You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The behaviour of the water is different when running on mobile hardware compared to what is seen on desktop h/w.
On desktop, the movement of the water is dampened over time and relatively quickly returns to a smooth surface.
On both ios and android the surface appears to be perturbed by additional velocities which stop the water from ever settling back to a smooth state.
One obvious difference is that the pixel ratio, when run on the desktop h/w, is 1.0 but on the mobile h/w it varies (on the examples I have tried it on) - 2.0 on an iPhone and 2.75 on an Android. Not sure if this is related - but it struck me that the "averaging of neighbour pixels" may be looking "further away" than expected and picking a velocity that is higher than an immediate neighbour - causing the wrong average to be calculated. But that's a wild stab in the dark.
I'm struggling to debug the shaders - to compare the actual calculated values on the different h/w - but I'll continue trying...
Any thoughts/fixes would be appreciated.
The text was updated successfully, but these errors were encountered:
The behaviour of the water is different when running on mobile hardware compared to what is seen on desktop h/w.
On desktop, the movement of the water is dampened over time and relatively quickly returns to a smooth surface.
On both ios and android the surface appears to be perturbed by additional velocities which stop the water from ever settling back to a smooth state.
One obvious difference is that the pixel ratio, when run on the desktop h/w, is 1.0 but on the mobile h/w it varies (on the examples I have tried it on) - 2.0 on an iPhone and 2.75 on an Android. Not sure if this is related - but it struck me that the "averaging of neighbour pixels" may be looking "further away" than expected and picking a velocity that is higher than an immediate neighbour - causing the wrong average to be calculated. But that's a wild stab in the dark.
I'm struggling to debug the shaders - to compare the actual calculated values on the different h/w - but I'll continue trying...
Any thoughts/fixes would be appreciated.
The text was updated successfully, but these errors were encountered: