Phoenix FD RAM usage in 2017

[crimea]

Hey, so a few notes:
- Reflection Glossiness ~0.84 really seems to suit the horizon better than ~0.9 or 1.0 - I even changed the ocean preset to use it as a default for tomorrow's nightlies.
- Turned on Refraction in your liquid material as well, and this allows to enable fog color as well. However, note that if you choose to use fog color in a V-Ray mtl, you should not use an open geometry sheet such as the ocean. This is why I plugged a VRayPlane underneath the ocean and set it to black diffuse. Without this, underwater foam coloring would not work - you will get the liquid mesh to render the exact same way, but the underwater foam will always remain white and will not be affected by the fog color until you close the volume with a plane (or just use a closed volume, but for the ocean I can't think of another way to do it). This is also new to the Ocean toolbar preset from about a month back
- Turned off the Horizon Roughness completely - we've introduced this parameter for compatibility with old scenes, but I don't think it produces good results and furthermore its effect varies drastically with changing the render resolution.
- Boosted the Ocean Tex's Level of Detail to 25. This will bring back the detail in the near ocean areas. This is another important change since 3.00.01 - it's added for cases where you have a strongly reflective ocean surface with objects dipped into the ocean and a moving camera - in these cases a very strong detail causes unpleasant flickering, so now by default the Ocean Tex with LOD of 10 produces a little less detail than before, along with the other fixes that went into this problem.
- And for the sake of completeness - the Ocean Subdivs were capped to 2 in 3.00.01 so no matter how high you boosted them, the RAM usage would not skyrocket, so now after this limitation is removed, 5 subdivs really mean 25 times more polygons per pixel..

Isn't it possible to have it working the way it used to in 2014 with phoenix 3.00.01 and vray 3.40.03? It is working really well there - low RAM consumption, great noiseless yet detailed horizon, great amount of detail up close.

Not sure I agree with having a general lower glossiness on an ocean, there would have to be a great amount of detail on the surface for that to make sense, but even then I think it's best to let it stay at 1. The current best solution I've found is using a distance blend falloff in glossiness channel, to make only the horizon blurry.

When you say turning off Horizon Roughness completely, do you mean setting it to 1.0 or 0? I find setting it to 1 is best.

The capping of 2 ocean subdivs I don't understand, as raising this does improve quality dramatically in 3.00.01 for us. There's a very noticeable difference between say 8 and 12.

Thanks!

[Svetlin.Nikolov]

I'm afraid it's not going back to the way it was in 3.0 and the 2.25 nightlies because the ocean surface generation contained quite a few bugs back then. The way 3.0 rendered the ocean was tuned for specific camera angles and almost anything involving different camera angles and animations was broken in 3.0 and produced flickering, jittering, or completely missing ocean surface.

The ocean subdivs were indeed capped - you'd get some difference increasing them because of secondary algorithms processing the surface, but the effect of the subdivs was limited and needed to be allowed to reach higher poly counts in order to eliminate flickering that would be especially visible at low camera angles into shadows of objects placed into the ocean, while still being able to keep high ocean texture detail. This was another reason the approach to ocean detail was reworked and can't be directly compared to 3.0 using the same values.

The horizon roughness would change dramatically if you reduce the image resolution by half or if you increase it too - it gradually starts losing its effect if you have large image resolutions and is quite strong on low-res images, so setting it to 0 would make scaling the image more predictable. Of course if you have a fixed resolution, this would not be a problem.

I'll keep looking into rendering the horizon at glossiness of 1 and will ping you if something can be done to improve that.

Cheers!

[crimea]

I see. Thanks for the detailed explanation. Appreciate it.

I rendered a video showing the three different setups available for us currently if it's of interest:

https://we.tl/dsiTKkaOPu

Password: phoenix

[Svetlin.Nikolov]

Hey, can't log in - do you know what credentials should I use, or should I register a new account?

Thanks!

[crimea]

Don't think you need a login, only a password to download the file. Password is phoenix

Rendering times:
2014, 12 subdivs: 1:16
2017, 5 subdivs fix on: 4:32
2017, 5 subdivs fix off: 3:17

RAM consumption is similar.

[Svetlin.Nikolov]

Hey,

Unfortunately seems like we do need some credentials:

[crimea]

Ah whoops, I just copied the URL from an email. It actually hotlinks to a different URL than what our eyes see so you'd have to copy+paste the URL as written in the post, but I'll edit the post to fix it, so now you'll be able to simply click it (hopefully!).

[Svetlin.Nikolov]

Regarding render times - it's important to separate the time spent constructing the ocean before the VFB kicks in - this duration is where Phoenix works, and during the actual rendering the speed is up to V-Ray. You can also make out the pre-process time as the difference between the "Rendering Time" in 3ds Max's status bar next to the MaxScript listener and the time displayed in the VFB (if there is no GI).

[Svetlin.Nikolov]

Ah, got the video, thanks!

[Svetlin.Nikolov]

Btw, asking just in case - is it the same V-Ray between Max 2017 and 2014?

[crimea]

True, and a big percentage of the time is in fact "Transforming Vertices" - constructing the ocean. About 50% in both cases. Always try to make the renders as usable at possible, so don't have much else than the sea and sky, and use only BF+BF GI.

max 2014 | phoenix 3.00.01 | 12 ocean subdivs - full time 1:08, buckets' time 0:29,3
max 2017 | phoenix 3.04.00 | 5 ocean subdivs - full time 3:37, buckets' time 1:16,3

Rendered both twice to make sure things were correct, without using the computer much to get more accurate results. To get the translation time only, of course just do [full time] minus [buckets' time].

max 2017+3.04.00 is noisier yet has a greener sample rate pass, the other one is mostly blue and smooth.

[crimea]

Btw, asking just in case - is it the same V-Ray between Max 2017 and 2014?

No,
2014 has vray 3.40.03
2017 has vray 3.60.02

[Svetlin.Nikolov]

This is a very important part of the comparison - I also used to compare Max 2012 versus Max 2016 with different V-Ray versions back in April and it turned out that V-Ray changed behavior, 3ds Max changed as well the way it needed the mesh normals to be passed so we had to fix that, and also Phoenix changed which was the only thing I originally was sure about.

[Svetlin.Nikolov]

Will be back to testing locally here as well, just have to put out some fires..

[crimea]

I see. Would it be worth it for us to see how phoenix 3.00.01 and vray 3.40.03 (what we have installed for max 2014) - behaves in max 2017? Hate downgrading though..