Video training for low-bandwidth users: Quality? Doable? MBs/min?

Hi, I've been spending a while browsing the forums and web and have found Qs and As that have touched on these questions, but not sufficiently answered these specific questions.  Thanks in advance to anyone willing to respond.

Context:

Half of my learners will be in the US, with high-speed Internet and half will be in Sub-Saharan Africa in low-bandwidth environments.  Some of those in Africa will be using a USB modem and paying for data by the MB.  They will be low-income individuals in low-income countries and will likely be paying for the data themselves.  Goes without saying that they don't have money to waste.  We certainly don't want to require them to do this training and then, "Surprise, you've used 4GB of data!"

We are planning a "choose your own adventure" video training.  In other words, watch Video A and then make a decision about what the character should do, either option 1 or 2.  Depending on the response, either Video A1 or A2 will play.  And, repeat, i.e., more videos, more decisions.  In total, we will have 20 videos (all with audio), and each will be 2 minutes.  Other than the videos (which we want to be good quality; not necessarily true HD, but "good"), we won't have much else.  Two buttons/decisions on each video with a little bit of text on each button, and that's really it.  No quizzes.  No PPT-like screens, etc.  We envision something like this (but slightly longer videos): http://www.youtube.com/watch?v=JFVkzYDNJqo.

Questions:

  1. Is Storyline the right tool for me?  We haven't purchased it yet.  If YouTube wasn't blocked where several of our learners will be, we'd probably consider using YouTube as they did with the "annotations"/hot-buttons in the example I provided above and embed them into a WordPress site.  But, we also have to publish this to DVD (some, not all users will be able to get this on DVD)... and one thing I really like about Storyline is that I'd be able to publish/burn the training to DVD.  Other tools for developing and delivering this type of interactive training online don't seem capable of easily transferring it to DVD.  
  2. I'm convinced now that the quality of the video could be "good" coming out of Storyline (my account executive sent me a sample), however, I'm concerned about overall file size.  I sent a URL of a Storyline sample to a colleague in Kenya and she reported about a 45 second wait time for the first slide to load.  And, this wasn't even a video training that I linked her to.  I've read that the compression capabilities of the Storyline publishing tool works wonders in getting file sizes down.  But, does it also lower quality substantially?  Is it possible to deliver good-looking video to low-bandwidth folks?
  3. I'm not sure how Storyline exports files... I've read that it exports a file structure... if that's the case, can someone tell me approximately how many MBs a "good" quality 1-minute video is (that has been exported/published from Storyline).   And, I can extrapolate from there to figure out approximately how many MBs, total, all of my videos will be.  
  4. Assuming all other things (other than video) are sparse and simple, how many MBs would the rest of the published project be, i.e., all the "junk" that Storyline exports, does that add up to be pretty large? (I know the "junk" is really necessary to make the training work properly).  I'm so concerned about total MBs, because, again, some of my users are going to be paying per MB used.  I'm guessing whether they stream everything or try to download it they will have to pay for the same amount of MBs used (for those using USB modems).
  5.  Anyone have any other tips, feedback, recommendations, stories to share, etc?

Thanks!

ms

10 Replies
Mike Enders

Marc,

It's really not an easy question to guesstimate the size of a course given all of the potential variables involved.  But for a frame of reference, here's what I found in a similar project I did last year. It was a demo in which I used Camtasia to edit and output the videos.  

Source Videos:  1280x720 .mov files captured at 29.97 fps and 16M colors

Videos were brought into Camtasia for editing and output to MP4 at pretty stock settings:

These MP4s were then inserted into Storyline and resized to 408x230.  

The initial and resulting file sizes before and after publishing (and video length) can be seen in this image:

I hope this helps!

Mike

Marc S

Thank you, Mike, it's great to get a feel for how that workflow resulted in those file sizes.  Do you happen to have a link to that training so I could see the quality of the videos?  I'm particularly interested in seeing the longer videos that started out on the larger side.  And, if it's possible to also see the test you did where you brought the level down to a 6, that'd be great, too, to see how they compare.  

Marc

Mike Enders

Hi Marc,

Unfortunately, I don't have a link (it was an NDA project for a former client).   However, an easy way to test (if you've downloaded the trial) is to grab a web cam and shoot a number of videos of varying length, insert, and publish them out to see how they're impacted at different compression settings.  In the past I've played with .WMVs, .MOVs, etc. to get a feel for how they'd be impacted.

Mike

Todd Thornton

If you could achieve the same great results as Mike I don't guess it would matter as those are fantastic numbers, but you could use the Ken Burns Effect (think some refer to this as Photeo) to reduce the file size if you don't need full motion all the time. Basically replacing video with still images. (as contained periodically in the original video in my example if you wanted a low bandwidth version)  

I recently came across a CSS script that allows for creation of this effect, (At least in some browsers) so who knows what will constitute "video" in the future. 

Todd

Phil Corriveau

Hi Marc.  You might want to check out an Articulate blog entry I wrote five years ago (which I can't is already that old!!). It focuses on Presenter, but the general concepts hold true today and in any authoring tool.  In particular, I highly recommend that you consider NetLimiter as a tool to simulate low bandwidth conditions

Todd's suggestion of  "simplifying" with the use of still images is a great one, as video is a bandwidth eater--especially in areas where bandwidth is at a premium. 

Hope that helps...

Marc S

Thanks, Phil.  I actually came across your blog entry when I was first researching this!  I'll check out NetLimiter... looks like a good resource.  

As good and logical as the idea of "simplifying" the video with still images is, a requirement for this project is the use of "real" video.  I do recognize the challenge... which is why I'm spending so much time researching size and quality!

Thanks all.