When amplitude is large water can go above faucet

[KatieWoe]

Test device:
Dell
Operating System:
Win 10
Browser:
chrome
Problem description:
For phetsims/qa#264
Likely small and may not be easily solvable. When water is in side view the water splash can be high enough to go over the faucet. This can appear a bit odd.
Steps to reproduce:

1. Go to Wave Screen, Water side view scene
2. Adjust the Amplitude to be max
3. Start waves

Screenshots:

Troubleshooting information (do not edit):

Name: ‪Wave Interference‬
URL: https://phet-dev.colorado.edu/html/wave-interference/1.0.0-dev.66/phet/wave-interference_en_phet.html
Version: 1.0.0-dev.66 2019-01-22 02:30:43 UTC
Flags: pixelRatioScaling
User Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:64.0) Gecko/20100101 Firefox/64.0
Language: en-US
Window: 1536x760
Pixel Ratio: 2.5/1
WebGL: WebGL 1.0
GLSL: WebGL GLSL ES 1.0
Vendor: Mozilla (Mozilla)
Vertex: attribs: 16 varying: 30 uniform: 4096
Texture: size: 16384 imageUnits: 16 (vertex: 16, combined: 32)
Max viewport: 32767x32767
OES_texture_float: true
Dependencies JSON: {}

[arouinfar]

Nice catch @KatieWoe. During the previous dev test (dev.51) the maximum amplitude was just behind the faucet, and was only occluded slightly for one frame. Without pausing and stepping through, it's not at all obvious that there is any overlap between the water and faucet.

In dev.56 through dev.62, the overlapped worsened, and began lasting for several frames.

Starting in dev.63 onward, the overlap looks like this:

@samreid this behavior is not ideal, though not particularly harmful for 1.0. Can you look into it, and see if you could return the scaling so it's more on par with dev.51?

[samreid]

Good catch @KatieWoe . This problem was introduced in #315 (comment). When the lattice size increased, I increased the source amplitude to match the prior decay behavior. It is currently set to 120%, noted here:

wave-interference/js/common/WaveInterferenceConstants.js

Lines 102 to 104 in b3a5cb2

	// The wave dies out more quickly on a larger lattice. At the initial calibration of lattice size = 101 = 61+20+20,
	// this was 1.0
	AMPLITUDE_CALIBRATION_SCALE: 1.2

@arouinfar can you please advise? Some options:

• Leave it as it is
• Move the faucet further away (may look odd when there are two sources). May take 1-2 hours depending on the assumptions in WaterDrop model, or less if we just increase the water drop speed accordingly.
• reduce the amplitude scaling back to 100% and compensate by recalibrating other factors such as the wave brightness factors and light intensity screen + graph scale factors.
• your idea here!

[arouinfar]

@samreid I'll go ahead and defer this for 1.0, but I think we should explore this option for 2.0:

reduce the amplitude scaling back to 100% and compensate by recalibrating other factors such as the wave brightness factors and light intensity screen + graph scale factors

[samreid]

Self-assigned for initial investigation.

[samreid]

If we like the rest of the calibration, perhaps this could be adjusted by simply changing the scale factor for the "side view" water height. @arouinfar do you think that could work? It could save a lot of time over re-tuning brightness/intensity screen/graph/etc.

[arouinfar]

If we like the rest of the calibration, perhaps this could be adjusted by simply changing the scale factor for the "side view" water height.

@samreid that sounds like a great idea! Let's give it a try.

[samreid]

getWaterSideY is currently:

return Util.linear( 0, 5, waveAreaBounds.centerY, waveAreaBounds.centerY - 80, waveValue );

If I change that to

return Util.linear( 0, 5, waveAreaBounds.centerY, waveAreaBounds.centerY - 47, waveValue );

Then the water stays right under the faucet, even at max amplitude. However, this squashes the rest of the wave, which may be undesirable. Here is a superimposed image showing original and reduced:

@arouinfar is acceptable, or do we need to keep up the wave scale out to the right? Another alternative would be just to rescale the wave values in the vicinity of the source--not globally.

[arouinfar]

@arouinfar is acceptable, or do we need to keep up the wave scale out to the right? Another alternative would be just to rescale the wave values in the vicinity of the source--not globally.

@samreid I think it will be acceptable. After all, earlier dev versions didn't extend above the faucet in side-view (for example, dev.55). I'm not sure we want to disproportionally scale the amplitudes, so I'd prefer to stick to the global approach.

Can you push this change to master so I can test it out?

[samreid]

Pushed to master, please review on phettest.

[arouinfar]

Looks good in master @samreid!

[samreid]

Thanks, closing.