Updated T. rex skull - W.I.P.

[Dinoreplicas]

The lower jaws of theropods weren't joined at the centre point (unlike mammals, the jaw was a number of bones and not a fused unit) so it makes sense to temporarily separate the two halves to make it easier to sculpt the hard to reach areas.

[Dinoreplicas]

For convenience I joined the meshes for the neck bones into one object but I still need to pull them apart to sculpt some detail into the areas which are occluded when in their articulated position. I can achieve this fairly easily by using a temporary 'blendshape' also known as a 'morph target'.



Due to me being so obsessive about getting the shape of the cervicals correct I put most of the detail into the base meshes. Consequently there's not much to add at the sculpting stage.



I'm using symmetry across the X axis to lay down the basic detailing. To finalise I turn the symmetry off to give individuality to the left and right sides. Being able to sculpt both sides simultaneously is one of the benefits of digital sculpting. At some point though you've got to break that computer perfection.

Because of the millions of years of crushing and stretching, fossils tend to be very distorted and asymmetric but I don't want to emulate the fossil look. I'm aiming to model the bones as they would occur in the living animal. When I examine the actual bone skulls of animals they're pretty symmetrical with most difference occurring mostly in the details.

[Dinoreplicas]

It's always fun making these little GIF animations but procrastination is one of my known flaws.



Just about done with the skull and neck now. I'm going to leave it alone for a couple of days then come back and look at the work with fresh eyes. That always helps me spot things I've missed. In the mean time I'm going to start the mesh work on the foot.

[Mini Minmi]

Looking good! Keep it up!
I'm enjoying witnessing your journey in all its little details.

[Lone Trike]

Looks really awesome! I totally agree with you on the part about the symmetrical sculpting, when working digitally. Most of the time you can clearly see, if something was made with the computer, just because this symmetry option was used way to excessively.

[Dinoreplicas]

Thanks for the comments guys, not been using the computer seriously for a couple of days but I made some progress last night.



To start the foot I've used a hexagonal prism (just a low poly cylinder in Blender) as a base mesh for the three main (weight supporting) metatarsals.



After pushing around vertices and adding/deleting edges and loops here and there the bones start to emerge.
Much viewing of the Sketchfab 3D Trix and the Brochu paper helped me to close in on the forms.

[Dinoreplicas]

Generic sub-D meshes for phalanx and ungual in 2 views. These two shapes form the basis of the majority of the toe bones. Substantial topology modifications will be required for the proximal phalanges though.

[Dinoreplicas]

The metatarsals seem to be fitting together OK after a bit of adjustment.



Tyrannosaurs are well known for their interlocked and compressed upper foot (known as an arctometatarsal). The middle bone (digit 3) is tightly pinched between the adjacent metatarsals. It's believed the extra strength and rigidity provided by this feature helped Tyrannosaurs to be faster than similar sized prey.

That degree of tight interlocking (like a 3D jigsaw puzzle) I'll have to complete at the sculpting stage when the mesh will have a much higher polygon density.



The right upper foot viewed from behind, left to right are metatarsals II, III and IV. The flat egg shaped feature on the back of MT II is the attachment area for digit I, kind of like a 'dewclaw'.

[Dinoreplicas]

Full set of base meshes modelled for the toe parts.
I'll spend some time getting all the foot bones fitting together properly in a neutral pose before I start sculpting details.

[Dinoreplicas]

Wahey! it's a foot!
Quick anim showing assembly of parts and one full rotation. Think I've got this to a point where it's ready for sculpting. Going to get it out of my head for a while so I can look at it with fresh eyes. Tomorrow I'll have a last check/adjust then I'll UV map the mesh before I start detailing, same reason as per the skull.



I can definitely say that so far, the foot has been a substantially easier (and quicker) build than the skull which of course is inevitable, it being a much simpler structure.

I'm now thinking maybe I should have dealt with the skull as individual bones. If I ever do another revision I reckon that's what I should do. So rather than have a couple of polymesh structures representing cranium and mandible which I have to carve the bones into, I'll make a separate mesh for Nasals, Maxilla, Lacrimal etc.

Bleh! maybe next time. Don't know why I love to torture myself with these crazy notions, I just do!

[Mini Minmi]

Now your t.rex can both bite and stomp! 

[Dinoreplicas]

Now your t.rex can both bite and stomp! 

Hehe, seems like it's getting more dangerous with ever post!

Made the UV map for the rex foot prior to sculpting (shown at subdivision level 2). As with the skull and neck, this will only be of use if I ever make some renders. The UVs are not that tightly packed but it still takes ages to unwrap and position the meshes, bit of a chore but it's done. Anyway I'm satisfied it's ready to sculpt, to get at the occluded areas I'll fall back on the pull apart/push together method I used for the neck (see Reply #61).



Also this is currently a right foot because I've mostly based the foot model on the preserved bones of SUE (see the Brochu 2003 paper). However I'm going to flip the mesh horizontally so it becomes a left foot. The reason for this is for consistency with the Allosaurus arm and foot 3D prints.

[Lone Trike]

I don´t know much about modeling in blender, but it´s quite satisfying how clean your meshes look. When I sculpt in zbrush, it takes less than a minute for me to make the topology a complete mess

[Dinoreplicas]

I don´t know much about modeling in blender, but it´s quite satisfying how clean your meshes look.

For organic, nature type 3D work I'm kind of old school. I use sub-D surface modelling with sculpted detail at higher subdivisions. In some ways, that's kind of a depreciated workflow these days as many modern methods make topology and poly-count irrelevant.

When I sculpt in zbrush, it takes less than a minute for me to make the topology a complete mess

With zbrush's multi-million poly capability, it seems clean meshes aren't always required. I think one of the popular methods is to sculpt freely without concern for topology, then re-topologise later for more efficient (even realtime) models. Higher frequency details can be baked back in, either topographically or as bump/normal maps. I've also heard zbrush has a great poly-reduction feature if you're model is intended for 3D printing.

[Lone Trike]

I've also heard zbrush has a great poly-reduction feature if you're model is intended for 3D printing.

Yes, it´s called Decimation-Master. It´s great, if you want to reduce the polycount without losing detail. But it also turns quads into triangles, which is ok for 3d printing. But I guess, if you want to animate your model it´s better to keep the quads.

[Minmiminime]

WOW this project is looking good 😻 Very impressive stuff!

[Dinoreplicas]

WOW this project is looking good 😻 Very impressive stuff!

Thanks, I'm enjoying the research, the work and the challenge but also it's rewarding to me that some of you have found it interesting.
I'd recommend anyone to pick up their preferred project (whatever that may be) and take it as far as they can.

[Dinoreplicas]

Foot bones separated prior to sculpting.

[Dinoreplicas]

Metatarsals now sculpted and detailed. Front and rear views.
Reference provided by Brochu 2003 paper and Sketchfab Trix model.

[Dinoreplicas]

A few views of the sculpted metatarsals in their connected positions. It's now a left foot (if you hadn't already noticed).