OOAK Table Tennis Forum
http://ooakforum.com/

DHS D40+ three star balls, good news
http://ooakforum.com/viewtopic.php?f=83&t=31070
Page 10 of 10

Author:  iskandar taib [ 09 Aug 2017, 23:08 ]
Post subject:  Re: DHS D40+ three star balls, good news

I'm sure it's being done already.. but they'll tell us about it.. :lol:

Iskandar

Author:  kaesees [ 10 Aug 2017, 00:16 ]
Post subject:  Re: DHS D40+ three star balls, good news

iskandar taib wrote:
I'm sure it's being done already.. but they'll tell us about it.. :lol:

Iskandar



Good point. When the 40+ ball came out the ITTF said it would have 5% less speed and 20% less spin. While these are suspiciously round numbers, I'm sure they came from somewhere.

Author:  iskandar taib [ 10 Aug 2017, 17:56 ]
Post subject:  Re: DHS D40+ three star balls, good news

Uh.. I meant to write "they'll NEVER tell us about it". I'm sure they've been high-speed-photographing these balls (and the seamless ones, and the Nittakus, AND the old DHSs for comparison) in the Butterfly Dojo, at the ITTF research facility in Singapore, at ESN's labs in Germany, among other places. We are but mere dilettantes.

kaesees wrote:
Good point. When the 40+ ball came out the ITTF said it would have 5% less speed and 20% less spin. While these are suspiciously round numbers, I'm sure they came from somewhere.


They did? Where and when exactly did they say this???? I don't remember anything of the sort. (Doesn't mean it didn't happen, of course, but there's so much fake info out there one has to ask..)

Iskandar

Author:  Baal [ 14 Aug 2017, 20:56 ]
Post subject:  Re: DHS D40+ three star balls, good news

I don't know if Pathfinder Pro is able to do this with his camera or not. I know he is getting set to do more testing though. Cameras capable of doing this used to be pretty expensive. I would love to help him. The problem is that half of a large continental land mass and a big ocean separate us, and he is the one with the camera. Also we both have jobs.

I am sure the companies have info on this that is fairly proprietary and of course the newer balls are larger so they pretty much have to spin less and move a bit more slowly also. That is a qualitative statement, not a quantitative one.

Quantitatively accurate data on how the various 40+ balls behave are not generally available so we are left with impressions and numbers pulled out of thin air (or other places).

My somewhat cynical view is that when someone writes (as has been repeated endlessly by a couple of posters over at MyTT for example) that one particular 40+ ball spins 30% less than another, it really means nothing more than that the writer likes it 30% less than the other ball. The like or dislike is a perfectly legitimate opinion of course, but the numbers can't be taken seriously.

Author:  GMan4911 [ 15 Aug 2017, 01:53 ]
Post subject:  Re: DHS D40+ three star balls, good news

What kind of frame rates are needed to do such a test? This http://www.sony.com/electronics/cyber-shot-compact-cameras/dsc-rx10m3/specifications might be able to do the trick and for some people, at $1400 MSRP, it's relatively affordable. Baal, you know you want it. ;)

Author:  kaesees [ 15 Aug 2017, 02:22 ]
Post subject:  Re: DHS D40+ three star balls, good news

GMan4911 wrote:
What kind of frame rates are needed to do such a test?


Back-of-the-envelope guess: a recent EmRatThich video put the number of 3000 RPM out there for the spin rate achievable in table tennis. That comes to 50 revolutions per second. You'd want some multiple of that to avoid aliasing, especially if you're using one of those balls with black bands that visually splits the ball into quarters.

Author:  Baal [ 15 Aug 2017, 08:10 ]
Post subject:  Re: DHS D40+ three star balls, good news

GMan4911 wrote:
What kind of frame rates are needed to do such a test? This http://www.sony.com/electronics/cyber-shot-compact-cameras/dsc-rx10m3/specifications might be able to do the trick and for some people, at $1400 MSRP, it's relatively affordable. Baal, you know you want it. ;)


I made the mistake of taking up road cycling to improve my table tennis fitness, which it certainly does. The problem is that is quite enjoyable and addicting and -- sadly -- insanely expensive once you start getting into it. Much worse than TT!

But yes. I know I want it. :(

Author:  Baal [ 15 Aug 2017, 08:18 ]
Post subject:  Re: DHS D40+ three star balls, good news

kaesees wrote:
GMan4911 wrote:
What kind of frame rates are needed to do such a test?


Back-of-the-envelope guess: a recent EmRatThich video put the number of 3000 RPM out there for the spin rate achievable in table tennis. That comes to 50 revolutions per second. You'd want some multiple of that to avoid aliasing, especially if you're using one of those balls with black bands that visually splits the ball into quarters.


I seem to recall from Nyquist's theorem (distant memory!!) that that means we would want to sample at > 100 frames/sec to avoid aliasing artifact in a ball rotating at 50/s. (Aliasing can make a wheel appear to move backwards when it is being filmed). I have no idea whatsoever if that would be easy to find in a cheap camera.

Edit: the one that was posted above by GMan can do 458 frames/second which might be fast enough. But I don't know enough to know if there might be other issues. If, er, someone ELSE wanted to buy it.

Author:  iskandar taib [ 15 Aug 2017, 14:04 ]
Post subject:  Re: DHS D40+ three star balls, good news

What sort of frame rates are they using here?



At 2:54, 3:46 etc.

Iskandar

Author:  darucla [ 15 Aug 2017, 21:45 ]
Post subject:  Re: DHS D40+ three star balls, good news

Difficult to check because the sequences are so short, but the video FPS is 25 (PAL: standard), and the slow-mo shots are maybe 1/16 normal speed, so perhaps 400 fps shooting speed. (This is after very little actual testing. :oops: )

PS: After another 20 seconds of research, I feel it is probably more than 16, but less than 32. My playback system replays 16x speed quite smoothly, and the half second or so of the action shots suggests they are still slowed down. The system jumps at 32x playback, making that useless.

Author:  darucla [ 15 Aug 2017, 22:32 ]
Post subject:  Re: DHS D40+ three star balls, good news

There are a number of compact cameras that offer high speed shooting. This one https://www.amazon.co.uk/digital-camera ... B00MVIDG6E sells for £220, so almost nothing outside of the UK :lol: and it apparently shoots up to 1000 fps. I suspect that the resolution at that speed will be very poor though. I am going to try to find actual footage somewhere.

Edit: Actual specs. Movies:FHD:1920x1080(30fps) / HD:1280x720(12fps/15fps/20fps/30fps)* / STD:640x480(30fps) / HS 1000:224x64(1000fps) / HS 480:224x160(480fps) / HS 240:512x384(240fps) / HS 120:640x480(120fps) / HS 30-240:512x384(30 to 240fps) / HS 30-120:640x480(30 to 120fps)

Author:  iskandar taib [ 16 Aug 2017, 12:41 ]
Post subject:  Re: DHS D40+ three star balls, good news

darucla wrote:
Difficult to check because the sequences are so short, but the video FPS is 25 (PAL: standard), and the slow-mo shots are maybe 1/16 normal speed, so perhaps 400 fps shooting speed. (This is after very little actual testing. :oops: )

PS: After another 20 seconds of research, I feel it is probably more than 16, but less than 32. My playback system replays 16x speed quite smoothly, and the half second or so of the action shots suggests they are still slowed down. The system jumps at 32x playback, making that useless.


I wouldn't call them useless. You can still estimate RPM even though the spots are blurred. And you can detect aliasing if you use spots of different colors or sizes.

Iskandar

Author:  darucla [ 16 Aug 2017, 19:23 ]
Post subject:  Re: DHS D40+ three star balls, good news

Trying to replay at 32x leads to the actual sequences sometimes not being shown, because at that speed my playback system does not play continuously but instead jumps from one place to the next, or shows just a quick image of the slo-mo shot. Also, it really isn't designed to do the job, and is only showing a sequence of stills at 32x, which makes my system useless in this case. In any case, the sequence is only about 8 seconds in normal playback, so half a second at 16x and quarter second at 32x. I'm only using my eyes to gauge the speed. But I'm fairly confident that the sequence is shot between 1/16 and 1/32 speed; i.e. 400 -800 fps.

Still haven't found any actual slo-mo footage from one of the cheaper cameras.

Page 10 of 10 All times are UTC + 9:30 hours
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/