ZBrushCentral

ZBrush beta testing

Back with another question :slight_smile:

Looking at the rear view of your leg, you have modeled skin folds behind the knee. Perhaps you haven’t built this mesh with the intention of animating it, but I am wondering what you think is the best route to take when detailing models to be used for generating displacement maps for use in animation.

For example, generally, would it be better to model the flexed pose (which probably has ‘more’ detail in the form of creases etc.), or the unflexed, stretched pose? Or maybe the procedure is quite different, perhaps requiring the medium frequency details to be modelled into both (and even more) poses?

In Pix’s dino animation thread he mentioned tweaking the map between morphs, so I am curious as to how that tweaking is done. Is the base mesh changed and a new map generated, or is the map itself worked on?

Perhaps I should ask Pix himself :), though I would be very interested in your professional opinion how you might go about this.

It would be something to keep in mind while I am making models in ZBrush 1.55, especially if I want to apply the new features to them later on.

Many thanks :slight_smile:

Good question, Bonecradle


Yes good question,

First in response to Kircho I have been modeling for about 7 years professionally. Ive seen quite a few tricks for producing the highest quality models for both automotive and entertainment purposes ( but I don’t know everything by far ).

But what I have been able to produce with Zbrush digitally, is something that could not be done before without the use of a high quality scanner or clay+lots of time. In fact it is the first time that I have been able to fully reproduce creatures/characters,
digitally that I could conceptualize on paper.

Currently the workflow for this kind of tool is as follows.

  1. Develop a good design.

  2. Build a high quality Base Mesh of your creature, character, weapon, armor
 etc.

  3. Continue with Zbrush to add MedRez surface detail. Things like bone through skin, muscle variation, fatty tissue, tears in armor,
    you know all that cool kind of stuff. This detail will give your models a photo realistic look to them ( not limited to photo realism ).
    If you want your models to look wicked then DO THIS!

  4. Create a 2d (normal/displacement) map that represents the offset between your LowFeq and MedFreq
    ( there are plugins that do this on the NET ).

  5. Paint a map for high frequency detail like skin pores, scratches, small veins, small wrinkles
etc.

  6. Render the original Base Mesh with the 2d map as a displacement or a normal map for games.

For an example of a similar workflow go see the Rings movies. Look at the Mountain Troll and Gollum.
The idea behind this is not new ( not mine ). Zbrush just allows you to do it in record time!

I will have more detailed tutorials for this available once the new software comes out.
Cheers!

Dave Cardwell
http://cube.phlatt.net/forums/spiraloid/viewtopic.php?TopicID=581
http://www.drone.org/

If you want all know about displacement map and other similar things
go to Olivier Renouard link given by the Skycastle’s Signature :cool:
Pilou

Very cool Dave! Very cool!!

I am happy to meet you skycastle !!
Tnx for sharing this news over zBrush!!!
:smiley: :+1: :+1:

Thanks a lot! Now I’ll have to calm myself down by chewing on a rag until the upgrade comes out. Can’t
 calm
 down
 gnngnn


Nice meeting you, Skycastle.

Hey Skycastle! :slight_smile:
Did you work on Gollum or the Troll?! :slight_smile:

Also, I’ve never heard these terms before. Can you (or anyone) confirm if I’m right? :

Low frequency = the mesh
Medium Frequency = Muscles and anomolies at that level
High Frequency = the main texture

Think of the frequencies in terms of real wavelengths. Low frequency waves can be a meter or more in their period. High frequency waves are very short in length, very small.

So skin pores and the fine wrinkles in the skin are ‘high frequency’ because they are small, and there is very little space between each pore or wrinkle. Then there are the ‘medium frequency’ details. Look at the palm of your hand. Up close, the fine wrinkles and palm-print ridges are high frequency, but the larger, deeper wrinkles cut across the entire palm or the insides of the fingers, and are generally ten-times (or more) deeper and longer. Now hold your palm farther away and look at it through slitted eyes. You should just be able to make out zones of light and shade that suggest that the palm is slightly concave, the fingers are plain cylinders, and the heel of the palm is rounded, merging with the cylindrical wrist. These very basic shapes are the low-frequency detail of the hand.

The finger-print ridges and skin pores are just too fine (high frequency) to model (yet). Even if we had the computing power, I don’t know if it would make sense to model this level of detail anyway.

The medium-frequency wrinkles, bone protrusions, tendons, etc. can be modeled, but that level of modeling is out of the reach of real-time rendering for games, and even in movie production, attempting to animate highly detailed models may not be feasible (current processing power and tight deadlines). But if a highly detailed model can be rendered from a low-poly (low-frequency) model combined with a normal map derived from a medium frequency model, then both real-time game engines and movie animators can benefit.

Great stuff Dave! Keep on pushing the package, looks like your having a lot of fun :+1: :sunglasses:

Thanks Jaycephus! That was very helpful! :+1: :slight_smile:

Hi all,

For those that are interested, I’ve written an a review article on this very topic that will be published in the next issue of Highend Magazine (www.highendmag.com)

It covers the current technology and techniques that film and games studio’s use to manage their high resolution characters designs.

Best regards,

Zendotheworld


oh,the old times :slight_smile: