...
3Delight – We have only one control for the general quality of the render. In the case of direct lighting, 3Delight "understands" that samples are best used for light sampling and that's what it does. As tests will show, those samples have a linear impact on perceived noise levels. Note that 3Delight doesn't care how many geo lights there are in the scenes : one one sample control is needed one sampling parameter acts on all lights adaptively. This is not the case with RenderMan/RIS and Arnold, where user intervention is necesseray necessary with each added geo light. The test scene contains only one geo light to simplify comparison
Results
The following graph gives a good idea on the convergence rate of the different light samplers.
| | Chart |
|---|
| width | 800 |
|---|
| domainAxisUpperBound | 300 |
|---|
| title | RMSE vs. RAYS |
|---|
| type | xyLine |
|---|
| yLabel | RMSE |
|---|
| domainAxisLowerBound | 0 |
|---|
| xLabel | Million Rays |
|---|
|
| Rays | 2.45 | 3.13 | 4.51 | 7.27 | 12.78 | 23.8 | 45.88 | 90.1 |
|---|
| 3Delight | 0.0933142 | 0.0658266 | 0.0441248 | 0.0290439 | 0.0185575 | 0.0117566 | 0.00752546 | 0.00449892 |
|---|
| Rays | 0.678 | 3.26 | 10.8 | 16.94 | 43.4 | 173.6 | 694.5 |
|---|
| Arnold | 0.15699 | 0.100115 | 0.0501787 | 0.0396399 | 0.0242515 | 0.0117413 | 0.00693426 |
|---|
| Rays | 1.47 | 2.94 | 5.88 | 11.7 | 21.1 | 47.02 | 94.14 | 376.3 | 751.3 |
|---|
| RenderMan/RIS | 0.151125 | 0.121487 | 0.0953649 | 0.0728148 | 0.0520606 | 0.0373876 | 0.0265388 | 0.0138148 | 0.00854045 |
|---|
|
|
The following graph shows the time required to achieve a certain quality. From user's perspective, this is an important quantity.
| | Chart |
|---|
| width | 800 |
|---|
| title | RMSE vs. TIME |
|---|
| type | xyLine |
|---|
| yLabel | RMSE |
|---|
| xLabel | Time (seconds) |
|---|
| | Time | 5.72s | 7.63 | 12.15 | 20.68 | 37.75 | 72.07 | 142.05 | 280.68 |
|---|
| 3Delight | 0.0933142 | 0.0658266 | 0.0441248 | 0.0290439 | 0.0185575 | 0.0117566 | 0.00752546 | 0.00449892 |
|---|
| Time | 1 | 2 | 6 | 9 | 21 | 81 | 492 |
|---|
| Arnold | 0.15699 | 0.100115 | 0.0501787 | 0.0396399 | 0.0242515 | 0.0117413 | 0.00693426 |
|---|
| Time | 6.74 | 7.23 | 7.99 | 9.42 | 18.51 | 29.40 | 98.08 | 383.39 |
|---|
| RenderMan/RIS | 0.151125 | 0.121487 | 0.0953649 | 0.0728148 | 0.0373876
| 0.0265388 | 0.0138148 | 0.00854045 |
|---|
|
|
The following graph shows how much time it takes to build the acceleration data structure depending on sample/ray count.
| | Chart |
|---|
| width | 800 |
|---|
| domainAxisUpperBound | 300 |
|---|
| title | Time to First Pixel vs. Rays |
|---|
| type | xyLine |
|---|
| yLabel | Time to First Pixel (seconds) |
|---|
| domainAxisLowerBound | 0 |
|---|
| rangeAxisUpperBound | 50 |
|---|
| xLabel | Million Rays |
|---|
| | Rays | 2.45 | 3.13 | 4.51 | 7.27 | 12.78 | 23.8 | 45.88 | 90.1 |
|---|
| 3Delight | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 |
|---|
| Rays | 0.678 | 3.26 | 10.8 | 43.4 | 173.6 | 694.5 |
|---|
| Arnold | 0 | 0.35 | 1.2 | 3.2 | 11 | 41 |
|---|
| Rays | 1.47 | 2.94 | 5.88 | 11.7 | 47.02 | 94.14 | 376.3 | 751.3 |
|---|
| RenderMan | 3 | 3.2 | 4 | 4.7 | 9 | 15.5 | 51.2 | 97 |
|---|
|
|
...