><https://groups.google.com/d/topic/e-prime/IV1KWxm3Q-U>https://groups.google.com/d/topic/e-prime/IV1KWxm3Q-U
>,
><https://groups.google.com/d/topic/e-prime/Azi36c7sXSI>https://groups.google.com/d/topic/e-prime/Azi36c7sXSI
>,
><https://groups.google.com/d/topic/e-prime/0Yqv3FHXm6k>https://groups.google.com/d/topic/e-prime/0Yqv3FHXm6k
>, and
><https://groups.google.com/d/topic/e-prime/Q6si9tksg4Q>https://groups.google.com/d/topic/e-prime/Q6si9tksg4Q

><<mailto:katie.j...@gmail.com><mailto:katie.j...@gmail.com>katie.j...@gmail.com>

>wrote:
>Thank you very much for these suggestions. I had downloaded the
>latest version of DivX

>(<<http://www.divx.com/>http://www.divx.com/>http://www.divx.com),

><<http://psychology.msu.edu/Workshops_Courses/eprime.aspx>http://psychology.msu.edu/Workshops_Courses/eprime.aspx>http://psychology.msu.edu/Workshops_Courses/eprime.aspx
>Twitter: @EPrimeMaster
>(<<https://twitter.com/EPrimeMaster>https://twitter.com/EPrimeMaster>https://twitter.com/EPrimeMaster

>)
>
>/----
>Stock reminder: 1) I do not work for PST. 2) PST's trained staff
>take any and all questions at

><<https://support.pstnet.com>https://support.pstnet.com>https://support.pstnet.com

>, and they strive to respond to all requests in 24-48 hours, so make
>full use of it. 3) In addition, PST offers several instructional
>videos on their YouTube channel

>(<<http://www.youtube.com/user/PSTNET>http://www.youtube.com/user/PSTNET>http://www.youtube.com/user/PSTNET

[Katie Jankowski]

Hi David,

Thanks for your explanation and advice.

Unfortunately, none of the following settings solved the frozen last frame issue of the 1st movie:

1) setting end movie action to none and stopafter to nextonsettime (original)

2) setting end movie action to none and stopafter to offsettime

3) setting end movie action to terminate and stopafter to nextonsettime

4) setting end movie action to terminate and stopafter to offsettime

Do you have any other ideas?

In response to your comment, my movies are all approximately the same length (12000ms-13000ms), but I want the durations to be slightly different for each (e.g., 11500,ms, 12000ms, 12500ms, 13000ms; which I have specified by creating a durations attribute). (However, the combined length of movie 1 + movie 2 is always 24000ms, I just use the durations attribute as a way to trim the movies, since Corel VideoStudio doesnt always offer ms precision.)

As a tester, movies 1 and movies 2 in my first trial are both exactly 12000ms, and the durations of both of these are set to 12000ms (so there should be no problem). Like all trials in the script, the last frame of movie 1 freezes, but the last frame of movie 2 (last object in each trial) does not. The prerelease is set to 0, so I don't think the prerelease is causing the frozen frame.

Thank you,

Katie

--
You received this message because you are subscribed to a topic in the Google Groups "E-Prime" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/e-prime/LHq3Niv1zfk/unsubscribe.
To unsubscribe from this group and all its topics, send an email to e-prime+unsubscribe@googlegroups.com.
To post to this group, send email to e-p...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/e-prime/533ed138.4294320a.31aa.17fdSMTPIN_ADDED_MISSING%40gmr-mx.google.com.

[McFarlane, David]

A long overdue update, sitting in my to-do pile for five years now ...
Katie and I worked this out off the list and I promised a summary, so
here it is in case anyone every stumbles across this ...

To review ...

Katie's specifications for her fMRI study:
- Each trial consists of a fixation, an introduction text, a short
movie, and then a second short movie which should follow seamlessly
after the first.
- In order to stay synchronized with the fMRI scanner, the duration of
each trial must remain constant at 60 s, with 12 s fixation, 24 s
introduction, and 24 s total for for both movies.

We got things worked out here on the Group with one remaining issue --
the final frame of movie 1 would remain on the display until the start
of movie 2.

Of course, Katie could achieve a smooth transition between movie clips
simply by combining the two clips into one file. That would also
simplify the programming tremendously. But that could involve
considerable labor (especially with more combinations & orders), and may
interfere with tight synchronization between the onset of the second
movie clip and the brain imaging acquisition.

So we worked out an alternative solution after first working out some
general rules:

(0) Loading movies can take considerable time (just as loading images
used to take considerable time on slower machines of a couple decades
ago) -- the longer the movie (e.g., 20 MB files) the longer the load time.
- Make your movie files at the same resolution as you will display
them (certainly no larger, as that would be wasteful).
- Make your movie files as small as reasonable, based on
compression, quality, etc.

(1) You should not load any movie while a movie is already running, this
will likely interfere with the running movie. As a result, pay close
attention to PreRelease and Generate PreRun (TopOfProcedure vs.
BeforeObjectRun). E.g., ...
- If you use a Movie as the last object in a Procedure, and use a
PreRelease and TopOfProcedure, and repeat the Procedure back-to-back,
then the next run of the Procedure will try to load the next movie while
the current one runs, resulting in a problem -- so don't do this!
- Similarly, if you run back-to-back Movies within a Procedure,
and use a PreRelease on the first movie and use BeforeObjectRun, then
E-Prime will try to load the second movie while the first one is still
running, resulting in a problem -- so don't do this!

(2) When movie playback stops (the movie file reaches its end, or the
Movie object ends and performs a Stop After), the last frame of the
movie remains showing until E-Prime displays something else. E.g., ...
- If you run a Movie as the last object of a Procedure, with a
PreRelease of 0 and TopOfProcedure, and repeat the Procedure
back-to-back, then the last frame of the Movie may "freeze" on the
display for a moment while the Movie loads for the next run of the
Procedure (depending on the size of the movie file and the speed of your
system).
- Similarly, if you run back-to-back movies within a Procedure,
with a PreRelease of 0 and BeforeObjectRun, then the last frame of the
first Movie may "freeze" on the display for a moment while the second
Movie loads (again, depending on the size of the movie file and the
speed of your system).
This is all exactly how things *should* work -- after all, what else
would we have the system do between the end of a movie and the start of
the next visual stimulus? If you want the screen to go blank, then add
something after the movie to do that. If E-Prime did that
automatically, then we would not have the option of *not* having the
display go blank, so the given behavior actually gives added flexibility
to us users.

Ironically, Katie's Procedure ran into special complexities precisely
because of its simplicity, running little more than two Movies
back-to-back with the second Movie at the end of the Procedure. With
the trial Procedure set to use a Generate PreRun of TopOfProcedure, it
would try to load movies either during or right after the second movie
(and before the fixation), which caused problems; alternatively, with
BeforeObjectRun, it would load the second movie either during or right
after the first movie, which caused other problems.

We needed some way to load *both* movies at some other time, e.g.,
*during* the fixation rather than before it. And so we found a way to
do that.

Instead of having one trial Procedure of
Fixation-Instruction-Movie1-Movie2, we split that up into two
Procedures, one for just the Fixation (let's call it FixationProc), and
the other with Instruction-Movie1-Movie2 (let's call it MoviesProc), and
then used an outer List to run these in a suitable sequence. We set the
Fixation object with a suitable PreRelease value, set MoviesProc to use
TopOfProcedure, and optionally gave the Movie2 object some suitable
PreRelease value (we might have still had to keep PreRelease of Movie1
at 0, I don't remember now so leave that as an exercise).

With that simple change, E-Prime would load both movies during the
PreRelease of Fixation, and thereafter run the two movies seamlessly
back to back!

Alternatively, we could achieve the same effect all in one Procedure by
resorting to a bunch of inline code to take care of loading the movie
files during or after the fixation. To figure that out we would look at
the movie loading & running code generated by E-Studio with a Procedure
set to TopOfProcedure. This would allow even finer control of when
E-Prime loads, or runs, movie files. We did not actually explore this
though, leaving it as an exercise.

Bottom line, we could definitely get E-Prime to accomplish the desired
ends here, it just takes some deep understanding and finesse (and a very
fast machine helps!).

-- David McFarlane