Crooked Stick Golf Club (LiDAR)

[sandsaver01]

I am starting this thread to both inform the lidar designers and ask for advice.

Thanks to whomever drew out the holes for Crooked Stick in OSM, creating the course file was reasonably straightforward. The course as brought down was before the 2015-16 modification of some of the holes, so based on the latest Google Earth overhead I have made the appropriate changes. Although I have no idea how to tell how accurate this lidar data is, the terrain seems reasonably accurate based on my remembrance of the course. Crooked Stick has six sets of tees and the designer seems to allow only five, so I had to eliminate one. I also had to flatten 80 tee-boxes since the lidar plot is not so good about that.
The plant meter was about 35% full when I brought the course down. Crooked Stick has a lot of trees, houses, and other stuff, so it will be interesting to see if I can getting it looking realistic before maxing out the meter.
A question: placing the tees was easy enough, but how do you decide where to put the pins? Do you just do like on a regular RCR and put them in, or is there a resource to find where they have been during a tournament?

[Deleted]

If you have not made many LiDAR courses before, I recommend you play around with the alignment before you make any significant changes to the land.

Did you import trees? If yes, they're not exactly the most accurate but at least they give an indication as to where they might belong.

If it's a course with a lot of tall grasses, you may have to resort to spam planting them to get a remotely dense coverage without maxing out the meter, but that depends on what else you plan on adding to the course's environment.

For pins you can do whatever you want or, if desired, you can try to find a tournament pin sheet from when a PGA event was played there (from roughly the same time that the LiDAR data is from in an ideal world). I'm not sure where people look to find these but I know it has been done before. For mine, I normally just look at the green and figure out where the hardest pinnable hole location is as well as 3 other interesting ones that use most of the green real estate and go from there (but none of the courses I pick are famous enough to host any big tournaments).

[sandsaver01]

Arielatom - I did look at the alignment and it looks pretty good, the terrain around bunkers, tees, and creeks lines up quite well. I did not import trees, since I think I can get a pretty good idea from the photos available online and from the Google earth view, and I was also worried about the meter. I do not think that tall grasses will be a problem, but if they are I will have to look at some videos of "spam planting" since all my previous designing has been on PS4 where that is not an option. As for the pins, I will try to find a pin sheet from one of the PGA tourneys that has been played there, or the USGA Sr. Open in 2009. If not I will do as you said, try to figure out the best locations.

Thanks for the input - I really appreciate it.

[sandsaver01]

I am making reasonable progress on Crooked Stick Lidar. The plant meter is at 29% so I may be able to do what I want to finish the decoration.

Here is a picture of Hole 1 IRL from the tee:



and here is a view from my Hole 1 tee:



It is not an exact match, but I think it looks OK.

[sandsaver01]

I posted the text below in the LiDar help forum and got no response, so I will try again here:

Looking for a little input from all you experienced LiDar designers out there. I am quite far along on a LiDar version of Crooked Stick Golf Club, maybe 90% complete with some grass planting and small stuff. I know an obvious anomaly when I see it; spikes, ridges in places where they shouldn't be, etc. but how do you tell whether other terrain features which maybe a little suspect are an anomaly? I hate to go smooth stuff that really should be there - kinda defeats the purpose of LiDar. Any suggestions appreciated.

On a related subject, this was one of Pete Dyes first course designs, and he may have gone overboard on the greens! I have not tried to set the four pin positions yet, but just looking at the greens as I play-test at 187 using the middle of the green I think it may be difficult if not impossible to find four placements. How much smoothing should I do, and what method would you recommend?

[Deleted]

so the best way to do this would be to use Google Earth 3D ground view if available (and if you can't find proof in pictures that a hump or mound-like feature should or should not exist). I would think 3d ground view is available for Crooked Stick considering its proximity to Indy.

For green smoothing I fall into more of the purist camp. I smooth them only if the contours are rough / bumpy / abrupt transitions which result in visible angles in the green grid. If I can't find 4 pins on most greens at 187 I reduce the green speed until I find something that suits the course better. So far I have published 4 LiDAR courses at anything form 132 to 172 and only one of them (the 172 one) had all pins 'legal' on 187. It's sometimes the nature of the beast that not all real courses are designed to accommodate 15 stimp speed, especially if the course in question has greens with a steady pitch to them. In this instance, too fast and the entire green is not 'TGCT Legal' for pins.

The other camp is to try to soften the green contours just a touch until they're playable on 187, ideally while changing the green contours as little as possible. Now how to smooth is something I'm not too well-versed in. I would say use the fuzzy flatten brush at either 0ft 0 or raise/lower to get a smoother contour. I use this when I have a green that is so clearly 'incorrect' that I have to redo most if not all of it. I have one of those on my current LiDAR WIP, assuming it had something to do with how tree-covered the particular green is considering it's QL1 data. Since I have no pictures of the green either and no 3d ground view, the best I can do is guess what it might be like.

The only smoothing I generally do is pretty time consuming and is done with the fuzzy brush in its smallest size. Mostly flatten but can be flatten or raise depending on what makes a certain spot smoother. In doing this I'm trying to preserve the green contours at all costs; the goal of this is to create a surface with smoother, more readable breaks that putts more 'true' and prevent 'triple breakers' that are mainly thanks to lower quality data. Knowing what I have seen from Indiana's main data as I plan on making Wolf Run in the near future, chances are that Crooked Stick is QL3 data as well, and QL3 generally has some artifacts / bumpiness to it which usually but doesn't always affect the greens.

[sandsaver01]

Sept 30, 2019 19:48:12 GMT -5 arielatom said:

so the best way to do this would be to use Google Earth 3D ground view if available (and if you can't find proof in pictures that a hump or mound-like feature should or should not exist). I would think 3d ground view is available for Crooked Stick considering its proximity to Indy.

For green smoothing I fall into more of the purist camp. I smooth them only if the contours are rough / bumpy / abrupt transitions which result in visible angles in the green grid. If I can't find 4 pins on most greens at 187 I reduce the green speed until I find something that suits the course better. So far I have published 4 LiDAR courses at anything form 132 to 172 and only one of them (the 172 one) had all pins 'legal' on 187. It's sometimes the nature of the beast that not all real courses are designed to accommodate 15 stimp speed, especially if the course in question has greens with a steady pitch to them. In this instance, too fast and the entire green is not 'TGCT Legal' for pins.

The other camp is to try to soften the green contours just a touch until they're playable on 187, ideally while changing the green contours as little as possible. Now how to smooth is something I'm not too well-versed in. I would say use the fuzzy flatten brush at either 0ft 0 or raise/lower to get a smoother contour. I use this when I have a green that is so clearly 'incorrect' that I have to redo most if not all of it. I have one of those on my current LiDAR WIP, assuming it had something to do with how tree-covered the particular green is considering it's QL1 data. Since I have no pictures of the green either and no 3d ground view, the best I can do is guess what it might be like.

The only smoothing I generally do is pretty time consuming and is done with the fuzzy brush in its smallest size. Mostly flatten but can be flatten or raise depending on what makes a certain spot smoother. In doing this I'm trying to preserve the green contours at all costs; the goal of this is to create a surface with smoother, more readable breaks that putts more 'true' and prevent 'triple breakers' that are mainly thanks to lower quality data. Knowing what I have seen from Indiana's main data as I plan on making Wolf Run in the near future, chances are that Crooked Stick is QL3 data as well, and QL3 generally has some artifacts / bumpiness to it which usually but doesn't always affect the greens.

Thanks, arielatom, this input is greatly appreciated. One question though; you mention "QL1" and "QL3" lidar data, how do you tell whats what?

[Deleted]

Oct 1, 2019 6:32:28 GMT -5 sandsaver01 said:

Sept 30, 2019 19:48:12 GMT -5 arielatom said:

Those data 'tiers' are somewhat arbitrary, although if you want to see information on the data set you're using, head back to the NOAA Digital Coast interagency page and it will list the given data parameters. Generally speaking, QL1 data is the 'most accurate with fewer artifacts and better data parameters,' or just 'good data' in general. Each quality 'tier' is a bit worse than the one before it with QL3 data not being 3DEP certified (if that even means anything). Sometimes a data set can have good specs but still be 'low quality' which was the case with the Massachusetts 2011 ARRA data set that I used for Black Rock (it also had all sorts of issues for Eastward Ho, so much so that no one has decided to make a LiDAR version of that course despite the heightmap being available); in this case I think it was labeled as QL3 because it had so many artifacts and inaccuracies.

In general I wouldn't worry about it as long as the data doesn't have lumps n' bumps all over the place that clearly don't belong there and also if there aren't large, clearly inaccurate contours i.e. green grids that are almost entirely red.

[sandsaver01]

Thanks again. I went to NOAA site and obtained the following data:

DATA SET NAME 2016 Indiana Central 3 County Lidar
DATA ACCESS The National Map
METADATA ACCESS The National Map
COLLECTION DATE Mar 11 - 18, 2016
QUALITY LEVEL 2
MEETS 3DEP Yes
REASON DATA DOES/DOES NOT MEET 3DEP QL2 or better quality lidar data with USGS Base Spec v1.2 products
PROJECT STATUS Complete
RESTRICTIONS Public
DATA TYPE Lidar-Topo
VERTICAL ACCURACY 4.7 cm
HORIZONTAL ACCURACY Not Provided
POINT SPACING 0.65 m
VERTICAL DATUM NAVD88
HORIZONTAL DATUM NAD83
PRODUCTS AVAILABLE TBD
NOTES Cooperative Agreement 2016 BAA

I assuming the lines in red show that the maps graphs out to about 2" elevation difference measured about every 2 feet. Am I interpreting this right?

[Deleted]

Ahh QL2. That explains why those 3 counties weren't part of the recent QL2 Indiana LiDAR project like Wolf Run was (that data still isn't available quite yet for Boone County).

You are interpreting it correctly, although it usually doesn't take artifacts into consideration. It's also claiming that the data is within that, somost of it should be far less than that. If anything those are possible QL1 specs if the data were extremely low on artifacts. If there are some spots you are not sure about you might as well check them with your sources. I'd be willing to take a look at the file sometime if you want or perhaps screenshots of things you're not sure about.

[sandsaver01]

Oct 2, 2019 0:35:33 GMT -5 arielatom said:

Ahh QL2. That explains why those 3 counties weren't part of the recent QL2 Indiana LiDAR project like Wolf Run was (that data still isn't available quite yet for Boone County).

You are interpreting it correctly, although it usually doesn't take artifacts into consideration. It's also claiming that the data is within that, somost of it should be far less than that. If anything those are possible QL1 specs if the data were extremely low on artifacts. If there are some spots you are not sure about you might as well check them with your sources. I'd be willing to take a look at the file sometime if you want or perhaps screenshots of things you're not sure about.

Thanks again. I am curious - what do you do when you "look at the file" - can you spot artifacts that way by looking at the raw data? What program do you use to view the data?

[Deleted]

I fool around with the lighting, usually needs quite a low angle to see that stuff. For 'looking' at the file I just meant the course in TGC. I don't have any special programs or anything like that. Nothing you can't do on your own as long as you know what to look for.