If it is the standard LFS car renders, you don't need to spend so much money for Brazil, the Mental Ray that came with 3ds max is just as good and is free.
Mmmm. Your BMW skin, the front grill and above the headlamp is just like what I created years ago. Even the mark on the grill looks the same...
Anyway, good skin.
Here are my old renders to proof the grills and lights look very close.
make sure vertices are welded together as one object. At the moment they may be 3 or 4 seperate polygons that made up the glass. You will have to weld those touching vertices to make them as one element, then apply auto smooth group.
you will be able to tell if the glass is one element or not by go to the sub-level, choose from (vertice, edge, face, polygon, element), select element as your sub-level, click on the glass object, if the glass object is highlighted in red, that mean your glass is a single object, but if only part of your glass highlighted, that means your glass was made up by fragmented elements, which they'll need to become one object, otherwise will not be able to smooth.
The glow effects can be acheived with Glare Lum in Mental Ray without going to post production or Photoshop. Other Rendering package may have simular feature too.
That is not HDRI you making. You are only making an ordinary environment map. HDRI stands for High Dynamic Range Illumination. Which means it has many different light exposure informations in one single file. Imagine that is like taking 12 photographs of the same image but with different stops of exposures (from very dark to very bright) and all those informations are all stored in the one file.
I have read some comments from the www.sektor-41.com about your header...
Becareful here, don't try to throw in too many effects just because you can or somebody ask you to. Step back a little and re-think if the effects will do harm rather then good to your nicely modelled object.
Here is the PSD file. The speed line is so easy, just do a long stretched rectangle fill with white, then create a layer mask and add black grads on both end. Merge the layer mask, apply a little amount of Gussian blur. Then change the layer's blending mode (in this case I use overlay) and reduce the layer's opacity as you needed.
Press Ctrl-T to scale transform the layer so you can stretch it, or compress it.
In my opinion, your header is better in terms of composition and colours.
His car and the word 'sektor41' are 2 seperate elements in the header and they are not working together. The model of that car may be excellent, but the composition is not. Your tight cropping into the car makes the whole header more interesting and providing more imaginations of what the whole car may look like.(sometimes less is more)
I have attempted a photoshop edit of your header based on b/w one (hope you don't mind) add a slight blue tint to it. To prevent the image gets to 'cold', I have added a slight wash of the orange on the right. a little work on the words to make it more blend in.
BTW, I don't think flame coming out will work, first you will have to decide to create the flame in 3d or by Photoshop, 3D flame is not easy to look convincingly real especially for a still image. (you may get away in an animation, but a still image is difficult to have a result that you will like. Water and Fire are both very tricky to look good in 3D)
This is the model of a head that I done few years ago based on Chow Yun Fat. I didn't have very good references of Chow, the front photo and the side photo of him is over 10 years apart.
Start off by low poly model with mesh smooth apply, the idea is that I can rig the low poly head for animation (which is easier to control because less vertices to worry), then apply mesh smooth for renders.
The only thing with shadow map is that it won't create transparent shadow effect (i.e. light pass through glass) it only give you transparent effects if you use fall off for your material's opacity. If your material has a 30% opacity for example, but not using fall off, the shadow map will not calculate the transparency value and the shade of your glass shadow will be same as a solid object. Where as Ray Traced shadow will give you transparent effect in both ways.
Without seeing your working file it is difficult to say. There could be a few things. What do you use to rig the mouth movement? using path for the mouth shape and then apply Skin modifier? If you do it this way, you will have to make sure the area of vertices selected are exactly what you wanted, e.g. lower jaw will only include vertices in the lower part, if some of the upper lips area are selected, when you move the lower jaw up, some upper lips will move up too. Also make good use of soft selections. This will help the skin deform smoother (it also depends on how many polys you have in the bending area. more polys will look smoother). From watching the AVI, I can see some of the upper mouth corner vertices were not selected as soft selection.
Another way to do is create morph targets. create clones of the object, bend the mouth shape manually by using edit poly/mesh with soft selection swithed on. Create clones of some basic gestures such as open mouth, closed mouth, smile, 'O' shape etc. once you have created them, you can apply morph modifier to your original object, then include the clones as morph targets, now you can blend between each targets to animate the mouth movement. (remember when creating morph targets, do not increase or delete any polys, the morph target has to have the same amount of polys as its original)
I personally perfer using Morph tartgets, because I get more control of what I want, and I can always create more new morph targets.
Adv. Raytraced is not supported by MR, however, you still can use it.
I'll normally use the Ray Traced shadow. The MR renders Ray Traced with difussed edge, unlike Default Scanline Renderer, which will render hard edge.
Please find attached 3 images of different settings of MR Ray Traced light that will give you softer edge by adjusting the size of the light. If your shadow spread alot, you will have to increase the U V samples to have smoother blend, otherwise your shadow will be too noisey.
The resolution of your bitmap will affect the displacement result too, if the image is low res, you may get ragged edges due to the pixelation of the image. Higher res will give a smoother edge.
Also, apply a small amount of bump map using the same map you used for displacement, this should enhance some small texture details that displacement map may not pick up.
The wheel you are using is not refined enough to render like a chrome wheel convincingly. edges are too sharp, need to be smoother for the reflection to work well (blend in better). Model your own wheel will be better. You can search in the internet for tutorial of how to create one. It is not that difficult.