Do screencasts suffer without footage of the lecturer?

If you have looked at a few of my screencasts, you will be aware that I am currently recording only the laptop screen, without including any footage of myself waving my arms around.
Now, the laptop I am using has has its own built-in webcam, and Camtasia offers me various options when it comes to screen capture. It may be possible to include footage of myself with my screencasts. Would this make my screencasts/videos more appealing?
I think that it would be impractical for me to edit together two video streams after each class, so I would be looking for a solution that was easy to arrange while lecturing. For example, one possibility might be to use the standard screen capture method, but to have a small window open showing the webcam picture. At suitable points, I could bring the webcam image to the front, and it would then be captured by the standard screen-capture system. I may well experiment with this.
Does anyone have any views on this?
Joel
January 9 2010

Advertisements

23 responses to “Do screencasts suffer without footage of the lecturer?

  1. I think it would make them more appealing. It gets kind of tedious to watch the screencast and only hear your voice – at least that’s my experience, my mind always starts to wander after a while. A video would probably make it easier to steady the attention.

    Like

  2. OK, I have investigated the “inbuilt webcam” approach, and that is not going to work. When I reach out with the tablet pen towards the tablet screen, the webcam shows a scary image of a huge hand and pen coming straight towards you! My next attempt will be with an external webcam.
    Watch this space for developments!
    Joel
    January 11 2010

    Like

  3. Using an external webcam (suitably positioned) looks promising. Now I need to determine the best settings to use at all stages (again!).
    Currently I am looking at the official tutorial page at http://www.techsmith.com/learn/camtasia/6/edit/pip.asp
    (PIP stands for Picture-in-Picture).
    It may be best simply to record both video streams using Camtasia, and then to display the results either with the webcam picture showing small in one corner, or with the two videos side by side. Clearly this will require experimentation.
    Joel
    January 11 2010

    Like

  4. The side-by-side option produced by Camtasia looks quite good, but is not ideal for uploading to YouTube: as far as I can see you end up with two separate mp4 files, only one of which has sound. It may be that there are other options here, but that is as far as I have got so far.
    Putting the PIP in one corner of the video may be a good system … but what if it blocks some important part of the material? With the default Camtasia system, you don’t see the webcam picture on the screen during screen capture. Indeed you can decide the dimensions and positioning of the PIP later, and even turn it off and on for certain sections. However, I am trying to minimize the amount of post-recording editing required. So perhaps I should go back to my original idea, and NOT let Camtasia record the second stream at all: just access the webcam directly, display the image on the screen (when I want to), and let the Camtasia screen capture do the work.
    More experimentation is required!
    Joel
    January 12 2010

    Like

  5. Possibly, if it was easy, it would be best to have a side-by-side version showing the two videos, but with everything combined into one mp4 file. I don’t know how to do this yet (see the previous comment for what I have obtained so far).
    As my next experiment, I will go for Picture in Picture (PIP) using a fairly small picture at the bottom right of the screen. I will then try to avoid writing there or displaying there anything that I am currently talking about. To help remind me about this, I have added in a margin rule at 175mm in from the left to my default windows journal note page.
    Some other possibilities I have in mind: avoid using Windows Journal in full screen, and resize the windows journal window to leave a column clear. That way, in the final video, the PIP won’t overlap the windows journal window. However, this would mean that I would lose more of the screen than necessary.
    In order to increase flexibility here, I have changed the option on my windows taskbar so that it is no longer always on top. This is useful, in any case, as it allows a bit more vertical space in non-full-screen windows journal.
    Joel Feinstein
    January 12 2010

    Like

  6. OK, I have now made some experimental recordings. Unfortunately the video and audio are now starting to go out of sync again (and this time I definitely DO have the laptop running on mains power!). So I think that I have reached the limits of my current laptop. When technology advances further, I can try again.
    I suppose that I could add a still photo of myself in the corner. Maybe that would help to personalize it a bit?
    Joel
    January 12 2010

    Like

  7. Or am I being too fussy?
    See
    http://wirksworthii.nottingham.ac.uk/webcast/maths/webcam-test/webcam-test.html
    for a synchronization test using my current setup. The synchronization is imperfect, but maybe it is good enough?
    Whether it would get worse over a longer video will need further testing.
    Joel
    January 12 2010

    Like

  8. It’s ok, I think it’s less than half a second out of sync. If it doesn’t get worse when recording longer, the video feed would be a nice improvement.

    Cheers,
    Thomas

    Like

  9. I have included three small webcam+screencasts in the combined 80-minute video of Chapter 9 from G12MAN Mathematical Analysis on Sequences of functions, uniform convergence and pointwise convergence. See http://wirksworthii.nottingham.ac.uk/webcast/maths/G12MAN-09-10/Chapter9/ and the associated blog post at https://explainingmaths.wordpress.com/2010/01/08/uniform-convergence-and-pointwise-convergence/
    Joel
    January 17 2009

    Like

  10. OK
    I now have my first “full-length” screencast including PIP obtained from a USB webcam attached to my laptop.
    This first session is an introductory class on totally ordered sets and partially ordered sets, available at
    http://wirksworthii.nottingham.ac.uk/webcast/maths/G14FUN-09-10/FUN-250110/
    There is a lot of blurring whenever I move. So perhaps I need to move less! Probably the further away from me the webcam is, the better. But it may be that I can change the settings to improve this. Or perhaps I will need a better webcam?
    Comments welcome!
    Joel
    January 26 2010

    Like

  11. Hi Joel,
    it’s nice. Not sure, if it’s the webcam, or the graphics not being able to catch up. Maybe it improves if you decrease the resolution you capture. In any case, it’s okay this way. I think it’s easier to concentrate on the maths for a longer period of time, if you shortly switch your attention from the board to the webcam feed and back.

    Cheers,
    Thomas

    Like

  12. Alan Mintey (thanks!) has suggested that the problem might be the “shutter speed” (exposure time) in my settings. With the Logitech webcam I am using, I appear to have the choice between automatic gain control (which sets the exposure to random amounts between 1/20 and 1/50 sec) or manual gain control. But when I set the exposure to 1/50 sec in the manual settings, it always switches to 1/10 sec as soon as I attempt to accept and OK the settings. So I am better off with automatic for now.
    Another theory is that I really am at the limits of the laptop, and the webcam capture is not getting enough processing power. I have tried going for a very LOW resolution capture (which would be good enough to produce a small picture in the corner), and the results are a bit better. Instead of blurring, the image is now a bit jerky. I’ll put a sample on the web at some point.
    Joel
    Jan 28 2010

    Like

  13. See http://wp.me/posHB-7t and http://wirksworthii.nottingham.ac.uk/webcast/maths/G14FUN-09-10/FUN-290110/ for my latest efforts. I think that I need to try slightly higher resolution, or to have the webcam a little closer to me. I should probably also try to stay on-screen!
    Joel
    January 29 2010

    Like

  14. Hi Joel – Have just come across your page via Peter Rowland’s tweet…

    Your screencast looks good with the video in the bottom right corner. You could start a session with full size (or at least a bit larger video) and give an overview of what you will talk about, then minimize in the corner? And go large again at the end?

    I’ve recorded maths lectures for a few years now (using Camtasia Studio first, now Camtasia Relay), and don’t usually include video of myself. This is for on campus and distance students.

    Like

    • I’m not yet sure whether I can easily vary the PIP size during one Camtasia production. Of course, I can combine different videos on the timeline, so could do it that way. However, for a 30-lecture module, I think that I would rather go for a reasonably automatic production process after each lecture.
      As you can see, I am experimenting with the PIP arrangements. From the PIP point of view, which works better, I wonder, out of Lecture 13, part b and Lecture 15, part a?

      Like

  15. I’m not quite sure what I have done right, but things seem to have dramatically improved!

    See
    http://wirksworthii.nottingham.ac.uk/webcast/maths/G14FUN-09-10/FUN-010210/

    This time the webcam recorded at 320×240 (and I then scaled down the resulting video a little).

    Joel
    February 1 2010

    Like

  16. Hi Joel,
    about the latest screencast: I found the webcam feed a bit laggy.

    Just one question concerning the lecture itself: Is there a reason why you don’t introduce a convergence theory, such as nets or filters? You mentioned both, but left them for interested students to look up. Granted, you won’t need any of them in normed spaces, but they sure come in handy later.

    Cheers,
    Thomas

    Like

    • Unfortunately there is a shortage of time, and I have to make some difficult decisions. In addition, subnets are very subtle and can cause a lot of problems for students unless treated rather carefully. For example, it still bothers me that the sequence x_n=n has a convergent subnet in the Stone-Cech compactification of the natural numbers, \beta \mathbb{N}, but does not have a convergent subsequence there!
      At least we do something on ultrafilters as part of the proof of Tychonoff’s theorem, so that gives the students a head start when learning the filter/ultrafilter approach approach for themselves.
      It looks like, with my current equipment, there will be an element of luck involved in the synchronization. However, adding light does at least deal with the blurring problem. I’ll keep working on it!
      Joel
      February 10 2010

      Like

  17. My latest screencast + PIP used a 40 watt bulb (rather than 60 watts used previously) and with the lamp a bit further away from me (Alan Mintey suggested that I had the lamp too close).
    See http://wirksworthii.nottingham.ac.uk/webcast/maths/G14FUN-09-10/FUN-190210/
    I used my own creative livecam optia this time, rather than the logitech webcam I used for most of the other PIP videos.
    The synchronization seems pretty good this time. However only time will tell whether this was a fluke!

    Like

  18. Just wanted to leave a comment of thanks for your documentation of this process. This answered several questions that I have been considering.

    Steve

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s