PDA

View Full Version : cue stick accelerometer measurement results



dr_dave
03-08-2005, 03:11 PM
FYI, I just posted on my website (http://www.engr.colostate.edu/~dga/pool/) some recent results from accelerometer measurements of cue stick reactions during strokes and object ball impacts. A photo, several plots, and a descriptive analysis can be found in TP A.9 (http://www.engr.colostate.edu/~dga/pool/technical_proofs/new/TP_A-9.pdf). I would be curious if others interpret the plots differently from what I've written. Please share any insights or comments you might have.

Regards,
Dr. Dave

PS: HSV A.34 (http://www.engr.colostate.edu/~dga/pool/high_speed_videos/new/HSVA-34.htm) might help provide some insight on the effect the hand grip has on the cue stick acceleration after impact.

SpiderMan
03-08-2005, 04:15 PM
<blockquote><font class="small">Quote dr_dave:</font><hr> FYI, I just posted on my website (http://www.engr.colostate.edu/~dga/pool/) some recent results from accelerometer measurements of cue stick reactions during strokes and object ball impacts. A photo, several plots, and a descriptive analysis can be found in TP A.9 (http://www.engr.colostate.edu/~dga/pool/technical_proofs/new/TP_A-9.pdf). I would be curious if others interpret the plots differently from what I've written. Please share any insights or comments you might have.

Regards,
Dr. Dave

PS: HSV A.34 (http://www.engr.colostate.edu/~dga/pool/high_speed_videos/new/HSVA-34.htm) might help provide some insight on the effect the hand grip has on the cue stick acceleration after impact. <hr /></blockquote>

Dave,

It's fascinating.

Do you think that the deceleration sometimes appears to be spread over a substantial time (relative to the millisecond-or-so accepted ball-contact time) because the shooter is "pulling" his stroke? That surprises me, as the shooter was of professional caliber.

Or are we seeing a measurement artifact because the response time of the recorder is too slow, essentially "spreading" the negative-going impulse?

Actually it appears that in the "endrunner gentle" and "central strong" graphs, he apparently was accelerating at the impact because the downward slope is nearly as steep in the positive region as in the negative. In the "drawback soft" graph, the acceleration drops slowly, so that would have to be a stroke pull rather than impact.

So, after thinking out loud, I'd say the response time of the recorder was not the culprit in plot #2, it was the shooter's fundamentals.

Again, fascinating material!

SpiderMan

wolfdancer
03-08-2005, 05:23 PM
Dr. Dave, I'm thinking you folks got too much free time on your hands over there at CU......you becoming the Michael moore of pool, old Ward Churchill even ticking off the republicans, and wasn't there some scandal with the football program?...And now poor old Liz Hoffman has thrown in the towel.
Anyways, do you have time from your busy schedule, shooting pool videos, to do a little teaching???
Hey, you need to add some time to these vids, to spice them up. Just about when they get interesting, it's over....which is why my first left me................
It's all good stuff though, enlightening...and I think Speilberg might be looking over his shoulder lol

Bob_Jewett
03-08-2005, 06:05 PM
<blockquote><font class="small">Quote wolfdancer:</font><hr> Dr. Dave, I'm thinking you folks got too much free time on your hands over there at CU.... <hr /></blockquote>
Perhaps Coloradans are less sensitive to this sort of thing than those in some states, but you may want to look into the difference between Boulder and Ft. Collins. No pool balls were harmed during football team recruiting at Colorado State.

wolfdancer
03-08-2005, 06:11 PM
bob, you mean I got the wrong school????
Damn!!!!!!!!!!!!!!!

wolfdancer
03-08-2005, 06:16 PM
"Or are we seeing a measurement artifact because the response time of the recorder is too slow, essentially "spreading" the negative-going impulse?"

"Actually it appears that in the "endrunner gentle" and "central strong" graphs, he apparently was accelerating at the impact because the downward slope is nearly as steep in the positive region as in the negative. In the "drawback soft" graph, the acceleration drops slowly, so that would have to be a stroke pull rather than impact."
So it IS rocket science?????

I was already confused before I read your post....it should be illegal for anyone with an engineering background to post on pool.

Bob_Jewett
03-08-2005, 06:41 PM
<blockquote><font class="small">Quote dr_dave:</font><hr>... Please share any insights or comments you might have.

Regards,
Dr. Dave
... <hr /></blockquote>
I think there is something screwy with the plots. The very sharp acceleration just before deceleration is particularly surprising and makes me wonder whether there might not be something loose in the mounting. It would have been better if they could have calibrated the setup by having the stick swinging like a pendulum on strings for a test impact. Another problem is that each of the "retreats" on the first plot has a glitch right in the middle. That also makes me think that something was loose, as I don't believe that a normal stroke will have such funny bumps on it. The setup with the pendulum could also have been tried against a very heavy steel block which has a well-known result.

On the "soft draw back shot" it appears that the stick hits the ball after the peak of acceleration. That is, the tip-ball collision is probably at the time of the sharp doublet. That is also the time that the acceleration has apparently gone to zero.

Also, whatever graphics package is being used to plot the traces should be trained to move time=0 to the moment of impact and to choose reasonable units for the time scale. Like 0.05 seconds between tick marks. The time scale as presented is pretty much useless. How far is a millisecond?

It should be possible to integrate acceleration to get velocity and to integrate that to get position. I think at least the velocity plot would be useful.

It would have been very, very useful to have a correlation between video stills and particular points on the graph, especially if the video time resolution were fine enough to see exactly when the ball had left the tip. The time resolution on the plots seems to be quite coarse compared to the time of interest which is considerably less than a millisecond. Longitudinal vibrations during tip contact, if they occur, are expected to have periods around 600useconds since the speed of sound in wood is about 4000 meters/second, and the sampling rate should be at least four times per cycle to begin to get a useful waveform.

Leviathan
03-08-2005, 08:09 PM
'Lo, Spiderman. Have you watched Dr. Dave's HSV A.36? It shows a "draw shot with large offset, light grip, good follow-through, slow speed, and about 2.25 feet of draw." It appears to show that the cue tip remains in contact with the ball (or possibly chatters on the ball) for a heck of a long time during such a shot. Is it possible that the cue's deceleration is slower here simply because it takes a while to accelerate the cb to its max speed with a soft draw stroke with a big offset?

On another subject, the cue tip often touches or slides over the cloth after contact with the cb in a draw stoke, and I wonder whether this could affect the cue's vibration pattern.

AS

dr_dave
03-08-2005, 08:23 PM
<blockquote><font class="small">Quote SpiderMan:</font><hr>Do you think that the deceleration sometimes appears to be spread over a substantial time (relative to the millisecond-or-so accepted ball-contact time) because the shooter is "pulling" his stroke? That surprises me, as the shooter was of professional caliber.<hr /></blockquote>I can't imagine that the shooter was "pulling" his stroke. I will ask the experimenters for feedback on this and other points people have brought up. I am hoping that they will participate in this discussion.

<blockquote><font class="small">Quote SpiderMan:</font><hr>Actually it appears that in the "endrunner gentle" and "central strong" graphs, he apparently was accelerating at the impact because the downward slope is nearly as steep in the positive region as in the negative. In the "drawback soft" graph, the acceleration drops slowly, so that would have to be a stroke pull rather than impact.<hr /></blockquote>In all 5 plots, it looks to me that the acceleration is fairly close to zero at impact, which would mean that the speed is close to maximum at impact. Cue tip impact is where the spikes appear in both curves. I'm not sure I answered your question or not. Let me know if I didn't.

Regards,
Dave

dr_dave
03-08-2005, 08:35 PM
<blockquote><font class="small">Quote wolfdancer:</font><hr> Dr. Dave, I'm thinking you folks got too much free time on your hands over there at CU<hr /></blockquote>
You've got the wrong university. /ccboard/images/graemlins/frown.gif I'm in Fort Collins at Colorado State University (CSU). GO RAMS!!! No offense taken.

<blockquote><font class="small">Quote wolfdancer:</font><hr>you becoming the Michael moore of pool, old Ward Churchill even ticking off the republicans, and wasn't there some scandal with the football program?...And now poor old Liz Hoffman has thrown in the towel.<hr /></blockquote>
Again, wrong school. People in Boulder are a little weirder than the people in Fort Collins. If you ever visit both places, you would know what I mean.

<blockquote><font class="small">Quote wolfdancer:</font><hr>Anyways, do you have time from your busy schedule, shooting pool videos, to do a little teaching???<hr /></blockquote>
I do teach (mechanical engineering), and I love it. I am also working on the 3rd edition of a textbook right now. I would describe what it is about, but I doubt anybody (except maybe Bob) would be interested. If you are really curious about what I teach and other stuff I work on, you can check out my various websites (http://www.engr.colostate.edu/~dga/dga/web_pages.html). All of that stuff is my work ... pool is my passion! /ccboard/images/graemlins/smile.gif

Regards,
Dave

dr_dave
03-08-2005, 08:50 PM
<blockquote><font class="small">Quote Bob_Jewett:</font><hr>I think there is something screwy with the plots. The very sharp acceleration just before deceleration is particularly surprising and makes me wonder whether there might not be something loose in the mounting.<hr /></blockquote>I don't think there is any deceleration before impact, just less acceleration. The acceleration seems to be very close to zero at impact in all 5 cases. I think impact occurs at the sharp spikes.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr>It would have been better if they could have calibrated the setup by having the stick swinging like a pendulum on strings for a test impact.<hr /></blockquote>
Agreed. I will ask the experimenters to try this out. It would be interesting to see the comparison.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr>Another problem is that each of the "retreats" on the first plot has a glitch right in the middle. That also makes me think that something was loose, as I don't believe that a normal stroke will have such funny bumps on it.<hr /></blockquote>
These mid-bump "glitches" could be a result of the hand grip interacting with the cue stick as the grip changes slightly or as the flesh flexes, but I'm not sure. Again, I'll ask the experimenters for feedback.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr>On the "soft draw back shot" it appears that the stick hits the ball after the peak of acceleration. That is, the tip-ball collision is probably at the time of the sharp doublet. That is also the time that the acceleration has apparently gone to zero.<hr /></blockquote>
I agree. That means the cue stick has maximum speed at impact, which would be expected with a good stroke.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr>Also, whatever graphics package is being used to plot the traces should be trained to move time=0 to the moment of impact and to choose reasonable units for the time scale. Like 0.05 seconds between tick marks. The time scale as presented is pretty much useless. How far is a millisecond?<hr /></blockquote>
That's a great suggestion.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr>It should be possible to integrate acceleration to get velocity and to integrate that to get position. I think at least the velocity plot would be useful.<hr /></blockquote>
Agreed. I'll ask them if they can provide these additional plots from the data.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr>It would have been very, very useful to have a correlation between video stills and particular points on the graph, especially if the video time resolution were fine enough to see exactly when the ball had left the tip.<hr /></blockquote>
Agreed. The Austrian group has sent me a DVD that they claim has video and acceleration data superimposed. When I get the DVD, I will ask them to let me post a video clip of this.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr>The time resolution on the plots seems to be quite coarse compared to the time of interest which is considerably less than a millisecond. Longitudinal vibrations during tip contact, if they occur, are expected to have periods around 600useconds since the speed of sound in wood is about 4000 meters/second, and the sampling rate should be at least four times per cycle to begin to get a useful waveform.<hr /></blockquote>
Agreed.

Bob, Thanks a bunch for your insight. I'm sure the experimenters will appreciate it.

Regards,
Dave

eg8r
03-09-2005, 06:27 AM
[ QUOTE ]
It would have been very, very useful to have a correlation between video stills and particular points on the graph, especially if the video time resolution were fine enough to see exactly when the ball had left the tip. The time resolution on the plots seems to be quite coarse compared to the time of interest which is considerably less than a millisecond. Longitudinal vibrations during tip contact, if they occur, are expected to have periods around 600useconds since the speed of sound in wood is about 4000 meters/second, and the sampling rate should be at least four times per cycle to begin to get a useful waveform. <hr /></blockquote> LOL, anyone else have a headache after reading this? /ccboard/images/graemlins/grin.gif

eg8r

Wally_in_Cincy
03-09-2005, 07:17 AM
<blockquote><font class="small">Quote eg8r:</font><hr> LOL, anyone else have a headache after reading this? /ccboard/images/graemlins/grin.gif

eg8r <hr /></blockquote>

I find it difficult to care enough to develop a headache /ccboard/images/graemlins/smile.gif

To each his own though. If I had finished engineering school I might have found it worthwhile too.

SpiderMan
03-09-2005, 08:48 AM
<blockquote><font class="small">Quote Bob_Jewett:</font><hr>
I think there is something screwy with the plots. The very sharp acceleration just before deceleration is particularly surprising and makes me wonder whether there might not be something loose in the mounting. <hr /></blockquote>
I don't see anything screwy with that particular data. Remember, this is acceleration and not velociity, therefore even a nearly-vertical trace could merely represent the beginning of a smooth ramp in velocity. Also, notice that the peak acceleration is only 3 G's, which only represents a force of less than 4 lbs being applied to the butt of the cue.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr> On the "soft draw back shot" it appears that the stick hits the ball after the peak of acceleration. That is, the tip-ball collision is probably at the time of the sharp doublet. That is also the time that the acceleration has apparently gone to zero.
<hr /></blockquote>
Yes, Dave stated that in his earlier response to my post, and I think I agree.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr>
Also, whatever graphics package is being used to plot the traces should be trained to move time=0 to the moment of impact and to choose reasonable units for the time scale. Like 0.05 seconds between tick marks. The time scale as presented is pretty much useless. How far is a millisecond? <hr /></blockquote>
Second that emotion!!! My thoughts exactly when I first looked at the plots last night. I had to pull out a calculator to figure out how much time I was looking at between nonexistent "grids". Dave, would you accept that in a lab report from one of your students? /ccboard/images/graemlins/grin.gif

<blockquote><font class="small">Quote Bob_Jewett:</font><hr> Longitudinal vibrations during tip contact, if they occur, are expected to have periods around 600useconds since the speed of sound in wood is about 4000 meters/second, and the sampling rate should be at least four times per cycle to begin to get a useful waveform.<hr /></blockquote>
Not included are any specifications on the measurement equipment. We don't have the bandwidth, sensitivity, linearity, etc of the accelerometer. It's possible some of the glitches could be measurement artifacts.

Despite the nitpicks, I really think Dr Dave is doing a great service. I'm glad he's putting all this effort into collecting data that can provide insight into things we've speculated and theorized about for years.

SpiderMan

SpiderMan
03-09-2005, 08:52 AM
<blockquote><font class="small">Quote Wally_in_Cincy:</font><hr> <blockquote><font class="small">Quote eg8r:</font><hr> LOL, anyone else have a headache after reading this? /ccboard/images/graemlins/grin.gif
eg8r <hr /></blockquote>
I find it difficult to care enough to develop a headache /ccboard/images/graemlins/smile.gif
To each his own though. If I had finished engineering school I might have found it worthwhile too. <hr /></blockquote>
I guess you hit the nail on the head, Wally. There are nerds /ccboard/images/graemlins/shocked.gif on here that are fascinated by this stuff /ccboard/images/graemlins/grin.gif

SpiderNerd

Bob_Jewett
03-09-2005, 11:38 AM
<blockquote><font class="small">Quote dr_dave:</font><hr>... Please share any insights or comments you might have. ... <hr /></blockquote>
It's unfortunate that velocity info is not available for the plots.

The "Central Strong" shot has the largest velocity of the shots shown. I calculate that the change in speed from going back to going forward on the final backward practice stroke is 1g*0.2seconds more or less, so the velocity is roughly +-1 meter per second on the velocity peaks of the practice stroke swings. The integral of the forward acceleration looks to be about the same or a little larger area for that shot, so the speed of the stick at impact might be 1.5 meters/second. For the usual ball/stick mass ratios, the stick will slow to 0.75 m/s during the 0.001s collision time for an average deceleration during that time of 750 meters /second /second. That's over 70g's of average deceleration, and the peak is expected to be about twice that. There is no large stick deceleration shown at the time of impact.

The measuring apparatus is broken.

[As for the seemingly inordinate interest some participants in this thread show in the physics of shots, is this any less worthwhile than the endless discussions of whether a Lucabushka is better than McShawn, or whether a TotalDestroyer jump/break cue is better than a Happy Toad? Maybe the geeks should invade those discussions and point out that the participants are risking Repetitive Stress Injury for nothing useful.]

dr_dave
03-09-2005, 11:44 AM
<blockquote><font class="small">Quote Bob_Jewett:</font><hr>It's unfortunate that velocity info is not available for the plots.<hr /></blockquote>
Agreed. I've asked the experimenters for this.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr>For the usual ball/stick mass ratios, the stick will slow to 0.75 m/s during the 0.001s collision time for an average deceleration during that time of 750 meters /second /second. That's over 70g's of average deceleration, and the peak is expected to be about twice that. There is no large stick deceleration shown at the time of impact.<hr /></blockquote>
I don't think their sensors can measure peaks that large. Also, I don't think their sampling rate is fast enough to capture the complex dynamics occurring during the extremely short impact period. Again, I've asked the experimenters to provide complete specifications of their systems. I'll report the data when I receive it.

Regards,
Dave

DickLeonard
03-09-2005, 12:52 PM
Dr.Dave the fellow I knew who put all nineballs in on the break was 5ft5 and 120 lbs, when he broke, his thumb and first two fingers dangled towards the floor and the cue rested on his ring and baby finger. He made no violent move with his body just delivered the cue and it sounded as if a bomb went off.####

SpiderMan
03-09-2005, 12:58 PM
<blockquote><font class="small">Quote Bob_Jewett:</font><hr> <blockquote><font class="small">Quote dr_dave:</font><hr>... Please share any insights or comments you might have. ... <hr /></blockquote>
It's unfortunate that velocity info is not available for the plots.

The "Central Strong" shot has the largest velocity of the shots shown. I calculate that the change in speed from going back to going forward on the final backward practice stroke is 1g*0.2seconds more or less, so the velocity is roughly +-1 meter per second on the velocity peaks of the practice stroke swings. The integral of the forward acceleration looks to be about the same or a little larger area for that shot, so the speed of the stick at impact might be 1.5 meters/second. For the usual ball/stick mass ratios, the stick will slow to 0.75 m/s during the 0.001s collision time for an average deceleration during that time of 750 meters /second /second. That's over 70g's of average deceleration, and the peak is expected to be about twice that. There is no large stick deceleration shown at the time of impact.

The measuring apparatus is broken.<hr /></blockquote>
Rather than "broken", I'd suggest that the measurement system probably just lacks the necessary frequency response and/or dynamic range to capture the peak that occurs at ball contact. The other events appear to provide reasonable data.

If you could view and integrate the entire sequence, the time integral of the entire acceleration plot should sum to zero. This is assuming that the recorder captures all events between the stick starting at zero velocity and eventually stopping after follow-through. And assuming you could accurately glean numeric data from the odd scale.

Since a large velocity change occurs at impact, this means that the area under that very-short-time portion of the curve would need to be significant, therefore the peak would be 'way off scale in this plot. That's the only event that I see where the data appears to not look reasonable, so most likely the measurement system clipped the magnitude, couldn't track the rate, or both.

Note that if you limit integration to the time frame of warm-up strokes, it (qualitatively) appears that the areas above and below zero acceleration do cancel. In the region of the "stroke", "hit", and "follow-through", there is "negative acceleration" area missing. I suggest that this is the lack of instrument response in recording the peak, but not necessarily a broken instrument.

<blockquote><font class="small">Quote Bob_Jewett:</font><hr> [As for the seemingly inordinate interest some participants in this thread show in the physics of shots, is this any less worthwhile than the endless discussions of whether a Lucabushka is better than McShawn, or whether a TotalDestroyer jump/break cue is better than a Happy Toad? Maybe the geeks should invade those discussions and point out that the participants are risking Repetitive Stress Injury for nothing useful.] <hr /></blockquote>
Damn right!

SpiderMan

dr_dave
03-09-2005, 01:15 PM
<blockquote><font class="small">Quote DickLeonard:</font><hr> Dr.Dave the fellow I knew who put all nineballs in on the break was 5ft5 and 120 lbs, when he broke, his thumb and first two fingers dangled towards the floor and the cue rested on his ring and baby finger. He made no violent move with his body just delivered the cue and it sounded as if a bomb went off.<hr /></blockquote>
All 9 balls on the break! That's awesome! I would love to see a video of that (and post it on my website for others to see). Any chance of getting a clip?

Some of the small players from the Phillipines also have sledgehammer breaks. Size and weight aren't everything. Things that really matter include technique, skill, muscle physiology, and having a cue weight well matched to the person.

Regards,
Dr. Dave

SpiderMan
03-09-2005, 01:15 PM
<blockquote><font class="small">Quote wolfdancer:</font><hr> "Or are we seeing a measurement artifact because the response time of the recorder is too slow, essentially "spreading" the negative-going impulse?"

"Actually it appears that in the "endrunner gentle" and "central strong" graphs, he apparently was accelerating at the impact because the downward slope is nearly as steep in the positive region as in the negative. In the "drawback soft" graph, the acceleration drops slowly, so that would have to be a stroke pull rather than impact."

So it IS rocket science?????
I was already confused before I read your post....it should be illegal for anyone with an engineering background to post on pool. <hr /></blockquote>

Sorry, I'm a nerd. But I still like to play pool.

SpiderMan

JohnnyP
03-09-2005, 04:12 PM
Hello, Bob.

Seems like we had a conversation like this last summer.

What is the rise time of the 100g deceleration? I have the accelerometers damped, to filter out those "pesky" high frequency signals.

(Don't want to enable Java to get those yellow smiley thingys)

Bob_Jewett
03-09-2005, 04:34 PM
<blockquote><font class="small">Quote SpiderMan:</font><hr>
Rather than "broken", I'd suggest that the measurement system probably just lacks the necessary frequency response and/or dynamic range to capture the peak that occurs at ball contact. The other events appear to provide reasonable data. <hr /></blockquote>
A good data acquisition system will have an anti-alias filter that will take care of the bandwidth issue. The plots shown lack a major, major feature of the event -- the tip hitting the ball. In fact, no large decelleration at the time of impact is shown. The large, narrow spike with ringing is positive (at first) as if the stick were accelerating during impact. I think that is not what occurs.

In the same way, people tried to draw conclusions about sticks hitting balls from standard video and slow playback. It's only when you can see the action at several thousand frames per second (as with the amazing video from Austria), that you can see what's really going on and start to understand it, such as the improved understanding of squirt that fast video has enabled.

SpiderMan
03-09-2005, 04:40 PM
<blockquote><font class="small">Quote Bob_Jewett:</font><hr> <blockquote><font class="small">Quote SpiderMan:</font><hr>
Rather than "broken", I'd suggest that the measurement system probably just lacks the necessary frequency response and/or dynamic range to capture the peak that occurs at ball contact. The other events appear to provide reasonable data. <hr /></blockquote>
A good data acquisition system will have an anti-alias filter that will take care of the bandwidth issue. The plots shown lack a major, major feature of the event -- the tip hitting the ball. In fact, no large decelleration at the time of impact is shown. The large, narrow spike with ringing is positive (at first) as if the stick were accelerating during impact. I think that is not what occurs.

In the same way, people tried to draw conclusions about sticks hitting balls from standard video and slow playback. It's only when you can see the action at several thousand frames per second (as with the amazing video from Austria), that you can see what's really going on and start to understand it, such as the improved understanding of squirt that fast video has enabled. <hr /></blockquote>
See the JohnnyP post - overdamping may be responsible for the frequency-response issue.

Can you post a link to the Austrian video you referenced?

SpiderMan

DickLeonard
03-09-2005, 05:25 PM
Dr.Dave that was 50 years or so before tv cameras. That was done before the Brunswick Centennials with the old clay balls. He also put 7 of 9 twice, once getting stitched with the nineball.

He was just before my era but he came into the poolroom I was running. I asked him if that story was true and he said yes and gave me his history.

Then he gave me a three hour breaking demonstration the most balls he made was 6 with the least being 2. All with staying inside his stance.####

wolfdancer
03-09-2005, 05:53 PM
Bob, it's all interesting....just over my head, but I get the general idea of what's being described.
In fact, since someone posted about a re-accelerating stroke
( which may be covered in the book "The Final theory: Rethinking Our Scientific Legacy" )....the thread may save players here endless hours of trying to duplicate that stroke.
So, in your professional opinion is the McShawn better'n the Lucabushka, and hopefully it's not corked.
I thought the "Limbsaver" was supposed to prevent the repetitive stress injuries caused by the numbing forces of off-center hits?

wolfdancer
03-09-2005, 06:13 PM
It may be nerdy.....but if you could compare a top player's stroke to, say mine, I think the graphs would show how they are able to move the cueball almost effortlessly, while I struggle to get any "action"

wolfdancer
03-09-2005, 06:16 PM
"LOL, anyone else have a headache after reading this? "

eg8r, no, this was one of your milder posts...it's only when you discuss politics, that I get a headache

Bob_Jewett
03-09-2005, 06:59 PM
<blockquote><font class="small">Quote SpiderMan:</font><hr>...
Can you post a link to the Austrian video you referenced? <hr /></blockquote>
That's the one with the high-speed color clips and the infrared shots that has been discussed elsewhere. It was made by a group in Vienna. It's a Windows Media Video format file of about 3MB:
http://www.bskunion.at/efler/Faszination_Dreiband_Windows_Media_9.wmv

SpiderMan
03-10-2005, 07:44 AM
<blockquote><font class="small">Quote Bob_Jewett:</font><hr> http://www.bskunion.at/efler/Faszination_Dreiband_Windows_Media_9.wmv <hr /></blockquote>

Pretty good photography. Coupla questions - (1) I could play that file but not save it to my hard drive. Any ideas? (2) The "rail running" clip just after the thermal imaging of the masse shot didn't look "real". Is it an animation?

SpiderMan

1robby
03-10-2005, 09:00 AM
Dear Spiderman!

The "rail running" clip is real! We have used a very special OEM-camera with a diameter of 5mm mounted on a teflon-plate. So we could film under the cushion!

Best Regards from Austria!

Robert

SPetty
03-10-2005, 09:40 AM
<blockquote><font class="small">Quote SpiderMan:</font><hr> I could play that file but not save it to my hard drive. Any ideas?<hr /></blockquote>Funny - I saved it to my hard drive before I read your post... I just right clicked on Bob's link and did "Save Target As...".

Fun video!

SpiderMan
03-10-2005, 10:36 AM
<blockquote><font class="small">Quote 1robby:</font><hr> Dear Spiderman!

The "rail running" clip is real! We have used a very special OEM-camera with a diameter of 5mm mounted on a teflon-plate. So we could film under the cushion!

Best Regards from Austria!

Robert <hr /></blockquote>

Robert,

How did you keep the vantage point so accurately at a constant distance ahead of the rolling ball? Did you write an algorithm to control digital zoom?

Surely you didn't physically slide the camera with such accuracy?

This video is very interesting, it really deserves to be discussed in a thread of it's own. Do you have more?

SpiderMan

JohnnyP
03-10-2005, 11:14 AM
Robert:

I'm running WMP free version 7 on my iMac, and cannot see your video. The player runs but the screen is black.

I checked the video encoding, and it says WMP version 9.

Does that mean it can't play on the free version?

1robby
03-10-2005, 11:50 AM
Spiderman!

I think you won´t believe me, but i really physically slided the camera under the cushion!! This was team-work with Andreas Efler (Austrian 3-C-Professional).

We have produced a DVD about Carom with a lot of Highspeed - and Infrared - Videos. All the background music (ambient/chillout-sound) was composed by Sascha Borovcanin from Swizerland. There is no speech on the DVD. We want to show the difficultest discipline of billiards from a new and different view. You can read a short note on www.caromtv.com (http://www.caromtv.com) (at FreeViews)

1robby
03-10-2005, 11:56 AM
JohnnyP!

You´re right! It only plays with the WM-9-version. I have cut and encoded it with the actual version from Adobe Premiere Pro (licensed version).

Robert

Wally_in_Cincy
03-11-2005, 06:48 AM
<blockquote><font class="small">Quote Bob_Jewett:</font><hr>
[As for the seemingly inordinate interest some participants in this thread show in the physics of shots, is this any less worthwhile than the endless discussions of whether a Lucabushka is better than McShawn, or whether a TotalDestroyer jump/break cue is better than a Happy Toad? Maybe the geeks should invade those discussions and point out that the participants are risking Repetitive Stress Injury for nothing useful.] <hr /></blockquote>

Geez Bob don't take it personal /ccboard/images/graemlins/smile.gif

I say if you want to discuss physics have at it. After all the whole game is nothing but physics.

BTW, I don't read the McJoss v. Furyasi or the Sludgebammer v. Ballcrusher or the "Are Sumatran Albino Pigskin Tips really worth it" threads either.

dr_dave
03-11-2005, 09:07 AM
<blockquote><font class="small">Quote DickLeonard:</font><hr> Dr.Dave that was 50 years or so before tv cameras. That was done before the Brunswick Centennials with the old clay balls. He also put 7 of 9 twice, once getting stitched with the nineball.

He was just before my era but he came into the poolroom I was running. I asked him if that story was true and he said yes and gave me his history.<hr /></blockquote>
I wish I could have been there.

Have you, or has anyone else, seen someone pocket all 9 balls on the break? I've seen 6 many times and 7 a few times.

eg8r
03-11-2005, 09:16 AM
[ QUOTE ]
Pretty good photography. Coupla questions - (1) I could play that file but not save it to my hard drive. Any ideas? <hr /></blockquote> Just right-click on the link he provided and choose Save target as.

Sorry if someone already answered this. /ccboard/images/graemlins/smile.gif

eg8r

dr_dave
03-11-2005, 10:52 AM
<blockquote><font class="small">Quote Bob_Jewett:</font><hr>I think there is something screwy with the plots. The very sharp acceleration just before deceleration is particularly surprising and makes me wonder whether there might not be something loose in the mounting.<hr /></blockquote>
Bob, Here's a response I got back relating to your concern:
<blockquote><font class="small">Quote Austrian_experimenter:</font><hr>
The sensor acceleration board was fixed with a dual-component-compound in the block made of carbon fibre reinforced plastic. The connection fitted very strong and exactly into the end of the queue, so there was nothing, really nothing loose at this mounting!<hr /></blockquote>

dr_dave
03-11-2005, 11:11 AM
<blockquote><font class="small">Quote dr_dave:</font><hr>I don't think their sensors can measure peaks that large. Also, I don't think their sampling rate is fast enough to capture the complex dynamics occurring during the extremely short impact period. Again, I've asked the experimenters to provide complete specifications of their systems. I'll report the data when I receive it.<hr /></blockquote>
Here's a follow-up to my earlier message:

I just found out from the Austrian group some specifications for their accelerometer system. They used the Analog Devices ADXL202 (http://www.analog.com/en/prod/0%2C2877%2CADXL202%2C00.html) evaluation board. The sensor guarantees accurate readings only in the +/- 2g range and the maximum sampling rate is only 275kHz, so the sensor is not capable of capturing the physics occurring during the 0.001 second impact time. For more information, refer to the online datasheet (http://www.analog.com/en/prod/0%2C2877%2CADXL202%2C00.html).

PS: The sampling interval used for the experiments used to create the top three plots in TP A.9 (http://www.engr.colostate.edu/~dga/pool/technical_proofs/new/TP_A-9.pdf) was 0.008 seconds (relatively slow)! I don't have details yet on the bottom two plots provided by John Pizutto.

SpiderMan
03-11-2005, 11:14 AM
That's what I was thinking, a frequency-response issue at impact. Still very interesting data without the "missing millisecond".

SpiderMan

SpiderMan
03-11-2005, 11:17 AM
<blockquote><font class="small">Quote dr_dave:</font><hr> <blockquote><font class="small">Quote Bob_Jewett:</font><hr>I think there is something screwy with the plots. The very sharp acceleration just before deceleration is particularly surprising and makes me wonder whether there might not be something loose in the mounting.<hr /></blockquote>
Bob, Here's a response I got back relating to your concern:
<blockquote><font class="small">Quote Austrian_experimenter:</font><hr>
The sensor acceleration board was fixed with a dual-component-compound in the block made of carbon fibre reinforced plastic. The connection fitted very strong and exactly into the end of the queue, so there was nothing, really nothing loose at this mounting!<hr /></blockquote> <hr /></blockquote>

Nothing broken, nothing loose, just a dynamic-range and frequency-reponse issue. Very usable results (for whatever us nerds tend to do with it).

SpiderMan

dr_dave
03-11-2005, 11:18 AM
Robert,

Thank you for participating in this discussion directly. Maybe you can respond to other concerns or questions that come up concerning your plots in TP A.9 (http://www.engr.colostate.edu/~dga/pool/technical_proofs/new/TP_A-9.pdf) or your sample high-speed clips at: HSVA.76.htm (http://www.engr.colostate.edu/~dga/pool/high_speed_videos/new/HSVA-76.htm).

Thanks again for you efforts and great results. Keep up the good work.

Regards,
Dave

dr_dave
03-11-2005, 11:20 AM
<blockquote><font class="small">Quote SpiderMan:</font><hr> <blockquote><font class="small">Quote dr_dave:</font><hr> <blockquote><font class="small">Quote Bob_Jewett:</font><hr>I think there is something screwy with the plots. The very sharp acceleration just before deceleration is particularly surprising and makes me wonder whether there might not be something loose in the mounting.<hr /></blockquote>
Bob, Here's a response I got back relating to your concern:
<blockquote><font class="small">Quote Austrian_experimenter:</font><hr>
The sensor acceleration board was fixed with a dual-component-compound in the block made of carbon fibre reinforced plastic. The connection fitted very strong and exactly into the end of the queue, so there was nothing, really nothing loose at this mounting!<hr /></blockquote> <hr /></blockquote>

Nothing broken, nothing loose, just a dynamic-range and frequency-reponse issue. Very usable results (for whatever us nerds tend to do with it).

SpiderMan <hr /></blockquote>
Well stated. /ccboard/images/graemlins/smile.gif

Dave

dr_dave
03-11-2005, 11:22 AM
<blockquote><font class="small">Quote SpiderMan:</font><hr>Still very interesting data without the "missing millisecond".<hr /></blockquote>
I agree completely! I found the plots very thought-provoking. I hope the Austrian group will continue to share their results.

Regards,
Dave

Bob_Jewett
03-11-2005, 01:40 PM
<blockquote><font class="small">Quote SpiderMan:</font><hr>... Nothing broken, nothing loose, just a dynamic-range and frequency-reponse issue. Very usable results (for whatever us nerds tend to do with it).
... <hr /></blockquote>
I still strongly disagree. The plot, because it is missing the most important feature, is very misleading. You cannot draw useful conclusions from it. With that obvious problem, I continue to wonder about the reality of the glitches in the warmup strokes that I've mentioned before. What other parts of the actual waveform are missing or distorted?

I hope Robby Leitner and company continue to do experiments, because their results so far are very interesting, but it is important to note that the ADXL202E is the wrong sensor to use for this purpose, due to its limited full-scale range. The bandwidth of the chip (around 6kHz) actually seems to be high enough to catch the peak during tip contact.

JohnnyP
03-11-2005, 04:08 PM
Bob:

I miss shots because my stroke goes crooked, especially when I try to draw the ball from a distance. You could say that I miss position because I don't know and can't control the magnitude of the hit.

The primary intent of my experiment was to show the straightness of the stroke, by means of the swerve trace.

The stroke trace was included as a reference, and to provide a trigger for the display.

Displaying the stroke trace proved to be a bonus, since it shows the timing of the hit. The screen capture called "straight stroke" shows that I hit the ball late. Obviously, this would affect the magnitude of the hit, but it's easier to correct the timing of your stroke, rather than try to achieve a certain magnitude of hit.

To paraphrase the textbook writers, I leave the magnitude of the hit as an exercise for the teacher.

1robby
03-11-2005, 10:22 PM
Bob!

In Summer we will do some more experiments, perhaps i can find a really good and better acceleration sensor for this application. Our main-intention of these recordings was to show the warm-up-phase of different players. I know that the sampling rate of this board is too low for the measurements at the hit, but it is a very good tool for beginners to optimize their motion. If i find a suitable sensor we will do the measurements at the same time with highspeed-video-recordings! Then i will use a digital-storage-oszilloscope or a faster computer.... (unfortunately we did the fist recording with a 500Mhz-laptop-computer)

rob

dr_dave
03-12-2005, 10:24 AM
<blockquote><font class="small">Quote 1robby:</font><hr>If i find a suitable sensor we will do the measurements at the same time with highspeed-video-recordings!<hr /></blockquote>
Robert,

That would be great to see. Please share some video clips with us after you complete these experiments. I would be happy to post them for you if it is not convenient for you to post them on your site. Keep up the great work!

Dave

dr_dave
03-12-2005, 10:36 AM
John,

I'm glad to see you have joined in on the discussion. Thanks again for contributing your plots to TP A.9 (http://www.engr.colostate.edu/~dga/pool/technical_proofs/new/TP_A-9.pdf). It is reassuring to see data from two different sources (you and Robert) that look very similar. It appears that your sensor may have had a better range and sampling rate than the one used by the Austrian group (see technical details (http://www.billiardsdigest.com/ccboard/showthreaded.php?Cat=&amp;Board=ccb&amp;Number=183786&amp;page =0&amp;view=collapsed&amp;sb=5&amp;o=&amp;vc=1)). Could you share with us some of the specifications of your system? Also, do you have any additional insight concerning the features in the plots that have not been discussed already?

Regards,
Dave

1robby
03-13-2005, 02:46 AM
Dave!

We have used the ADXL 202 Evaluation kit, RS232 Interface, software and part included from Analog Device. The maximum sample rate for this kit is 275 Hz! Our main-intention of these recordings was to show the warm-up-phase of different players. I know that the sampling rate of this board is too low for the measurements at the moment of the contact with the cue-ball, but it is a very good tool for beginners to optimize their motion (you can easliy see possible side-deflections on the online-graph during the warm-up-phase...)!

You find the descripion of the set on
http://www.analog.com/en/prod/0%2C2877%2CADXL202%2C00.html (link to Evaluation Boards).

The software we have used was from Crossbow technology www.xbow.com (http://www.xbow.com) (ACCEL-VIEW v2.0)

Robert

dr_dave
03-16-2005, 07:57 AM
<blockquote><font class="small">Quote dr_dave:</font><hr>I just found out from the Austrian group some specifications for their accelerometer system. They used the Analog Devices ADXL202 (http://www.analog.com/en/prod/0%2C2877%2CADXL202%2C00.html) evaluation board.<hr /></blockquote>
FYI, the evaluation board used is described in detail at: evaluation board documentation (http://www.analog.com/UploadedFiles/Evaluation_Boards/Tools/350433521ADXL202EB-232A_c.pdf).

<blockquote><font class="small">Quote dr_dave:</font><hr>The sensor guarantees accurate readings only in the +/- 2g range and the maximum sampling rate is only 275kHz<hr /></blockquote>
FYI, the maximum sampling rate is 275 Hz (i.e., 275 samples per second), not 275 kHz (275,000 samples per second), as I originally reported.

JohnnyP
03-16-2005, 04:08 PM
Dave: I used ADXL311's in my experiment, and limited the bandwith to less than 500 Hz.

Outside of the basics, I haven't used it enough to know what the traces mean.

I did manage to figure out that an off axis hit shows up on the swerve trace as oscillations after the hit, and the stroke trace can show the timing of the hit.

The accelerometers are very sensitive, and can pick up the slightest hesitation or twitch in your stroke.

I don't have a reliable stroke, so the traces are all over the place. Someone with a good stroke will have traces that tend to look alike. But you know that already, because he keeps making balls.