Bushnell Launch Pro Device with Foresight

[Copacetic]

I still don’t get why people think it needs to be so expensive, it’s purely a business call on how they price it, the components won’t add up to more that 500 or so, since quad was designed there have been massive improvements in camera sensor technology’s. Camera chips with on ultrafast readout built in are now cheap, led by phone development. Assuming they can get and are putting the latest chips in them, they can sell this at 1000, make 100 a unit and sell millions. Or they can get it at 2000, make 1100, sell thousands, reduce to 1500 sell 100k of them,  it meanwhile the r10 from garmin is taking their whole market away. As it’s ‘good enough’ for 95% of people and much cheaper. 
 

their shots at garmin and radar in the ad means they must be going after the market. 

[karstens_ghost]

1 minute ago, Copacetic said:

I still don’t get why people think it needs to be so expensive, it’s purely a business call on how they price it, the components won’t add up to more that 500 or so, since quad was designed there have been massive improvements in camera sensor technology’s. Camera chips with on ultrafast readout built in are now cheap, led by phone development. Assuming they can get and are putting the latest chips in them, they can sell this at 1000, make 100 a unit and sell millions. Or they can get it at 2000, make 1100, sell thousands, reduce to 1500 sell 100k of them,  it meanwhile the r10 from garmin is taking their whole market away. As it’s ‘good enough’ for 95% of people and much cheaper. 
 

their shots at garmin and radar in the ad means they must be going after the market. 

 

Their shots at radar are simply setting people up to accept camera sensors as equals to radar, and not inferior.

 

The high speed camera tech that you're talking about is much better at the phone level, but I'm not sure 240fps cuts it when it is needed to estimate 300m over less than 1m. 

[jagmanjoe]

1 hour ago, Copacetic said:

I still don’t get why people think it needs to be so expensive, it’s purely a business call on how they price it, the components won’t add up to more that 500 or so, since quad was designed there have been massive improvements in camera sensor technology’s. Camera chips with on ultrafast readout built in are now cheap, led by phone development. Assuming they can get and are putting the latest chips in them, they can sell this at 1000, make 100 a unit and sell millions. Or they can get it at 2000, make 1100, sell thousands, reduce to 1500 sell 100k of them,  it meanwhile the r10 from garmin is taking their whole market away. As it’s ‘good enough’ for 95% of people and much cheaper. 
 

their shots at garmin and radar in the ad means they must be going after the market. 

I agree with you Copacetic.  Especially when they can also attach an annual fee or two to it for use and perhaps the second one for golf simulator course software.  The annual recurring revenue for even 250k to 500k units would be significant to their bottom line.

[whumber]

4 hours ago, karstens_ghost said:

The high speed camera tech that you're talking about is much better at the phone level, but I'm not sure 240fps cuts it when it is needed to estimate 300m over less than 1m. 

 

Cell phone cameras doing 720p at 1000fps have been around for quite a while now.

[Simpsonia]

And the GC2/GC4 use 10,000 FPS cameras. Now you can argue that Foresight might have been able to improve their software enough to rely on slower speed cameras (though they still have to get around Garmin's recent patent application for a launch monitor using slower speed cameras), but at that point it's all software and calculations. Now most characteristics beyond the most basic of ball speed and launch angle, such as spin rate and axis, club head data, etc will be much more calculated rather than measured, and will almost certainly not be as accurate as the GC2/GC4 is able to deliver with 10k fps cameras. Skytrak uses two 3,000 FPS cameras and is notoriously less accurate than the GC2/GC4, especially with high ball speeds. I very much doubt that software can be improved far enough that two (third is for club head) 1,000 FPS cameras (or below) can deliver the accuracy that Foresight/Bushnell is promising. 

 

I really do hope that this comes out in a price point that delivers all that for a song and dance, but I think it's incredibly unlikely that it comes in at under $3k, let alone the lower numbers people have been tossing around. 

Edited August 9, 2021 by Simpsonia

[whumber]

2 hours ago, Simpsonia said:

And the GC2/GC4 use 10,000 FPS cameras. Now you can argue that Foresight might have been able to improve their software enough to rely on slower speed cameras (though they still have to get around Garmin's recent patent application for a launch monitor using slower speed cameras), but at that point it's all software and calculations. Now most characteristics beyond the most basic of ball speed and launch angle, such as spin rate and axis, club head data, etc will be much more calculated rather than measured, and will almost certainly not be as accurate as the GC2/GC4 is able to deliver with 10k fps cameras. Skytrak uses two 3,000 FPS cameras and is notoriously less accurate than the GC2/GC4, especially with high ball speeds. I very much doubt that software can be improved far enough that two (third is for club head) 1,000 FPS cameras (or below) can deliver the accuracy that Foresight/Bushnell is promising. 

 

I really do hope that this comes out in a price point that delivers all that for a song and dance, but I think it's incredibly unlikely that it comes in at under $3k, let alone the lower numbers people have been tossing around. 

At what resolution is GC2 (and I assume GCQuad) reading out at 10Kfps? Given the size of the assembly and the state of camera technology when the GC2 was released it must be at a very low resolution. My understanding of the GC2 was that it is actually taking a single long exposure and then using high speed strobes to "stack" multiple exposures at very high speed and analyzing that composite. No idea about that GCQuad.

 

EDIT: FWIW I fully expect this thing to come in between $3-5K.

Edited August 9, 2021 by whumber

[GolfTurkey]

13 hours ago, karstens_ghost said:

 

Their shots at radar are simply setting people up to accept camera sensors as equals to radar, and not inferior. 

I’m not sure that’s a thing. Camera is far superior to radar indoors and PGA tour driving ranges are now full of GCQuads.

[Karazy]

The partnership with Bushnell is clearly to get into physical retail stores. How many stores will carry a 4 or 5 thousand dollar product? I would guess this launch monitor will be driven by monthly subscription fees. I don’t have any inside knowledge so just guessing here, $3,000 with a monthly/annual subscription model.

[RGlaze978]

Proprietary infrared and three camera system. 

[Stuart_G]

1 hour ago, GolfTurkey said:

I’m not sure that’s a thing. Camera is far superior to radar indoors and PGA tour driving ranges are now full of GCQuads.

 

It's a common opinion among these forums.  But the fact is that it has never been validated in any proper study.    The one proper study that was done actually showed the difference was not statistically significant (done between gc2/HMT and TM IIIe).

 

GCQuad's are more convenient to use, not more accurate - especially on an outdoor range with no face stickers to help with the club data.

 

Edited August 9, 2021 by Stuart_G

[GolfTurkey]

6 minutes ago, Stuart_G said:

 

It's a common opinion among these forums.  But the fact is that it has never been validated in any proper study.    The one proper study that was done actually showed the difference was not statistically significant (done between gc2/HMT and TM IIIe).

 

GCQuad's are more convenient to use, not more accurate - especially on an outdoor range with no face stickers to help with the club data.

 

 

Agree re. using them outdoors. I don't have any scientific studies to reference, only my own experience with radar and camera LM's indoors.

 

Radar needs way more space. Then you need even more space for sensible results if you have a ball speed above room temperature. A couple examples why something like a Mevo+ sucks indoors (and my distances, tee heights etc. are meticulously measured and inputted):

 

I was doing a combine this weekend. I hit consecutive shots with a 7i and PW, both straight and well struck and worth at least 3 points. The 7i spin registered as > 11K and the PW was 5K so both got zero points with the 7i pulling up well short and the PW airmailing the target. Makes it useless for skills test and Sim play IMO.

 

There is a Flightscope training video on the web somewhere with Alex from FS hitting 7i shots on a Mevo+ in an ideal looking environment. Almost every single spin rate is in italics. Ever seen that on a GC2?

[karstens_ghost]

5 hours ago, GolfTurkey said:

I’m not sure that’s a thing. Camera is far superior to radar indoors and PGA tour driving ranges are now full of GCQuads.

 

I know that. The question was towards the marketing statement. Most laymen feel radar is superior, and it is in the sunshine...

[Stuart_G]

8 hours ago, GolfTurkey said:

 

Agree re. using them outdoors. I don't have any scientific studies to reference, only my own experience with radar and camera LM's indoors.

 

Radar needs way more space. Then you need even more space for sensible results if you have a ball speed above room temperature.

 

More space, yes.   But the rest is an exaggeration.   Indoors, enough room for 2 revolutions of the ball is the ideal.  A little less will give reasonable data.  Only distance issues tend to be with driver (assuming they remember to use the metallic dots).  

 

No experience with Mevo+  but based on it's cost and size, can't say I'm surprised, you do get what you pay for.  Smaller means less power and more susceptible to RF interference.   I've seem posts of some tring to set them up pointed right into a metal garage door and still expect them to work perfectly (it's kind of funny).  If your judging all radars based on the mevo+ then your feelings might be understandable, even if misguided.  But I haven't had any higher percentage of misreads with my Xi Tour than I have with my GC2.

 

 

 

 

 

[Stevie989]

The slow rolling of the teaser stuff via email is funny. 
 

mygolf spy have been pretty chatty about the foresight but radio silent on the r10. I know from what’s he’s said before Tony covey reckon that the small Doppler radars are limited in how good the can actually get. 

[hammersia]

On 8/9/2021 at 10:00 AM, karstens_ghost said:

 

Their shots at radar are simply setting people up to accept camera sensors as equals to radar, and not inferior.

 

The high speed camera tech that you're talking about is much better at the phone level, but I'm not sure 240fps cuts it when it is needed to estimate 300m over less than 1m. 

And yet simple maths tells you even a bog standard 240fps camera will capture four images of a 180mph ball speed over one metre. And my old Vector Pro didn’t even need to do that as it just used one camera per image, which these days would be a five dollar sensor. 
 

We obviously won’t know for sure until specs and a tear down are done, but it’s perfectly possible to make and sell this CONFIGURATION profitably for under a thousand bucks. 

[Stuart_G]

42 minutes ago, hammersia said:

And yet simple maths tells you even a bog standard 240fps camera will capture four images of a 180mph ball speed over one metre.

 

But not very accurately.  At the high end for ball speed (which the units have to be designed for), it's maybe 3 (but more likely 2 images) and the ball could move 1/3 of a meter over the course of one frame being captured.    Trying to measure anything on such a fuzzy/distorted ball would not be very useful.

 

 

Quote

And my old Vector Pro didn’t even need to do that as it just used one camera per image, which these days would be a five dollar sensor. 
 

 

can't say for sure but the vector likely used a global shutter.  Even at lower frame rates, that would still be a lot more than that, even now.

 

Regardless, you still have to look at a lot more than the hardware cost when pricing.  Lots of R+D costs in any new unit like this as well as algorithm development and firmware costs.

 

 

Edited August 10, 2021 by Stuart_G

[whumber]

16 minutes ago, Stuart_G said:

 

But not very accurately.  At the high end for ball speed (which the units have to be designed for), it's maybe 3 (but more likely 2 images) and the ball could move 1/3 of a meter over the course of one frame being captured.    Trying to measure anything on such a fuzzy/distorted ball would not be very useful.

 

I'm not sure why you think the images would be fuzzy (blurry?). Even with a framerate of 240fps you can have much shorter exposure times to reduce motion blur. Rolling shutter distortion would be the larger concern but with at worst a 1/240s scan time motion blur should be relatively low and can be corrected for to an extent.

[JMG71]

Why team up unless aiming for a new business plan and at the very least x100 unit sales? Also, the way to go these days is a low entrance point with annual or monthly fees.

Why even bother about radar in their e-mail ads unless positioning themselves for some low hanging fruit on the Garmin R10 potentially high volume market?

My bet is 1-1.5k but with recurring fees to unlock features.

If so, they will make loads of money. If not, they will likely still be successful fighting for scraps in the Mevo+ and Skytrak semi premium segment.

[Stuart_G]

14 minutes ago, whumber said:

 

I'm not sure why you think the images would be fuzzy (blurry?). Even with a framerate of 240fps you can have much shorter exposure times to reduce motion blur. Rolling shutter distortion would be the larger concern but with at worst a 1/240s scan time motion blur should be relatively low and can be corrected for to an extent.

 

If yours cant', find a friend with one of these cell phone camera that can take 240 fps video - see for yourself what the ball looks like when it's in motion.

 

Edited August 10, 2021 by Stuart_G

[whumber]

12 minutes ago, Stuart_G said:

 

If yours cant', find a friend with one of these cell phone camera that can take 240 fps video - see for yourself what the ball looks like when it's in motion.

 

 

That's because the software in cell phones generally only allow the 180 degree rule for shutter speed in order to maintain a "cinematic" look which is exactly the opposite of what you want for and kind of movement analysis. Shutter speed (or angle) and frame rate are mostly independent with the exception being that you can't go any longer than 1/framerate.

Edited August 10, 2021 by whumber

[hammersia]

40 minutes ago, whumber said:

 

That's because the software in cell phones generally only allow the 180 degree rule for shutter speed in order to maintain a "cinematic" look which is exactly the opposite of what you want for and kind of movement analysis. Shutter speed (or angle) and frame rate are mostly independent with the exception being that you can't go any longer than 1/framerate.

👍Top Bearding that

[Stuart_G]

47 minutes ago, whumber said:

 

That's because the software in cell phones generally only allow the 180 degree rule for shutter speed in order to maintain a "cinematic" look which is exactly the opposite of what you want for and kind of movement analysis. Shutter speed (or angle) and frame rate are mostly independent with the exception being that you can't go any longer than 1/framerate.

 

Frame rate limite is determined by the shutter speed.  So max frame rate is a pretty good indication of what that speed potentially could be.

 

Here is (one of) my sources:

 

https://core.ac.uk/download/pdf/78638051.pdf

 

What's your source?

Edited August 10, 2021 by Stuart_G

[whumber]

9 minutes ago, Stuart_G said:

 

Here is (one of) my sources:

 

https://core.ac.uk/download/pdf/78638051.pdf

 

What's your source?

Your source is fine, but to be blunt it doesn't appear that you understand it. I suspect you're confusing the distance moved between frames (a function of framerate) with the motion blur (a function of shutter speed). Again, like I said before, framerate and shutter speed are fundamentally independent from each other. Your own source has an explanation of this in Figure 4(a). Saying a camera has a max framerate of 240fps doesn't actually tell you anything about the amount of motion blur you'll see, the most you could take from that is the slowest possible shutter speed you could shoot is 1/240s.

Edited August 10, 2021 by whumber

[hammersia]

On 8/6/2021 at 6:26 PM, radiman said:

10 times out of 10 I would choose GC2 like accuracy over a skytrak even if it were $2k more expensive.  I have hit on skytrak plenty and found it to be incredibly inaccurate.  I don't trust the ball data and certainly wouldn't trust the club data provided.  I haven't owned one so I am not sure if that can be improved upon.  Main reason being I found it to not really be worth a $2k investment.  Foresight ball data indoors is unrivaled.  To me, that alone would make it worth more $$ to me than a skytrak.  

That really isn't the experience of the many Skytrak owners in here.

 

It's gets a little criticism for its sides pin algorithms particularly at the top end ball speeds, but that isn't really germane to the ball measurement question anyway. 

[Stuart_G]

19 minutes ago, whumber said:

Saying a camera has a max framerate of 240fps doesn't actually tell you anything about the amount of motion blur you'll see, the most you could take from that is the slowest possible shutter speed you could shoot is 1/240s.

 

But it doesn't tell you that there wont be blur either - which is what you seem to be arguing and fairly certain of (unless I'm mistaken).   The only difference here is that you seem to making assumptions about the max shutter speed and I'm not.  

 

Also, the generic home simulator setup is not going to come close to laboratory lighting conditions - so expecting performance to be close to the spec'd max shutter speed isn't really reasonable.  Likely the reason GC2/Quad went to non-visible lighting and strobes - and also why skytrack struggles in some lighting conditions.

 

 

 

Edited August 10, 2021 by Stuart_G

[BSTRONG1313]

Wow! This is all going over my head. I just want it to give me reliable club data at a semi-reasonable price.

 

But as someone with no experience in thid space I can tell you there is no way this retails for $1,000. Margin is almost always preferable to high volume from a business perspective. You don't see Bushnell dabbling in the $150 rangefinder space. 

[radiman]

58 minutes ago, hammersia said:

That really isn't the experience of the many Skytrak owners in here.

 

It's gets a little criticism for its sides pin algorithms particularly at the top end ball speeds, but that isn't really germane to the ball measurement question anyway. 

Sure, I get that.  It's just my experience.  I know a guy who had a skytrak for a few years.  Last year he bought a GCQuad and said it is hands down more accurate.  He has used his skytrak much more than I ever had used one.  He said that as he used more club the less accurate it would be.  Again, just my experience and what I have heard from him.  Assuming at worst we will get the same ball data that a GC2 or Quad would produce, I think this is simply going to be a much more accurate unit. 

[Copacetic]

What I was trying to say above was that sensors have moved on massively in last few years. These new cameras won’t have a ‘shutter’ they will be digital sensors most likely with a global shutter, meaning you can read all the data in one go, the only ‘limit’ on shutter speed is the processing power you have behind the sensors. The phones don’t have 240fps sensors, they have sensors that the phone can only read from at that rate as they keep power for everything else. 
 

basically it’s a whole new world since GCquad was developed and piggy backing on phone and mirrorless camera development means a big opportunity as the units do nothing else. So they can be made pretty cheaply and foresight know well someone else will do it if they don’t. 
 

a good explainer for people who don’t work with camera sensors for a living 

 

https://www.digitalcameraworld.com/au/news/what-is-a-global-shutter-and-why-is-it-so-important

[Simpsonia]

All of this a moot point now. The marketing materials above clearly show that the hardware in the new LM is capable of "capturing thousands of images a second", which is clearly well beyond the 240 fps or even 1K fps cell phone grade sensors being debated. 

 

 

[Rory4Pres]

It's nice some new guys are coming into the market place.  We've been waiting for ages.  Skytrak is still selling the same consoles they were selling in late 2014.  I full expect this new bushnell/foresight unit to be much superior to it.  Maybe it'll finally force Skytrak to come out with a new model.  Maybe it'll force Mevo+ to unlock all of it's features.  

 

Either way, the killer is how much E6 and others are charging for course simulator modes.  I'd love to see more competition in that sector.  I shouldn't have to pay anything close for a fake course as I do for the real thing.  Also, if short shots, chipping, bunker play and putting aren't close to the real thing, then the only thing it's really good for is practicing (aka range time).  I get unlimited range balls on real grass for under $200 a year so E6 wanting as much as they are asking for with some consoles is hysterical.