Please help me to improve!


#3

Hi Andre,

thanks for watching and replying.
I am allready on the lookout of internships. Pixomondo turned me down so. Didn’t had Acht Frankfurt on my list. These here right?
http://www.acht-frankfurt.de/

Will look at them and might apply. I first applied only at studios for film and advertisment. But not all replied. I want to improve and then apply again until I get an intership spot.

So thanks for the suggestion. Any critiques about my works and ideas what to model next with extra attention and motivation?


#4

best would be an object from you apartment. like a tv remote or anything that is near you, so you can have closeup look, inspect the details, edges and so on. its not important what it is, is just about modeling.
spent time… a lot of time and model even the small bits.

right now, your models are looking a bit “blobby” with a lack of “character”.

there is a amazing tutorial from “Grant warwick” on vimeo, about the topology and controlling the meshflow


#5

I lived 35 minutes from Frankurt. :slight_smile:

I work on 3D modeling forautomotive as well, but not the same as what normally posted here. LOL.

Anyway, I agree that the dinosaur is the best of the lot.

If you are interested in character modeling, then a huge amount of time is required to observe & study anatomy.

Sketching is always good. I find one quick & good way is to take a medium tip marker & outline people in magazine or newspaper. It can be done while you are watching TV or waiting for someone, & i do it once in a while. Drawing over the main lines to to study clothes foldings, figure study, & expression. Quick & effective.


#6

Wow, thanks for the reply! Now I have a way to go. Will search for that tutorial, do something that I have near me as detailed as I can and will outline people in magazins and then try to model them. Really thanks. I’ll post what I model here.

Do you also have some tips for uv-layouting. I found out, that even when the topology is mirrored and therefor the same that if I do unfold on the parts the unfold is differently - like one part gets tiny while the other stays fairly large.
I use Maya.


#7

Nice work so far, Monica! Perhaps i’d try to push the presentation with one of the characters further, like for example showing different poses with the second one to convey the quality of the design. And it’s also much better when showing a model in default pose to line it up at different viewing angles.

I don’t think that your models were too basic (except maybe the woman and the monkey at their current state), but the presentation is very shy. Sell your models like when they were the best characters in the world. Convincing others needs self-confidence - it’s an illusion to expect enthusiastic reactions from strangers when even the creator doesn’t seem to be fully convinced. Draw on what you already have first, instead of jumping from one model to the other. Convey the value of your work!


#8

They point of modeling is getting the details + keeping the mesh as simple as posible.

For example your dino has way too much points. You cant handle that mesh.
You want to know/to have control of every point of your mesh.
The ear is a good example of how the lines follow the shape. This way you can have less points to achive the desired form.

I personaly like the monkey but is too simple.
Try to add details on the shape of the feets.
The belly is a circle, add shapes there.
For the eyes you could make buttons and sew them to the body. That tension will make some wrinkles on the face to the eye.
Mouth is a line, as before, add shapes to it.

Always show your meshes.


#9

Thanks Argon for your advice.

I am really a lil bit confused. I always hear like you say: as less polygons as possible. And then they like most the dino - honestly the lines you see on the dino is the low poly version with smooth applied to it, so maya could render out the toon-lines cause mental ray can’t do that - I didn’t modeled that much detail (the dino is from a course from Digital Tutors, also the naked woman). And my friend in games industry sayed that they want high poly (high quality) meshes and look out for people who can do those.

I don’t have time this weekend to make large updates and progress - I will model a detailed mesh on monday, and then I will add detail to the monkey - he was actually made in a few hours before going to the long learning process of rigging.
And when I model and post theses meshes I will work also on the presentation!

Also thanks to zokana

And it’s also much better when showing a model in default pose to line it up at different viewing angles.

I know what you mean by looking at lianlais WIP! I’ll try to do that with my characters.
You are kind of right, it’s more like that I do things to accomplish something very tiny - like with the second char I just wanted to prepare myself for organic modelling since I did much hardsurface modelling - and so I am not too convinced that the design is good, cause I just modelled at the PC without reference and without thinking too much. Just what looked good to me - and I am not so much of a designer.
All meshes are like firsts.


#10

The reason people like to see the low poly version also is so they know how it was made before turbosmoothing or similar was applied. It’s nice to see under the bonnet/hood and look at the basics, as well as the outer shiny parts.


#11

Hey SundayNights,

The points made about polycounts is highly subjective - Polycount doesn’t come into the equation unless you’re in Games Art like myself. VFX’s only restriction is render times/meeting the deadline/your workstations limitations.

The best suggestion so far is to model something you can have in front of you and examine in fine detail as you model.

Modelling from tutorials can only teach you so much, I personally found that reading the Topology thread on this website to help the most - The most important thing you need to learn is the theory behind edge-flow. Understanding why the mesh acts the way it does when smooth/deforming will allow you to get things right first time - and fix it when it doesn’t quite work out :stuck_out_tongue:

Never hurts to sketch what you’re abouts to model - paying close attention to the planes and defining features. Before Zbrush I found drawing over my sketches a rough-edgeflow often helped me avoid issues with poor topology!

Look forward to see’ing some high-res stuff


#12

Hello there,

so, I modeled today and tried to cope with very high poly. First I started a basketball-shoe but this was somehow too difficult for me, so I did a (japanese) fan. So I am not sure if this is high poly enough. I also spended more time rigging and skin-weighting it than modelling… I will try again something more difficult - but maybe not my old shoe (might be better for zbrush/mudbox). The render is with maya-software.


fächer_show von Sundaynights auf Flickr


#13

Ok, here a better attempt to model something high poly.

I tried to model after this video: http://www.youtube.com/watch?v=LFqopkUTO0Q&feature=relmfu

I am sure I got some features wrong, but I hope the edgeflow is right. I think I did better with the transition between eye and outer part of the scull - but I certainly had problems with the fold of the eyes and where they shout start and end, as the video didn’t showed that so well. I should buy also a table-mirror. ^^
I am also unsure of the inward folding of the mouth-corner.


female_head_18.09.2012_mentalray von Sundaynights auf Flickr


female_head_18.09.2012_viewport_wire von Sundaynights auf Flickr

As I followed the video the style should be livelike but still symmetrical. ( I think symmetrical is best for rigging and animation -the symmetrie can get broken then with a blendshape)

Critiques are really welcome! I am looking forward to see which parts are ok and which should get improved.


#14

The mesh looks more or less ok. The only problematic parts are the outer sides of the eyes (the temples actually with that abrupt change of density plus that connection with that transverse quad right in that curvy shape), and the loop around the corners of the mouth should get fully continuous (regular curve) - the curve should follow the nasolabial fold all around the mouth (your loops seem to be pressed in toward the mouth corners).

Personally i don’t agree that much with the alignment of the wire at the jaws, having there mostly a troublesome area. IMO better to fully follow the jaw line. But i have seen many models done like yours, working fine at the end when raising the resolution.

The overall density isn’t flawless from a theoretical point of view. The rule is: same curve requires same amount of loops (exception: areas, which will bend even more by morphing or rigging, needing then slightly higher density). So the double density on the cheeks (compared to the forehead for example) might be somewhat too high (even if planning strong cheek action…), but perhaps you’re right with preferring continuous loops there.

As for the shape:

  • the bridge of the nose is sticking out too much (at eye level).
  • chin and mouth area are somewhat pushed forward (in side view).
  • ears seem to be somewhat worn out… plus small and thin
  • that ‘double shape’ at the outer brows is questionable (should be continuous)
  • nose could be much wider, the sides getting almost below the inner eye corners (also on cute faces with very wide eye positions).

#15

Hi Zokana,

thanks for your advice. I will make the adjustments and try to apply them also to my next models. Here first a quick paintover.


female_head_19.09.2012_wire_paintover von Sundaynights auf Flickr

So, if I understand you right: if I want to have the actually density on the cheeks I should also add density on the forehead?
I used the ear as shown above in my first post but was unsure how big I should make them… will make them a lil bit sicker and fleshier!


#16

Yes, the ideas for corrections seem to be fine. Except perhaps the widening of the nose - the width of the bridge is ok while it’s more the lower part needing volume.

The other way, i’d say. (i’d lower the density on the face) But you’re right when keeping the current density on the face.


#17

My suggestion
Get the topology right first, then get proportion right. Post your updated image, & we can look into the next step.


#18

Oh too bad they turned you down and yes that’s the Acht Frankfurt, fingers crossed :slight_smile: Only apply with your best stuff and watch video tutorials about modeling, texturing, shading and rendering. There’s plenty of stuff out there.
Critic wise a lot has been said by the others already. Hmm pick something you can get a lot of reference photos about as well as texture material (you might just shoot by yourself).

Cheers
Andre


#19

I also thought, that just modelling along with a video without any reference isn’t the best way to get good results, but it helped to get the process down and see how and where the edges flow. I will work on the first head and make the adjustments tonight, but for now I want to show you a new version of a human head/face that I did with reference.
Forgive me that I hadn’t the patience to set up lights and render it out.

The reference: I had a hard time lining them up and the ear isn’t in the exact same position.

xray:
[url=http://www.flickr.com/photos/sundaynights/8009326791/]

wires:
[url=http://www.flickr.com/photos/sundaynights/8009335072/]

[url=http://www.flickr.com/photos/sundaynights/8009327117/]
viewport view:
[url=http://www.flickr.com/photos/sundaynights/8009326893/]

[url=http://www.flickr.com/photos/sundaynights/8009326979/]
[url=http://www.flickr.com/photos/sundaynights/8009326979/]

what do you think?
I tried to let one edge stay where the jaw line should be, and I broke up the symmetry.

I was at the main station, sending someone off, and was stunned how many different faces there are. I tried to imagen how I would model them and where to put the lines and so on.
I definetly will do more faces!


#20


female_face_WIP21.09.2012 von Sundaynights auf Flickr

and here the worked on version of the previous female head. I think she looks like a gypsy now and I want to get her black long hair, with wavy eyebrows, big eyelashes, flowing garmets and big earrings.
I used the reference from PKD to match proportions. Thank you PKD!
Is there a good side for modeller reference like you posted? Or just big brother google?


#21

I like your new model (the first you’ve posted now). The only problem with the mesh:

  1. that ‘cobweb’ formation above the ears
  2. the very straight loops at the inner brow area
  3. and that pulling up of the wire to the ears at her neck. The lower res at the sides of the neck would cause problems when rigging (the loops have also to follow the shape of the planned bending)

The shape has some nice attempts but it’s partially overdone - especially cheeks and jaw remind slightly on a potato (especially in 1/4 view) - get more subtle there and also your victim (your real model) will like it much more.

An other point is the rear half of the head, needing more attention.(the widest area of a skull is right behind the ears when looking from above) And keep an eye on the back of the head also in 1/4 views, looking somewhat strange at the moment. Plus give the model those nice ears instead of such large things (not reaching down that low at all).

An other tip: better to use tele angle lenses when photographing someone for ref - the center of the picture is otherwise enlarged due to distortion, getting then large noses, large cheeks, large ears, etc. If you don’t have such a camera then do the shot from further away and crop the head later.


#22

Thanks Zokana,

you have good points there!
I’ll work on her and adjust her as soon as I can.

An other tip: better to use tele angle lenses when photographing someone for ref - the center of the picture is otherwise enlarged due to distortion, getting then large noses, large cheeks, large ears, etc. If you don’t have such a camera then do the shot from further away and crop the head later.

I actually first startet with a very close-up taken photo and realized when it came to the ear that something isn’t right (because they actually got cut off). I took a shot from farer away (the ones from reference are just zoomed in), but it might be still somewhat distorted as I made them with my smartphone. I’ll test my digicam - maybe it makes better reference photos.
I fear that I can’t affort to buy tele angle lenses for my cam - that’s costy phew.