Hi!
I am new to ELAN and I am using it for behavioral coding of videos. I want to code touch behavior on a frame-by-frame basis and for this I edited the preferences to “Create annotations aligned with video frames”. But how does this work, exactly?
Or in other words: I wanted to export the annotations to a text file and I would need those annotations to be aligned with frames, meaning that the text file would list each video frame with the corresponding time stamp and annotation.
Is this possible, and if so, could somebody tell me how?
Thanks in advance!
Fanny
Hi Fanny,
When you select “Creates annotations aligned with video frames”, the annotations that are created
are aligned with the video frames ie begin and end time of the created annotations are rounded to nearest video frame.
As you mentioned you will be coding the touch behaviour frame to frame, I assume you will be creating the annotations per frame . So when you export these annotations to the text file, all the annotations of that tier with the duration, begin time and end time can be exported to the file.
Hope I understood the scenario properly!.
Best,
Divya
Hi Divya,
thank you for your reply, I appreciate it a lot!
What you are saying makes total sense, that should give me the kind of text output that I want.
I guess I am having a somewhat different issue then, because right now I am making annotations for time sequences, meaning I select a time period from my video and make an annotation for that sequence.
If I want an annotation for each and every frame, does that mean I have to select and annotate a different way? And if so, could you maybe tell me how?
Thank you so much in advance, I am totally new to using this kind of software.
Best,
Fanny
Hi Fanny,
It depends on the length of each annotations. If your aim is to annotate each and every frame( ie with each annotation having equal length throughout), then the annotations can be created with the help of Tier → Create Regular Annotations, linked here
In the Create Regular Annotations dialog , you can select the begin and end time, the duration will be automatically filled. You can fill the Annotation size as your frame size.
You can explore this option and see if this serves what you are trying to achieve.
Best,
Divya
Hi Fanny,
Here are some options you could try:
Based on the requirement of getting list of each frame in the output file:
Option1 : You could create a Tier1 and create annotation by selecting a time period that you wish and adding the annotation value to it. Then create Tier2. In Tier2, the annotation can be created with the help of Tier → Create Regular Annotations, the begin and end times for this will be the Tier1’s annotations begin and end time. And the annotation size will be your frame size.
A new tier ( Tier3) can be created by using Tier → Create Annotations From Overlaps
Steps for this would be
Step1 : select Tier1 and Tier2
Step2 : select criteria ‘regardless of their annotation value’
Step3 : enter the name of the destination tier ie Tier3 and select the tier type.
Step4 : select option ‘Value from a specific tier’ to be Tier1
Now you can export this Tier3 as tab-delimited text and in the output options select 'Separate column for each tier.
Option2 : You could create a Tier1 and create annotation by selecting a time period that you wish and adding the annotation value to it. Then create Tier2. In Tier2, the annotation can be created with the help of Tier → Create Regular Annotations, the begin and end times for this will be the Tier1’s annotations begin and end time. And the annotation size will be your frame size.
Go to Tab-delimited export and Select Tier1 and Tier2 .
in the output options select ‘Separate column for each tier’ and in that select ‘Repeat values of annotations spanning other annotations’
Also for both above options instead of creating the Tier2 with the begin and end time of the annotation on Tier1, you can create an annotation for the whole file ie with begin time to be the beginning of file and end time to be the end of file. Then your output would consists of all the frames of the file. Output will be very big in this case. You can choose the options according to your requirement 
Best,
Divya
Hi Divya,
thank you so so much for your efforts to help me!
I will try these options out and I am sure, one of them will work for me.
Thank you!
Best,
Fanny
