Talk:Blender 3D: Noob to Pro/Creating Models With Photo Assistance

From Wikibooks, the open-content textbooks collection

Jump to: navigation, search

To take a picture along the X, Y and Z axes at the same time, you could use two mirrors. One is placed next to the puppy at 45 degrees to the camera and 45 degrees to the puppy. Another is placed above the puppy, also at 45 degrees to the camera and 45 degrees to the puppy. This produces three images, one of the puppy, one of its reflection seen 90 degrees to the right, and one of its reflection seen from overhead.

This must be a joke... What if we had a title 'Turning a Cube into a Bigfoot'? Will you take a picture of him? This tutorial must be without reference pictures. Make some simple puppy model or change title in 'Turning a Cube into a photorealistic Puppy'--Popski 21:04, 18 February 2006 (UTC)
Generally, if one were making a model of a Bigfoot, one would have reference drawings to go by, I'd think - sketches, at least. And where this is the second tutorial intended for complete neophytes, I'd think that it's entirely appropriate to leave the photos in here. Moofus 00:28, 3 July 2006 (UTC)
Would someone who has a puppy mind posting some orthogonal photos (and providing a link to them, wherever they are) My lack of the former and hence the latter is making it really quite challenging for me to follow this tutorial, must as I would like to :) Sandothegrate 21:06, 1 March 2006 (UTC)
I've posted 6 pictures of a toy wolf to get this tutorial rolling. I'm trying on my own to do it the way I've seen in 3d Studio Max, with images mapped to 3 planes (top front side) but even though I can get the images to line up, it doesn't show in the ortho views, even when I turn on textured view. Must be a button I'm missing. Back to the manual I guess. --Dmwick 03:25, 17 March 2006 (UTC)
Well, I took the wolf pictures, thanks Dmwick, and made a wolf using them, and wrote up a fairly comprehensive tutorial. That was before I read this page and discovered that Popski thinks reference pics are a joke. No, Popski, that ain't nice, and it ain't assuming good faith to tell a newcomer to Wikibook contributions that their work is a joke. Have you got a better idea for taking three photos of a squirmy dog in x,y and z axes? Then, write it up.

Well, be that as it may, I need some guidance on getting the screenshots uploaded. Any advice would be appreciated: there's a bunch of them, and they're large. Are there any size limits? I've got Irfanview, so I could cut the size, if that's necessary. I'd like to get the pics up before I add the text, since I don't want the text up without the pictures - it won't make sense. Also, since I am new here, is this Popski fellow someone I need pay any attention to? Bill 03:56, 23 April 2006 (UTC)

It'd be good if it was possible to post said tutorial.. I'm currently clueless as to what to do. as for images, maybe you should use imageshack? or something? --124.178.120.214 08:16, 8 June 2006 (UTC)

Shouln't this chapter rather explain cube modelling? The pictures look (to me as a noob) like some "curve and path modelling" (see some chapters below) or polygon-to-polygon modelling. I have expected something like extruding and splitting cubes. Correct me if I'm wrong.

The wolf is based in cubes in the first several steps, but more basic cube modeling is available (see Quickie_Model and Modeling_a_Simple_Person). --75.33.63.136 02:06, 1 January 2007 (UTC)

Contents

[edit] Views

The text says

To line the new endpoints up in three dimensions, open three view windows, put one in top view, one in side view, and the third in front.

How do I open three views?

it is explained in a previous tutorial. There is a link to get there in the setting up section.Badalia

[edit] New Tutorial

I have created a new tutorial that I think is much clearer for this topic. Is anyone open to seeing what it looks like?


I like the idea of starting with a simple pyramid instead of the dog, but your tutorial needs some editing. You didn't bold the keystrokes and it needs to be made more into steps and less in small multi-step paragraphs.
This should also probably be split into 2 sections on the categories page and not called 'turning a cube into a puppy' as you don't really use the cube in either besides the 1 vertex point. Something like 'creating models with photo assistance' with a few parts.
It might be a fun idea to use some schematics instead of photos since the photos aren't easy to take.. there are a lot of neat schematic websites out there that have cars and mechs and all of that, scaled and ready to be created. I haven't done it myself yet though, just an idea. Once I get a better handle of Blender I'll give that a shot. --Darkdaven 19:09, 7 August 2006 (UTC)


I do not know how to change the font in the tutorial. If you know how to do it, please do. I will break the tutorial in to smaller steps. (The "cube/vertex" was how the original writer of the tutorial started their object, I think because they (and I ) don't know how to simply create a vertex from nothing, but the point is well taken)

...later...I figured out how to change the font. I split the directions down some, but I don't think they can go down any further. Let me know what you think.

I found this website that had schematic drawings for every starship in startrek. Its pretty silly but I think that it will do the trick-Badalia
I like the tutorial on the pyramid one.. makes me understand precisely the technique of modelling with pictures as guides. But I cannot follow the second part of the tutorial... Anyway, now I am trying to model a puppy (from photos of my puppy soft toy), not sure exactly how I should be doing it. Can I post screenshots of my progress and then you all help me see if my method is correct? (my model has really a lot of vertices! i think..)

Would it be possible to choose a slightly easier photo to model than one of a wolf? I've spent three days or so at it, and it's coming along miserably... Perhaps something inorganic and not unexpectedly curvy, like a lightswitch, chair, bottle, or computer monitor? The curvy edges of the wolf make everything so confusing. I find myself playing the "which vertex goes where?" game... Some suggestions as to how to keep track of everything and keep the edges created by subdivisions straight (I'd particularly like some advice on when they're necessary, or if it's simply better to have as many surfaces as possible? I get lost looking at my model) --MyOwnLittlWorld 02:54, 3 January 2007 (UTC)

Like the guy above me, I'm really struggling with the wolf, the problem for me is the lack of direction at this point : " Let's start refining the model starting with the tail. Line up the verts over the wolf in each viewport by lasso selecting multiple vertices (CTRL LMB, Drag). then move to the right location with grab (GKEY)." Then in the next pictures, they have a decent looking tail I can't get at all. I'd be grateful if someone could spell it out step by step.

[edit] Moved discussion from module page.

(user comment) I too was doing this tutorial when I realized the pictures were a bit off. So I decided to do my own photos of my own object. However this turned out to be more difficult then I thought at first. While taking your own photos be aware of several things. The camera you are using has to have its lens parallel to the object you are taking a picture of. Also when you take the 3 different photos (front, side, top) or even more you need to keep the camera the exact same distance from the object as well as the exact same zoom on the lens for all your pictures. Once you have your photos, rather then lining up key points in the photos you should center your object in the center of the photo.
(user comment) It's not necessary to center the object in the photo. In the Background Image window, you can tweak the center point with X Offset and Y Offset.]
(user comment fr. random9q) I'm using v2.4.2a -- don't know whether this is a bug or something else -- but after selecting the background image for a viewport it fails to display the image. That, ah, makes this page in the tutorial a bit abstract. But still worth the time, as -- basically -- you're learning how to take a very plain object and whittle it down to what you want. Notice that this is different than everything done with the stick figure, where what you did was additive? Two very different techniques, and I'd imagine that in production you and I will be using both interchangeably. If you can get the background imaging to work then you're doing better than I am, and you get to practice whittling down an object to a specific reference image.
By the way, it doesn't really matter whether you choose to use the background image's X and Y bias settings to offset the image or choose to simply drag your object around to align it with the image. The important thing is this: once you've used the first image and gone on to the second and the third, the edges of your puppy-to-be had best match up against the image you're about to work against. If you don't, your puppy is going to be a bit distorted, which for most folks kind of defeats the purpose of using a reference image. (If you're like me, however, it might make a cool way to devise a weird art-piece that only looks like the original object from very specific angles yet, otherwise, looks nothing like it.) (end comment) (Moved by --Herraotic 22:42, 6 January 2007 (UTC))

[edit] Could use some clarification

"Subdivide the appropriate faces" (In figure 2.3.2) may be a bit too vague for a newbie.. at least, it definitely is for me because I'm a little stuck here. I know how to subdivide faces, but I don't have an exact idea of how I'm going to achieve what I need here with subdivisions.

[edit] Agreed

I have no clue what the "appropriate faces" are here. I tried a couple different faces, but have no clue as to how form the parts of the body from the new faces. - 216.6.179.63 18:43, 18 July 2007 (UTC)

[edit] Horse Images

Would like to contribute these images for use with this excercise:

Horse Back,  Horse Front,  Horse Top,  Horse Bottom,  Horse Right,  Horse Left

I found that the following paramaters sized & lined up the images pretty well:
front -   size: 4.87, x-offset: 0.32, y-offset: -0.13
  top -   size: 5.00, x-offset: 0.00, y-offset: 0.00
right -   size: 5.22, x-offset: 0.46, y-offset: -0.35

I plan to improve these (shooting from much farther away, better lighting, better background) when I can borrow my friend's Nikon. But until then, this is the best my camera can do.

I'm having trouble loading up the images on Blender and the guide is really vauge on how to do it.

[edit] Vertex at cursor?

Is there a way to create a vertex at the cursor? I know Ctrl+LMB creates a vertex, and then you can send it to where your cursor is using Shift+S, but it would be nice to be able to skip a step and just create it there in the first place. Any thoughts?

[edit] pointer to finished model?

Could someone provide a url for a finished wolf model from this tutorial as done by the original author or an experienced user? I think it would be helpful to be able to compare my end result to what might be 'expected' from someone who knows what they're doing.

Excellent tutorial, otherwise!

[edit] Two vertices in the same spot

What's a good way to select the right vertices, e.g. to move the right hind leg back? I always end up selecting some of the vertices of the left hind leg also, somehow...

Do I really need to switch between face, edge, and vertice mode and rotate furiously until I'm sure I got just the right vertices, and nothing but them?

I've learned a lot with this turorial, but it's given me a lot of headaches too. There's got to be a better way to learn this...

[edit] modelling the tail

Does anyone know how many cubes does the tail consist of? In figure 2.3.2, the picture shows that it consists of 2 cubes. However, in figure 2.3.3 it seems that the back cube has been divided into two cubes. I can't understand how the top and side views reflect the same lower part of the tail. Any help would be greatly appreciated.

[edit] Appearance of background image depends on zoom!

I tried the section on making a pyramid using photo assistance. I used the supplied Trianglexemplar.png and created a corresponding square. I find that the triangle is skewed left/right and clipped, and the amount of skew depends on how closely the view is zoomed in. What's going wrong?

[edit] View Windows

NUM1 NUM7

NUM3 NUM0


Is not: (at least in v2.47)


Top Side

Front Camera


I don't know if it was different in older versions, this confused me at first, so it should be corrected.

> Definitely agree with this.

[edit] Help needed as to which vertices to move to form the tail and legs.

There is no clarification for moving the vertices. It just says, "Now move the vertices to form the legs"..."Now continue on to the hind legs." Alright considering that I'm farely new at this program it is quite difficult to decide which vertices to move. I've tried this tutorial 4 times now and always end up with multiple vertices in the same place making it hard to form the other legs. Plus moving the wrong vertices make lines in a place not needed. So how do we decide which vertices move? Also sometimes when I move a vertex, it becomes dis-aligned in a different veiw. And when I try to fix that vertex in the veiw it became dis-aligned in, it becomes dis-aligned in the veiw that I was just working with. This happens even when I use axis restriction making it impossible to align that vertex. So my question is, how do I align a vertex in a way that it doesn't become affected in the other 2 veiws? One more question, Could I just extrude faces of the square over the body, head, and tail excluding the legs, then just extrude or use Ctrl + LMB in edit mode with a vertex selected to form the legs then add faces over the non-faced parts of the model?


No wonder it said, "This tutorial might frustrate you," and, "Step three in a whole make take serveral hours." I'm not sure if this is some kind of test for us to learn something on our own, or if the writer of this tutorial left out information, accidently or not, or if he became lazy and expected us to have the mental knowledge to figure this out for ourselves. Either way, this part was not well put together and needs fixing or a different example to teach us "Photo Assistance." This isn't the first time I noticed incomplete information...thus far into the book making it look sort of, unprofessional. (no offence)

Retrieved from "http://en.wikibooks.org/wiki/Talk:Blender_3D:_Noob_to_Pro/Creating_Models_With_Photo_Assistance"