Trackman.. An engineering perspective

[SocketRocket]

Oh well, the forum just popped in my earlier post. Hope that explains it a bit.

There is a bit of a bug in the Forum software at the moment that puts some post on hold. A Mod will find it and post though.

Your explanation of the tracking systems and their methods was very interesting, thanks for that. I had read a bit on Trackman and considered that it took samples of the balls changing height over the time spent and could use Trig to calculate the distance travelled. I found it difficult to work out how it calculated the spin rate but put it down to somehow using the Magus effect against the previous ball speed and height. I had guessed that measuring the movement of the ball right or left during flight would enable the rotational axis tilt to be calculated which in turn would make it possible to work out the swingpath. I was stumped on how it could work out centredness of strike though.

Anyhow, after reading your explanation of how the Doppler Radar uses the returning frequencies I can understand it better., Again, thanks.

 

[palindromicbob]

Thanks nosevi. Firat rime ive read a long post in ages it was so interesting to me. Espeacially as i was fitted on a cam system.

 

[JustOne]

I had guessed that measuring the movement of the ball right or left during flight would enable the rotational axis tilt to be calculated which in turn would make it possible to work out the swingpath.

Measuring the ball in flight makes no sense to me whatsoever, that would totally ignore any wind variation, hitting a slight draw into a left/right wind would look like you hadn't put any draw spin on the ball at all, totally inaccurate. I can understand using radar for height, initial direction, speed, but not spin unless you got that measurement within the first 6ft but other than that I don't buy it... or it's just not accurate. Further to your comment Socket it would be wrong to work outr a swingpath from an already (potentially) incorrect spin axis.

(note... I'm going to have to re-read this post considering I've written something arse about face already on this thread) :rofl: The Rubik's Cube was easier

on this pdf it does say they track the entire flight... http://trackman.dk/getmedia/d37a0a85-b5ea-4511-a6d0-35d087fbfbec/TrackManProducts.aspx

EDIT: but on this page it says it tracks the spin within the first 10ft of flight...
http://trackman.dk/Products/TrackMan-Pro.aspx

....don't forget to click on Tiger Woods and listen to his views.... :thup:

 

[Piece]

In a nut shell a radar system will always say where THAT last shot went, it may however get the 'how did it get there' a little off. A camera system such as the GC2 (at least should) give you accurate spin numbers and launch condidtions. Whether your ball goes exactly where it predicts is dependent on whether you're at sea level in a given temperature, pressure, wind, etc. I went the GC2 route because I didn't want to know where the last shot went, I wanted to know how I hit the ball. Both are pretty good systems though, it really depends on what you want them for.

Sorry strayed a tad off topic with the GC2 but does demonstrate the only minor down side in a doppler radar system.

Great first post Nosevi! Welcome!

I totally agree with your analysis on Doppler systems and your point comparing Doppler v camera systems is excellent. My feeling is that outdoor, full track Doppler systems track the ball and then reverse engineer some of the stats to fit the flight. In almost all cases this is fine, but if there's a strong right to left wind/gust, the ball is going to be tracked moving that way, so how does it stop the data from implying draw spin when in fact you may have hit it straight or faded it slightly? Intrigued to know how it handles wind in the calculations...from experience not very well?

The GC2 photographs the ball at least 10-20 times in a very short time frame to give the very accurate launches conditions. It's not surprising that GC2s are now very popular on tour and also with the manufacturers, often replacing the Doppler systems by beating them on price and accuracy.

 

[Nosevi]

Measuring the ball in flight makes no sense to me whatsoever, that would totally ignore any wind variation, hitting a slight draw into a left/right wind would look like you hadn't put any draw spin on the ball at all, totally inaccurate. I can understand using radar for height, initial direction, speed, but not spin unless you got that measurement within the first 6ft but other than that I don't buy it... or it's just not accurate. Further to your comment Socket it would be wrong to work outr a swingpath from an already (potentially) incorrect spin axis.

And therein lies the problem. This is how Flightscope say they calculate spin axis on the X2:

"Total spin is always measured directly and spin axis calculated using 3D movement of the club and tracking curvature of the ball. Wind has little effect. With no wind and a controlled environment there shouldn't be any tolerance on spin axis, because the horizontal azimuth is so low."

The word to notice is that wind has "little" effect. It does have some. How much effect is obviously going to depend on the length of ball flight it takes you to determine the spin axis and the strength of the wind. It's not perfect though.

People make the mistake on a radar system of seeing that it 'predicts' the ball flight correctly and so assuming it gets the spin numbers etc right. The thing is it doesn't predict ball flight, it tracks it so that's always correct on a range and you can't check it inside. Whether or not the spin numbers are right has nothing to do with whether it puts the correct ball flight up as they are calculated partially from it.

 

[Nosevi]

Thanks Piece.

Don't get me wrong all the radar system do some things very well. The thing is many people misunderstand the limitations (although to be fair both flightscope and trackman are upfront about what they measure and what they calculate). Where I think they can fall down especially is when they are used inside where you end up working out things like spin axis based on the direction the club is moving in and the launch angle. Also people sometimes 'test' them outside, then use them inside not realising the numbers are calculated in a slightly different way.

 

[bigslice]

Hi all, new to the forum and just registered. I do know a little about this subject and I hope you don't mind if this answer is a little protracted. As well as having a little experience of active array doppler radars (I've tested one for the military a few years back and had to have a knowledge of how they worked and how different moving objects are effected/tracked by a doppler radar and the 'software fixes' employed) I've used several launch monitors, mainly Flightscope. Personally I own a GC2. Having looked at the X2, upcoming Xi and GC2 I went the camera route, I'll explain why later. I posted up an explanation of why I decided to buy a GC2 on a US golf forum, however after an online 'debate' with one of the 'moderators' in which he spouted rubbish and displayed a complete lack of understanding on the subject (other than the fact that he used Flightscope in his business) and I pointed out the pros and cons of both approaches, all of my posts comparing GC2 to a radar system were deleted - interesting definition of 'Forum'.

Anyway, in answer to the OP's question, doppler radars such as Flightscope can read spin but it doesn't really do it by looking at the dimples as such. It does it by monitoring the harmonics of the beams scattered back from the ball. In layman's terms (not for the OP, clearly) a regular radar sends a 'ping', listens to it coming back and from the time delay gets a distance to the target. It then sends another 'ping', again listens of the returned energy, gets another distance and can work out the speed from a difference between the two. A doppler radar however works in a slightly different way. It sends a very precise frequency of energy to the target. When that energy is 'bounced' back it's frequency will have changed depending on the speed of the object it is 'tracking'. If the object is coming towards the radar the frequecy is increased, if it's going away, such as with a golf shot tracked on radar, it is decreased. An example of this effect is when a car comes towards you and the engine note is high but when it passes and goes away the engine note appears to be lower.

This is a far more accurate way of tracking speed but it also allows for another reading on a golf launch monitor. As the ball flies away from the radar, if it is spinning, let's assume backspin, there is a difference in speed between the top of the golf ball and the bottom. With backspin the bottom of the ball is moving away from the radar quicker than the top. When the radar listens to the returned pulse it doesn't get just one frequency of returned energy, the pulse frequency now covers a range which spans from the fastest part of the ball and the slowest. They can use the average to accurately give ball speed and the difference between the two, as they know the diameter of a golf ball, will give them the spin speed.

There is however one minor snag. Due to the dimple pattern which scatters the beam, size of a golf ball, number of elements in the radar array, it's not possible to determin which bit of the golf ball is going faster and which bit is going slower with any accuracy - they therefore can only measure total spin speed, they can't give the direction of that spin and so spin axis (or sidespin component and backspin component if that's how you prefer to see it). Because of this maths and observation of the effects of spin have to take over. When it comes to Flightscope being used outside it looks at the overall flight shape and the motion of the club (albeit the back of it) through impact. Inside it uses a slightly different and less accurate way of determining spin axis by using the club motion (again the back) and the ball launch vector. Outside they say that wind has a very small effect on the calculation, it's clear that the inside model is less accurate as they would use it outside if it wasn't.

In addition there are some other calculations that have to be made, not measured as the radar is looking at the back of a club not the front. It has to derive face angle and dynamic loft from looking at the club trajectory and ball launch while bearing in mind that they have total spin not spin axis. All of that said it's very clever maths and they are very bright chaps. It's probably pretty close.

When looking at the accuracy of different systems most people (including moderators on US golfing forums apparently) fall into a very simple trap. They hit a ball on a range and see which system gives the correct flight. The radar does as it simply tracks the ball so they conclude that a camera system is inaccurate in comparison, buy a radar system and put it in their shop. What they fail to understand is that a radar system is always going to give an accurate ball flight if it can see it - that's not in debate and is pretty obvious. It may or may not give accurate spin numbers. It should be pretty close on total spin but spin axis (so backspin and sidespin), while it may always appear correct if you can see the flight (as this is what it uses to calculate it) may be a little off. In indoor mode radars use different methods to determin spin axis and use that to predict ball flight. They determin total spin in a similar way though although with a shorter flight a metal dot helps.

In a nut shell a radar system will always say where THAT last shot went, it may however get the 'how did it get there' a little off. A camera system such as the GC2 (at least should) give you accurate spin numbers and launch condidtions. Whether your ball goes exactly where it predicts is dependent on whether you're at sea level in a given temperature, pressure, wind, etc. I went the GC2 route because I didn't want to know where the last shot went, I wanted to know how I hit the ball. Both are pretty good systems though, it really depends on what you want them for.

Sorry strayed a tad off topic with the GC2 but does demonstrate the only minor down side in a doppler radar system.

cool answer and sounds like someone who knows wot he is talking about, i assume your targets in the army were bigger

 

[SocketRocket]

Measuring the ball in flight makes no sense to me whatsoever, that would totally ignore any wind variation, hitting a slight draw into a left/right wind would look like you hadn't put any draw spin on the ball at all, totally inaccurate. I can understand using radar for height, initial direction, speed, but not spin unless you got that measurement within the first 6ft but other than that I don't buy it... or it's just not accurate. Further to your comment Socket it would be wrong to work outr a swingpath from an already (potentially) incorrect spin axis.

(note... I'm going to have to re-read this post considering I've written something arse about face already on this thread) :rofl: The Rubik's Cube was easier

on this pdf it does say they track the entire flight... http://trackman.dk/getmedia/d37a0a85-b5ea-4511-a6d0-35d087fbfbec/TrackManProducts.aspx

EDIT: but on this page it says it tracks the spin within the first 10ft of flight...
http://trackman.dk/Products/TrackMan-Pro.aspx

....don't forget to click on Tiger Woods and listen to his views.... :thup:

JO, I did explain in my post that those were my 'initial' considerations and that the post by Nosev1 had enlightened me.

 

[JustOne]

JO, I did explain in my post that those were my 'initial' considerations and that the post by Nosev1 had enlightened me.

I realise that SR so no worries, I'm just guessing this stuff anyway as frankly I don't really care how they get the figures only that the results are accurate (i prefer to talk ABOUT the numbers and not how they get them) however what I'm getting the vibe of is that you could potentially get fitted for clubs that'll be just perfect in a left/right wind but rubbish on a calm day

 

[Nosevi]

cool answer and sounds like someone who knows wot he is talking about, i assume your targets in the army were bigger

The targets were much bigger - aircraft in the RAF - but the theory is the same. With some aircraft you have props on the front which can cause doppler radars problems if left to their own devices, but the software we had to 'resolve' the target from the raw data could work around that. We were very interested in tracking multiple targets and the radar being able to work out one from the other, two light aircraft close together can look very similar to a radar to one large one and as we were now working with processed radar we found that the software could be fooled and aircraft could disappear if one got too close to another. At the station I was at, a train line went under the approach and it was found that if an aircraft went over a train on the mainline the software could get confused and your tracked target would take a 90 degree turn and head off to Scotland for the weekend. Where before as controllers we saw raw data we now had to consider radar resolution, ie how many targets can be resolved in one patch of sky which we had to consider when formulating new regulations. We also had to consider how our regulations were effected by implementing different software modes (which we called filters as that is what they were on our old raw data radar systems) which I guess is similar to switching to 'indoor mode' in that certain things are changed and it alters how targets are displayed and how accurate that is.

Overall, the controllers testing the kit needed a 'working understanding' of how the kit worked which is what I have. The techies had to know all about what we refered to as the "wiggly amps" and all the really clever stuff like how the radar could 'spin' the signal so it could look though rain etc. I just needed to know how that effected it's accuracy.

Anyway that's probably why it irritated me when a Flightscope guy on a US forum kept telling me how I knew nothing about it, radars are the only way forward, are not in any way effected by wind etc (even though even Flightscope say they are) and camera systems make assumptions but radars don't. As I said, camera systems know where the ball is initially going and what it is doing, they make assumptions and calculations about where it will end up. Radars know where a given shot went, they make assumptions and calculations about what the ball was doing to get there. On indoor mode a radar makes a few more calculations as it doesn't know where it has ended up either.

All of that said the radars are very clever and the maths they use is very complex - they do a fairly good job and will tell you where your ball went more accurately than a GC2 (if the radar is on outdoor mode).

 

[Nosevi]

I realise that SR so no worries, I'm just guessing this stuff anyway as frankly I don't really care how they get the figures only that the results are accurate (i prefer to talk ABOUT the numbers and not how they get them) however what I'm getting the vibe of is that you could potentially get fitted for clubs that'll be just perfect in a left/right wind but rubbish on a calm day

That all depends a bit on how quickly the kit works out the spin axis and also how clever they are at discounting additional movement caused by the wind. Regarding the numbers and accuracy, ironically it's very difficult to disprove the accuracy of the camera system unless you can be sure you're hitting a generic ball in the exact atmospheric conditions that the system was set to replicate. However a GC2 SHOULD report the spin numbers correctly as it takes up to 24 pictures in about the first 8 inches of ball flight. From that it really is quite easy to see what just the spin is, it's simply measured. By hitting a ball on both Flightscope and a GC2 you should be able to check the Flightscope spin axis or side spin vs back spin and even compare indoor to outdoor mode. On the 28th I'm going for a lesson and have both a Flightscope (albeit a Kudu) and obviously my GC2. Can do a few tests if you like?

 

[vkurup]

Nosevi.,.. Welcome to the forum... And what a way to announce your entry. You definitely did not fly under the radar on this one.

The posts were very informative, and probably qualifies for one of the best in 2012... Great work.

 

[Nosevi]

Nosevi.,.. Welcome to the forum... And what a way to announce your entry. You definitely did not fly under the radar on this one.

The posts were very informative, and probably qualifies for one of the best in 2012... Great work.

Thanks mate. Funny that saying nothing different got all my post on the subject 'edited' (ie removed) from a US forum as the administrator used Flightscope. Like I said before, I'm far from anti-Flightscope, I think it's a good system, and the guys who make it don't try to hide it's limitations. Clearly they're not going to highlight them but they are upfront about what they measure and what they calculate. If you understand how the systems work you can make an informed decision about which system you want based on how you want to use it.

Anyway, glad it was of interest.

 

[daymond]

Great post Nosevi. If you need any help I've still got my five figure log tables and a set of scale rules. lol.

 

[Nosevi]

Great post Nosevi. If you need any help I've still got my five figure log tables and a set of scale rules. lol.

Thanks Daymond. To be honest was thinking about it and a comparison between a GC2 and a 1st or 2nd generation Flightscope (I only have access to a Prime and a Kudu I think) would probably be unfair as Flightscope had some known issues relating to the club tracking on earlier devices. As some of the spin numbers are calculated with reference to this it would probably be unfair to post up findings. Although the kit at my range has the latest Flightscope software, I don't know whether the issue was a software one or a hardware one. If I was to find the spin numbers are out, ie the reverse engineering of the data as a previous poster described it as was flawed, it could well have been improved upon by Flightscope.

It is worth noting though that the issue was, I believe, inaccurate reading of the angle of attack. If you consider that the spin axis is determined by reference to a 3D tracking of the club, as is dynamic loft, face angle etc, you can quickly see how an error in one reading can quickly translate into erroneous readings throughout the system. That's why I think the camera systems (GC2 with HMT for example) are better for analysing how a shot is struck whereas a radar system is better for saying exactly where a given shot flew.

 

[Piece]

Do you have HMT for your GC2?

 

[Nosevi]

Do you have HMT for your GC2?

Not yet although eventually I would plan to add it. Thought it best to get one item past the 'boss' at a time!

 

[Foxholer]

Nosevi.

Thanks for the extremely informative explanation of how backspin is established. Sounds like an old style BSG 'discussion' thread experience you had with the US Moderator!

 

[Nosevi]

Nosevi.

Thanks for the extremely informative explanation of how backspin is established. Sounds like an old style BSG 'discussion' thread experience you had with the US Moderator!

No probs. To be honest it was just a tad annoying, no big deal. Anyway, could have been he just remembered my helpful and subtle comments in their 'Ryder Cup' thread. And I tried SO hard not to gloat ........