The hair division has always been a topic that makes groomers discuss, and its by no means easy, as it will have big repercussions on the structure of the groom, and if on production, on the other departments. Planning a good groom division will require considering things as length, shape, area, colors, groom quality, texture, and simulation. There are other areas that you may want to consider but we will look into the basic 5 on this tutorial.

1 – By Shape or Type

If the hair is visibly different and in different areas, the easiest solution is to divide them.
It may sound simple, but it’s important to understand that sometimes more grooms mean more control and optimization. You will not mix a “bun” with “dreadlocks” they are just different beasts with different requirements.
A Groom division can be found on hairstyles with a lot of defined flows and intricate designs.
For example, dividing clumpy areas, from fluffy areas, or on a more complex groom, braids from dreadlocks. Or even a ponytail.

Ponytails are normally divided into two, and the loose part it’s grown from a disc. The other areas are mostly related as the grooming quality is different and will require specific details.

Animals, on the contrary, have a lot of different types of traits that can define a groom’s division by shape. This is an example of division by SHAPE only, that’s why the groom on the leg and the mane are selected as the same. In normal circumstances ( Adding by simulation ) we will divide this groom also as furLeglong and furMane, not just as fur mane.

Here you can see that even if the furBody has so many characteristics, we are dividing into just one groom as most of them can be hit by maps.

The small tufts that you see on the hair can be divided further, and place manually for art direction. But we will just mention and not look into it.

2 – By Length

The longer fur will get a different groom because it takes a lot of curve points.
Length, is one of the most important characteristics of the groom, as it defines the behavior of the hair. Longer fur will have different properties over short fur in most scenarios, from clumping, shape, color, and simulation parameters. That’s why we should always try to divide by length when the “blending” process is not that complex, and we can read a clear division on its characteristics. A mane is a perfect example of this.

Another reason is that the longer fur will require more curve points, and this, by definition, requires more resources, ram, motion blur, and space. Longer hair it’s heavier, and on some grooming software, you can not divide the groom and assign different curve points by length. That’s why we divide them, as we don’t want to have a short fur that needs three curve points, on the same groom as long hair with 20+ curve points.

Support us on Patreon

3 – By Density

The density balance is one of the main areas that artists struggle in groom systems. If your density is super high in one area, and super low on another, then it can be possible that you will have a map that gets tilted towards one side. Either low-density values and you are painting super dark values or super white values in one small area.
This is the reason why it could be better to divide the groom. But take in mind that the blends are always hard to execute and take time.
For example, on a squirrel, the tail and the body will have different densities and may be easier to control them just by dividing the groom.

The problem that you may find with non-divided grooms, is that the balance of the density map is too tilted towards one side [HISTORIGRAM]. 

For example, on this groom, you can see that the face is super dense while the other areas of the groom are sparse, making it hard to balance the values and, a small change will mean the increase of more than double the density of the peak places.

4 – By Colour

Are they going to have different materials/look/color?
If the answer is yes and you can divide them fully by color or material, then that will make it easier to lookdev.
For example, eyebrows, having a different color, or beard on digidoubles.

On animals, it can get slightly tricky, as some of them have patterns that make them unique, and the best way to control this will be with maps. For those cases, we stick with the one groom technique. But at the same time, most animals have several coats of fur, and some of them don’t present the patterning. The look technique on animals will depend on how concentrated the qualities of the color are on the fur.
For example, on this Koala, the ear tufts have a different color, gradient, and pattern than most of the hair, so its probably easier to create them as a different groom.

Watch the tutorials On PATREON

5 – By Simulation

Are they going to be simulated, if so differently?

Simulation is not always necessary on a groom, you may get away with basic deformation of the skin if the hair/fur is short, or you may not even need simulation at all because of the complexity of the shot. 

But in the cases where simulation is needed, you will need to think about how to divide your grooms to give proper control to the CFX department later on. You don’t need to simulate the whole thing under normal conditions, and just specific areas of the groom may need guides or controls for that, so it is easier and friendlier from the groomer’s point of view to divide the grooms, by SIMULATED and NON SIMULATED grooms. 

For example, on a Lion, the Mane will require simulation most of the time, but the short fur rarely. So we divide the groom into those specific areas.

In a digidouble, the flyaways sometimes require more control, so probably on a complex windy shot, it may be good to have them as a different groom.
(on a hero shot, may need to be simulated differently so we divide them to simulate them apart)


For example, lets do the excersize of dividing this complex digidouble groom. 

The first question will be, what are the visible types or shapes of the hair? 
We can see 3 types, braids, interweabed braid, and bangs. 
The front braid will be the more complex one, but we could divide it in two, because the groom quality is different on the loose hair and on the braid. 

it will be complex to control the yellow and pink area together, mostly because the characteristics are extremely differently from one another, thats why its the base choice to divide the hair there.

Also its a common technique to grow the braids out of geo, so if possible try always to use geo to define the braids, plats or buns.

in general we will divide this character in approximately 9 smaller grooms

  1. MainHair
  2. MainHairFlyaways
  3. Bangs
  4. Braid1
  5. Braid2
  6. Braid3
  7. Eyebrows
  8. Eyelashes
  9. Peachfuzz

But there are possibilities depending on how hero the groom can be that the braids will requiere single flyaway grooms, as the braids have different structures and behave as a character on their own. This will add 

  1. Braid1_Flyaways
  2. Braid2_Flyaways
  3. Braid3_Flyaways

Ornatrix Specifics 

1. Ox allows you to divide an existing groom into as many parts/stacks as you need using strand channels or vertex color channels. In the specific example of flyaway hairs, in Ox, those are added using the Frizz node, which allows you to export the flyaway hairs into a strand channel that you can then use to create a new hair stack and simulate on its own.

2. Same as one but with Edit Guides>Strand Channels.

3. We already implemented this on the Width, Frizz, and Clump nodes. If you have two or more strand groups you can blend the effects of those operators between them. We’ll see if this can be implemented in hair distribution as well. Video:

4. Ornatrix can sample by length. The Length node has an option that allows you to add cvs/strandPoints/vertices based on the strand length. You can also sample by curviness so strands that have less curviness will need fewer details and will get fewer points.

5. You can separate grooms to different stacks using branches or merge them using the Merge node.