Couple questions before buying.

[artfabrique]

Hi. I'm a developer from Crazypanda Games. We want to fire up a couple of new mobile products and we plan to use your assets.

1. Do you have performance tests on those devices? It is the most popular devices in certain regions in 2019.
-  Xiaomi Redmi Note 4
-  Xiaomi Redmi 4X
-  Samsung Galaxy S7
-  Samsung Galaxy A5
-  Samsung Galaxy J5
-  iPhone 6s
-  iPad Mini 2
If not - is there a way to conduct tests before buying and rushing into the integration?

2. Do you have a bundle of all your assets (ropes, cloth) or they intersect with each other in some manner?

[LordCommander]

I am also interested in this asset, but I am concerned about performance on mobile devices. 


They need to provide some samples on mobile devices. Videos would be great, but also some reports.

[josemendez]

Hi there,

We cannot possibly test Obi on all exiting mobile devices, so we've only tested in a handful:

- iPhone 4S
- iPhone 6
- iPhone 7
- iPhone XS
- iPad Air
- iPad 2019
- Samsung Galaxy S3
- Samsung Galaxy S4
- Samsung Galaxy Note 3

Also, performance metrics can be pretty misleading since performance depends on a lot of parameters:

- Amount of particles per soft body
- Amount of shape matching clusters
- Mesh density
- Amount of contacts/frame
- Amount of dynamic colliders in the scene
- Timestep
- Iteration count for all constraint types

We have measured performance on desktop for a very specific set of parameters:
http://obi.virtualmethodstudio.com/performance.html

Most of the time, the only way to tell if it will work for your particular purpose is to try it. We have a very loose refunds policy, so if you find it just doesn't work for you ask for a refund and we'll have no problem granting it.

[artfabrique]

Hi there,

We cannot possibly test Obi on all exiting mobile devices, so we've only tested in a handful:

- iPhone 4S
- iPhone 6
- iPhone 7
- iPhone XS
- iPad Air
- iPad 2019
- Samsung Galaxy S3
- Samsung Galaxy S4
- Samsung Galaxy Note 3

Also, performance metrics can be pretty misleading since performance depends on a lot of parameters:

- Amount of particles per soft body
- Amount of shape matching clusters
- Mesh density
- Amount of contacts/frame
- Amount of dynamic colliders in the scene
- Timestep
- Iteration count for all constraint types

We have measured performance on desktop for a very specific set of parameters:
http://obi.virtualmethodstudio.com/performance.html

Most of the time, the only way to tell if it will work for your particular purpose is to try it. We have a very loose refunds policy, so if you find it just doesn't work for you ask for a refund and we'll have no problem granting it.

Sounds great! I'll try then! I think it's a good idea to roll the same test on the lowest of the most popular android phone. Then it can be extrapolated through Antutu benchmark data with 5% precision. We did similar stuff for our google-like map scaling feature (realtime kd-tree matchmaking search) and it worked out pretty good for us.