Discussion in 'Tennis Tips/Instruction' started by andrehanderson, Aug 1, 2014.
Not sure how to do it.
Need air temps, court surface, type of balls used and how old, exact frame per second for you techno types (not the camera's theoritical running speed), and WHICH of your serves, since you hit fast ones, slow ones, second serves, and some slight mishits by sound.
I'd generalize, mid 90's and very good for 4.0 play.
The balls were brand new Penns. Weather was about 80 degrees. Not sure about the rest.
You already know you have a good serve, for 4.0.
Court surface determines bounce height. Abrasive courts, like new HarTru, or clay can have flat first serves bouncing 40" high at the baseline, and it still going only 100 mph.
Slice old painted cement, that same 100 mph serve might only bounce to 24" heights at the baseline, because the ball skids.
Penns are good middle of the road bouncers, but exact model is nice to know, as some Marathons bounce like Dunlrocks, while other Champs bounce like Wilsons.
Either way, you serve is fast enough to win most any 4.0 tourney/match, if it has good placement and it's reliable.
you're right...a lot of factors involved. We were playing on really slick painted cement. Uneven in spots, as well.
I should have guessed by the lean angle of the net posts.
Not exactly country club material.
You hit every serve with a spin component, so you can increase ball speed by flattening out your serves.
I think you're topping out at about 100 mph. Nice serves!
+1 on the speed. Right about 100 give or take a few.
Done a lot of video kinematic analysis.
Best angle for serve speed is from the side and down to see clearly when the lines are crossed. A bit of trig gives distance between frames. Distance over time (between frames) gives speed. Pretty good accuracy can be had with this method.
It is much harder estimate speed with ball approaching or moving away from camera.
I say about 30 miles per hour............. just kidding nice serving..
90-100, Imo. Next time put the camera behind you for all your serves.
Thanks president. I'll try that angle, plus the angle math geek mentioned.
Good to see you on the courts again!
Nice serves also! Was a person holding the camera? Looks like they got nailed at the end of the video haha
idono, sorry I haven't gotten back to you man. Want to hit anytime on weekends? Andreh, good serving and nice to see you back on courts. I would say definitely 90+ but hard to tell from the camera.
No problem, I've been out of town for most of the summer anyway. Won't be back until the 22 or so, but I'll send you a message!
Hey andre, agree with the avg estimate - right around 95, 100.
For speed you need to measure distance and time.
Take a video of your serve from the side using a 30 fps camera. The more accurately you view perpendicular to the ball's trajectory, the more accurate the measurement will be. Video camera on a tripod so that it does not move. (Better not to use a smartphone video camera as under some circumstances their frame rates may vary from 30 fps. But they are probably accurately 30 fps in sunlight. )
An object moving at 100 MPH is traveling 1760"/sec.
A ball at 100 MPH will travel
58.7" between frames (= 100 MPH).
1) Measure between the first and and second frames that catch the ball after it is struck.
2) You need a scale in the video. Hold up your racket before the serve and point it straight toward the center of the service court. Video a short segment to get the length of the racket to use as a scale. Or use some other clear scale. (a PVC pipe cut to 58.7" will indicate 100 MPH.)
I forgot - you need to walk the scale out along the trajectory while videoing to account for scale changes across the frame especially from wide angle lenses.
To get unknown speed, X, in MPH with measured length between frames of L, in inches, use
X = L X 100 MPH/58.7"
There will be less motion blur in direct sunlight but you can still work with motion blur on the ball. Just use the center of the ball's blur smear. More accurate if shadows do not affect the appearance of the ball's blur smear.
On your camera's video screen or on your computer, measure how far the ball moved between those first two frames (ignore impact) using the scale that you videoed.
If this is done carefully with a 30 fps camera, it can give very accurate serve speeds. I would trust it more than low-cost radars, if carefully done.
For credibility post two frames from your side videos.
Velocity measurement with 120 fps.
Thanks psv and Chas!
nice serving, but I am not seeing anything over 90 IMHO - in the 80's I would say
Mid 80's is about the speed the pros serve at, right?
Mid 80's is a pretty weak second serve for men pro's I would say, but good for rec non pro levels.
Thanks...I was just kidding about the pro speeds, though.
ah ok Silly me
I just used a stop watch to measure the time between the hit and the bounce. Let's say it's 0.4 seconds. I'm sure with simple software you can measure this more accurately, using the Youtube video. Estimate the distance between hit and bounce (the court dimensions are known). Easy maths. I'd say around 95 mph if it's 0.4 seconds.
Here is a thread where they estimate the serve velocity based on the frame count.
I believe that there a smartphone applications that uses this frame counting principle.
These techniques must give some kind of average velocity from impact to bounce. There may be an approximate correction to get the serve speed right off the racket, which is what is measured at tournaments. [There is a considerable to significant speed loss from air resistance. The amount of speed loss would vary depending on the wind direction and speed.]
If you measured an average serve speed of 95 MPH from impact to bounce ?? then there should be a correction to get the serve speed off the racket.
When measuring with a stop watch there is a considerable error for short times such as 0.4 second. You could also view the video in Quicktime and count single frames between impact and bounce as the thread above describes. For 30 fps a 0.4 seconds flight time is caught in about 12 frames. But you will have an error in counting frames of at least one frame and probably worse since you are unsure of impact and bounce with 30 fps. One frame uncertainty would give and uncertainty of about +/- 8% in serve speed. For a 100 MPH serve the uncertainty would be +/- 8 MPH.
Vimeo & Youtube Video Compression - If a video is taken at 60 fps instead of 30 fps and uploaded I know that Vimeo and probably also Youtube only save 30 fps and discard half the frames. Take care to check the time between frames for any frame counting. This may change as the sites improve their capabilities.
I'm thinking 90+, maybe not quite up to 100.
If you want to measure frames and such, that can give you a pretty good estimate. It will tell you the overall time and then you can use an average speed vs. time equation to estimate the starting velocity. Of course, variations in temperature, humidity, wind, etc. will affect that, but it will give you a pretty good estimate. Remember, that you will get quantized answers. If you have a really fast frame rate, it might not make a lot of difference, but at 30 fps, counting one frame more or less might make a 10mph difference.
100 mph is the baseline serve speed for a ttw poster; 110+ if you've ever thrown a football.
I don´t know your service speed.
What I know is that you are forcing your shoulder and making something strange in that motion.
I played your youtube video in VLC and counted frames using frame advance (e hotkey fwiw) for the first serve you hit and got 15. I put that into this web application, http://donthireddy.us/tennis/speed.html, and came up with initial speed of 96.41 mph.
Based on the video that speed looks reasonable. Nice service motion!
Incredible, Beernutz and I disagree by 1.6 mph! What's wrong with us? Did we even bother to watch the first vid presented?
Really? From the selected clips shown it is an effective rec serve; decent pace.
It is certainly not nice or graceful even by rec standards. Shoulder's relatively flat, elbow low, hand way off to the side, body falling back...
I probably would suggest that he work on his toss as that appears to be affecting his posture and balance especially on his first serve in the video where he seems to be falling backward to catch up with the toss. Even with those issues he managed to generate a nearly 100 mph serve and his later ones look better balance wise though the speed calculates about the same. It looks to me like he gets good upward momentum and leg push even with a bit of back leg "flailing" and that he has pretty good pronation which helps him generate speed.
Disclaimer: his serve is harder than mine.
I don't mean to come across as a tough crowd... heck it's better than more than half I see but it looks muscled vs graceful. (I don't believe it's near 100 but not going to get into another inflated speed thing.) Still, he generates nice pace but with that strike height I wonder if he's hitting 25% on firsts. I know he's got a one hander but he even tosses with two balls in tossing hand. And I see very little leg input; a strong guy who uses his shoulder to muscle in a better than average paced serve.
Disclaimer: I'm just as critical when viewing my own serves....
Looks like a reasonable estimate to me.
Nobody is talking high percentage when serving fast.
30% is liveable and useable for match play, if that 30% always garner's a weak return, and the server has a reliable 2nd serve that doesn't get attacked.
LeeD are you sure about these percentages? I have a very bad tendency to have a low first serve percentage and it gets me killed when I play. When I serve with higher percentages, my percentage of holds go up.
Does your first serve, when it goes IN at 30%, always give you a sitter to start the point?
Is your second serve good enough to NOT be attacked, so you start the point at least even?
Gotta mention, played some dubs with President today. I double faulted FOUR times in one service game, Josh missed two volley putways at net, and we still held every time I served.
I kept thinking "twist", and just tossed the ball up and swung with nothing working upstairs. That can happen when it's TOO easy to hold serve.
Guys! Thanks for all of the discussion. I really appreciate and enjoy it.
I think my first serve percentage is upwards of 50%. I'm super bored at the moment so I'm going to upload all of the serves from that session so I can get a better idea of what my first serve percentage is.
By that logic, all preconditions, I suppose any % will do. Why bother with 30%!?
o.k., just cut up the video so that I only included my serves for the whole session (as long as the recording lasted).
16 first serves were in
16 second serves were in
7 double faults
About a 41% first serve percentage.
Will link the video as soon as the upload is complete.
Thank you for taking the time to do that!
Is that double fault rate normal for you? If so, I think you definitely need to try to bring it down, that's nearly 1 out of every 5 serve points. Obviously getting some more first serves in will help with that, but 7 faults compared to 16 second serves in suggests the second serve needs more margin
I never really tracked how many, so that may be normal for me. It's definitely something I need to work on. I tend to go for it with my second serve and almost nerve "just get it in."
U92626, there has to be some percentage of first serves IN, or it's difficult to count the winners vs out first serves, right?
BTW, I double faulted FOUR times in one service game today, missed most of my first serves (except the aces), President missed a couple of close to net sitter volleys, and I still held serve..
On my other 5 service games, didn't double fault and held easily.
Ignore tennis ocd. The serve percentages have to go up though. I'll look at your form and see if I can spot anything.
Thanks Topspin Shot!
Here is the link, although the video might still be processing.
I think I miscounted. Just counted again and I had 17 first serves in.
Though I may have miscounted this time. lol
Separate names with a comma.