Margaret Fisher

Hi all, I am looking for a patch for the gametrak v 1.3 - a motion sensor for golfers - to have it trigger audio. The Laptop Orchestra uses these devices and I'm sure there's a Max4Live patch out there. Can anyone help?

Hi all, I am looking for a patch for the gametrak v 1.3 - a motion sensor for golfers - to have it trigger audio. The Laptop Orchestra uses these devices and I'm sure there's a Max4Live patch out there. Can anyone help?
Thank you, Margaret


Hi all, I am looking for a patch for the gametrak v 1.3 - a motion sensor for golfers - to have it trigger audio. The Laptop Orchestra uses 

hi-device

If so, why don't you connect it to computer and report what it

sends and what you would convert that data into...

Do you have one of that devices ?
If so, why don't you connect it to computer and report what it
sends and what you would convert that data into...


Yes I have the GameTrakdevice, with the circuit board ‘adjusted’ to bridge info over to pc. (I have another GameTrak but it turned out to be an earlier model). Thanks for your reply.

Was able to make a workable patch in pdextended but pd crashes on my pc.

Yes I have the GameTrakdevice, with the circuit board ‘adjusted’ to bridge info over to pc. (I have another GameTrak but it turned out to be an earlier model).  Thanks for your reply. 

Main thing is to know if it communicates via usb to serial

Original devices were PS2 only and then there were

If it is HID, you can use max hi object to capture

From the brief look into infos, Gametrak should have 6 potis, so you would expect 6 values. 

If you send that pd patch that works, I would check it to see how to convert that.

Main thing is to know if it communicates via usb to serial
or as HID device.
Original devices were PS2 only and then there were
different mods to make it usb capable.
If it is HID, you can use max hi object to capture
the data, otherwise serial object.
From the brief look into infos,  Gametrak should have 6 potis, so you would expect 6 values. 
If you send that pd patch that works, I would check it to see how to convert that.


Source Audio, thank you. I don't know if it is communicating as serial or HID. Here is a jpg of the patch in pd extended (left) and an earlier one in pd. I'm pretty sure the one on the left worked - the one on the right was part of the process. I appreciate your offer to convert! Hoping to use this with dancers.

GameTrak-PDextendedHID.jpg

Source Audio, thank you. I don't know if it is communicating as serial or HID. Here is a jpg of the patch in pd extended (left) and an earlier one in pd. I'm pretty sure the one on the left worked - the one on the right was part of the process. I appreciate your offer to convert! Hoping to use this with dancers.
Margaret     www.mafishco.com

Here's the actual file but it seems to be in text. Could not upload the mxt or pat versions.

PDExtendedGolf_HIDIN.pd

If that PD patch was fully working with gametrak, than it is all easy.

Use hi object in max, take few parts from help file

in order to auto activate it on patch start,

activate polling and insert route 1 2 3 4 5 6

If you need any further help with that just post.

If that PD patch was fully working with gametrak, than it is all easy.
Gametrak is connected as HID device.
Use hi object in max, take few parts from help file
to get exact name of the gametrak device
in order to auto activate it on patch start,
activate polling and insert route 1 2 3 4 5 6
as in pd to route the 6 values.
If you need any further help with that just post.


Hi again, a few parts from help file? u-menu is one? how to activate it? loadbang at top? What does polling mean? How do I then map within Ableton Suite to control audio clips (not midi)? I may be too far down the learning curve, but if you can help, I’m willing to soldier on . . . Many thanks! M

I think this is very clear description...

Screen Shot 2020-08-13 at 09.35.45.png

The idea is to first populate umenu with available devices.

One could copy gametrak exact device name in a message

and use it to auto load it on next start.

If device is present, one could bang poll message to activate

That way all would be done without user action, if gametrak is present

here is a patch to start with, do what it tells inside.

gametrak.maxpat.zip

I am not Live user and have no interest in it.

But maybe one or the other hint should be possible

Did you really look at the help file ?
I think this is very clear description...

The idea is to first populate umenu with available devices.
One could copy gametrak exact device name in a message
and use it to auto load it on next start.
If device is present, one could bang poll message to activate
data polling.
That way all would be done without user action, if gametrak is present
when patch or max4live device starts.
here is a patch to start with, do what it tells inside.


---------
I am not Live user and have no interest in it.
But maybe one or the other hint should be possible



Many thanks for the zip file. Will pursue this with a little more insight, thanks to you.

Did not realize data polling was a ?message? "poll", so No, did not read help file to extent I needed/need to. Will do! Opening the patch now.

Many thanks for the zip file. Will pursue this with a little more insight, thanks to you.
Did not realize data polling was a ?message? "poll", so No, did not read help file to extent I needed/need to. Will do! Opening the patch now.


Results! but on a mac laptop, not on the pc. It prints out "Game-Trak V1.3" and an example of the numbers on poll are 15 2140 (both strings active, I get 16 and 17 also as the first 2 digits). Will the instrument find and playback audio clips or map to controls?

Left: resting state as low as 15 526, or 15 770; gets into 15 3-4000s with action of 'joystick / string'

Right: not as much differential resting at 15 2025 but also can be 15 4089, mostly reads out at 15 2067-2070

Is right 'joystick/string' programmed in this small patch you sent?

Variations in data
Left: resting state as low as 15 526, or 15 770; gets into 15 3-4000s with action of 'joystick / string'
Right: not as much differential resting at 15 2025 but also can be 15 4089, mostly reads out at 15 2067-2070
Is right 'joystick/string' programmed in this small patch you sent?


I don't know what you mean by joystick/string

usual output of HID Game Device in Max and hi object

In your case 15 16 17 are IDs for the axis or whatever

Now tell me how could that Pure Data patch work for you

with routing IDs 1 2 3 4 5 6, which GameTrak seems not to send ?

No success on PC ? means GameTrak does not get detected ?

Did You look in the hardware- game devices or whatever that is called on Windows ?

There are many variations on what people used to communicate over USB port.

Arduino - serial based, HID compliant using VUSB or newer Arduino HID compatible boards,PIC , Teensy ...

Something is preventing it from working on windows...

Might be a missing driver, or not well done USB Init .

If you would check what it reports as Product and Vendor ID

in Mac system Profiler, it would maybe throw some light on that matter.

When you start the patch I posted, you could click on 

That will list all available HID devices and number of their elements.

So you can find out how many elements are expected from GameTrak.

Usually the same kind of element have consecutive IDs

one typical Joystick could have 32 buttons IDs 0 - 31

I read at some point that GameTrak has 6 Potis ?

If that is so you would have 6 elements to report and use.

To separate elements, one uses route object at HI output.

route 15 16 17 would create 4 outlets which output data

belonging to IDs 15 16 17, and passing the rest at rightmost outlet.

Then you can split, scale or whatever incoming data to create switches,

I don't know what you mean by joystick/string
usual output of HID Game Device in Max and hi object
is ID and then value.
In your case 15 16 17 are IDs for the axis or whatever
GameTrak sends.
The following number is value.
Now tell me how could that Pure Data patch work for you
with routing IDs 1 2 3 4 5 6, which GameTrak seems not to send ?
-------
No success on PC ? means GameTrak does not get detected ?
Did You look in the hardware- game devices or whatever that is called on Windows ?
Maybe it needs a driver ?
There are many variations on what people used to communicate over USB port.
Arduino - serial based, HID compliant using VUSB or newer Arduino HID compatible boards,PIC , Teensy ...
Something is preventing it from working on windows...
Might be a missing driver, or not well done USB Init .
If you would check what it reports as Product and Vendor ID
in Mac system Profiler, it would maybe throw some light on that matter.
-----------
Back to the device itself.
When you start the patch I posted, you could click on info message.
That will list all available HID devices and number of their elements.
So you can find out how many elements are expected from GameTrak.
Usually the same kind of element have consecutive IDs
one typical Joystick could have 32 buttons IDs 0 - 31
than 6 axis 32 - 37 etc.
I read at some point that GameTrak has 6 Potis ?
If that is so you would have 6 elements to report and use.
--------
To separate elements, one uses route object at HI output.
route 15 16 17 would create 4 outlets which output data
belonging to IDs 15 16 17, and passing the rest at rightmost outlet.
Then you can split, scale or whatever incoming data to create switches,
controls etc.


P.S. by the way that values are really unusual for Hid joystick,

which usually has max resolution of 0 - 1024

P.S. by the way that values are really unusual for Hid joystick,
which usually has max resolution of 0 - 1024

Viewing the Max Console, I find ID 15, 16, 17, 18, 19, 20

From the mac:
product ID 0x0982
vendor ID 0x1467
In2Games Ltd.
Viewing the Max Console, I find ID 15, 16, 17, 18, 19, 20
# elements in Game-Trak V1.3:   25


Vendor ID is probably 14B7 - which is registered to In2Games Ltd.

That would mean that this is a genuine xbox version from manufacturer itself

That could help you to check the driver problem on Windows.

This also confirms your readout - it should be 12bit, means range 0 - 4095.

6 Pots: Left X,Y,Z Right X,Y,Z are received with prepended IDs

15 - 20, and Footswitch with ID 3 if one connects one.

Now the next question is what exactly do you want to do with the values ?

Vendor ID is probably 14B7 - which is registered to In2Games Ltd.
That would mean that this is a genuine xbox version from manufacturer itself
and not some arduino mod.
That could help you to check the driver problem on Windows.
This also confirms your readout - it should be 12bit, means range 0 - 4095.
6 Pots: Left X,Y,Z Right X,Y,Z are received with prepended IDs
15 - 20, and Footswitch with ID 3 if one connects one.
Now the next question is what exactly do you want to do with the values ?


Yes, it is genuine. No foot switch. I tried to find a driver online. No luck and I did not keep the CD that came with the unit.

One thing I want to do is trigger audio clips by identifying parameters that are paired to the clips. I don't know if I'll have subtle enough control to, e.g., identify Left, Right, Upper Left, Lower Right (these are just obvious examples). I'm looking at this as a dance (gesture) + theater (text) tool to be used within a 6 channel audio installation. 

I paired your patch to one I found online that creates separate data streams for the Game-Trak It's an image so I reconstructed it as a new Max patch (with errors, I'm sure).

That patch worked for me as a hybrid with yours, and only partly. It registers the changing "parameter identifier sequence" - which doesn't always include all 6 pots. It continues to log these sequences regardless, even as unit is unattended for an hour. Not sure those identifier sequences are useful (it's a little slow). 

Thanks so much for your helpful responses! M

Yes, it is genuine. No foot switch. I tried to find a driver online. No luck and I did not keep the CD that came with the unit.
One thing I want to do is trigger audio clips by identifying parameters that are paired to the clips. I don't know if I'll have subtle enough control to, e.g., identify Left, Right, Upper Left, Lower Right (these are just obvious examples). I'm looking at this as a dance (gesture) + theater (text) tool to be used within a 6 channel audio installation. 
I paired your patch to one I found online that creates separate data streams for the Game-Trak   It's an image so I reconstructed it as a new Max patch (with errors, I'm sure).

That patch worked for me as a hybrid with yours, and only partly. It registers the changing "parameter identifier sequence" - which doesn't always include all 6 pots. It continues to log these sequences regardless, even as unit is unattended for an hour. Not sure those identifier sequences are useful (it's a little slow). 
Thanks so much for your helpful responses! M


In future please open the patch in editor use select all and copy compressed.

I am not using Live, have to convert amxd to maxpat for my old max 6 that I use mostly.

You seem not to understand the purpose of the route object.

route 0 4 8 etc takes a list or stream of numbers and

outputs any number that follows the set route value.

You stated that GameTrak has 15 16 17 18 19 20 IDs

expected input range is 0 - 4096, not 0 65536 ?

If you look at some unknown patch, take a look in the help file

of each single object inside to understand what it is doing.

To make it short - that patch makes no sense in any of it's parts,

ending with outputing non signal numbers to plugout~ signal output.

Here is the patch you could use to initialise hi object

Than you have to decide what this values need to do.

it gives one no idea what is meant with that.

that you want audio clip 1 to start playing if

value following ID 15 is higher than 2000,

Or to scale input range to 0 127 and control volume

You need to check how GameTrak behaves in real life,

what is real min max values it outputs, does it allways send

First to that patch.
In future please open the patch in editor use select all and copy compressed.
Than paste here.
I am not using Live, have to convert amxd to maxpat for my old max 6 that I use mostly.
-----------
You seem not to understand the purpose of the route object.
route 0 4 8  etc takes a list or stream of numbers and
outputs any number that follows the set route value.
You stated that GameTrak has 15 16 17 18 19 20 IDs
prepended to left and right xyz axis.
So how should that work ?
And what is scale 0 65535 0 4096 doing ?
expected input range is 0 - 4096, not 0 65536 ?
If you look at some unknown patch, take a look in the help file
of each single object inside to understand what it is doing.
To make it short - that patch makes no sense in any of it's parts,
ending with outputing non signal numbers to plugout~ signal output.
---------
Here is the patch you could use to initialise hi object
and poll the raw values as they come in.

That is the starting point.
Than you have to decide what this values need to do.
I can't help you with statements like
"dance (gesture) + theater (text) ..."
it gives one no idea what is meant with that.
Proper info would be for example
that you want audio clip 1 to start playing if
value following ID 15 is higher than 2000,
and stops if it is lower than that.
Or to scale input range to 0 127 and control volume
of audio Track 1.
You need to check how GameTrak behaves in real life,
what is real min max values it outputs, does it allways send
or only if one really moves etc etc
to develop something usefull.


Your new patch did not work right away. I grafted it to elements from your first patch (see below). SUCCESS! Finally! I have values for each axis. 

I can set parameters. I will probably need help with triggering audio at discrete values. I might be able to figure out scale of input range 1 127 for volume. . . might. . . maybe. . . "You need to check how GameTrak behaves in real life" - yes, now, we have a meaningful relationship (gizmo + my imagination) and I can make a chart of values related to gestures. Thank you so much. I am more than grateful. 

Your new patch did not work right away. I grafted it to elements from your first patch (see below). SUCCESS! Finally! I have values for each axis. Now I can set parameters. I will probably need help with triggering audio at discrete values. I might be able to figure out scale of input range 1 127 for volume. . . might. . . maybe. . . "You need to check how GameTrak behaves in real life" - yes, now, we have a meaningful relationship (gizmo + my imagination) and I can make a chart of values related to gestures.  Thank you so much.   I am more than grateful.  
Here is the working hybrid:

Good news - and don't worry about mistakes - we all make them,

I forgot the main thing in the last patch - to connect menu outlet to hi

Screen shot 2020-08-19 at 8.34.54 AM.png

I find it important to be able to auto activate stuff, in first place

for the stage use, when one has enough other things to look after.

There are many techniques to smooth data output, avoid repeating values,

scale or turn into multifunctional split ranges and so on.

To be able to analyse data flow from the GameTrak, It would 

be helpful to record gestures over a length of time.

I could send you a patch to do so if you want

Good news - and don't worry about mistakes - we all make them,
like myself.
I forgot the main thing in the last patch - to connect menu outlet to hi
object, see the red patch cord:

I find it important to be able to auto activate stuff, in first place
for the stage use, when one has enough other things to look after.
-------
There are many techniques to smooth data output, avoid repeating values,
scale or turn into multifunctional split ranges and so on.
Feel free to ask, I'll be glad to help.
To be able to analyse data flow from the GameTrak, It would 
be helpful to record gestures over a length of time.
I could send you a patch to do so if you want



This is so generous. I'd like all that you suggest! Is that too much? Record gestures over length of time with data flow analysis. Smooth data output. Avoid repeating values 

repeat value x number of times? Create multifunctional split ranges. I've just begun mapping the ranges of the values . . . Looking forward. What can I contribute?

This is so generous. I'd like all that you suggest! Is that too much? Record gestures over length of time with data flow analysis.  Smooth data output. Avoid repeating values and repeat value x number of times? Create multifunctional split ranges. I've just begun mapping the ranges of the values . . . Looking forward. What can I contribute?

Just tried the fixed patch - yes, having the Game-trak device load automatically is very helpful. Thanks again!

I don't have that device, and so can't tell how it behaves,

getting only min max range means not much, to start with I would

say let's record 6 data outputs in 10 ms interval at different distance to the device itself.

Maybe you could as first give me few infos about mac you use, screen size as first.

I would make big sliders or numboxes, so that you can play with

Gametrak and check it's sensitivity before recording.

that would make things more simple for me.

I have a question - are Z axis inverted ?

I don't have that device, and so can't tell how it behaves,
getting only min max range means not much, to start with I would
say let's record 6 data outputs in 10 ms interval at different distance to the device itself.
Maybe you could as first give me few infos about mac you use, screen size as first.
I would make big sliders or numboxes, so that you can play with
Gametrak and check it's sensitivity before recording.
Are you using Max without Live too ?
that would make things more simple for me.
-------
I have a question - are Z axis inverted ?
Z is vertical measurement or ?

Here is max patch to show and capture GameTrak output.

GT-Capture.maxpat.zip

Here is max patch to show and capture GameTrak output.
Everything is self explantory inside.
Please run this in Max, forget Live.


Screen size 7" high, 11.25" wide (inches)

The device has clips that attach to rotating ball on device on one end and hand gloves on other, with retractable red line - like fishing line - that triggers the motion sensing. I took values for each of the indentations you can barely see on the silver ring in which the ball sits, plus distance values from floor to 9 feet (extent of my reach).

image.png

Your first patch with colors working - a little oversize for lil' mac. More response later today. Thank you! M

LeftRightValues_Charted copy.jpg

Screen size 7" high, 11.25" wide (inches)
The device has clips that attach to rotating ball on device on one end and hand gloves on other, with retractable red line - like fishing line - that triggers the motion sensing. I took values for each of the indentations you can barely see on the silver ring in which the ball sits, plus distance values from floor to 9 feet (extent of my reach). 

The day flew past and I'll have to wait until tomorrow to do more. Have a good evening, Margaret

that is 13" screen, but could range from 1280 x 800 on old macbook

GT-Capture-2.maxpat.zip

Nice graphic and infos.
Could you post the size in pixels,
that is 13" screen, but could range from 1280 x 800 on old macbook
up to 1440 x 900 or more on retina ...

still waiting for CD (hopefully with driver for pc)

display: 1280x800
still waiting for CD (hopefully with driver for pc)
meanwhile, many distractions in Calif 


I hope you are safe and that fire inferno would get soon under control 

Is this device varying much depending on the user,

like if a taller person might hit the min/max range

something to capture one person's movements

One anyway has to scale that possible max 0 - 4095

to reliable range, few numbers could also get lost

One would capture 6 axis like in the last patch,

(max - min per axis) and pick the smallest one.

than scale input of all 6 axis to that range.

That could be for example left Z axis 4064 - 1098 = 2966

each axis would map it's measured min max range to 0 - 2900

Using that base, one would do scaling, splitting etc

One could store calibration per person ..

I hope you are safe and that fire inferno would get soon under control 
--------
Ok, so the last patch should fit well.
Is this device varying much depending on the user,
like if a taller person might hit the min/max range
values easier than a shorter one ?
That would maybe call for a calibration,
something to capture one person's movements
and adjust all 6 axis accordingly.
One anyway has to scale that possible max 0 - 4095
to reliable range, few numbers could also get lost
to filter out jitter if it exists.
One would capture 6 axis like in the last patch,
then request min/max values from 6 axis,
measure the range size
(max - min per axis) and pick the smallest one.
than scale input of all 6 axis to that range.
That could be for example left Z axis 4064 - 1098 = 2966
so one would zmap 1098 4064 0 2960.
each axis would map it's measured min max range to 0 - 2900
Using that base, one would do scaling, splitting etc
One could store calibration per person ..


Thank you for your thoughts re fire - no control north and northeast of SF yet. Maybe less intense heat in some areas.

I appreciate your detailed description. I don't yet understand when or where the discrete parameters appear (e.g., 0 - 1500 audio clip A on, audio clip B, C, D off; 1501 - 2900 audio clip B on, Audio Clip A, C,D off). I think the confusion is with scale which shows up early and late in your description: "scale that possible max 0 - 4095 to reliable range" and " Using that base, one would do scaling, splitting etc". Does the scale 0 - 4095 appear in the patch GT-CAPTURE-2.maxpat? Or that patch is just a readout of all and any numbers?

I found one of the more effective visual perspectives for using these 2 lines (and not just repeating the performance moves of e.g., Stanford's laptop orchestras) involves putting the performer on the horizontal. 

I will have to have 2 people to collect the new data: A is horizontal, and B records the values of A's movements to arrive at a new set of basis numbers for all 6 axes.

Do you have suggestions regarding this collection? Do you think once a few numbers are collected, the difference between the original data set and the new one can be applied to the rest of the numbers (e.g., -90)? or will the values vary individually because of gravity ?

IMG_8832.jpg

Thank you for your thoughts re fire - no control north and northeast of SF yet. Maybe less intense heat in some areas.
------------------------- 

I appreciate your detailed description. I don't yet understand when or where the discrete parameters appear (e.g., 0 - 1500 audio clip A on, audio clip B, C, D off; 1501 - 2900 audio clip B on, Audio Clip A, C,D off). I think the confusion is with scale which shows up early and late in your description: "scale that possible max 0 - 4095 to reliable range"  and " Using that base, one would do scaling, splitting etc".   Does the scale 0 - 4095  appear in the patch GT-CAPTURE-2.maxpat? Or that patch is just a readout of all and any numbers?
--------------------------
I found one of the more effective visual perspectives for using these 2 lines (and not just repeating the performance moves of e.g., Stanford's laptop orchestras) involves putting the performer on the horizontal. 
I will have to have 2 people to collect the new data: A is horizontal, and B records the values of A's movements to arrive at a new set of basis numbers for all 6 axes.
 Do you have suggestions regarding this collection? Do you think once a few numbers are collected, the difference between the original data set and the new one can be applied to the rest of the numbers (e.g., -90)? or will the values vary individually because of gravity ?

Xbox CD was all video - nothing there by way of a driver for the pc. too bad.

I am out for a while, will post all details and additions

to the basic patch, which till now has no calibration included later today.

You will need to provide some answers to my questions in the previous post

1- if difference in person's dimensions play a role

2- is output of GameTrak steady or it jitters ( jumps few values up and down)

It is all simpler than it sounds.
I am out for a while, will post all details and additions
to the basic patch, which till now has no calibration included later today.
You will need to provide some answers to my questions in the previous post
regarding
1- if difference in person's dimensions play a role
2- is output of GameTrak steady or it jitters ( jumps few values up and down)


one has to set usable range of all 6 axis.

By measuring their min and max values, than finding which one has the smallest range,

and then scale all 6 axis from their min max to 0 till that common range.

That would make it easier to use all 6 axis in same way.

The other thing I was talkin about is to create "profile" for each user, dancer 

or whatever you call that person hanging on that 2 lines,

in case smaller person's output would differ from taller one that much

If there is no real difference in that aspect, than one con forget it.

After the ranges of 6 axes are set, then one can start to use the values.

It makes no sense here to list all possible ways to split and use values

and ranges, it is up to you to say what you want the values to do,

Shortly back.
To repeat :
one has to set usable range of all 6 axis.
By measuring their min and max values, than finding which one has the smallest range,
and then scale all 6 axis from their min max to 0 till that common range.
That would make it easier to use all 6 axis in same way.
------
The other thing I was talkin about is to create "profile" for each user, dancer 
or whatever you call that person hanging on that 2 lines,
in case smaller person's output would differ from taller one that much
that it needs adjustment.
If there is no real difference in that aspect, than one con forget it.
--------
After the ranges of 6 axes are set, then one can start to use the values.
It makes no sense here to list all possible ways to split and use values
and ranges, it is up to you to say what you want the values to do,
then it would be easy to manage it.


hi device