PreShow :  ( Warm up) →  00:00 – 08:03  :::   Show: Skip to → 08:13  :::   Q&A : Skip to → 20:28

 

In this episode of ( officially CGalterTv).

You’re going to learn:

  • How Technology is limiting us from taking advantage of it.
  • Why you should stop learning from photographers and what to do instead.
  • What is Post Production?
  • Q&A (Facebook and YouTube): How to avoid fringing and hello edges when you start color correcting your work.
  • Q&A (Gmail): What to do when you work with maps like Normal maps, displacement maps, specular maps under a linear workflow.

 


 

SHOW NOTES

 

Post Production starts in the early stages of your work, not at the end.

We need to reconsider the fact that post-production is a finishing step, post production starts in the early stages of your work and not at the end.

In order to have the necessary data to work with while color correcting and adjusting, a few steps are required.

In fact, it’s mandatory for post-production to work properly.

Why you should stop learning from photographers and what to do instead?

Photographers methods for post production aren’t useful as they limit themselves while taking photographs because they rely on the machine ( SLR camera) with its 10 ko algorithms to decide for them what is the proper, sharpness, saturation, exposure etc …

For instance:

If the camera decides that a corner is going to be black then there is no way back in the post to fix it as it’s just a color, details are clipped, some applies in the case of the highlights.

So if we apply their post-production methods to our CGI work we will be limiting ourselves as we’ll. it will be reflected in our work

What to do instead?

We should learn from filmmakers, people who deal with video in general as their methods are far more advanced.

Why?

Because they broadcast live and need extra security, clipping is not an option for a live show!

It’s like a radio station, they have those devices etc that allow them to monitor sound and prevent it from clipping ad going nuts.

 

Why we need to take control and prevent technology, digital software and algorithms from guessing on our behalf.

Digital software use a gazillion of hacks to make the decision on our behalf, that’s Awesome!

But when Softwares works on Automatic a few things happened:

  • As in the example of what happened with a Digital Cameras, algorithms spend so much time calculating and guessing what is the proper Sharpness, Exposure etc. and this result in Rendering that is clipped or damaged.
  • Data that we can’t use in post production in a correct mathematical way ( Non-Linear workflow ).
  • Making the software on Automatic will in some cases increase the render time.

 

You Should Take Control and Stop relying on Technology.

 

You can guess Better, and the post will be far more fun.
.
 

 .

Q&A

.

How to avoid fringing and hello edges when you start color correcting your work? Question by Simon and Thanos.

An alpha is a black and white image, what is Black is equal to 0 and what is white is equal to 1.

Which means that anything that you multiply by 0 ( Black ) is going to be deleted and what you Multiply by 1 ( White ) is going to be preserved.

that how we get the Transparency.

An alpha should be applied only once while operating several color corrections and adjustment to a specific layer.

because if you zoom a little bit on the shift between the black and white color of the alpha you’re going to notice that there is a new color ( Grey ).

Ex: Let’s say that the grey is, 50% of the 0 ( Black ) and 50% of the 1 White, When we apply an “S” curve trick to alter the Shadows and highlights of the given image we are altering our Friend ( The grey ) color as well.

it happens every time you add a new color correction with that some alpha.

so for every CC you multiply by the alpha, this will result in shifting the edges.

What to do to fix this,

apply the alpha only once

How?

3ds max export images with an alpha already multiplied, it’s called a premultiplied image.

You need to work on an image without the alpha applied to it, it’s called Unpremultiplied image, just another way to say that the alpha is not multiplied.

You can use ProExr plugin to load the images without the alpha applied.

it’s like we reverse that temporarily, make our CC and then reapply it at the end of the chain.

In photoshop, I use groups and you can make an adjustment to your passes in one Groupe and once you finish you apply the alpha to the entire group.

that’s it

Recap:

Load your images as Unpremultiplied images using the Proexr plugin ( AE and PS ), then make CC, Finaly Apply the Alpha at the end.

 

What to do when you work with maps like Normal maps, displacement maps, specular maps under a linear workflow?

Question by Manny.

If you Create textures in photoshop in a linear space and export your maps as a 32 bits maps.
then back in 3ds max, you don’t need to reverse the gamma as it’s already linear. its gamma is 1. 3ds max will automatically read it’s bit depth and ignore it.
Another example will be of a displacement map from Zbrush or a vector from Mudbox.
A Displacement map or a normal map have micro details that can be easily lost if exported as a 8 bits maps.
they are calculated in ZBrush and Mudbox in a linear space and saves as such, to keep those data and details, back in 3ds max we need to make it’s gamma to 1,
3ds max will ignore it anyways and assume that it’s Linear on auto.

 


Links and Resources Mentioned in this Episode