General discussion about anything even only vaguely club or rowing related
|Message board > General Discussion > Rowing video quality|
|Rowing video quality by RTT - Sun 14th Mar 2010, 10:56pm|
|I suspect this issue has been tackled by many of us before, so I was hoping to gather suggestions and information into one place. |
The problem is simple to frame: What is the best and most practical way to take, process and display video of people rowing, assuming the camera is being used on a bike?
I know that using a tandem can have very positive effects on the raw video, but even then there is lots of shaking involved. How can this be reduced at time of capture? What tools do you / have you used to process the footage? What file types lend themselves best to storing and playing footage of rowing?
Please fire away with suggestions / further questions etc. and hopefully we'll gather a bit of a guide for anyone trying to video rowing in future.
|by Simon - Mon 15th Mar 2010, 9:21am|
|I used to use Digital8 tape with firewire (sure it has a prper name) to get it onto the laptop from where I just used moviemaker to splice things if I wanted to make a montage. I did also find that using the viewfinder rather than the little screen helped a lot, but with the viewfinder held slightly further away from you so you could also look through the camera and get an idea of what was just out of shot.|
|by Mark - Mon 15th Mar 2010, 11:57am|
|Many people use deshaking software and the one I've used in the past (on the video of FaT W1 bumping back Jesus last Lents) was the plugin Deshaker for the program VirtualDub.|
Obviously the shake is transferred to the black rectangle surrounding the picture which can be annoying but I think the greatest problem with that program is that if you don't know how to use all the software functions properly you can lose a lot of video quality - which is what happened with the video I tried to edit. I'm sure with practice and a good guide you could edit very well with it and I think it is quite simple to then output the video in a standard format like .avi or .mpeg.
|by Aaron - Mon 15th Mar 2010, 2:05pm|
|Having taken a lot of footage during bumps in a variety of ways I would say the steadiest raw footage of bankpartying I have is when I was on my bike filming freehand as it were. The tandemcam had potential but Mark tended to get a bit excitable and waved the camera around a lot. I was quite surprised how steady the footage was when I was filming & steering at the same time, I think having a proper camcorder helped as I could get a firmer hold on it.|
In terms of video format I didn't have much choice, the camera recorded in 1080p .mts which I presume is a Canon proprietory format, this isn't too much of a problem if the footage is being used solely for coaching, VLC will play .mts with no problems however it won't upload to Youtube properly and file conversion is where I've encountered the biggest problems (and why W1 still don't have their montage). The answer seems to be to buy some professional video decoding/editing software or do the whole lot in Linux. Ideally you would want the output from the camera to be in mpeg, alternatively avi would be ok and a lot of smaller cameras do output this however I haven't found one that does a large resolution and doesn't use an obscure format. The videos I've put up on youtube already are from my dSLR which outputs 720p .mov, this format works ok but needs conversion before it can be edited with any free software on windows (or mac it seems).
More thoughts once I've converted the files to something workable.
|by jmg - Mon 15th Mar 2010, 6:40pm|
|Had a lot of success with this problem on skis this year, by attaching a camera to a ski pole rather than holding it in the hand. The camera has to be small and light, e.g. GoPro, which we used and one of which JPD now owns. Once it's set up on the pole, it's pretty easy to know where you're pointing it.|
The major advantage is that involuntary movements of your hand have less effect on the shot, provided the camera is set up to point along the axis of the pole: Movements to and fro along the axis don't result in visible shake in the shot, while small movements perpendicular to that axis also don't significantly change the framing. Rotation of the camera, which is the key factor in camera shake, is reduced by flexibility and inertia of the pole, combined with the damping inherent in its being handheld. Only axial rotation is not stablised at all.
This worked well on skis, and I don't think the towpath is much bumpier than a red run at the kind of speeds we're talking about. The major issue I can see would be not impaling anybody as you go along. A shorter pole with extra mass at the end, or held at the midpoint with mass at either end, would maybe work just as well