Speed Hazer, Taron...


Could you just give us a hint of what the different inputs and outputs do :curious:, I haven´t succseeded in rendering anything but a constant black colour out from it, even though it tried to feed it with noises and stuff. :blush:



I just had a quick play with ‘Speed Hazer’ today for the first time, it looks like you should be able to grass and stuff with it. I’ve managed to get a kind of snow after playing for 5 minutes attached is my file for you to play with.

Hope it helps



Thanks!!! :thumbsup:


Just realised that I’m using Thomas’s AoN shaders in the file, you’ll have to change ‘AoN_Crumple’ to something else if you haven’t got the shaders.



No probs I got AoN as well as TLHpro :stuck_out_tongue:


Just thought I would post some more tests with SpeedHazer . The first is just plain sphere, next I added the speedhazer node connecting length…


…then finally the Noisette node to the opacity.



Looks nice :slight_smile:


Thought I’d post this Speed Hazer information from Taron here as well;-

Opacity - it’s what it always is, of course…(overall “alpha” of the shader)

Value - it’s now called Density, I’ve renamed it for the next version. This is where you would hook in the textures, like procedurals for instance (noise, noisette, AoN things, but also any form of density source like meta-effectors or textures…anything)

Shading - this is where you hook in actual shading nodes like the basicShader for instance. When ever the density threshold is reached to determin the shell or surface of your volume, it also generates normals, which will then get passed to the attached node (basicShader) to get shaded by lighting, reflections, anything you’d want it to look like. This is what will make the solid volume appear like geometry. Translucency won’t work with it really that well, though, because it still is just virtual geometry, so to say…but extending of the lighting to get part of the translucent effect does work fine…but, uhm, let me know if you find it necessary and I’ll see what I can do.

Min and Max - those inputs can modulate (or set) the min and max distance settings for further definition or animation of these values!

Threshold - modulate (or set) the threshold that you want for the density readout. Kind of curious to make that texturable, too…it allows for a different access to the control of the solid volume shape and animation.

Density - now called Global Density…adjusting the overall density through animation or texturing…I am not completely sure why I was even putting that in, but I really figured: why not!

Top and Fade Top just like Bottom and Fade Bottom allow the texturing of these parameters, which could localize different heights and feathering of the volume…tricky one to think about. Using Y mapped texturemaps or meta effectors would be mostly recommendable for that. It’s the “control-freak” parameter set…hihi…one for me for instance.

Random - Ah…funny…I guess I made it so you could texture the random offset of the raymarch sampling…why? Hmm…keep it organic under your percisely defined circumstances…more control-freak things! :smiley:

Well, so much for the overview!

The next release is coming closer, so you can savely get used to the new input naming (value-density, density-global density)

edit. added some examples from Taron.

Here are some older clips I had done during the creation of the SpeedHazer…
<http://www.projectmessiah.com/~taron/Clips/SpeedHazer_sm01.avi> (DivX) (this one shows solid volume rendering with noisette and basicShader attached for shading the volume!)

<http://www.projectmessiah.com/~taron/Clips/SpeedHazer_rm01.mov> (Quicktime) (this is just fog with raymarching and a goofy little robot-leg I did real quick to show a friend…hehe…wicked expressions for the mechanisms, too, but not too careful.)


Moved so info is not lost.

" By the way, after people were talking about their confusion about the “SpeedHazer” and it’s ability to deal with solid volume, I’ve decided to outsource a portion of it and call it “Voxelini”, because it does such magical things…hihihi…SpeedHazer still retains the potential and all the same features as it has them now, but the specialized Voxelini will come with a few more specific features."

As someone who has been wrestling with ‘Speed Hazer’ and struggling to get solid volumes from it, I can at least attribute some of my problems to using the procedural shapes, which made testing quick but inconsistent and floored (my bad!). I’ve being trying unsuccessfully to produce a shape out of haze. But this might be possible in ‘Voxelini’ is this the case?



Bit more experentation with Speed Hazer, the first is a cube with AoN noise shader attached to main material, the second is the same set-up with Speed Hazer added.




The “repaired” SpeedHazer brings back the simplicity and immediate rendering “success” that it had right in the beginning (which is before anybody knew…doh), but it still is geared towards open-space volumes. Voxelini is optimized for enclosed volumes, defined by any geometry! It intergrates the geometry’s normals in that it has the ability to blend between them and the procedural surface inside. I’ve planned to eventually integrate another curious concept, which would allow the use of other geometries on the inside to further define the internal volume and also help animating it, but that will have to wait a few weeks…

Both shaders require only very little understanding of raymarching, where you will have to determin the macro and micro step length for the procedure to get faster, yet, accurate renders. I could have just made some heuristical quality level, but figured I could simply give you exactly what’s supposed to happen as a parameter. So you can tell it to check the volume every 5mm for anything and once the ray has found something it would go back one step and progress only every 1mm (for example) to find exactly where that surface has started (roughly speaking). The amount of precision is therefore perfectly up to you. I really like it that way, but in the future I have planned to offer several different methods to either optimize the procedure or simplify it for the user. There are some different ways of detecting procedural surfaces in the raymarching world and I’m trying to figure out the best possible ways. I’m very happy with my current solution, but I do sense that there are more ideal ways of doing it. I’m trying to figure out what’s the best quality at the fastest speed, but still to allow the highest level of accuracy, if demanded. I’m not too happy with the idea of fatal trade-offs…hehehe…

It’s all on its way! :slight_smile:


Thanks to Taron’s. :bounce: I’ve been able to produces solid volumes with Speed Hazer, The example below uses the cube for consistency sake.



This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.