Jump to content


Photo

3D Grass Observations


  • Please log in to reply
115 replies to this topic

#1 DPRoberts

DPRoberts

    Advanced Member

  • Members
  • PipPipPip
  • 1,610 posts

Posted 09 June 2019 - 04:14 PM

So, the 3D grass train is finally moving out of the station. Albeit, slower than most would probably prefer. The speed has been a combination of real life interference and unfortunately, my nature to not want to get all the 3D grass lemmings going in one direction, only to tell you next week that you are heading in the wrong direction.

 

I wanted to use this post here to get a consolidated location for some opinions on the grass without cluttering some other threads. Especially, if people happen upon this in the future and aren't privy to the other conversations.

 

Here is my main conclusion that I am coming to (subject to change, of course)...

1. A short visibility distance with a sharp falloff may be best. This appears to be what they are doing in the EA Rory game from the few videos that I've watched online.

 

rMdMQ1t.jpg

 

    Advantages -  Densely planted grass with a reasonable FPS, better blend with the still visible texturing in near proximity (a course texture or course detail texture), elimination of the little bits of grass in the distance, handles more lighting scenarios without obvious color mismatches

   Disadvantages - limited view distance of the 3D grass (may not be seen on other side of green) or in the rough beyond a tee box

 

All images below are the same settings for the grass. It is all in the same course build.

JWT532Y.jpg

 

thYkUs2.jpg

 

DpdejPC.jpg

 

4fUJMqw.jpg

 

I really think this concept is getting close to what we will be happy with. I will be releasing this style today on Orchard Valley for v1_1 so that I can maybe bring some closure to this. 

 

Ultimately, it will be every designer's choice on what to do with the grass. I am only hedging because I think some basics for understanding will help us in the long run. I would hate to see a flurry of releases only to discover over half the courses aren't playable by most.

 

I am anxiously waiting to see what can be done with Kiawah or Chamber's Bay or any gruffy links course. I am only worried there on the viewing proximity but hoping with longer blades you can actually plant less densely and push that further out than a course like Orchard or Scioto. 

 

Comments are welcome. Pictures can help me see what you like and don't like. Look forward to improving this process with the community!


  • Charles, zmax - sim, Birdie and 1 other like this

#2 Stingreye

Stingreye

    Advanced Member

  • Members
  • PipPipPip
  • 341 posts
  • LocationSouthern California

Posted 09 June 2019 - 04:46 PM

Thanks for all the work! I think you are on to something getting rid of thinning.

#3 clubcaptain

clubcaptain

    Advanced Member

  • Members
  • PipPipPip
  • 1,310 posts

Posted 09 June 2019 - 04:49 PM

The technicalities are way beyond my understanding (so looking forward to real easy implementation) but I trust tou implicitly to do a greta job for the community as you have always done. The pics look great but the proof of the pudding will be in the playing so waiting eagerly for the next iteration.


PC specs...

Intel core I7 9700k 3.6 GHZ-Turbo 4.9 GHZ
64 GB Corsair vengeance  LPX DDR4 2400 MHZ
Asus PRIME Z390-P 
Nvidea GeForce RTX 2060 6GB
X box 360 wired controller
Windows 10 PRO NA 64 bit


#4 Buck

Buck

    Advanced Member

  • Members
  • PipPipPip
  • 7,530 posts

Posted 09 June 2019 - 04:54 PM

Man the lighting engine in Rory is just so drool-worthy..

On their grass - it really does seem to reside with making the fall off not noticeable (blends in).

What was the deal with the “black tips” on Orchard?
Was that there on Scioto? Seemed to look a lot worse in the distance to me than on Scioto?

#5 DPRoberts

DPRoberts

    Advanced Member

  • Members
  • PipPipPip
  • 1,610 posts

Posted 09 June 2019 - 05:17 PM

I was getting 80-110 FPS on Scioto pre grass. I was getting 50-75 on Orchard pre grass.

So, I was able to push viewable grass much further. I'll have to go back and look because I know the bits are there, they may just match my underlying texture better in the distance so essentially you can't see as many of them. I also removed the fresnel effect on Scioto so the distance texture didn't lighten when you are lower to the ground. I should probably do that for Orchard.

I'm guessing we are also getting more used to the issues and they are more obvious to us now.
  • Buck likes this

#6 Buck

Buck

    Advanced Member

  • Members
  • PipPipPip
  • 7,530 posts

Posted 09 June 2019 - 05:22 PM

Thx again for your efforts DP

I only mention the black tips because the whole 3D grass thing for me, and it being enjoyable and a net “win”, is really highly dependent upon how well it visually blends in.

If one is noticing it all the time - at least for me - and always mentally seeing and taking note of the transitions and where it “is there” and “not there” and “that looks odd - what is that?” And all of that creeps in... it ends up sort of squashing the experience a bit.

Is this the sort of thing that is binary moving forward?
Will designers have to decide from the get go about going “3DG” or not?
Or is it possible we could have non-3DG versions, sort of like how Reflective water played out?

#7 DPRoberts

DPRoberts

    Advanced Member

  • Members
  • PipPipPip
  • 1,610 posts

Posted 09 June 2019 - 05:40 PM

I think you build and complete a course with thoughts for doing a grass version. Grass mostly should come at the very end. Even after beta testing in my mind. So, why not finalize the non 3d grass version and decide from there.

A lot will depend on if you plan for the grass at the start. It will make your tweaking of all the splines much easier to do so.

Knowing what I know now and how I use Blender to retopologize my meshes, I didn't have to change any splines to do Scioto 3d grass.

I know some people hate 2 versions but if you name them correctly, they can both exist side by side in the carousel.
  • axe360 and Buck like this

#8 jspirate

jspirate

    Advanced Member

  • Members
  • PipPipPip
  • 487 posts

Posted 09 June 2019 - 07:35 PM

I think you build and complete a course with thoughts for doing a grass version. Grass mostly should come at the very end. Even after beta testing in my mind. So, why not finalize the non 3d grass version and decide from there.

A lot will depend on if you plan for the grass at the start. It will make your tweaking of all the splines much easier to do so.

Knowing what I know now and how I use Blender to retopologize my meshes, I didn't have to change any splines to do Scioto 3d grass.

I know some people hate 2 versions but if you name them correctly, they can both exist side by side in the carousel.

Yeah, I am torn on this.  As a player, I don't like having multiple versions of a course, but as a designer it makes sense to finalize and then do a 3DG version.

 

I still have too much left on current WIP, but I want to play grass!!!!


i7-8700K
GIGABYTE Z390 AORUS Pro
64 GB DDR4 3200
MSI GeForce GTX 1080 Ti Gaming X
LG 32GK850G
Win 10 Home Premium on 1TB SSD

 

Courses:

Bradford Links - released

Ironwood - released

Cypress Landing - released (also available in a 3DG version)

Crow's Nest - released


#9 M Rose

M Rose

    Advanced Member

  • Members
  • PipPipPip
  • 1,918 posts
  • LocationMelbourne, Australia

Posted 09 June 2019 - 09:05 PM

Should we try and add it to Pebble?

I think my update is about ready to go; just mulling over whether I should beta test it again.... apart from all the cosmetic changes, it's not really any different save for the 13th green, a new tee on #9 and some reshaping of a few bunkers. I only realized this morning that the Open is next week and that I have a bit less time than I thought, so I'm not really sure what my next move is.


  • Charles likes this

VG2 • PG • 2k21 • 2019 • Links • GBC • JN5


#10 axe360

axe360

    Advanced Member

  • Members
  • PipPipPip
  • 6,219 posts
  • LocationSo Cal U.S.A.

Posted 09 June 2019 - 09:32 PM

M Rose, IMO it can't hurt to throw up a beta for a quick run thru by a few of us. :)


Done with designing.

Released Courses: Real

The Golf Club @ Dove Mnt. AZ

Aronimink PA

Amana Colonies Iowa

Fictional:

The Grinder Anytown U.S.A.

 

 

                   


#11 zmax - sim

zmax - sim

    Advanced Member

  • Members
  • PipPipPip
  • 3,183 posts

Posted 09 June 2019 - 09:35 PM

Just by looking at the pics above, I think 1.1 looks better than 1.0

#12 Birdie

Birdie

    Advanced Member

  • Members
  • PipPipPip
  • 456 posts

Posted 09 June 2019 - 09:47 PM

DP, maybe i should ask first if i should post here... as i don't know if a lot of other people can take benefit from findings or testings which i do on my particular course or if my settings are easily transferable to other courses as i'm for example (nearly) exclusively building and optimizing my course for dynamic sky off (so far).

And to be true atm i'm also not to much focussing to make things work on every PC. It's simply my first attempt in CF and i'll still have to learn a lot on optimisation and so for now i mainly look that things work on my PC or PCs with similar specs like mine.

But if this will be the definite place for every aspect or finding on 3d grass, i'm happy to contribute here. 



#13 RoobsNY

RoobsNY

    Member

  • Members
  • PipPip
  • 18 posts

Posted 09 June 2019 - 09:59 PM

I'm playing your course now, on a 65" UHD (3840 x 2160) monitor.  It looks absolutely magnificent!  Unbelievable.  Damn this game looks good.  Getting 34 FPS on a nice Alienware rig with a GTX 1070.

 

And its not just the 3d grass.  Your whole design is magnificent.  Thanks for this amazing course.



#14 Sliceapottomus

Sliceapottomus

    Advanced Member

  • Members
  • PipPipPip
  • 1,152 posts
  • LocationRhode Island

Posted 09 June 2019 - 10:00 PM

Like the look just don't like the performance hit. left you extensive testing stats over at ogt under scioto all data is for orchard valley though..

Strange that Dynamic Skies Help me with the 3d grass.... if I play on lower settings with the sliders down I can get by with the rtm swing if i have to the video just plays back a tad choppy... I did run same test on arts version w/o the 3d grass and was locked @ 60 fps the whole time no matter where I was on the course with skies on and off...

 

https://www.onlinego...hp?f=269&t=4200


Ncccrca.jpg
Operating System - Windows 10 Pro 64-bit ,  CPU - Intel Core i5 7600K @ 3.80GHz ,  RAM - 16.0GB, MotherboardASRock B250M Pro4 (CPUSocket) , Graphics card4095MB NVIDIA GeForce GTX 1050 Ti ,  Storage256GB (SSD) , 931GB Seagate , 232GB Maxtor USB Device

#15 zmax - sim

zmax - sim

    Advanced Member

  • Members
  • PipPipPip
  • 3,183 posts

Posted 09 June 2019 - 10:41 PM

Slice,

Maybe it’s time to upgrade from a 1050 ti? I have a 1060 OC and avg around 50 FPS (some hit from reshade) which is fine for a simulator. No avatar to deal with. But I feel a 1060 is the minimum.

#16 Charles

Charles

    RTSM forever!

  • Members
  • PipPipPip
  • 869 posts
  • LocationHolland, Europe.

Posted 10 June 2019 - 04:49 PM

Orchid Valley 3DG plays around 35 FPS with dynamic skies on here. Compared to Scioto it's definitely less FPS (around 40% less) but I guess that has to do with 'course design stuff' which can vary per course? 35 FPS is still very playable for me so no problems there. For me around 20% is the bottom line before getting jitters and lags in RTSM. So anything above is good. I am looking forward to how Kiawah is going to turn out. Once courses are going to be made with 3DG from scratch (as compared with the present courses which are 'rebuild' to 3DG) and designers are getting better and better, the FPS should be no problems right? That's what I hope anyway...


  • DPRoberts likes this

RTSM / Pro / In the bag: D,3/7W,1H,5i-60W,P.

* Favorite Designers: Skunky - DoGgs - RobC - IanK - PGA Pro - Gary Norman - DrTwist - Bortimus -

DPRoberts - High Fade - Classic Renditions - Matt Rose - Mikahenrik and ... Mike Jones! *

"Golf is deceptively simple and endlessly complicated; it satisfies the soul and frustrates the intellect.

It is at the same time rewarding and maddening - and without a doubt the greatest game mankind has ever invented".

= Arnold Palmer =

 

 

 


#17 Sliceapottomus

Sliceapottomus

    Advanced Member

  • Members
  • PipPipPip
  • 1,152 posts
  • LocationRhode Island

Posted 10 June 2019 - 04:49 PM

Slice, 

Maybe it’s time to upgrade from a 1050 ti? I have a 1060 OC and avg around 50 FPS (some hit from reshade) which is fine for a simulator. No avatar to deal with. But I feel a 1060 is the minimum.

 

 

yeah and maybe it's time for a raise :D


Ncccrca.jpg
Operating System - Windows 10 Pro 64-bit ,  CPU - Intel Core i5 7600K @ 3.80GHz ,  RAM - 16.0GB, MotherboardASRock B250M Pro4 (CPUSocket) , Graphics card4095MB NVIDIA GeForce GTX 1050 Ti ,  Storage256GB (SSD) , 931GB Seagate , 232GB Maxtor USB Device

#18 DPRoberts

DPRoberts

    Advanced Member

  • Members
  • PipPipPip
  • 1,610 posts

Posted 10 June 2019 - 06:41 PM

@Charles,

I think, if a designer has an understanding of ways to optimize and reduce Tris (large outer spline with less Tris, reduced resolution of terrain as main terrain, controlling Tris and materials on peripheral objects), then he/she will be able to add grass and in a much better manner than we do now.

Everything can always improve up until a point. There are Unity limitations which will always win out.

But, as seen with DrRobs bunkers and Slewin's bunkers, some people can figure out how to get more out of the programs.

We will all be drooling when someone puts out a version with long grass. I'm hoping to save room to have it be blowing in the wind as well. The tee box views will be remarkable!
  • Charles likes this

#19 zmax - sim

zmax - sim

    Advanced Member

  • Members
  • PipPipPip
  • 3,183 posts

Posted 10 June 2019 - 08:00 PM

Slice, 

Maybe it’s time to upgrade from a 1050 ti? I have a 1060 OC and avg around 50 FPS (some hit from reshade) which is fine for a simulator. No avatar to deal with. But I feel a 1060 is the minimum.

 

 

yeah and maybe it's time for a raise :D

 

FPS and raise aside, which one looks better to you?   1.0 or 1.1?



#20 Sliceapottomus

Sliceapottomus

    Advanced Member

  • Members
  • PipPipPip
  • 1,152 posts
  • LocationRhode Island

Posted 10 June 2019 - 08:07 PM

not a big enough change but scioto's grass looks awsome compared to orch3d... maybe it is the light green grass I don't know but all i can do is compare to scioto...


Ncccrca.jpg
Operating System - Windows 10 Pro 64-bit ,  CPU - Intel Core i5 7600K @ 3.80GHz ,  RAM - 16.0GB, MotherboardASRock B250M Pro4 (CPUSocket) , Graphics card4095MB NVIDIA GeForce GTX 1050 Ti ,  Storage256GB (SSD) , 931GB Seagate , 232GB Maxtor USB Device




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users