Jump to: Board index » General » Fusion

Help to Understand World Position Pass

Learn about 3D compositing, animation, broadcast design and VFX workflows.
  • Author
  • Message
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Help to Understand World Position Pass

PostFri Nov 14, 2014 5:12 am

I'm new to Fusion or any serious composite software so please bare with me. The interface feel very intuitive and I see myself learning very fast but I need to get familiar with some fundamentals of using software of this type. I am very interested in using the World Position Pass functions. My aim is to use fusion for post pro work in 3D rendering including animation. My main 3d tools are Sketchup Pro 2015, Thea Render, Blender and Xfrog.

First off I am still developing an understanding of how volume masking works with WPP. I'm not 100% sure that my render software supports the required camera depth information to be used in Fusion. I am using Thea Render which does have a depth channel. Taking a guess from reading some posts around the web, I saved out an EXR render of the depth channel of a simple scene. When brought into Fusion, it seems correct but I get an error in the console saying "no z channel found" after attempting to use the Booleans or z to world position nodes. And I am stuck on the 3D camera and how to import the camera from the scene. The original model was done in Sketchup where I am able to export a FBX file. I tried importing that and uncheck everything but the camera but the only thing that comes into the flow window is a merge node and not a camera3D with it.

Can someone explain exactly what is required from a 3D or rendering package before importing in Fusion? Or why would a "no z depth channel found" would show up in the console?
Offline
User avatar

Thomas Milde

  • Posts: 318
  • Joined: Tue Aug 28, 2012 11:27 am
  • Location: Augsburg, Germany

Re: Help to Understand World Position Pass

PostFri Nov 14, 2014 6:51 am

Hi.

I can't help you with all the problems, but the missing z-channel might be a wrong designation by your 3D-software. You said you saved th z-buffer as an EXR. So you should have some kind of grayscale image representing your buffer. Try to use the "channel-booleans" node, set it to copy and then copy the luma channel to the z-buffer. Maybe this is a first workaround to get Fusion to use that buffer.
Tech:
Win 10 pro, Ryzen 9 5900X, 128 GB, M.2 for media
RTX 3080 ti, Intensity Pro, DVR Studio + Fusion 18.x, Desktop Video 12.x
Offline
User avatar

Stefan Ihringer

  • Posts: 257
  • Joined: Mon Nov 10, 2014 10:40 pm

Re: Help to Understand World Position Pass

PostFri Nov 14, 2014 8:32 am

With Z channels, it's always a thing of what data they actually contain. Fusion wants negative distance units from the camera position (from 0 up to -100 and beyond). Some renderers might just render a gray-scale image with values between 0 and 1 (or 0 and 255 if you're thinking in 8 bit terms). This is no good for Fusion but it was the way it used to be done and some plugins (like Frischluft's Lenscare for z blur) still expect z channels like that.

But your issue seems to start earlier since you're saying that the Z channel doesn't even show up in Fusion. There are two possible reasons:

  • Your EXR doesn't have a Z channel. You might have to set up more than just exr output in your renderer and unfortunately I can't help you with that since I don't know Thea or blender. But there might be options in the render settings where you can define "passes" or "AOV" buffers. You might need to enable Z depth there first.
  • Your EXR has a Z channel but it's not called "Z" so Fusion doesn't detect it. In your Loader there's a tab where you can see all kinds of channels in your EXR. R, G, B and A should already be preselected. Can you check the contents of the dropdown next to "Z"? Are there more channels present in addition to rgba?

Then we can go on from there...
blog and Fusion stuff: http://comp-fu.com/2012/06/fusion-script-macro-collection/
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Re: Help to Understand World Position Pass

PostFri Nov 14, 2014 2:08 pm

I see nothing but R,G,B in the drop down lists of the loader. I was looking in the format tab. Was that the right place to look?
In Thea the channel only says depth not "z-depth" so maybe some information is absent and I only get a greyscale image. I brought this issue up in their forums. They are pretty good at listening to the users.
The other issue I'm having is the 3Dcamera. I exported a FBX file from Sketchup and attempted to use only the camera information in Fusion. But after pressing OK, only a merge node shows up and no camera. I think I'm missing something.

Thomas- I will try that method today.
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Re: Help to Understand World Position Pass

PostFri Nov 14, 2014 3:11 pm

Is Lightwave, Max and Maya the only 3D scene types supported in Fusion? What if I wanted a camera from Sketchup or Blender?
Offline
User avatar

Marek Subocz

  • Posts: 66
  • Joined: Mon Nov 10, 2014 9:21 pm

Re: Help to Understand World Position Pass

PostFri Nov 14, 2014 4:23 pm

Fusion also supports FBX scene.
One thing to rememember is to bake camera animation, especially if it's parented or in some rig.
Then just export to fbx, make sure You check the stuff You want (camera, meshes, lights etc) and Fusion should read it without the fuss

hth

EDIT:
one more thing - use File>import>(3D)scene...
when You drag'n'drop an fbx file, it will show only one node, not the whole structure.
Also if You add Camera / transform 3D / merge 3D, there is "import transform" button, to get the data from other 3d apps

Happy Fusioning
Last edited by Marek Subocz on Fri Nov 14, 2014 9:39 pm, edited 1 time in total.
Offline
User avatar

Stefan Ihringer

  • Posts: 257
  • Joined: Mon Nov 10, 2014 10:40 pm

Re: Help to Understand World Position Pass

PostFri Nov 14, 2014 9:34 pm

Dennis Beswick wrote:I see nothing but R,G,B in the drop down lists of the loader. I was looking in the format tab. Was that the right place to look?


Yes, that's the place. Seems like your exr doesn't have a Z channel. Can your renderer export to a separate file?

By the way, since you want a world position pass... why are you dealing with Z? Can't you render a WPP directly?
blog and Fusion stuff: http://comp-fu.com/2012/06/fusion-script-macro-collection/
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Re: Help to Understand World Position Pass

PostFri Nov 14, 2014 10:14 pm

Thanks for the tips.
I been exploring Thea Render and Blender trying to find the right export. I think the developers of Thea are working on the passes required in the next version which will be released very soon. I got closer with Blender and I'm still studying to figure it out. I see some users were able to create a WPP file and I'm doing some reading now. I did make a simple 3D animation from Blender exported as an OpenEXR sequence. When I loaded that one I got a "z" channel in the format tab. But the z channel does not look correct in the viewer. The background is white and all the objects are black. No variation in greyscale.

I feel like I'm getting closer. This is making me get to know my 3D software better. I will report back.
Offline
User avatar

Stefan Ihringer

  • Posts: 257
  • Joined: Mon Nov 10, 2014 10:40 pm

Re: Help to Understand World Position Pass

PostFri Nov 14, 2014 10:22 pm

Looking at Z in the viewers has some pitfalls. For one, z channels come in various formats and like I said before, the one that is suitable for Fusion is the one that comes out of its Renderer3D. Also, vray renders one that can be used without any conversions if I remember correctly. The Z channel contains floating point distances that are negative and can reach from 0 to -100 and on to negative infinity.

If you move your mouse over the black and white image in the viewers, what does it say in the bottom left corner of the Fusion window? (there should be numbers of the current pixel's rgba and z value). If it's something between 0 and -1000 you're probably all set. The viewer just can't display that range. There's a button below each viewer that looks like a vertical bars with a gradient. If you press it, the viewer will normalize its contents (smalles value becomes black and largest value becomes white). Does this display a gray scale image?
blog and Fusion stuff: http://comp-fu.com/2012/06/fusion-script-macro-collection/
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Re: Help to Understand World Position Pass

PostFri Nov 14, 2014 10:48 pm

I see many values as I move the mouse over the black and white image (after pressing "z" in the viewer). I see position X and Y and then color channels for RGB, A and Z. Should there be a "z" value in the positions?
Offline
User avatar

Stefan Ihringer

  • Posts: 257
  • Joined: Mon Nov 10, 2014 10:40 pm

Re: Help to Understand World Position Pass

PostFri Nov 14, 2014 11:50 pm

That's good, the Z channel seems to be working (position X and Y are just mouse coordinates in the image).

It's just a viewer thing that you're seeing a black and white image. There's a special setting in the viewer preferences that deals with how those large (and possibly negative) Z values are turned into a grayscale image. You could change it but you could also press that normalize button (the vertical bar with a gradient below the viewer). That should create a grayscale image for temporary viewing.
blog and Fusion stuff: http://comp-fu.com/2012/06/fusion-script-macro-collection/
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Re: Help to Understand World Position Pass

PostSat Nov 15, 2014 6:00 am

Yes, I can see the difference working with that footage. The great news is that I got official word that Thea Render's next version will contain a world position pass export along with more channels. It's great timing. So in the mean time I'll be working through Blender for my exports.
Offline

Kirill Kryukov

  • Posts: 24
  • Joined: Wed Nov 12, 2014 6:09 pm

Re: Help to Understand World Position Pass

PostSat Nov 15, 2014 3:21 pm

Looks like depth channel comes inverted from Blender exrs. Black is closer. What's the simplest way to invert it in Fusion?
Offline
User avatar

Stefan Ihringer

  • Posts: 257
  • Joined: Mon Nov 10, 2014 10:40 pm

Re: Help to Understand World Position Pass

PostSat Nov 15, 2014 9:29 pm

It would be really useful to look at an example file. I don't know if you can upload here (you can at steakunderwater.com - just register over there in case it doesn't work here).

I'm sorry that it seems like I'm complicating things... but with Z it's never easy to give definitive answers without seeing the footage. There are different formats from different renderers, sometimes accidental clipping, the viewer does its own interpretation of the Z channel which it doesn't do if you copy Z into RGB using a channel boolean...

The answer on how to invert a z pass would be a custom tool (z formula = 1-z) or a brightness/contrast (gain = -1, brightness = 1). But that's only valid if your Z values are actually between 0 and 1. Are they? (you can check with the mouse or by enabling the pixel inspector subview).

:!: A depth pass with values between 0 and 1 is a nice mask for color corrections and fake fog. But it's not a valid Z channel for Fusion's depth tools (WPP, SSAO, depth merge, z blur...)
blog and Fusion stuff: http://comp-fu.com/2012/06/fusion-script-macro-collection/
Offline
User avatar

michael vorberg

  • Posts: 943
  • Joined: Wed Nov 12, 2014 8:47 pm
  • Location: stuttgart, germany

Re: Help to Understand World Position Pass

PostSat Nov 15, 2014 10:36 pm

in blender with cycles you can create a WPP pass :
http://blender.stackexchange.com/questi ... ition-pass
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Re: Help to Understand World Position Pass

PostSun Nov 16, 2014 2:11 am

I'm having the same issue as Kirill. My Blender Z channel comes inverted. I also notice the values go greater than 1.0 in the z as I move my mouse in the viewer.
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostSun Nov 16, 2014 11:41 am

Dennis Beswick wrote:I'm having the same issue as Kirill. My Blender Z channel comes inverted. I also notice the values go greater than 1.0 in the z as I move my mouse in the viewer.


make sure that you doing rendering in linear space and 32bit float.
I did a quick example about Z and WPP, also I showed how you can invert Z depth.
https://cloud.mail.ru/public/c095a209faf8/wpp.rar
Offline

Kirill Kryukov

  • Posts: 24
  • Joined: Wed Nov 12, 2014 6:09 pm

Re: Help to Understand World Position Pass

PostSun Nov 16, 2014 3:00 pm

Thanks Michail for your example. It's turns pretty easy to invert depth.

When you talk about liner space, do you mean exr files should be saved/outputted with gamma 1 (as they usually are by default in any renderer) and should be looking darker in Fusion than they are in Photoshop for example that displays them with a correction for typical 2,2 monitor gamma ( as do Blender and Max) and Fusion looks like not doing so?
Offline

Kirill Kryukov

  • Posts: 24
  • Joined: Wed Nov 12, 2014 6:09 pm

Re: Help to Understand World Position Pass

PostSun Nov 16, 2014 3:18 pm

Stefan Ihringer wrote:
:!: A depth pass with values between 0 and 1 is a nice mask for color corrections and fake fog. But it's not a valid Z channel for Fusion's depth tools (WPP, SSAO, depth merge, z blur...)


Is there a way to remap depth value from 0-1 that comes from general depth b/w image into a valid one?

By the way, Z values in Blender exrs looks like goes from 0 to infinity or something. I see Z: 1.0e+11 on the background.

While from 3d max exrs G-buffer "Z" I can't get anything reasonable, z channel shows just some weird random stripes across the screen, normalized or not. It allows to implement render ellement(pass) into the file but it came the same as simple 0-1 image. Do somebody know how output depth from 3d max properly?
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostSun Nov 16, 2014 8:31 pm

Kirill Kryukov wrote:
Stefan Ihringer wrote:
:!: A depth pass with values between 0 and 1 is a nice mask for color corrections and fake fog. But it's not a valid Z channel for Fusion's depth tools (WPP, SSAO, depth merge, z blur...)


Is there a way to remap depth value from 0-1 that comes from general depth b/w image into a valid one?


I'm really recommend to look at this DVD http://www.thegnomonworkshop.com/store/ ... ompositing
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostSun Nov 16, 2014 8:33 pm

Kirill Kryukov wrote:Thanks Michail for your example. It's turns pretty easy to invert depth.

When you talk about liner space, do you mean exr files should be saved/outputted with gamma 1 (as they usually are by default in any renderer) and should be looking darker in Fusion than they are in Photoshop for example that displays them with a correction for typical 2,2 monitor gamma ( as do Blender and Max) and Fusion looks like not doing so?


yes, with gamma 1.0. to see linear images with Gamma 2.2 you can use LUT in the viewport settings or connect your end of the chain with Gamut node, choise sRGB in a dropdown list and check Add Gamma.
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostSun Nov 16, 2014 8:43 pm

and one more thing about wpp outputed from 3ds max.. keep in mind that 3ds have Z as up axis.. so if you render wpp with mentalRay or, maybe, vRay - you will need to flip axises in Fusion. Cebas FinalRender outputs correct wpp by default.
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Re: Help to Understand World Position Pass

PostSun Nov 16, 2014 9:29 pm

Mikhail- I tried your example but I'm missing something. The EXR file that came with your example says wpp but there's no z channel in it. In the viewer I can see the city, the noise and fog separately. Other than that not sure what to look for.
Offline

Kirill Kryukov

  • Posts: 24
  • Joined: Wed Nov 12, 2014 6:09 pm

Re: Help to Understand World Position Pass

PostSun Nov 16, 2014 10:55 pm

Thanks for the tips Michail. As for viewing right gamma I should set the gamma to 2,2 or sRGB output manually inside LUT button options. Right? I don't see any changes by default when push LUT button.

ps. Denis. looks like Fusion doesn't see extra channels inside exrs unless you manually appoint them in loaders "Format" tab
Michail's file does have embedded Z rgb pass in 0-1 scale + world position (also rgb) pass.

But looks like it see "true" g-buffer depth things automatically. Like in Blender exrs.

Wonder why I can't get something out of Max g-buffer. Tried rla and rtf, no lack, they come RGBA only. Anything from Max exrs is coming as pure mess. Starting to love Blender even more :)
Last edited by Kirill Kryukov on Mon Nov 17, 2014 12:17 am, edited 2 times in total.
Offline
User avatar

Stefan Ihringer

  • Posts: 257
  • Joined: Mon Nov 10, 2014 10:40 pm

Re: Help to Understand World Position Pass

PostSun Nov 16, 2014 11:05 pm

Mikhail Korovyansky wrote:and one more thing about wpp outputed from 3ds max.. keep in mind that 3ds have Z as up axis.. so if you render wpp with mentalRay or, maybe, vRay - you will need to flip axises in Fusion. Cebas FinalRender outputs correct wpp by default.


You can use ColorMatrix for flipping.
I once had a macro on vfxpedia for this... hope it resurfaces soon.
blog and Fusion stuff: http://comp-fu.com/2012/06/fusion-script-macro-collection/
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostMon Nov 17, 2014 11:33 am

Kirill Kryukov wrote:Thanks for the tips Michail. As for viewing right gamma I should set the gamma to 2,2 or sRGB output manually inside LUT button options. Right? I don't see any changes by default when push LUT button.

ps. Denis. looks like Fusion doesn't see extra channels inside exrs unless you manually appoint them in loaders "Format" tab
Michail's file does have embedded Z rgb pass in 0-1 scale + world position (also rgb) pass.

But looks like it see "true" g-buffer depth things automatically. Like in Blender exrs.

Wonder why I can't get something out of Max g-buffer. Tried rla and rtf, no lack, they come RGBA only. Anything from Max exrs is coming as pure mess. Starting to love Blender even more :)


yes, LUT in viewport is empty by default, you need to set it by yourself (hit little triangle and set color gamma to 2.2 in Fusion View LUT).
and yes again, when you load exr in Fusion you need to assign extras by hands.
when I setting up exr I'm manualy did saving for rgb in extra passes only and set them to 32float.
Offline
User avatar

Stefan Ihringer

  • Posts: 257
  • Joined: Mon Nov 10, 2014 10:40 pm

Re: Help to Understand World Position Pass

PostMon Nov 17, 2014 12:11 pm

What's the name of the z pass generated by the renderer? I think it gets detected automatically if it's named "z".

These extra channels are always 32bit float by the way, even if the loader is set to integer bit depth.
blog and Fusion stuff: http://comp-fu.com/2012/06/fusion-script-macro-collection/
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostMon Nov 17, 2014 12:33 pm

Stefan Ihringer wrote:What's the name of the z pass generated by the renderer? I think it gets detected automatically if it's named "z".

These extra channels are always 32bit float by the way, even if the loader is set to integer bit depth.


it callls Z.. nope, not detected.. anyway..
about channels bit depth - I think it depends on 3d package.. 3ds puts extras with different bit depth and you need to set it in saving settings.

this is how exr settings looks by default: http://postimg.org/image/dfwyle9bh/
Offline

Kirill Kryukov

  • Posts: 24
  • Joined: Wed Nov 12, 2014 6:09 pm

Re: Help to Understand World Position Pass

PostMon Nov 17, 2014 7:57 pm

Mikhail Korovyansky wrote:
yes, LUT in viewport is empty by default, you need to set it by yourself (hit little triangle and set color gamma to 2.2 in Fusion View LUT).
and yes again, when you load exr in Fusion you need to assign extras by hands.
when I setting up exr I'm manualy did saving for rgb in extra passes only and set them to 32float.



Thanks for making it clear. Those multiple complicated color management setups do confuse me a bit.

By the way, I solved my problem with Max. Just had a render preset where main beauty pass was set as tga and other passess as exrs. Now after I set it as exr, rla or rpf everywhere it comes to Fusion just fine with all set automatically.

Now I have another question. I see every g-buffer channels aliased. I get edge artifacts with SSAO because of this. Wonder how should I cope that? What if my renders have a lot of hairs and small details? Render everything in much bigger resolution? I have games background mostly and remember when the deferred rendering approach appeared ( that looks similar to Fusion one) firstly AA was a huge problem but by now its all set somehow. Guess should be some tricks?
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Re: Help to Understand World Position Pass

PostMon Nov 17, 2014 8:50 pm

Mikhail- I opened your example and really took a look. Everything seems to be working ok now. I must of done something wrong when opening it the first time. Thanks for this.

Did you import the camera or was that made locally? I saw in a video tut, they used a 3D camera and said we needed to get the camera info from the 3D package. My issue is that my main renderer (Thea) does not export camera info at this time but it does export EXR and soon wpp. I have no problem getting a camera via FBX from Blender but it is not my renderer of choice and would not want to be going back and forth to change camera angles if I want to change the shot at times. I can export from Blender to Thea via a plugin, but once I leave Blender, it's best for me to not have to go back since I cannot get the camera out of Thea. Is this a big issue if I want to work from Thea to get results like your example?
Offline
User avatar

Stefan Ihringer

  • Posts: 257
  • Joined: Mon Nov 10, 2014 10:40 pm

Re: Help to Understand World Position Pass

PostTue Nov 18, 2014 2:34 am

Mikhail Korovyansky wrote:
Stefan Ihringer wrote:What's the name of the z pass generated by the renderer? I think it gets detected automatically if it's named "z".

These extra channels are always 32bit float by the way, even if the loader is set to integer bit depth.


it callls Z.. nope, not detected.. anyway..
about channels bit depth - I think it depends on 3d package.. 3ds puts extras with different bit depth and you need to set it in saving settings.

this is how exr settings looks by default: http://postimg.org/image/dfwyle9bh/


I see, I misunderstood. I was just talking about the Fusion side, where these channels end up being stored as 32bit float even if you render just half float data for those passes.
blog and Fusion stuff: http://comp-fu.com/2012/06/fusion-script-macro-collection/
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostTue Nov 18, 2014 2:08 pm

Dennis Beswick wrote:...Is this a big issue if I want to work from Thea to get results like your example?

as for now I don't know how to read camera transformations from EXR in Fusion.. If you can export all that you need from your 3d app so I don't see any problems.. If not - you will need to find a way how to get all data that you need (wpp+cam). BTW - if you have a camera and Z you can convert it to wpp using Fusion.. anyway I think you definetly always need to export your camera (FBX, *.ase, *.lws, *.ma or so).
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostTue Nov 18, 2014 2:13 pm

One more example:
how to get WPP in 3ds with MentalRay and how to setUp it inside Fusion (axis flips and so on):

https://cloud.mail.ru/public/549d23602575/WPP_mr.rar

If someone will have a better way - feel free to share your tips.
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Re: Help to Understand World Position Pass

PostWed Nov 19, 2014 8:09 pm

I finally figured out how to get my cameras from Sketchup in Fusion :D

The confusion for me was FBX being at the root of the import title. Through the file menu, I loaded a 3ds file and was able to import my cameras from Sketchup. I wonder why FBX is used as the title for importing all the 3D formats? It should simply be "Import 3D scene" and the tool should be called "3DMesh Tool". FBX is a format among many others. Maybe the developers had a good reason for this labeling?
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostThu Nov 20, 2014 12:51 pm

Dennis Beswick wrote:I finally figured out how to get my cameras from Sketchup in Fusion :D

The confusion for me was FBX being at the root of the import title. Through the file menu, I loaded a 3ds file and was able to import my cameras from Sketchup. I wonder why FBX is used as the title for importing all the 3D formats? It should simply be "Import 3D scene" and the tool should be called "3DMesh Tool". FBX is a format among many others. Maybe the developers had a good reason for this labeling?


haha, it was called "Import 3D Scene" till to 6th version.. ))) and then it was renamed to "FBX Scene" ))
Offline

Dennis Beswick

  • Posts: 57
  • Joined: Tue Nov 11, 2014 7:22 pm

Re: Help to Understand World Position Pass

PostThu Nov 20, 2014 3:03 pm

Now I want to figure out why the cameras do not face the model in Fusion as they do in The 3D package. The cameras are facing away from the scene when imported. Any ideas?
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostThu Nov 20, 2014 4:47 pm

Dennis Beswick wrote:Now I want to figure out why the cameras do not face the model in Fusion as they do in The 3D package. The cameras are facing away from the scene when imported. Any ideas?


maybe you need to turn on\off "Flip Y" when importing?
Offline
User avatar

Blazej Floch

  • Posts: 191
  • Joined: Tue Nov 11, 2014 12:48 am
  • Location: Toronto, ON

Re: Help to Understand World Position Pass

PostSun Nov 23, 2014 8:34 pm

I guess the reason is that everything under the FBX Menu item is imported through the FBX libraries which also support 3ds, obj etc.
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostMon Nov 24, 2014 11:09 pm

one more example (about "how to use wpp for volume masking?", "how to use wpp as Zdepth?", "how to get Z from wpp and flip it?" and finally "how to get Z from wpp and make it normalized (from 0 to 1)?")

https://cloud.mail.ru/public/23a6290de13c/_WPP_03.rar
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostMon Nov 24, 2014 11:35 pm

Kirill Kryukov wrote:Is there a way to remap depth value from 0-1 that comes from general depth b/w image into a valid one?


solved? connect your Loader with AutoGain node and check "Do Z". It should to remap your values for Z depth.
Offline
User avatar

Stefan Ihringer

  • Posts: 257
  • Joined: Mon Nov 10, 2014 10:40 pm

Re: Help to Understand World Position Pass

PostTue Nov 25, 2014 8:03 am

The Z AutoGain is a bit dangerous because it will be recalculated each frame. This means that if your smalles and largest Z values change over time you will distort your Z buffer during the animation.

To turn a 0-1 z buffer into one that is understood by Fusion's depth tools, you need to turn it into the 0 to -infinity range. If that Z pass came from 3dsmax for example, there were normalization values in the render settings which you need to "undo".
blog and Fusion stuff: http://comp-fu.com/2012/06/fusion-script-macro-collection/
Offline
User avatar

Mikhail Korovyansky

  • Posts: 140
  • Joined: Mon Nov 10, 2014 9:59 pm

Re: Help to Understand World Position Pass

PostTue Nov 25, 2014 1:31 pm

Stefan Ihringer wrote:The Z AutoGain is a bit dangerous because it will be recalculated each frame. This means that if your smalles and largest Z values change over time you will distort your Z buffer during the animation.

To turn a 0-1 z buffer into one that is understood by Fusion's depth tools, you need to turn it into the 0 to -infinity range. If that Z pass came from 3dsmax for example, there were normalization values in the render settings which you need to "undo".


ah, ok, didn't know that, never used it for sequences.. it looks like the same flickering problem if you will render exr in 3ds and include Z from frame buffer and not from render elements, as minimum with mentalray it gives flickering in Z.

Return to Fusion

Who is online

Users browsing this forum: No registered users and 52 guests