what's here

more . . .

  

[ under construction ] 

 


how to shoot video of skating technique

 

capture device

analog versus digital:  There's lots of advantages to digital for editing and sharing and archiving. But for one-time analysis of a skater's current motions to help them improve technique, analog videotape (like played in a VCR machine) should work fine. I've been happy using it for technique analysis.

digital resolution:  320x240 is sufficient for many purposes, provided the person using the camera has some skill, the environment is good -- and it definitely helps if the camera has a zoom capability. 640x480 (or more) makes analysis a lot easier if capture is less than ideal - (first couple of day shooting technique video normally produces less than ideal results).

If you're going to try to measure an angle or length from the video, then you want at least 640x480.

frame rate:  25 frames per second seems sufficient. Some cameras will capture at 60 frames per second. Which could be nice for some purposes, but my experience is that 30 frames per second is sufficient for normal technique analysis. (I'd rather have 640x480 resolution at 30 fps than 320x240 at 60 fps).

zoom:  Optical zoom is definitely useful - (but not essential if the goal is only one-time analysis of a skater's current motions to improve technique). A problem with some digital cameras which offer optical zoom is that they do not allow the zoom to be changed while the camera is in video capture mode -- so if you try the camera before you buy it, specifically check for this capability.

Effective capture without zoom takes more skill by the camera operator, more careful set-up of the environment, more time on checking and re-shooting, and fewer skating strokes available for best analysis. No matter what you do about all that to handle it for your own analysis and learning, I think you'll find that if you later want to share it more publicly it's not going to look as good as if it had been shot with optical zoom available.

optical viewfinder:  Lots of digital cameras I've seen for sale (as of 2007) do not have an optical viewfinder -- I think this is to make the camera more compact. But outdoors in bright sunlight, an optical viewfinder can be very helpful for capturing video because: (a) Typically you want the sun behind you for best lighting of the shot -- which is worst for glare on the non-optical viewer; and (b) When shooting video you don't have time while re-framing the moving subject and changing zoom, so glare on the viewer is a hindrance. Also, whether indoors or outdoors, aiming and re-aiming the camera without optical viewfinder is going to be more jerky and less accurate.

Especially for video of cross-country ski skating on snow, I'd be sure to have a camera with an optical viewfinder. Sunshine on snow is a big glare problem.

color: There are useful things for technique which you can see in a grayscale video, but for seeing things more accurately and easily and decisively, it definitely helps to capture and display technique video in color -- and have varied color markings on key parts of the skater's body (e.g. hips).

sound:  Being able to make comments describing the technique variation which you're shooting could be nice, but I've usually found it easy to figure out afterward without sound.

putting it together: A digital camcorder is great if you have one, but I think most digital cameras for sale in 2007 which offer video capture capability can be used effectively for capturing video for technique analysis. Key features I'd look for are in a camera for technique video would be: optical viewfinder, optical zoom active during video capture, 640x480 resolution.

Phone? The technique video I saw captured with a mobile phone in early 2007 did not seem very useful -- but future phones might be better. 

environment

indoors

It should be possible to shoot useful technique video indoors, and it's done all the time for ice speedskating. But for some reason I have not yet seen video of inline shot indoors which was useful for technique -- but with some care in set-up it should be doable.

I suspect it's difficult to get useful technique video from an indoor inline race, because where you want to place the camera for technique purposes is also where the racers need to go. (Also a problem at public sessions at rinks)

Key thing to watch out for indoors is lighting. Indoor roller rinks sometimes prefer a dimmer ambience. But if it does not seem real bright, it's going to be difficult to get useful video. Even with good indoor lighting, I suggest being sure to mark key parts of the skater's body (e.g. hips) with high contrast colors.

outdoors

There can be lots of freedom to set up good shooting outdoors.

Usually there is enough light outdoors during normal daylight hours for useful technique video. Cloudy skies may not be good for "pretty" videos, but they're not a showstopper for technique video -- provided that key parts (e.g. hips) on the skater's body are marked with high-contrast color.

For skating on snow, cloudy sky might even offer an advantage. by reducing glare.

Lots of people enjoy skating on a tree-lined off-road trail. But capturing video on this environment often runs into a problem: It can be difficult to get the camera far enough away from the trail to shoot a good side view.

lighting configuration

Generally you get best results with the camera between the main light source and the skater - (sun behind the camera). Bad results with the skater between the camera and the main light source - (sun behind the skater). Sounds simple, but when combined with other environmental constraints it can get tricky. So when planning out where to shoot, it's important to think about where the light will be relative to the skater and the camera.

If you don't have much freedom in choosing your location, might have to be careful about what time of day you shoot.

background

Best results for technique video is to have:

  • background with strong contrast to the skater's clothing (light versus dark, or bright color versus dull)

  • background still -- no motion other than the skater

  • background simple, non-distracting

side view

Usually there's no problem finding enough space to shoot a good front view, but sometimes (or often?) it's difficult to get far enough away to shoot a good side view. And a good side view with a good lighting configuration. Scout possible locations with this in mind.

shooting from a car

I've never done it. From viewing several car-viewpoint videos of technique, I notice there's a tendency to use mostly a front view. But this strikes me as kind of a waste, since a front view is the easiest to shoot from a non-moving viewpoint.

Seems to me that a better way to take advantage of a car is to use it to shoot difficult viewpoints: especially foot-aim view -- difficult you get at most two leg-pushes (usually only one) on one side from a video clip. With traffic-free pavement two lanes wide, could have the car go diagonally ahead and to one side of skater, and get 5-10 leg-pushes in a row from a foot-aim view. With much wider pavement, could also get a longer segment in nearly exact 90-degree side view, instead of rapidly changing angle from a non-moving viewpoint.

laboratory treadmill

There are treadmill machines wide enough to skate on, which are set up with video and position sensors -- perhaps also force sensors in the skates. People with access to that sort of thing also have access to advice about how to use it.

Other than cost, I guess the main concerns are: (a) How much the skater modifies their style to work in that environment; and (b) There's no actual air resistance on the treadmill, yet air resistance is critical for normal skating.

For snow-skating, hills are important for some techniques (e.g. V1 skate). I've seen one example where a skier's V1 on rollerskis on a treadmill looked different from their V1 on rollerskis climbing up an actual hill on asphalt.

marking the skater

You can get lots of very useful observations from video without special marking -- if the lighting is good and camera equipment and technique are sound. If conditions might be questionable (e.g. weak light, 320x240 resolution), then marking helps. For observing difficult things (e.g. hip position), then marking helps. If you want to measure things, then marking is important.

I generally do not find that the ankle and knee joints are difficult to locate accurately in video. What I do find tough is the hips, and sometimes the shoulders.

What I suggest trying is 2 - 2.5 cm (0.75 - 1 inch) wide tape. Colored either red, blue, or white -- whichever contrasts well with the skater's clothing.

Priorities for marking:

Front view:

  • Hip: vertical tape strip on front of leg at the center of axis of motion of hip adduction / abduction.

  • Hip: horizontal tape strip on front of leg at the same height off the ground as the "ball" of the hip joint (located on the outside). (So the vertical + horizontal hip strips make a "cross" on the front of each hip).

  • Shoulder: vertical tape strip at the center of axis of motion for using the shoulder to "wing" the upper arm outward and inward.

  • (lower priority: front of knee joint.)

Side view:

  • Hip: vertical tape strip over the "ball" of the hip joint.

  • (lower priority: outside of knee and shoulder joints.)

viewpoints

front view

This is the easiest view to shoot, and has useful observations. But for some important observations for leg motions, it's not as accurate as the foot-aim view.

The usual shortcoming of video capture strategy is to shoot lots and lots of strokes from a front view, and almost none from any other viewpoint.

rear view

So far I have not found this to add much useful to the front view for analysis of biomechanics and physics of propulsion.

side view

Nearly as valuable as the front view for analysis of biomechanics and physics of propulsion.

Several difficulties with side view:

  • Lots of people don't know what to look for in observing and analyzing from it.

  • The speed across the camera view is so fast that usually pausing and single-frame-advance funtions are needed to accurately observe.

  • Normally (from a stationary camera position) you get at most two strokes that are shot approximately straight from the side, sometimes only one.

Fortunately all you need is one complete stroke which is both representative of the skater's motion and well captured on video. Usually it makes sense to shoot side views of multiple skating "runs", in hope that at least one will

Could get more stokes with the camera position moving -- like from a car (on a very wide area) or on a special track around the perimeter of an indoor rink. But since you only need one or two strokes for analysis, the extra set-up might not be worth it.

If you suspect (perhaps from the front view) an asymmetry between the motions of the two legs, you might have to shoot separately from both sides.

foot-aim view

This is with the camera "diagonally" in front of the skater, so that the gliding line of skater's foot during the main outward push is aimed roughly at the camera.

More accurate than the front view for observing the "Sweep Out" leg moves (e.g. hip abduction, ankle pronation). Especially at slower skating speeds -- like up a steep hill or in a starting sprint -- the foot-aim view is very different from the front view. At higher speeds the difference is not so large.

Main difficutly with the foot-aim view is that normally (from a stationary camera position) you get at most two strokes that are shot with the foot aiming at the camera, sometimes only one.

The obvious way to capture the foot-aim view is to do it together with the side view, using the same camera location and the same skating "run". Start shooting while the skater is farther away, and continue shooting as the skater passes closer in to the camera. Usually one (or two) strokes while the skater was farther away will be in the foot-aim view configuration, and one or two strokes while the skater is closest will be in the side view.

? foot-side view

The most accurate way to observe some of the leg extension moves would be from perpendicular to the line of forward glide of the skater's foot in the main outward push.

But so far I haven't thought that it added much value beyond the other views, for analysis of biomechanics and physics of propulsion.

The obvious way to capture the foot-side view is to do it together with the side view, using the same camera location and the same skating "run". Start shooting while the skater is farther away, and continue shooting as the skater passes closest to the camera, and continue shooting as the skaters finished farther away. Usually one or two strokes while the skater is closest will be in the side view, and one (or two) strokes while the skater is moving farther away will be in the foot-side view configuration.

doing it

Capturing useful video for technique is trickier than it sounds. Before inflicting video on a group of skaters, I suggest first practicing and working through some of the tricky parts at least one time on your own -- definitely with the same camera and the same video viewing software you will use with a group -- and preferably in the same shooting location.

If not the same location, then at least do your practice shooting outdoors if your group work will be outdoors; and practice shooting indoors if your group work will be indoors.

?? the kinds of mistakes you can make with using the camera -- especially in bright light outdoors.

?? how much skating and shooting space and what angles you need for effective capture for technique analysis.

?? what video shot

 

First time, do it with just you and one person who is patient and forgiving -- and set aside lots of time for mistakes. I suggest going through this sequence:

  1. both of you practice operating the camera in video capture mode.

  2. you shoot the other person in both front view and side view.

  3. the other person shoots you skating in both front view and side view.

  4. go indoors and transfer the captured video to a computer, view the video files, do only very quick basic analysis -- see what went wrong with the camera work, what could have been done better.

  5. go back out and have the other person re-shoot both front and side views, applying the lessons from the first round of shooting.

  6. go indoors and transfer the captured video to a computer, view the video files, do some serious analysis -- but stop when you've identified two or three significant improvements which you could make to your skating. (No more than three).

  7. During the following week or two, try to practice those two or three improvements.

  8. Try another video session (after a week or two) with the other person shooting you. Then observe and analyze to find out how successful you were in changing your skating motions.

 

how much time?

 

how much experience? Do it just on yourself the first time, capture thru analysis thru trying to improve based on analysis, and shoot again.

should the skater consciously try to avoid certain faults? or just "skate natural"? or try different variations of styles and speed and cadence?

 

zoom + framing

problems seeing indicators and view in bright light (esp sunshine on snow)

digital camera: prefer optical viewfinder

 

on-site or near-site quick review

 

solo -- you skating with nobody else holding camera

It can be difficult some people who are serious about working on their technique to find someone else to shoot video of them very often.

I have seen some decent results with nobody holding the camera, though I've never tried it myself.

Analysis non-solo: Just because you succeed in capturing your skating with no help from another person does not mean it's smart to analyze and learn from it that way, with no help. It's very easy to trick yourself about what you're seeing in your own skating, and very easy to get (unnecessarily) discouraged about seeing your own skating. Or you see some valuable things, and work on them, but then you get bogged down in your own rut.

Get somebody else to give you another perspective on your skating, and sometimes get a new person to give you a fresh perspective. The most important reason to get another person involved is not that they're smarter than you or a faster or better skater than -- but that they're just different from you, different pair of eyes.

It should be possible to find experienced coaches that you can pay to review some video clips of your skating that you've put up on the web somewhere.

Or you can post to a discussion forum for unpaid people to give you suggestions -- often I've gotten some good advice that way, but mixed in with lots of confusion -- so if you try that approach, be prepared to sort things out.

Doing the capture solo

Here's what I think is needed to have a chance of success:

  • tripod.

  • camera that captures at least 640x480 pixels.

(not strictly necessary, but a remote control device to start and stop the camera would help a lot)

  • storage in the camera that can hold at least 5 minutes of video (at 640x480 resolution at least 25 frames per second) -- preferably 10 minutes, and more is better.

  • video editing software that lets you easily cut out the many wasted segments that you shot.

front view

key is to start far enough away. If you don't have a long-range remote control, I think you just have to let the camera run while you're skating out far enough (which is why you need lots of storage in the camera, and editing software to cut it out after you've capture it).

The problem for observation is that while it looks very dramatic to skate close into the camera, it's hard to tell in analysis what aspects of the changes from moment to moment are due to changes in motions of the skater, and what aspects are just from getting closer to the camera. Farther away it's easier to sort this out.

Think in terms of starting to shoot at least 30 meters or 100 feet away for skating at reasonable speed on level ground, and allow some space to get mostly up to speed before reaching the point where you want to start capturing. For climbing up a hill, the distance can be shorter. For a very steep hill, even shorter.

side view

This is tough for shooting solo. While I think 8 meters or 25 feet is enough distance for shooting with another person holding the camera, I'm guessing that double that is more what you need for shooting it solo from a tripod: at least 16 meters or 50 feet distance between camera and skater. (and here's where more resolution than 640x480 pixels could help). For higher skating speeds, even more distance.

So you need a big wide space if you're going to capture more than a stroke of direct side-view skating.

If you don't have a long-range remote control device for the camera, expect to burn lots of time and storage in the camera to skate out to the start and get up to speed.

foot aim view

basically the same problem as the side view, but might take more trial-and-error to find the camera angle that best captures the foot-aim view.

 

 

 

 

   

back to Top | more Video | more topics | Resources | Go by Skating index

how to view + analyze video

 

editing and sharing

  • problems with YouTube

 

viewers

 

single-frame advance

 

measuring

 

back to Top | more Video | more topics | Resources | Go by Skating index

more . . .

?? [ to be added ]