Golf USA
Results 1 to 1 of 1

Thread: Do 3D Systems Like 'MYSWING/GEARS /AMM 3D' reflect reality?

  1. #1
    Beginner
    Join Date
    Sep 2018
    Posts
    43
    Member
    50395
    Handicap
    15
    Thanks Thanks Given 
    0
    Thanks Thanks Received 
    9
    Thanked in
    8 Posts

    Do 3D Systems Like 'MYSWING/GEARS /AMM 3D' reflect reality?

    I've seen a few old articles that have knocked my confidence in the integrity of 3D systems , specifically in accurately monitoring the hand release actions in the golf swing through the impact zone (ie. clubhead 18 inches 'pre/post' impact).

    I would have thought that the most important area of the golf swing that we should be monitoring is impact and maybe using golf instruction to improve impact dynamics (maximising clubhead speed and optimising clubface direction 'just before/at/just after' impact).


    Here are some of the comments:
    Note that Dan Carragher and FWP(ie. Michael Martin) are PGA qualified instructors who use these 3D system tools. The issue is that they are using 3D system data to formulate best guideline instruction to their pupils (maybe comparing them with the actions of PGA pros). But if the 3D systems are showing data that is not reflective of reality , then surely their instruction could be flawed.
    Further , it seems that they have to use 2D Phantom camera images to check whether the 3D software is accurately recreating most of the actual positions and movements of the golfer/club.

    MYSWING

    "I recently had a very interesting 1 hour telephone conversation with Peter Gauthier (President of MySwingGolf 3-D motion analysis systems). He was refreshingly candid about the limitations of his 3-D system and he didn't try to irrationally argue that it could accurately measure the degree of left wrist extending and left forearm supination happening between P7 and P7.2 (as argued by Dan Carraher on behalf of his 3-D graphs produced by the MySwingGolf 3-D system and as also argued by Michael Neff for his GEARS system [which doesn't even have sensors on the left hand]).

    Here are some of the issues that we discussed.

    1) He stated that his 3-D system operates at 450 FPS compared to Phil Cheetham's 3-D system, which captures data at 240 FPS. Therefore, he has nearly twice as many data points per unit time, which makes it more accurate. However, he readily admitted that his 3-D graphs are based on algorithms, which draw "best fit" lines between data points and he stated that different 3-D systems use different algorithms. He stated that any algorithm cannot be perceived to be perfectly reflective of reality because they are basically "smoothing" algorithms needed to create a graph when there are significant time gaps between data capture points.

    2) When I talked about the problem of accurately measuring the amount of left wrist extending and left forearm supination happening between P7 and P7.2, he stated that his 3-D system's data capture rate was too slow to accurately measure those two biomechanical events over such a short time period. He stated that to do research on that specific issue, that one would need a much faster capture rate. He stated that his 3-D system was designed to be practically useful in the field for golf instructors, and that it was not designed to be used in a laboratory setting for obtaining a very accurate assessment of those two biomechanical events happening between P7 and P7.2. To my surprise, he even volunteered the opinion that his left forearm supination graphs are very inaccurate, and he stated that he intends to remove 3-D measurements of left forearm supination from future versions of his software. I introduced the idea of accurately measuring left forearm supination by placing sensors on the radial bone at the level of the left wrist and in the middle of the left antecubital fossa, which he agreed would be a theoretically possible approach. However, he stated that it would probably cost him $100,000 to modify his software and he didn't believe that there is enough consumer-demand to warrant such an endeavour. He thought that his 3-D system was the most practically useful 3-D system on the market because it is based on inertial sensors (that doesn't require an elaborate setup procedure and that it is independent of measurements based on the x-y-z axes) and that it could be used in the field by golf instructors and serious golfers.

    3) I brought up the issue of whether ball collision would cause slowing of the rate of left wrist flexion through impact and I asked him whether he knew of any scientific evidence to support Dan Carraher's wild claim that ball collision would slow down the rate of left wrist extending between P7 and P7.2. He stated that he knew of no scientific evidence to support Dan Carraher's claim - although we both agreed that ball collision could produce deflection and twisting of the peripheral clubshaft (especially in the situation of an off-center hit).

    4) I then asked him whether he would endorse the teaching ideology of rapidly and massively extending the left wrist through impact (as promoted by Dan Carraher and FWP). He stated that he was agnostic when it comes to golf instructional teaching and that his 3-D system was designed to be an objectively neutral tool that could be used to teach golfers different golf swing techniques. He was sympathetic to my alternative golf instructional idea of teaching a DH-hand release action and he stated that his 3-D system could be used to measure the amount of left wrist extending happening P7 and P7.2 with a reasonable degree of accuracy, but he stated that one could not rely on his left forearm supination measurements as being adequately accurate. "

    GEARS

    "I had a 1-hour Skype conversation with Michael Neff, owner of the GEARS 3-D system, yesterday. I was hoping to discover how his GEARS system could measure the amount of "left wrist extending" and "left forearm supination" happening between P7 and P7.2. What I discovered is that the standard GEARS 3-D motion analysis system cannot even measure those biomechanical events because they do not have any markers on the left hand or left lower forearm's radial bone.

    GEARS-avatar system "invents" some biomechanical events that do not correlate with reality

    In other words, the GEARS avatar's left wrist/hand motions have no correlation with reality. The GEARS system accurately measures the clubface angle because there are 3 markers on the clubhead, but it artificially "invents" the position of the left wrist/hand.

    I spoke to Michael Neff about placing more markers on the back of the left hand, over the left lower radial bone at the level of a watchband and over the middle of the left antecubital fossa, so that the GEARS system can accurately measure the amount of "left wrist extending" and "left forearm supination" happening between P7 and P7.2. "Time" will tell whether he is responsive to my suggestions and whether a future GEARS motion analysis system has this capability.

    An interesting tidbit of information. I asked Michael Neff how he checks to see whether his GEARS system can accurately measure the position of the clubshaft, clubhead, clubface and body parts at every moment of the golf swing - by using another measuring device - and he stated that he uses a Phantom camera operating at 20,000 FPS as a "gold standard" check."

    AMM 3D (Phil Cheethams 3D tool)- Is nearly 50% slower than MYSWING so I imagine its accuracy in measuring hand release biomechanics through impact is inferior.


    --------------------------------------------------

    PROOF OF ABOVE

    Here are comparison capture images of a 2-D capture image of Jamie Sadlowski at impact versus the GEARS avatar.



    At impact, Jamie Sadlowski's ulnar border of his left hand faces the target because he uses a very strong (4-knuckle) left hand grip, while the back of his left hand is roughly parallel to the inclined plane while it faces the ball-target line. Note that the clubface is square to the target at impact, which means that the clubface is angled about 60 degrees closed relative to the back of his left hand (as it was at address). However, note that the back of the GEARS-avatar's left hand faces the target and it is parallel to the clubface.

    Here is another comparison image at P6.


    Note that Jamie Sadlowski's left wrist is cupped at P6 and the clubface is about 45-60 degrees closed relative to the back of his left hand and this scenario is fully expected because Jamie Sadlowski adopts a very strong left hand grip. Note that the GEARS-avatar's left wrist is bowed and the clubface is straight-in-line (parallel) with the back of his left hand.
    Here is a DTL capture image of Jamie Sadlowski at P6.




    Note that the clubface is slightly tilted beyond vertical, but it is about 45 degrees closed relative to the back of his left hand (blue line).
    In other words, the GEARS avatar's left wrist/hand motions have no correlation with reality. The GEARS system accurately measures the clubface angle because there are 3 markers on the clubhead, but it artificially "invents" the position of the left wrist/hand.
    I spoke to Michael Neff about placing more markers on the back of the left hand, over the left lower radial bone at the level of a watchband and over the middle of the left antecubital fossa, so that the GEARS system can accurately measure the amount of "left wrist extending" and "left forearm supination" happening between P7 and P7.2. "Time" will tell whether he is responsive to my suggestions and whether a future GEARS motion analysis system has this capability.

    Last edited by WILDTHING; 06-14-2019 at 07:02 AM.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •