If you are curious about using the surface RT for teaching, take a look at this great blog: http://eiuitc.blogspot.co.nz/2013/07/surface-rt-for-teachers-glass-half.html
Click on the home button and take a look at some of Tom's past posts, as he also covers the surface pro.
Friday 20 September 2013
Video editing sucks
My previous laptop (an aging DELL that was high spec when first purchased) seemed to take forever when editing videos. Unfortunately the surface makes it look speedy by comparison. Video editing is a good way to max out the RAM and the hard disk access at the same time (CPU doesn't seem to max out as much though).
I know exporting and saving video is notoriously expensive computationally but waiting tens of minutes is frustrating.
This is the first instance where I've been really frustrated with the surface performance. Usually it trucks along fairly well with its 4 Gb of RAM but video editors should opt for using something with more grunt.
I know exporting and saving video is notoriously expensive computationally but waiting tens of minutes is frustrating.
This is the first instance where I've been really frustrated with the surface performance. Usually it trucks along fairly well with its 4 Gb of RAM but video editors should opt for using something with more grunt.
Experiments with screen casting and powerpoint
This week I put my surface through its paces as a presentation machine (partly because the screen on my dell died that day, so it was my only option!). Unfortunately not everything went according to plan (in fact just about nothing did). Fortunately every cloud has a silver lining!
On Tuesday I ran an information evening for prospective mathematics students which included talks from seven different people (using five separate power point presentations). I was curious to see how the surface would go as a presentation computer, where rapidly switching between different power point presentations was required,. I also needed to record the entire event (both video and audio).
I decided to install bb flashback to record the audio and video, as I have a little experience with it and know several colleagues who use it. It turned out that bb flashback comes in three different versions (express, standard and pro). The express version is free and you can trial the other two for a month, free of charge. If you are fortunate your institution may well have a site license for the pro version (which I since discovered was the case for us). I know from past experience that bb flashback is capable of creating much smaller recordings, while still maintaining adequate quality. As a back up I also used the university's own recording facilities but this generates very large files (approx 200Mb for an hour of recording). Obviously large files can be painful if you have to download the recording over a slow connection (or you have a limited amount of data included in your broadband plan).
The results were less than stellar. I had plugged the projector in via a vga adapter and opted to duplicate the displays. Powerpoint 2013 had different ideas though and when I started the full screen slide show, I wasn't given a duplicate view on the surface and the projector. Instead the surface showed a control panel with a smaller view of the slide (allowing you to navigate easily through slides) and only the projector showed the full screen slide view. This normally wouldn't be a problem but I was using bb flashback to capture the screen and unfortunately it captured the surface screen, not the projector screen. Even worse it didn't capture the whole screen (that would have been salvageable). Instead it only captured a portion of it, leaving me with not quite usable video.
The good news is that I got pretty reasonable audio quality via plugging in an external microphone into the single line in/line out port on the surface. I was worried that audio quality might have been an issue but the mic did a pretty good job. The mic I used was a pretty simple one that would normally be worn on a lapel, however I attached it to the e-lectern mic. It was fairly good at picking up audio from all directions, which was helpful as some of the presenters roamed a bit. The downside was you picked up a lot of noise from the audience too, resulting in a fair bit of low background noise. Fortunately the pro version of bb flashback includes functionality to remove background noise (although you end up with a slight echo on the recorded voice at times).
Later experiments with bb flashback seem to indicate there are bugs when capturing regions and windows on the surface. Even with no projector plugged in, it does not capture regions and windows correctly. Instead it grabs a slightly offset area from the region or window selected (it captures most of what you want but the area captured is to the left of the one indicated in red when you set up the area to record). Fortunately full screen capture works (but only if you avoid powerpoint run using a projector).
BB flashback also seems to have a buggy save dialogue, as several times when I went to save a file I got a busy icon that lasted for a minute or two, before I was able to enter a filename to save the recording as. This is NOT a good look when you are in the middle of a presentation, so after the first hang took several minutes to recover from I decided to ditch my original plan of making separate recordings for each speaker. I figured I would be able to split things up in post processing (which turned out to be doable with the pro version).
So how did I get around the problem of not quite usable video? Fortunately I had used the university recording system as a backup and it DID record the video from the projector feed (plus audio). Interestingly the audio recorded through the e-lectern mic was much more directional, which meant less background noise. It also meant that if the speaker wasn't talking to the mic the sound was significantly attenuated (and hence hard to hear). Between the bb flashback recording and the university recording system I had good audio and acceptable video for all speakers but needed some way of combining the two and still producing a small file.
For the one speaker who roamed a lot, I ended up using bb flashback player to play back the captured audio (since my external mic tracked him pretty well). I simultaneously used bb flashback recorder to capture the played audio and new video (a fullscreen view of the powerpoint, which I manually navigated through to sync up with the audio). This worked surprisingly well. I ended up with very acceptable video and audio once I had trimmed unnecessary material from either end (and filtered the noise from the audio track).
For the remaining speakers (who had talked to the mic) I initially tried played the university recording in full screen mode (with audio) and set bb flashback to record the entire thing. This wasn't a particularly elegant solution, as it took an hour to run through. Fortunately it was the kind of thing that you could just set up and leave running. At the end I was left with a HUGE bb flashback file (over 7 Gb) which contained adequate audio and video. Note the video was NOT as good quality as would have been obtained if bb flashback had captured the correct screen in the first place (as the recorded video from the university system is downsampled somewhat). I suspect the HUGE size was due to the display continually changing just a little bit, as the video played. This means that the recorded wasn't just grabbing frames when the powerpoint slides were changed, it was continuously grabbing frames. This left me with the option of going with the slides that were cut off or redoing it manually. After several hours of playing with software in the wee hours of the morning I decided to go with the first option (sleep was starting to appeal).
I definitely want to be able to use powerpoint coupled with screen casting software, so need to find a workaround for bb flashback or need something more reliable (I'll take a look at camtasia, in case that works better).
There were plenty of other bugs on the night too. Powerpoint crashed and starting a full screen powerpoint presentation always generated an exception on the e-lectern computer (as it was accepting the display signal from my surface). I suspect this was due to powerpoint overriding the duplicated display mode.
The silver lining to all this was that I had anticipated the possibility of difficulty in transitioning between powerpoint presentations, so had prepared several discussion points to throw out to the audience. At the close of each presentation I simply gave the audience the next discussion point, before setting to work on working around the bugs and setting up the next presentation. This meant it didn't matter if it took me two to three minutes to get things working again (their focus was on the conversation rathe than on me dealing with the technical issues). It also had the unexpected side effect of keeping the audience more alert and engaged throughout the entire evening Incorporating some discusison points or "active learning" exercises could be a REALLY useful strategy to have up your sleave if you think there might be technical hitches.
On Tuesday I ran an information evening for prospective mathematics students which included talks from seven different people (using five separate power point presentations). I was curious to see how the surface would go as a presentation computer, where rapidly switching between different power point presentations was required,. I also needed to record the entire event (both video and audio).
I decided to install bb flashback to record the audio and video, as I have a little experience with it and know several colleagues who use it. It turned out that bb flashback comes in three different versions (express, standard and pro). The express version is free and you can trial the other two for a month, free of charge. If you are fortunate your institution may well have a site license for the pro version (which I since discovered was the case for us). I know from past experience that bb flashback is capable of creating much smaller recordings, while still maintaining adequate quality. As a back up I also used the university's own recording facilities but this generates very large files (approx 200Mb for an hour of recording). Obviously large files can be painful if you have to download the recording over a slow connection (or you have a limited amount of data included in your broadband plan).
The results were less than stellar. I had plugged the projector in via a vga adapter and opted to duplicate the displays. Powerpoint 2013 had different ideas though and when I started the full screen slide show, I wasn't given a duplicate view on the surface and the projector. Instead the surface showed a control panel with a smaller view of the slide (allowing you to navigate easily through slides) and only the projector showed the full screen slide view. This normally wouldn't be a problem but I was using bb flashback to capture the screen and unfortunately it captured the surface screen, not the projector screen. Even worse it didn't capture the whole screen (that would have been salvageable). Instead it only captured a portion of it, leaving me with not quite usable video.
The good news is that I got pretty reasonable audio quality via plugging in an external microphone into the single line in/line out port on the surface. I was worried that audio quality might have been an issue but the mic did a pretty good job. The mic I used was a pretty simple one that would normally be worn on a lapel, however I attached it to the e-lectern mic. It was fairly good at picking up audio from all directions, which was helpful as some of the presenters roamed a bit. The downside was you picked up a lot of noise from the audience too, resulting in a fair bit of low background noise. Fortunately the pro version of bb flashback includes functionality to remove background noise (although you end up with a slight echo on the recorded voice at times).
Later experiments with bb flashback seem to indicate there are bugs when capturing regions and windows on the surface. Even with no projector plugged in, it does not capture regions and windows correctly. Instead it grabs a slightly offset area from the region or window selected (it captures most of what you want but the area captured is to the left of the one indicated in red when you set up the area to record). Fortunately full screen capture works (but only if you avoid powerpoint run using a projector).
BB flashback also seems to have a buggy save dialogue, as several times when I went to save a file I got a busy icon that lasted for a minute or two, before I was able to enter a filename to save the recording as. This is NOT a good look when you are in the middle of a presentation, so after the first hang took several minutes to recover from I decided to ditch my original plan of making separate recordings for each speaker. I figured I would be able to split things up in post processing (which turned out to be doable with the pro version).
So how did I get around the problem of not quite usable video? Fortunately I had used the university recording system as a backup and it DID record the video from the projector feed (plus audio). Interestingly the audio recorded through the e-lectern mic was much more directional, which meant less background noise. It also meant that if the speaker wasn't talking to the mic the sound was significantly attenuated (and hence hard to hear). Between the bb flashback recording and the university recording system I had good audio and acceptable video for all speakers but needed some way of combining the two and still producing a small file.
For the one speaker who roamed a lot, I ended up using bb flashback player to play back the captured audio (since my external mic tracked him pretty well). I simultaneously used bb flashback recorder to capture the played audio and new video (a fullscreen view of the powerpoint, which I manually navigated through to sync up with the audio). This worked surprisingly well. I ended up with very acceptable video and audio once I had trimmed unnecessary material from either end (and filtered the noise from the audio track).
For the remaining speakers (who had talked to the mic) I initially tried played the university recording in full screen mode (with audio) and set bb flashback to record the entire thing. This wasn't a particularly elegant solution, as it took an hour to run through. Fortunately it was the kind of thing that you could just set up and leave running. At the end I was left with a HUGE bb flashback file (over 7 Gb) which contained adequate audio and video. Note the video was NOT as good quality as would have been obtained if bb flashback had captured the correct screen in the first place (as the recorded video from the university system is downsampled somewhat). I suspect the HUGE size was due to the display continually changing just a little bit, as the video played. This means that the recorded wasn't just grabbing frames when the powerpoint slides were changed, it was continuously grabbing frames. This left me with the option of going with the slides that were cut off or redoing it manually. After several hours of playing with software in the wee hours of the morning I decided to go with the first option (sleep was starting to appeal).
I definitely want to be able to use powerpoint coupled with screen casting software, so need to find a workaround for bb flashback or need something more reliable (I'll take a look at camtasia, in case that works better).
There were plenty of other bugs on the night too. Powerpoint crashed and starting a full screen powerpoint presentation always generated an exception on the e-lectern computer (as it was accepting the display signal from my surface). I suspect this was due to powerpoint overriding the duplicated display mode.
The silver lining to all this was that I had anticipated the possibility of difficulty in transitioning between powerpoint presentations, so had prepared several discussion points to throw out to the audience. At the close of each presentation I simply gave the audience the next discussion point, before setting to work on working around the bugs and setting up the next presentation. This meant it didn't matter if it took me two to three minutes to get things working again (their focus was on the conversation rathe than on me dealing with the technical issues). It also had the unexpected side effect of keeping the audience more alert and engaged throughout the entire evening Incorporating some discusison points or "active learning" exercises could be a REALLY useful strategy to have up your sleave if you think there might be technical hitches.
Subscribe to:
Posts (Atom)