The Perfect Simulator

[Par2Pro]

Hey All,

In 2012, I would like to put together the perfect simulator - or as close as possible to perfect

What I would like from you is what you would like to see in YOUR ideal simulator.

Any feedback is good feedback; however please keep the feedback constructive.

I will also work with some of our top contributors on prototype systems.

The key to this perfect simulator is not only building it, but creating a support structure to ensure that everyone can buy one with confidence.

Looking forward to hearing your thoughts,
Cory

[ZMax]

Not sure how you're going to accomplish this but I'll give my opinion anyway.

My ideal golf simulator would be an all camera system that measures ball speed, ball path, ball LA, spin, and spin axis tilt. No need for club sensors. Need to keep it simple and cheap. It just needs to measure the ball, which is what's really needed to accurately simulate the ball flight.

With an all camera system, one should be able to hit from any surface. Putting can be done on a putting surface. No floor sensors or triggers.

Should be able to use Trugolf's E6 or Red Chain's simulation. Actually, Links LS would be perfect if possible. Must have lots of adjustability. I.e. The option to select the club or not, ball speed adjustments for each club, ball flight, spin, bounce, roll, etc...

The price for the hardware and software alone should be no more than $7K.

Cory, I would love to get involve with a prototype system.

[bubba22]

Now we are talking. I am 100% with ZMax. Count me in on testing.

[steb]

The perfect simulator doesn't have to perfectly simulate but the ball flight (direction, trajectory, speed and spin) must be believable to the top golfer.

I suggest three models, allowing the user to upgrade as money allows:

1) "Simulator" that has multiple cameras analyzing the ball for speed, path and spin.

2) "Analyzer" upgrade that adds more cameras for analysing the clubhead. Clubhead path (including attack angle), clubface angle (including dynamic loft and bulge recognition).

3) "Biomechanics" upgrade that adds more cameras for analyzing the kinematic sequencing of the body, specifically hips, shoulders, arms and hand motion. Possibly two options here, a marker based camera system requiring maybe 3 cameras and the ultimate, a full markerless camera system like 4dSwing with 5, 8 or more cameras.

Unacceptable

• Optical IR sensors
• Marking clubs
• Special balls (however marking own balls would be acceptable)
• Aligning balls
• Having to touch a console. Any navigation should be gesture or at least voice controlled.
• Screen bounce spin determination
• Face angle or spin calculation
• Entry level price > $5k.

Software

• Totally hands-free
• No unacceptable waiting for screen redraw
• Free course designer to encourage community designed courses, allowing people to recreate their own tracks for home practice and release famous tracks into the public domain without the royalties.
• Course designer to import unlocked Links courses for modernizing.

Luxury

• Analyzer module portable and self-calibrating for taking outside to a range.
• Minimal fussing around when going from left-hander to right-hander.

I could go on forever, plus I have a few killer ideas I'll keep close to my chest. If someone with the budget is seriously wanting to develop a system that would have even the professionals flocking to, feel free to PM me as I'm just coming to the end of a IT-based golf project and would consider relocating to develop the ultimate simulator system. I'm an Electrical Engineer, 20 years experience (mostly developing real-time embedded software), a single figure golfer and a good, broad knowledge of many areas involved, including the science of golf, swing mechanics, fitting, machine vision, motion tracking, biomechanics, radar theory, image processing and speech recognition.

[bubba22]

The perfect simulator doesn't have to perfectly simulate but the ball flight (direction, trajectory, speed and spin) must be believable to the top golfer.

I suggest three models, allowing the user to upgrade as money allows:

1) "Simulator" that has multiple cameras analyzing the ball for speed, path and spin.

2) "Analyzer" upgrade that adds more cameras for analysing the clubhead. Clubhead path (including attack angle), clubface angle (including dynamic loft and bulge recognition).

3) "Biomechanics" upgrade that adds more cameras for analyzing the kinematic sequencing of the body, specifically hips, shoulders, arms and hand motion. Possibly two options here, a marker based camera system requiring maybe 3 cameras and the ultimate, a full markerless camera system like 4dSwing with 5, 8 or more cameras.

Unacceptable

• Optical IR sensors
• Marking clubs
• Special balls (however marking own balls would be acceptable)
• Aligning balls
• Having to touch a console. Any navigation should be gesture or at least voice controlled.
• Screen bounce spin determination
• Face angle or spin calculation
• Entry level price > $5k.

Software

• Totally hands-free
• No unacceptable waiting for screen redraw
• Free course designer to encourage community designed courses, allowing people to recreate their own tracks for home practice and release famous tracks into the public domain without the royalties.
• Course designer to import unlocked Links courses for modernizing.

Luxury

• Analyzer module portable and self-calibrating for taking outside to a range.
• Minimal fussing around when going from left-hander to right-hander.

I could go on forever, plus I have a few killer ideas I'll keep close to my chest. If someone with the budget is seriously wanting to develop a system that would have even the professionals flocking to, feel free to PM me as I'm just coming to the end of a IT-based golf project and would consider relocating to develop the ultimate simulator system. I'm an Electrical Engineer, 20 years experience (mostly developing real-time embedded software), a single figure golfer and a good, broad knowledge of many areas involved, including the science of golf, swing mechanics, fitting, machine vision, motion tracking, biomechanics, radar theory, image processing and speech recognition.

steb it sounds like you should just pick up the GC2. It probably fullfills most of your wishes and is about 6-7 grand. I personally think that there is a strong role for ir reflective technology in the home sim market. I do agree that the camera based systems especially one that measures spin will be more accurate. Have you checked out the GC2?

[jagmanjoe]

I am new to the group but with my limited knowledge aren't the sensors about the best and most effective way to trigger the cameras you want to utilize? Also, for what its worth, if someone does get something going, I would be happy to contribute my knowledge relative to casting plastic housings for circuit boards, cameras, etc.

[steb]

steb it sounds like you should just pick up the GC2. It probably fullfills most of your wishes and is about 6-7 grand. I personally think that there is a strong role for ir reflective technology in the home sim market. I do agree that the camera based systems especially one that measures spin will be more accurate. Have you checked out the GC2?

Heh bubba, the GC2 is a decent attempt at the entry level 'simulator' I mentioned above but it puts portability before accuracy. Being as side-on as it is leads to an inaccurate horizontal launch direction. It's quoted ±1° horizontal launch accuracy would mean identical perfect shots down the middle would finish somewhere in a region about 10 yards wide at pro driving speeds. Add in the left-right spin error and you're up to about 17 yards. That's 1/2 to 2/3 of the typical fairway width. On a simulator that would mean if you aimed for the left half of the fairway to leave a good good approach angle and hit it perfect, you could find yourself in the rough.

Plus the bit they've made portable is the simulator bit. At the range, I can watch the ball flight with my own eyes. I can see within few yards my horizontal direction and see the most minute amounts of side spin. A ball readout is definitely helpful in windy conditions, but it has to be better than showing a 17 yard horizontal dispersion.

Also ±1° is the manufacturer's claim, which always seem to be optimistic. It could be a theoretical accuracy from their camera resolution or it could be a confidence interval derived from where the balls actually land. That confidence interval could be 95% of shots within ±1°, it could be 68% or some other less used %. Who knows?

What you want to be portable to take to the range is the 'analyzer' and 'biomechanics' bit. I'd want to see clubhead parameters and how my body is moving. Am I missing the sweetspot? Did I draw the ball or did the face bulge cause it to come back? Did I swing more through the centre of the ball to get the indicated ball speed or did I actually have more clubhead speed? Did I generate that extra clubhead speed via a more optimal, balanced chaining of body segments or did I swing out of control and just happened to catch it good?

GGS's hcam being directly above is not portable of course but it allows for the horizontal launch accuracy you need for an indoor simulator. Otherwise it's an arcade game, not a simulation.

[steb]

I am new to the group but with my limited knowledge aren't the sensors about the best and most effective way to trigger the cameras you want to utilize?

For triggering, not so bad. It's being used for measurement I'm not keen on. Still with high fps camera prices dropping so quickly, soon we should be able to rely completely on camera without breaking the bank.

[syhlif32]

Steb: One degree off theline on a 300 yard drive is 5.2 yards!

There are two suggestions out there for pricing $5,000 and $7,000
I would add a vote for $5,000 but could go for $7,000 too. That is for the sensors and software alone.
Ifone degrees accuracy is not good enough then add 2-3 zeros to the price.
Most of us have limited space so it will need to fit in aspace maybe 5-6 meters (17-20feet) long. That more or less rules out radar based systems, I think?
While I would not care about taking it to the range it would still be nice with a system that could be packed away andsetup in less than 10 min in case the room is used for other activities. Home theater or a car garage.

Voice control is easy and already done. I never use thekeyboard with my Optishot anymore.
Ability to make own/ use community developed golf courses would be high on my list. Free new/extra golf courses should be looked at as sales pitch not a source for extraincome for the software provider.

Simulators without swing feedback are for me pointless. Sothere will need to be some kind of club sensors. That is the only advantage a simulator has overgoing to outside playing
Ball type selection is something I would like to see. Herethinking ball types like low /high spin balls.

Select real worldclubs that behaved like your clubs does. This could be a way to generate incomeby charging golf equipment manufactures a fee for having their balls and clubsin the selection.
Online play/ tournaments would be great! Even better ifthere was a voip system so you could chat (banter) during the game.

Since we are talking about a commercial simulator there isthe aspect of service. I have been looking into upgrade my simulator.
Of the two companies I have approached. One has answeredevery email inside a few hours. The other company has not answered emails thelast 9 days. It was the other company’s simulator I really wanted but now notso sure. If you put down good money there need to be service!

Nice thread Cory, Nice to read what other people think is important.

[steb]

Steb: One degree off theline on a 300 yard drive is 5.2 yards!

Correct, hence ±1° leads to the 10 yard region I mentioned above, which expands to around 17 yards when ball spin error is also added in, or ±8.5 yards.

I think pictorially shows it better. Aiming for the dead center of the left portion of an average US Open fairway and hitting it exactly there could have the simulator showing your ball as anywhere across the red rectangle.

2d6r7lk.gif

[bubba22]

Well I play to a 4 handicap and am very happy with my Protee. If you can develop a cam based system at hat price then go for it and count me in. I think it will be a challenge for sure. Where are the GC2 users? Those with the GC2 have the benefit of using the sim as a launch monitor on the real course and can judge the realism of the ball flight etc. Unfortunately sims like the About golf, although perhaps extremely accurate, one will never know what the real accuracy is because they are not portable and you have to trust what is seen. I practice on the About golf sim once every few weeks and frequently go on a launch monitor so have some idea about my ball flight, spin, launch etc and can tell you that my Protee is very, very reasonable. Is it perfect, absolutely not. For m it is good enough an provides entertainment and decent practice. If a cam based system at that price is developed however, then let's go for it.

[tarmactrr]

Not sure how you're going to accomplish this but I'll give my opinion anyway.

My ideal golf simulator would be an all camera system that measures ball speed, ball path, ball LA, spin, and spin axis tilt. No need for club sensors. Need to keep it simple and cheap. It just needs to measure the ball, which is what's really needed to accurately simulate the ball flight.

With an all camera system, one should be able to hit from any surface. Putting can be done on a putting surface. No floor sensors or triggers.

Should be able to use Trugolf's E6 or Red Chain's simulation. Actually, Links LS would be perfect if possible. Must have lots of adjustability. I.e. The option to select the club or not, ball speed adjustments for each club, ball flight, spin, bounce, roll, etc...

The price for the hardware and software alone should be no more than $7K.

Cory, I would love to get involve with a prototype system.

This is basically what I'm looking for. I don't even care about the putting and the golf courses, I'm simply looking for something I can continue to work on my swing, and actually be able to see the changes in ball flight when I make adjustments.

[tarmactrr]

Ball type selection is something I would like to see. Herethinking ball types like low /high spin balls.

If you had it measuring ball spin then there is no reason for this. That's why ball spin measuring is a need for me, as I need to know how it's spinning with the ball I currently play. I also need to know how changes in my swing will affect the ball spin rate on different clubs.

[syhlif32]

Yes there is!
You might be using a 'almost golf ball' The soft ware will then translate that spin into the spin the ball you would prefer playing with would have had.
The last I should have added and not just been thinking. I forgot that some people are able to play with real golf balls.

[rdh]

My two cents is that if anything is to happen here, someone needs to determine the focus. Otherwise you'll have a Martin situation on your hands where you try to develop all sorts of options for price vs. accuracy/features. Lots of people are talking about some magic $5K-$7K system that can somehow be better than the $50K systems. I don't think that's a reasonable expectation unless you want to set yourself up for disappointment. I think it would be great for a plan to come together for a $5K-$7K system that is BETTER than the current $5K-$7K systems....but stay within reason. If you want to build/sell a system for $5K-$7K and can't live with +/- 1 degree accuracy on most measurements (which yes, all put together means there's some dispertion in how shots may come up) then I don't think you really want to build a simulator...you want to recreate real life somehow.

For the record, my Pro 7 is all shadow-sensor based (with reflective sensors for launch angle) and I've always been extremely impressed with what i PERCEIVE to be the accuracy. I can play a round in real life, and come home and count on seeing the same shot shapes, distances, etc. I saw on the course. When I hit a ball solid vs. thin, the ball flight does just what I'd expect. But, is the ball going a few yards left or right of where it really would have? I'm sure it is. But the overall experience is in line with what I can do in real life, and my scores mirror the same range of scores I have in the real world. You can create a system that may not be 100% verifiably accurate all the time, but still works great and serves its purpose.

[bubba22]

Glad you piped in rdh. I completely agree. I love my Protee. It plays like I play my real game give or take. If there is a push for perfection at the 5-7 grand range then go for it. I am not sure how easy that will be.

[steb]

A few yards left or right is fine, rdh. ±1° is too inaccurate for my wishes, but halve that to ±0.5° and I'm smiling. So it's not about being perfectly accurate, it's about being sufficiently accurate and of course that is personal. I define sufficiently accurate as if I aim for the left side of the fairway and hit it perfect, I want to see it somewhere on the left side. That's my precision on the course. Someone else may have yard precision, someone else may aim center and be happy to see it hit fairway anywhere.

[tarmactrr]

Yes there is!
You might be using a 'almost golf ball' The soft ware will then translate that spin into the spin the ball you would prefer playing with would have had.
The last I should have added and not just been thinking. I forgot that some people are able to play with real golf balls.

Ahh sorry, I read it as you if you were eluding to different types of REAL golf balls (IE Pro V1 vs Pinnicle etc). I was stating that ball spin will take care of that aspect.

[steb]

Also don't forget these $50k systems are enclosure, projector, screen, PC, installation... and the price of high fps CMOS/CCD sensors is moving rapidly down. I'm tinkering with a 375fps camera setup that goes for $60 new which was unheard of 5 years ago. It's by no means sufficient, but serves to illustrate accuracy won't always mean huge bucks.

[rdh]

What is the brand and model of that camera steb? I've been wanting to get one to record my swing (I've only had it done by a pro in a few lessons I took previously) and that one sounds pretty good.

[bubba22]

Do tell steb? What cam is that?

[steb]

The camera you're quite fond of, bubba, the PS3 Eye but three of them, each running at 125fps, time interleaved. I'm playing with them for biomechanic swing analysis rather than launch analysis, which would require a few more cameras added to either increase fps or resolution by stitching (they're only 320x240 @ 125fps).

[bubba22]

Nice steb. It truly is a great cam. Your right that I am fond of it. My favorite cam which I have been praising for a long time.

[Asics]

Following up on steb's posts, does anyone know how the GC2 specs compare with the published data of other systems such as Flightscope or even Trackman when those systems are used indoors? I'm sure FS and TM are more accurate when used on a range, but what are their specs if hitting into a net w/ limited ball flight of 10-15 feet? My guess is that the GC2 would be pretty competitive. I couldn't find published info from FS or TM on their websites regarding their specs when used indoors. Interesting to see that the indoor TM now costs only $13,995. I think it was around $20k a year or two ago.

[syhlif32]

While I do not know which one is the most accurate I do also find that it isn't that important!
There isn't any golfer around that could detect the 1 degree 'off'.

Could we get back to the 'perfect' golf simulator that could be made for $5-$7,000

I feel it is the software that is lacking behind, not the hardware if you look away form spin measurement.

[rdh]

I agree on the software topic. I think E6 is probably the most impressive software out there today, but I'm hoping red chain's next engine advances things forward. Really, the graphics and level of reality in the various software packages hasn't changed much since Links 2003. The main change has been making things 3D models with better lighting and movement (and the ability to render in real-time so the software can fly with the ball and stuff) but if you'd asked me 8 years ago where we'd be by 2011, I would have been sure that software would be able to render (not use photographs) with near-photo realism.

[bubba22]

I agree RDH. However having said that, the hardware with cams have and will continue to come down in price. I thinn the E6 is a top notch software. Hopefully the new Red Chain will follow as well.

[golfnuts]

the perfect simulator for playing a round on a golf course greatly improves the putting. otherwise, the score is really skewed as the putting that represents 40% of your score is a totally different beast to learn/conquer.

as for all the other stuff, here is my quick take:

figure out what you want- teaching, improvement, recreation, combination?

I use my GC2 for grooving ball striking or improvement. Before this machine, for 15 years I had hit thousands of balls in my basement without knowing distance. Mostly with the irons. Since I set up my system in March, I have hit 800-100 ball indoors. the first third were in full simulator mode mostly on a driving range. the last 2/3 were hit mostly with the simulator off. I just read the GC2 for "carry" distance. I ignore the spin readings. if I hit it extra long, I check to see if I pulled it. If I am "short of my expected yardages, I tweak the swing, set up, etc. Thereafter,I begin consistently hitting a 7 iron my target distance of 156-162. Personally, I do not need to see my club head data. just carry distance. and remember, your hitting surface,matt is simulating a fairway . . .not the rough.

the GC2 carry numbers are very accurate. However, if they are off a couple of yards, it does NOT matter. I want to see 20 balls with my 7 iron going distances within 8 yards. and this machine tells me that.

as for curving the ball, IMHO, that would require taking the GC2 to the range on a perfectly calm day and making notes. the GC2 says the ball went left "1030L", but until I can see what the ball does outdoors, it is a meaningless number. maybe a project for next summer . . .

as for playing 50 different world class golf courses, it is honestly a bit disappointing. as I already mentioned, in real golf putting is 40% of your score and simulation putting is very unreal. thus shooting any score on these great courses doesn't mean much to me. recreation in the dead of winter, maybe. more than likely, I will have closest to the pin contests from 100 yards and in. and that just might lower my 5 handicap . . .

as I began, know what you want to accomplish . . . I love the GC2 for my purpose.

[golfnuts]

good luck Cory

[rdh]

On the topic of putting...I actually find the simulator to be fine. Granted that I've tweaked my interface so a ball that rolls 4 feet on my turf does in fact go 4 feet in the software (assuming a flat green) but really putting is just different than real life and always will be. When I got my simulator I was three and four putting a lot, and figured the putting just wasn't realistic. The fact is, it's just a different beast. Putting is so much feel and sight-based in real life, that there's no way you can simulate that. Working against a static computer image and interpreting grid lines for break is a much different thing. After playing quite a bit, though, I generally putt fairly well on the simulator...or at least similar to how I putt in real life. I've gotten to know how hard to hit it for different distances in the simulator (just based on reading how many feet away I am), I've learned how much the break grid influences putts, and I've learned how much uphill/downhill readings (in the inches the software tells you) will affect things. So, my mindset when I set up for a putt on the simulator is just a different process than in real life where I am more working on just getting a feel for the break, speed, etc. In some ways it's a different skillset than in real life, and you need to work at it.

But...in the end the stroke is the stroke. You still need to hit a putt on line and with the speed you intended. When I do that on the simulator, it generally works out okay.

My bottom line acid test has always been...if I play a simulated round on a course similar to those I play in real life (parkland style) with similar length tees, is my score in line with my real life game? I can absolutely say it is. Almost every round this year I shot between 80 and 84 (a few blow up rounds, and one 78 that was on a short course)...typically on the bubble for breaking 80 each time but generally not doing it. When I play Cog Hill, Bro Balsta or Sunningdale on the simulator (my most frequented courses), I am in the exact same range. Sometimes I manage to shoot 79, but my best round is 78 and 90% of the time I am between 80 and 84. For me, that's a solid simulated experience.