Nov 24 2015

Launch Survey Results

25 respondents to the post-launch survey wasn’t too bad. Okay actually it’s 21 respondents since I submitted three from my devices and one I know came from a friend who did some late beta testing. But still! Data! I’m going to address everything from the survey here.

What was your overall impression of the launch experience?

Overall the response was very positive, with the majority voting 4/5. I think everyone read the question carefully and caught on to the fact that the “experience” included the website design & functionality and didn’t simply focus on the fact that it was a small sounding rocket flying a short suborbital trajectory.

What was your local launch time?

As I expected, the majority of viewers came from Europe, which is why I scheduled the launch when I did to straddle the European continent as well as the western US. Good to know for future launches, which I generally do in the mornings to have the entire day to carry out mission ops.

How would you describe video playback…

I was very pleased to see the majority of people rate the video as smooth for the pre-launch segment. I found that the video can run out of sync with the countdown, so every second I manually set the video position to match up with the countdown – this can cause noticeable jerks to the video. Once the telemetry countdown begins (anywhere from 5-10 seconds prior to launch) I stop syncing the video and just let it run so there are no more chances for stutters (at least not caused by my code).

In general video sync is a huge problem for me right now and the main reason why I initially did not plan to include video and when I decided to I only included the first few seconds of launch. As I said, video can run out of sync after a while so keeping it smoothly in sync during ascent isn’t something I looked into how to do. Furthermore, just re-flying the launch to make things happen similarly to how they happened during the capture of the initial telemetry data is a challenge unto itself. Finally, my video capture introduces numerous duplicate frames and stutters that have to be painstakingly manually removed otherwise it exacerbates the issue of the video going out of sync with the telemetry data.

So for the foreseeable future no launches will have more than 15-20s of video after the actual launch.

Would audio really make the launch video better?

The results slightly favor audio by 16% but it would need a much bigger margin to convince me to include it. To be quite honest I’m not all that fond of the audio for KSP. They don’t even have a real sound designer as far as I know, and it shows. Until they get some quality audio or one of the community audio projects gets completed, I really don’t plan to put any sound in the video and increase the file size.

What type of device did you use to view the launch?

As expected, mostly desktop. But there was one iPhone (in addition to mine) and an Android (submitted as Other with a Win10 desktop because I’m dumb and didn’t allow more than one answer to be selected). I was hoping for a more varied device environment to be honest. Probably should have asked people to watch it on multiple devices if they had them available. I watched on my computer, iPhone and iPad at the same time. Was pretty cool.

Please leave any suggestions or detail any problems you experienced

Let’s take a look at some feedback!

The page reloaded a few seconds before splashdown, not sure if that’s supposed to happen, but I didn’t get to see the splashdown. 🙁

So yea, that was supposed to happen because the battery power ran out so there was no more telemetry and thus no reason to stay on a page that no longer holds relevant data. Maybe the user didn’t read the craft description update in the new event?

A couple of seconds before splashdown is also a thing that never existed. When the page reloaded after telemetry expired the payload was at 2.4km descending at ~5m/s so it was several minutes from splashdown, which happened here.

Onboard camera feed would be sweet 😉

Yep it would. Obviously the rocket has to have an onboard camera for there to be an onboard camera feed however, and in this case there wasn’t one. Future launches could have one (past launches have), but due to the limited video I’m not sure how much use they will see.

Map was nice, but it would be useful to be able to zoom in on it

Hmm… I wonder if it’s worth leaving the zoom controls visible yet disabled so people can see that the map starts fully zoomed in. And for anyone asking, no I can’t make it go any further that’s as far as Kerbal Maps goes unless I generate new imagery for a closer zoom layer and I don’t know nor have time to do that for now.

I would like to be able see a live graph of (some?) of the data

I’ve actually considered this, and its totally possible for some of the data for me to do this absurdly easily (spoiler – take the Graphotron plot image and just scroll it out from under a hidden element so it appears to draw itself onto the screen). The problem then though is I won’t be able to easily expand upon it. But maybe I’ll use it as a placeholder to see how people like it.

Video did not load…

So it turned out the majority of viewers were Chrome users, which sucks because Chrome for some reason doesn’t want to load the video. I knew this from testing with a friend and supposed I should have taken the time to install Chrome and check it out, but I didn’t really have the time so I definitely will for the next launch. It was also one case of failure. But all the Firefox users came up fine, even one Chrome user managed to get video, tho one Safari user had trouble.

I have more work to do analyzing the debug data but overall – great first launch with the Flight Tracker v3.0!

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>