System Updates & Development News
by Martin Paul Gardiner
Electronics and software development engineer and founder of GSA Golf
The GSA Golf range of golf 3D graphics software and electronic hardware is being updated with new features continuously.
New Features are being added and mods being made on nearly a daily basis so please check here regularly for your free updates.
Click on site map for all installation links
Click above for all installation videos
Save even more money with GSA Golf DIY kits
GSA Control Panel (CP) update V.188.8.131.52 - Windows 7 / 8 version
February 26 2015
Please be sure to read the download and update instructions before updating your GSA Golf CP (Control Panel)
Note: The above download is just an update for existing customers and not the full version Control Panel
Download Visual Studio Windows 8 x86 runtime components here
Click the above image to download the full version of the GSA Golf Control Panel (V.184.108.40.206)
Click the above image to download the GSA Golf single course software.
Course can be played with the Control Panel via either a mouse or manual launch parameter setup for evaluation.
The above API / SDK can be used to interface GSA Golf products to other golf game software
Sales or Support ?
Due to the heavy work load on at the moment with development and fulfilling orders we may not be able to get to the phones.
If you have questions please e-mail and we will get back to you within 24 hours.
Please note that we are closed on weekends and national holidays
(we need this time for new developments and testing)
Note! All GSA Golf software and hardware updates (yes! product hardware updates too) are free of charge for life
GSA Golf are constantly researching and working on improvements to their product line.
If we find better methods and improvements for any processes we pass them on to our customers free of charge to ensure that their investments are secured.
February 26 2015
More LX Spin dot ball test results with Beta version of the software
This was a tough one to beat and cost me a good 4 hours of brain jogging this morning to solve.
The above is LX Pro frames 2 and 3 showing how some dots near the ball image perimeters come in and out of the valid FOV.
The dot matching software has to identify these dots and eliminate them from the equations.
This update is included in today's CP.
In addition: It was noted that soft triggering sometimes caused images to swap. i.e. you'd soft trigger frame 2 and frame 3 would appear. Now fixed.
New PX2 installation video
February 25 2015
Spin dot ball test results with Beta version of the software
I'm publishing the test results of the new spin dots balls on a separate page now.
Improved spin axis detection in today's CP for LX users.
February 24 2015
Multiple line scan cameras
You can now use multiple line scan cameras to increase the Left / Right ball path range.
Cameras are connected to a link box that comes free of charge with multiple line scan camera purchases.
February 20 2015
New spin dot ball software beta test version mods
Further testing today revealed a couple of issues that have now been resolved in todays' CP
Issue 1: Dot pattern matching not always matching correctly. Now fixed.
Issue 2: As camera gain is reduced for spin dot balls, ball itself wasn't aways being detected. Now fixed.
The new default for camera gain when using spin dot balls is 10db instead of 15db. New min gray scale is 100 instead of 120.
When using the new spin dot balls, ensure that "Spin dots on ball" has been selected on the CP's main window and not "Ball line or Logo on ball".
New: When using spin dot balls, the system is not calculating ball path by comparing ball image sizes anymore so the ball perimeter cross hairs are no longer being displayed. (Looks are lot cleaner now without all those other cross hairs)
Note: Ball path is now being calculated by comparing the distances the two dots pairs are apart.
New: When in composite frames mode, only the two dots that are used in the spin calculations are marked with red and green cross hairs.
In addition, the dots on frame 1 are shown both in the original frame and again in the second frame but smaller.
February 18 2015
New spin dot ball software beta test version now available
The beta version of the new spin dot ball software is now featured in today's CP update.
I've shipped the new spin dot balls to a number of LX customers now to test and will ship to all other LX customers soon.
February 17 2015
February 16 2015
New spin dot ball software - Spin Axis
The above shows the method used to calculate spin axis. The two sets of bright red and green cross hairs show the rotation. The smaller green and red cross hairs show where the two spin dots were in frame 1 and the two larger red and green cross hairs show the position of the dots in frame 2. The intersection of the two sets of perpendiculars denotes the center of rotation (shown as a blue cross hair). If there was no spin axis tilt, this intersection would be exactly in the center of the ball. The example shows the dot perpendicular intersection is above the center of the ball so spin axis tilt was positive in this case.
February 15 2015
New spin dot ball software
The above two images show a ball spinning on a tilted axis. In this situation, dots on the ball's perimeter may go out of the valid ball radius range and not be detected in one frame but are in another.
In the above example Dots D5 and D6 have been detected in frame1 but not in frame2. In order to measure spin rotation and spin axis we need to identify two dots in one frame and the same two dots in the other frame.
Ideally we'd select the two dots that are furthest apart. In frame1, these would be D1 and D6 but D6 is not detected in frame 2 so we have to find two sets of dots that are in both frames and then select the two that are furthest apart.
In this case, D1 and D2 are in both frames and are furthest apart.
The spin dot software in today's CP matches all dots in frame1 with their corresponding dots in frame2 and measures all distances they are apart in order to solve this problem
Note that the two dots that are used to calculate spin rate and spin axis are colored red and green. All others are colored aqua.
February 14 2015
Trajectory animation and carry is also now in the "Show Club Angles" window.
You can use this window to setup launch parameters graphically. Just left or right click in the parameter boxes to change values.
February 13 2015
New spin dot ball software
The first version of the new spin dot ball software for the LX and CX Bcam has been completed today. Ball dot pattern matching works very well.
If you are an LX or Bcam customer you won't be able to use it until you have received the spin dot balls though. I'll be shipping them out to you next week.
I also added carry and trajectory animation to the CP's main window today.
You can now set up launch parameters and test them without the need to run any game software. Carry distances are also now shown in meters and speed in kph if you have selected "Metric" in the Setup window.
February 12 2015
New spin dot ball testing
Tests today revealed that ball path detection for the LX using the new GSA Golf spin dots works really well. CP update for these balls will be coming this weekend.
BTW. The balls can also be used for CX Bcam.
February 11 2015
New spin dot ball testing
Testing of the new spin dot balls is well underway and software modifications for these balls should be completed in a week or so.
In order to detect the dots, the camera gain shouldn't be too high otherwise there won't be sufficient contrast between the black dots and the white ball for the software to detect them.
The above far right image shows that the dots have all but disappeared when the camera gain is set to 18db. The center and left images - on the other hand - show that all dots are detected with just 10 db gain.
However, the problem with reducing the camera gain is that the ball image size can't be so precisely detected. With just 10db, the size is only measured at 148 pixels, whereas at 18 db the size is 160 pixels.
In order to detect ball path, we compare the ball image sizes in the frames so correct ball size detection is important.
There are two solutions to this problem though:
1. the ball size is consistently measured too small so ball path image size comparisons will be as before.
2. we don't measure the ball image size at all but the distance the dots in the pattern are apart. i.e. if they stay the same in all frames then the ball path is straight, if they are smaller in frame 2 then the ball is moving away from the camera and thus the ball path is left and if they greater in frame 2, the the ball has traveled towards the camera and thus the ball path is right (assuming a right handed player).
Further testing this week will reveal which method is more accurate but I have a feeling the dot distance method will be the winner here. And as the software is already measuring the distances all the dots are apart from each other, this solution will be easy to implement.
I have 500 of these new 360 degree spherical printed spin dot balls in stock and will send 12 free samples out to all LX customers as soon as the image software has been completed.
February 9 2015
Modifications made in today's CP
For CX systems only
1. Putting position X in the Vcam is now separate from putting position X in the Hcam.
Putting position X in the Vcam has no effect anymore other than to indicate in the Vcam where the putting position is
2. Putting path trace is now seen in the Hcam
3. Carry distances in the CP now being calculated when LA under 10 degrees
February 7 2015
New camera left / right ball speed adjustments.
Assuming the Vcam is mounted on the right side of the enclosure, if ball path is left then left adjustment should be positive and vice versa, if ball path is right then right adjustment should be negative.
The below images explain what the issue is with measuring accurate ball speeds when ball path is off center.
If we disregard the focal length of the camera lens, then the above formula will always calculate the exact distance the ball has traveled within the camera's exposure time.
Knowing the exact distance the ball has traveled and exposure time, we can calculate the exact ball speed.
In practice, however, the focal length of the camera distorts these distances.
If the ball path is left then and the camera is mounted right, then the trace length will appear shorter as it is moving away from the camera and vice versa, if the ball path is right then the trace will appear larger and thus longer.
If adjustments for this are not made for this then ball speeds of balls going left will be too low and ball speeds calculated of balls going right will be too high.
As there isn't any set mathematical formula to solve these deviations I've added a simple left right ball speed adjustment that adds or subtracts a percentage of the measured ball speed to correct the deviation.
In the above example I've set the adjustment to be a max of 10% (to increase ball speed) if the ball path is left and a max of -7.28 % (to decrease measured ball speed if the ball path is right.
Note that these settings are max adjustment settings. The max will be reached at +/- 20 degrees of ball path.
So, in this example, if a ball path is 5 degrees left and the max adjustment is set to 10%, then there will only be a 5% increase in normal path measured ball speed.