Hey Ritchie
Good to see an update posted.
I checked out the model you sent me and it is in good condition and has the makings of a solid model. I was thinking that if you don't have much time left to rig all the snakes you could do some quick blendshapes to get them positioned, then if you need to animate/rig them later you can morph them back to their original straight pose. I would personally try and get the snakes looking all intertwined and natural before spending too much time rigging, just in case there isn't enough time to get everything completed and you get stuck with an electrocuted looking Medusa :)
Also, if you reduce the poly count on the snakes body and rely on the natural smoothness of subds you will be able to quickly position the snakes and get a look you are happy with.
I would also have all the snake heads facing the same direction similar to the ones on the back of the head. You have rotated the snakes on the top of the head to face forwards, but I would match them to the other snakes, then apply the body twist when you have rigged or are doing your blendshapes.
Looking forward to an update.
cheers
Andrew
Tuesday, 21 August 2007
AM - Biker Dan
Looking good as always Dan. Sounds like you've been playing around with your displacement maps a lot which is great. Don't be afraid to use bump mapping, it's obviously not a s effective as displacement but you can use it for pores if necessary.
The displacement on his face is looking a little strong and in sharp around the nose and on his cheeks. Tone it down a little and let the shader take care of the rest.
I'm not sure what your plan is for the dog, but it would be great to pose him slightly. Maybe have him leaning away from the biker, but restrained by the bikers hand on his back plate. You could even have his head turned towards the cigar with a worried expression...
Anyway, it's looking great, so keep it up...
cheers
Andrew
The displacement on his face is looking a little strong and in sharp around the nose and on his cheeks. Tone it down a little and let the shader take care of the rest.
I'm not sure what your plan is for the dog, but it would be great to pose him slightly. Maybe have him leaning away from the biker, but restrained by the bikers hand on his back plate. You could even have his head turned towards the cigar with a worried expression...
Anyway, it's looking great, so keep it up...
cheers
Andrew
Friday, 17 August 2007
Ritchie's OBJ
Hello everyone! I've put a few pics for you to have a look at andrew. The connections for the snakes has turned out well. So far shes modeled, rigged up and I'm on with the texturing. I'm doing a 4k map for the body and a 4k map for the snakes and i'll downsize them when it comes to rendering. I've sent you another obj also if you could have a look for me its much appreciated. Thanks again.
Thursday, 16 August 2007
Dan - Biker WIP
Hey All,
Right.......having played around with ZBrush for a bit, my attention has turned to trying to recreate the displacement in XSI with varying success. Mental Ray keeps running out of memory and crashing when you try to get the geometry resolution to go low enouth to get any of the really fine details. Either that or the wrinkles that I have added seem to inflate at rendertime.........it looks pretty wierd! At the moment I am struggling to get the pore detail on the biker's chin and head to come out. I have tried using a combination of displacement and normal maps, as well as exporting out a higher res mesh from ZBrush. The problem seems to be a Mental ray one, but it is all to possible that i have over looked something.
At the moment I am thinking of sticking with the dispacement that i have and trying to add a bum map to the mesh to hopefully get those elusive pores........does that sound like a good idea???
Also, whilst trying to give my brain a break from pondering displacements,I have started to add in some textures.......so here is some of that as well:
mesh with displacement and a normal map (normal map just taken from the last few levels of subdivision, in order to try and pick up the fine detals only.
Right.......having played around with ZBrush for a bit, my attention has turned to trying to recreate the displacement in XSI with varying success. Mental Ray keeps running out of memory and crashing when you try to get the geometry resolution to go low enouth to get any of the really fine details. Either that or the wrinkles that I have added seem to inflate at rendertime.........it looks pretty wierd! At the moment I am struggling to get the pore detail on the biker's chin and head to come out. I have tried using a combination of displacement and normal maps, as well as exporting out a higher res mesh from ZBrush. The problem seems to be a Mental ray one, but it is all to possible that i have over looked something.
At the moment I am thinking of sticking with the dispacement that i have and trying to add a bum map to the mesh to hopefully get those elusive pores........does that sound like a good idea???
Also, whilst trying to give my brain a break from pondering displacements,I have started to add in some textures.......so here is some of that as well:
mesh with displacement and a normal map (normal map just taken from the last few levels of subdivision, in order to try and pick up the fine detals only.
Wednesday, 15 August 2007
AM - Ollie
Hey Ollie
Malcolm is coming along nicely!
I've got a few notes for you regarding your model...
The bulge under his pectoral muscles seems a little sharp.
The waviness on the bottom of the pants seems to be incorrect. Remember that when it was made in the local sweat shop the length of the fabric would have been the same. It seems to me that you have tried to give it a natural wavy fabric appearance but in doing so you have changed the overall length instead of just scrunching up and pulling in/out. Once you apply some UV you can slap on a checker texture and make sure that there is not stretching. Also, it could really look great if you made the elastic on the short a lot tighter so that it sticks into the skin and creates an indentation.
The ankles need a little work. They are really close, but I think they would benefit from a little more attention and comparison to reference. Don't be afraid to add a little resolution here.
I have attached an image of the head with some of the areas I feel could do with some tweaks.
In order to make your life easier try and spend some time matching the camera better in XSI. Don't forget to look at the meta data on your photo and check the focal length you used, then use this in your 3D scene.
Overall though, he is looking great. There has been some good thought put into the topology and I'm sure this guy is going to look really impressive with the lighting setup you are considering.
cheers
Andrew
Malcolm is coming along nicely!
I've got a few notes for you regarding your model...
The bulge under his pectoral muscles seems a little sharp.
The waviness on the bottom of the pants seems to be incorrect. Remember that when it was made in the local sweat shop the length of the fabric would have been the same. It seems to me that you have tried to give it a natural wavy fabric appearance but in doing so you have changed the overall length instead of just scrunching up and pulling in/out. Once you apply some UV you can slap on a checker texture and make sure that there is not stretching. Also, it could really look great if you made the elastic on the short a lot tighter so that it sticks into the skin and creates an indentation.
The ankles need a little work. They are really close, but I think they would benefit from a little more attention and comparison to reference. Don't be afraid to add a little resolution here.
I have attached an image of the head with some of the areas I feel could do with some tweaks.
In order to make your life easier try and spend some time matching the camera better in XSI. Don't forget to look at the meta data on your photo and check the focal length you used, then use this in your 3D scene.
Overall though, he is looking great. There has been some good thought put into the topology and I'm sure this guy is going to look really impressive with the lighting setup you are considering.
cheers
Andrew
Tuesday, 14 August 2007
AM - Emma WIP
Hey Emma
Your scene has the makings of something great and you are on the right track.
The approach you took with the ground is correct but you could save a lot of polygons by subdividing more locally to the areas you want resolution. This approach can leave the other areas very low res and light weight.
Another quick method other than using Zbrush or Mudbox would be to use procedural maps and generate a displaced ground plane to use as a starting point. You could even paint a complete environment map in Photoshop using everything from procedural maps to real world geographic maps and aeraial photographs and use this to generate a displaced ground.
If I was to do this I would create a base mesh in Maya/XSI using the local subdivision approach, then take it into Zbrush/Mudbox for additional detail. I would export a resolution that represented the overall volume and shape of the ground and output this for my scene. I would then use the remaining finer resolution divisions to use as a displacement map.
The important thing is that there are a number of ways to approach this, both procedurally and artistically but the most important thing is to put detail where you see it. At the moment your camera doesn't get too close to the ground and you could get away with just a procedural displacement on top of your current mesh.
Are you going to be matte painting in some background mountains to add depth to the final frame of your scene? It will really help the overall look and scale of the scene.
Don't be discouraged by how long the ground has taken, it is looking good and you've obviously learnt a lot from the task. Spend some time now adding a nice subtle displacement and working on an impressive shader which really helps describe the surface eg: wet, dry, mossy, dirty...
cheers
Andrew
Your scene has the makings of something great and you are on the right track.
The approach you took with the ground is correct but you could save a lot of polygons by subdividing more locally to the areas you want resolution. This approach can leave the other areas very low res and light weight.
Another quick method other than using Zbrush or Mudbox would be to use procedural maps and generate a displaced ground plane to use as a starting point. You could even paint a complete environment map in Photoshop using everything from procedural maps to real world geographic maps and aeraial photographs and use this to generate a displaced ground.
If I was to do this I would create a base mesh in Maya/XSI using the local subdivision approach, then take it into Zbrush/Mudbox for additional detail. I would export a resolution that represented the overall volume and shape of the ground and output this for my scene. I would then use the remaining finer resolution divisions to use as a displacement map.
The important thing is that there are a number of ways to approach this, both procedurally and artistically but the most important thing is to put detail where you see it. At the moment your camera doesn't get too close to the ground and you could get away with just a procedural displacement on top of your current mesh.
Are you going to be matte painting in some background mountains to add depth to the final frame of your scene? It will really help the overall look and scale of the scene.
Don't be discouraged by how long the ground has taken, it is looking good and you've obviously learnt a lot from the task. Spend some time now adding a nice subtle displacement and working on an impressive shader which really helps describe the surface eg: wet, dry, mossy, dirty...
cheers
Andrew
AM - Phil WIP
Hey Phil
Your rider model is looking cool, there are some really good signs of understanding topology and anatomy. I agree with you that the head still needs a little work. I would be interested in seeing what photographic reference you are using. From the looks of things I would say you are just using the wrong camera focal length or reference which uses a wide angle lens eg:28mm
When selecting reference its best to try and find images that feel flat and have been taken with longer lenses. Most photographers will shoot with a prime lens and not a zoom which means your best options for matching a camera lens would be 28mm, 35mm, 50mm, 85mm.
When we shoot digital doubles we have a lot more control over our images and measurements are taken of everything and we create Canon lenses in Maya to match the on set camera. We also de-warp our images to completely flatten our photos to remove any distortion from the lens. It's a completely different approach when you just grab images from the web, but it's important to find images that will make your life easy.
Anyway, if possible try and get a few more images that show different angles of a female head, then try and lattice the general proportions to get a more realistic model.
The approach you are taking for the wings is a safe approach. Modeling in the veins will help if you have any close-up shots. Plus the resolution of the mesh isn't going to hurt at render time anyway. Just make sure that when you create your texture maps that you concentrate and painting a seamless blend between the modeled veins and the wing so they feel combined and natural.
cheers
Andrew
Your rider model is looking cool, there are some really good signs of understanding topology and anatomy. I agree with you that the head still needs a little work. I would be interested in seeing what photographic reference you are using. From the looks of things I would say you are just using the wrong camera focal length or reference which uses a wide angle lens eg:28mm
When selecting reference its best to try and find images that feel flat and have been taken with longer lenses. Most photographers will shoot with a prime lens and not a zoom which means your best options for matching a camera lens would be 28mm, 35mm, 50mm, 85mm.
When we shoot digital doubles we have a lot more control over our images and measurements are taken of everything and we create Canon lenses in Maya to match the on set camera. We also de-warp our images to completely flatten our photos to remove any distortion from the lens. It's a completely different approach when you just grab images from the web, but it's important to find images that will make your life easy.
Anyway, if possible try and get a few more images that show different angles of a female head, then try and lattice the general proportions to get a more realistic model.
The approach you are taking for the wings is a safe approach. Modeling in the veins will help if you have any close-up shots. Plus the resolution of the mesh isn't going to hurt at render time anyway. Just make sure that when you create your texture maps that you concentrate and painting a seamless blend between the modeled veins and the wing so they feel combined and natural.
cheers
Andrew
Subscribe to:
Posts (Atom)