A few months back i found an awesome creative STmap tutorial by hansama. Search on UA-cam 'hansama ST_map fire' He will animate a STmap (x,y) and with different chained Time offsets (roto'ed) and blurred on top of each other --> he will create a smooth wavy effect 👍 And thanks for your tutorial! It made much clear and the noise blend is a neat effect.
@@CompositingAcademy just throwing out an idea - going over zdepth passes and how to properly use them. I find they are super finicky and hard to use, as they come out differently in every render engine and differently sized scenes. Both for fog and defocus! Just a thought ;)
By having zip compression (the default) sometimes you will see slight wobbling from the ST map that's rendered out. It needs 32 bit but also no compression to have a clean result.
@@CompositingAcademy Hello! The gumroad link isn't free for me, sorry because my parent doesn't allow me to spend money online yet. May I ask if you can set it to free? I would really appreciate it. Thank you.
7:09 --> STmap affects the coordinate system of the pixels. If you warp or translate (x,y) your stmap --> your source will be affected the same way. All these things are 2D. The creative way is shown at 3:22, the Blending Morphs. Instead of warping and animating your image by hand, you can dissolve between two key-positions. This is a fast and dynamic solution :).
Hey there, There is a bit of filtering that can happen on the image when it's generating the pattern. The center of the bottom left pixel on the grid is [0.5, 0.5]. My understanding is that using a .5 offset sort of fixes a bit of blurring that can happen through cubic filtering when generating this pattern. In general I just keep the expression memorized though, I couldn't fully explain to you the technical details of this image filtering, I only know that that it's necessary in the expression.
Since you already have the geometry... why would you project the stmap and not just let it wrap around the object? (normal geo input instead of projectnode); depending on the UVlayout of the object it might get split? would it matter though?
The reason projecting it is useful in this case is because we can stick things from the cameras perspective. If we were to just use the UVs, we would run into a few problems if we wanted a similar result. There are many pieces of geometry which have different UV layouts for one, so it would be difficult to have things "overlap", for example. Also, we would have to mess around a lot with the placement of that alpha or texture to figure out the right spot to work with out camera angle. By projecting it we can remove all of the unnecessary problems and work directly from the camera's view, stick things where we need quickly, etc. It's a really great technique especially for sticking color grades to different areas of CG environments. Cheers!
@@CompositingAcademy true! the figuring out, though, could be done via the UV map you would get from the scanlinerender node? But I guess the coordinates of the stmap would just fall into crumbles and lets say pixel 1 wouldnt be placed on pixel2. What about projecting it from top instead of cameraview? wouldn't you "work around" the objects?
@@bliccer9319 Yep you definitely could! The problem with UV space comes in though when there's multiple pieces of geometry though, because their UV spaces don't align in resolution or placement. For example this hill if it had many pebbles or other objects on it, and you want to wrap an alpha over all of them together. I suppose you could project this ST map from the top down as well. Any projection could be "baked" this way, essentially! The only problem sometimes is when objects have other objects behind them. Nuke's project 3d doesn't do the best occlusion unfortunately.
Thanks Charles! At the moment there's some information about UV manipulation in regards to remapping textures with a CG render in my NK303 class, but that's about it. However there's a variety of other CG techniques taught in the classes, that go well into detail as well as a variety of ways to combine techniques. Also, more advanced CG classes are coming in the future that I'm working on currently! Cheers!
![UVs and UV Passes in Nuke: PART 1 [Beginner]](http://i.ytimg.com/vi/wb1WjHlXbn0/mqdefault.jpg)
![UVs and UV Passes in Nuke: PART 1 [Beginner]](/img/tr.png)
![Re-lighting Real Footage | Nuke Compositing [Advanced]](http://i.ytimg.com/vi/VYjmvB6d9NA/mqdefault.jpg)
![Compositing in UV space with Projections | Nuke [Advanced]](/img/n.gif)
If you guys have other creative or useful ways that you've used a UV coordinate system, feel free to share in the comments!
A few months back i found an awesome creative STmap tutorial by hansama. Search on UA-cam 'hansama ST_map fire'
He will animate a STmap (x,y) and with different chained Time offsets (roto'ed) and blurred on top of each other --> he will create a smooth wavy effect 👍
And thanks for your tutorial! It made much clear and the noise blend is a neat effect.
These videos are great, really hope you keep uploading. There aren't too many good Nuke channels!
Thank you! Yes definitely, I have a lot of ideas for other videos, so more will be coming.
@@CompositingAcademy just throwing out an idea - going over zdepth passes and how to properly use them. I find they are super finicky and hard to use, as they come out differently in every render engine and differently sized scenes. Both for fog and defocus! Just a thought ;)
You are just simply amazing! Keep on the good work
Such an awesome video. Congrats!
great video.This opens many possibilities..Like we can use this instead of position pass mask or use for texturing in nuke..Thanks a ton
great Tut as always :-D
Great
14:00 - Why no lossless compression?
By having zip compression (the default) sometimes you will see slight wobbling from the ST map that's rendered out. It needs 32 bit but also no compression to have a clean result.
@@CompositingAcademy Hello! The gumroad link isn't free for me, sorry because my parent doesn't allow me to spend money online yet. May I ask if you can set it to free? I would really appreciate it. Thank you.
nice
Some segments in the video are stamped not adjacent to each other
that was great tip, How about 2d? How we can use this technique in 2d?
7:09 --> STmap affects the coordinate system of the pixels. If you warp or translate (x,y) your stmap --> your source will be affected the same way. All these things are 2D.
The creative way is shown at 3:22, the Blending Morphs. Instead of warping and animating your image by hand, you can dissolve between two key-positions. This is a fast and dynamic solution :).
To create the UV/ST map in the expression you use "(x +.5) / width". why do you add the 0.5?
Hey there,
There is a bit of filtering that can happen on the image when it's generating the pattern. The center of the bottom left pixel on the grid is [0.5, 0.5]. My understanding is that using a .5 offset sort of fixes a bit of blurring that can happen through cubic filtering when generating this pattern.
In general I just keep the expression memorized though, I couldn't fully explain to you the technical details of this image filtering, I only know that that it's necessary in the expression.
@@CompositingAcademy Very cool!
Since you already have the geometry... why would you project the stmap and not just let it wrap around the object? (normal geo input instead of projectnode); depending on the UVlayout of the object it might get split? would it matter though?
The reason projecting it is useful in this case is because we can stick things from the cameras perspective. If we were to just use the UVs, we would run into a few problems if we wanted a similar result. There are many pieces of geometry which have different UV layouts for one, so it would be difficult to have things "overlap", for example. Also, we would have to mess around a lot with the placement of that alpha or texture to figure out the right spot to work with out camera angle.
By projecting it we can remove all of the unnecessary problems and work directly from the camera's view, stick things where we need quickly, etc. It's a really great technique especially for sticking color grades to different areas of CG environments.
Cheers!
@@CompositingAcademy true! the figuring out, though, could be done via the UV map you would get from the scanlinerender node?
But I guess the coordinates of the stmap would just fall into crumbles and lets say pixel 1 wouldnt be placed on pixel2.
What about projecting it from top instead of cameraview? wouldn't you "work around" the objects?
@@bliccer9319 Yep you definitely could! The problem with UV space comes in though when there's multiple pieces of geometry though, because their UV spaces don't align in resolution or placement. For example this hill if it had many pebbles or other objects on it, and you want to wrap an alpha over all of them together.
I suppose you could project this ST map from the top down as well. Any projection could be "baked" this way, essentially!
The only problem sometimes is when objects have other objects behind them. Nuke's project 3d doesn't do the best occlusion unfortunately.
Great video! do you cover UV manipulation in any of your udemy courses?
Cheers!
Thanks Charles!
At the moment there's some information about UV manipulation in regards to remapping textures with a CG render in my NK303 class, but that's about it. However there's a variety of other CG techniques taught in the classes, that go well into detail as well as a variety of ways to combine techniques.
Also, more advanced CG classes are coming in the future that I'm working on currently!
Cheers!
@@CompositingAcademy Thanks for the speedy reply, really enjoying 202/303/404, great refereshers and learning lots of new little tricks
Thanks again!