Spline layout...

[MikeV]

So, as a test project, I'm working on recreating the smaller market stand I talked about in this post.

 

I got the upright legs done easily enough, just created the basic shape of the legs and extruded and scaled. I'm now working on the larger cross-beam that you can see peeking out from beneath the canopy. Being the most "complex" piece of that particular model, it provides a new challenge for me, in that it has an area recessed into the "wood" and isn't one big flat piece. With polygonal modeling, it's a matter of some extruding and scaling. With patches, of course, it takes a bit more planning.

 

So, started thinking about how to best reproduce that piece. I decided to test myself with a bit of an experiment to see how well I understand spline/patch construction, by seeing how many ways I could go about getting the same result. In the end I came up with 4 different ways. The result of the 4th looked indistinguishable from one of the others, so I left it out. It was just a different way of getting the same spline layout.

 

In the end, I'm kinda curious if there's any one approach that's "better" than the others. I mean, the first one is using a dead-end spline to simulate a typical polygon reduction technique in poly-modling. So that's probably a no-go just on principle. But of the remaining two, would one be preferable over the other in terms of "proper splinemanship"?

 

Might seem like I'm making a big deal about a little thing, but it's one of those deals where experimenting can help me learn things that will help me in future projects. And it was a fun little "self-test" to give myself .

 

The screenshot shows the same 3 objects side by-side.. I rendered one set to show the result, and kept the other set in their normal wireframe/shaded view.

 

 

My other question about that is, if there's any tricks to getting that shape so the profile of it curves a bit more naturally, while keeping the sides "flat". I tried setting the CPs running along the edge splines to "curved" and it kinda gets the result I want, but it's very uneven and there's some "bulging" where I don't want it. I'm thinking maybe I need to work with the curve handles to tweak it?

[Fuchur]

Hi Mike,

 

the first approach with the 3-point-patches is the one I would not use. It has dead-end-splines and may give you trouble lateron.

 

The other two are more or less equally good, BUT hooks can be trouble-some in some situations too... nevertheless in 99 of 100 times it will be a good solution and it will result in lower patchcounts which is (in most cases) a good thing. Here you just need to decide, if a lower patchcount is important. (can have impacts on rendering-times, export, animation, etc.)

 

There still is a dead-spline in all of the approaches at the "window-like" structure (the area which is stamped embossed into the object) but it will very likely be okay in this situation.

 

The other question about the smoother edges, etc.:

There a several approaches:

1.) Bevel the edges.

2.) Use the bias-handlers to make it smooth.

3.) Do both and get the best of both.

 

For mechnical modelling I would recommend to use peaked splines and AFTER peaking, change the bias-handlers. This will give you a huge amount of control and will very likely result in better results and less afterwork than using curved-splines and bias-handlers. The trick is to use the bias-handlers in a way, that the resulting direction of both of them will become a tangent for the point you are changing the biases for. That will result in a smooth, but still mechanical look.

 

Of course if you bevel the edges and then make the beveled areas smooth (if this is look is what you are looking for) can create very neat looking edges.

 

Hope this leads in the right direction.

 

See you

*Fuchur*

 

PS: Just a hint: If you bevel the edges, the problem with the deadend-spline can be solved too.

[MikeV]

The other question about the smoother edges, etc.:

There a several approaches:

1.) Bevel the edges.

2.) Use the bias-handlers to make it smooth.

3.) Do both and get the best of both.

 

For mechnical modelling I would recommend to use peaked splines and AFTER peaking, change the bias-handlers. This will give you a huge amount of control and will very likely result in better results and less afterwork than using curved-splines and bias-handlers. The trick is to use the bias-handlers in a way, that the resulting direction of both of them will become a tangent for the point you are changing the biases for. That will result in a smooth, but still mechanical look.

 

Of course if you bevel the edges and then make the beveled areas smooth (if this is look is what you are looking for) can create very neat looking edges.

 

Hope this leads in the right direction.

 

See you

*Fuchur*

 

Thanks for the feedback!

 

Ack! I never even thought of that, the other dead-end splines in the window. So, I failed my self-test in other words. And here I thought I'd gotten it figured out. Well that's a downer. I guess you can't take anything for granted when it comes to using splines. Even in the simplest tasks, something's always going to pose a potential problem.

 

I'm not able to envision what you mean by beveling getting rid of the dead-end spline though? Is there an example you can point to to demonstrate that?

 

As for the curving I actually went back into A:M after posting that and tried it out for myself. The idea of working with the CP handles occurred to me literally while I was typing that post, so I hadn't actually tried it yet lol.

 

Here's the results. Could use a bit of tweaking, but it's pretty much what I'm after.

[Fuchur]

The other question about the smoother edges, etc.:

There a several approaches:

1.) Bevel the edges.

2.) Use the bias-handlers to make it smooth.

3.) Do both and get the best of both.

 

For mechnical modelling I would recommend to use peaked splines and AFTER peaking, change the bias-handlers. This will give you a huge amount of control and will very likely result in better results and less afterwork than using curved-splines and bias-handlers. The trick is to use the bias-handlers in a way, that the resulting direction of both of them will become a tangent for the point you are changing the biases for. That will result in a smooth, but still mechanical look.

 

Of course if you bevel the edges and then make the beveled areas smooth (if this is look is what you are looking for) can create very neat looking edges.

 

Hope this leads in the right direction.

 

See you

*Fuchur*

 

Thanks for the feedback!

 

Ack! I never even thought of that, the other dead-end splines in the window. So, I failed my self-test in other words. And here I thought I'd gotten it figured out. Well that's a downer. I guess you can't take anything for granted when it comes to using splines. Even in the simplest tasks, something's always going to pose a potential problem.

 

I'm not able to envision what you mean by beveling getting rid of the dead-end spline though? Is there an example you can point to to demonstrate that?

 

As for the curving I actually went back into A:M after posting that and tried it out for myself. The idea of working with the CP handles occurred to me literally while I was typing that post, so I hadn't actually tried it yet lol.

 

Here's the results. Could use a bit of tweaking, but it's pretty much what I'm after.

 

I would say, just try it out That will help you learning the most.

If it is to hard, could you upload your model?

I'll show you.

 

See you

*Fuchur*

[robcat2075]

This sort of machined-object surface is challenging in A:M. Your solution with the dead-end spline will probably be fine in most cases since the patches are sitting on flat planes, especially if this is a small prop.

 

Here's a topology that bevels the edge of the indent and avoids a dead-end spline. Much bias fiddling would be needed to make it finished looking.

 

 

 

Forum member Rob_T posed a similar problem. You can see my suggested solution here...

 

http://www.hash.com/forums/index.php?s=&am...st&p=367635

[Fuchur]

This sort of machined-object surface is challenging in A:M. Your solution with the dead-end spline will probably be fine in most cases since the patches are sitting on flat planes, especially if this is a small prop.

 

Here's a topology that bevels the edge of the indent and avoids a dead-end spline. Much bias fiddling would be needed to make it finished looking.

 

 

 

Forum member Rob_T posed a similar problem. You can see my suggested solution here...

 

http://www.hash.com/forums/index.php?s=&am...st&p=367635

 

I would bevel the inner part too. Like that you avoid that large 3-point-patch you've got there robert.

 

See you

*Fuchur*

[MikeV]

Hey guys.. Thanks for the replies and help!

 

I went and tried to replicate the beveled edge like Robert shows in his picture, but by beveling the outside and inside of the area as Fuchur suggested and, after 15 minutes of fussing with CPs and lines and going cross-eyed in the process... forget it lol. That's *wayyyyy* more trouble than it's worth.

 

At least I have an idea of how to tackle it in a future project, that is if I can't avoid having to do it altogether.

 

Thanks again for the help guys. It's appreciated.

 

 

Edit: Said I was going to take a break 'til tomorrow. Didn't happen :-p

[robcat2075]

What is that thing you are making at the top of the thread, anyway?

[MikeV]

What is that thing you are making at the top of the thread, anyway?

 

It's basically supposed to be a "market stand". It was one of 3 different sizes I was creating for a prototype of a game I was working on before deciding game design isn't really what I wanted to do.

 

I figured 'cause I know what the end result is supposed to look like, and I can do a literal 1-to-1 comparison of the two, it would be a good exercise of my own to learn A:M.

 

It's turned out to be a huge headache, though and I'm wondering if it's worth doing it at all.

 

Even though I said I was done for the weekend, being a stubborn person who hates leaving things remain unresolved/unfinished, or letting things "beat me", I decided to give it another go.

 

I was going along nicely, until I hit a 5 point patch that just will not cooperate. I've tested the splines out and fixed one. Every one of them now is continuous. No dead-ends. No 3-splines into one CP. Every spline forming the 5 sides is its own continuous spline. In fact, A:M would actually highlight the "Make 5 Point Patch" button, but nothing happened when I would press it. That happened a few times. Now it won't light up anymore, no matter what I do. Click selecting, lasso selecting, selecting and pressing "." twice.. Nothing. I've closed the program and re-opened it. Nothing.

 

So... Twinkies and gratitude to who ever can help me solve this one.

 

Anyway.. here's the model. I'm beginning to feel like a "special needs" case here, but if someone wants to take a look at it and can tell me what I'm doing wrong this time, that'd be awesome. Will be nice when I can complete something on my own without having to run to the forums for help.

 

I think I'm gonna go kill a bunch of stuff in Torchlight now or something lol.

MarketStand.prj

[robcat2075]

Hey that's looking real good!

 

I was going along nicely, until I hit a 5 point patch that just will not cooperate.

 

I'm stumped on that one. Do any of our spline gurus have any theories?

[mtpeak2]

I was able to fix it by deleting the vertical spline on the right side of the 5 pointer, breaking the splines, extruding the spline on the right (to the position where you deleted the spline) and then reattaching. After doing that, I was able to create the 5 point patch.

 

I don't know why you couldn't create the five pointer with what you had.

[MikeV]

I was able to fix it by deleting the vertical spline on the right side of the 5 pointer, breaking the splines, extruding the spline on the right (to the position where you deleted the spline) and then reattaching. After doing that, I was able to create the 5 point patch.

 

I don't know why you couldn't create the five pointer with what you had.

 

Me, either.

 

Robert noting that it had him stumped made me feel better about it at least. Like, "oh so actually *isn't* me this time?" lol.

 

It's stranger still that for a little bit, it *would* light up the 5-Point Patch button. It just wouldn't do anything when I clicked it. It would gray out and I'd have to select it again. After a few times it wouldn't even light up anymore.

 

Your fix is pretty much what I did prior to it working the time before that.

 

See, at first it didn't work because I had a "broken spline". It wasn't dead-ended, but it ended at one CP and a new one started from the same CP (if that makes sense). I detached and deleted the points, then reattached them to make it one continuous spline. And then it worked.

 

I then did an extrude operation on the "inset" part to bring it in so I could create the flat portion of it. After I did that extrude is when that 5-point patch disappeared and all the trouble started. Oddly enough, the opposite 5PP on the bottom of that section wasn't affected.

 

I'll give what you did a try and see what happens.

 

Thanks!

[MikeV]

Hmm.. Tried what you suggested and it didn't work. Still won't let me make a 5 Point Patch there.

 

Maybe something really bad will happen if I succeed and A:M is simply trying to protect my life.

 

 

 

 

It could happen!

[mtpeak2]

Try this, break the top horizontal spline of the 5 pointer, then reattach.

[robcat2075]

Ah.... HA!

 

ProblemCP.mov

[mtpeak2]

If I break the top spline of the 5 pointer and reattach, I can make a 5 point patch. I have no issue with the rogue cp.

[robcat2075]

If I break the top spline of the 5 pointer and reattach, I can make a 5 point patch. I have no issue with the rogue cp.

 

You must be doing it in a vary particular way.

 

If I do that, the Five-pointer button will turn on but it still doesn't make the patch fill in.

[mtpeak2]

It works fine here. Did it multiple times reverting the project, in various ways.

[MikeV]

Ah.... HA!

 

ProblemCP.mov

 

Ha! Something you've seen maybe only once before, and I managed to recreate it. Go figure.

 

I seem to have a knack for finding bizarro issues. I did QA at a software development company for a time and the programmers there hated me 'cause I would find these obscure bugs that they could never find or reproduce themselves. However rare they were, because they *could* happen, the programmers had to address them. And so they hated me.

 

Anyway!

 

As far as how it was modeled, I basically just made a cross-section and extruded it out up to the point where the inset begins, which is one spline after the one with rogue CP.

 

Now, after that I may have made changes that affected that particular CP/location, but I couldn't begin to tell you what it was or at what point or what I was trying to do specifically in the process.

 

Regardless.. that is crazy.

 

Let me see if I can fix it myself now.

[MikeV]

And voila... found the rogue CP, deleted it and now everything's fine. How crazy.

 

Rob and mtpeak, thank you both for taking a look at that!

 

Though now I'm curious why mtpeak is able to make it work even with the rogue CP there. That's interesting.

 

I promise I'll try not to make a habit of finding oddball issues like that lol.

 

Unless you want me to that is :-p.

 

Thanks again! Now I can actually continue on and finish the thing!

[mtpeak2]

It's this cp that is causing the issue. If I break either of these two splines and reattach, I can create a 5 point patch. When Robert deleted the spline, he fixed the problem.

[MikeV]

It's this cp that is causing the issue. If I break either of these two splines and reattach, I can create a 5 point patch. When Robert deleted the spline, he fixed the problem.

 

Hmm... I re-downloaded the broken project, deleted that CP and rebuilt it. The 5 point patch button lit up, but no patch was made when I clicked it.

 

So I opened the file again, with the issue still there, deleted that CP again and rebuilt it, and the second time, it worked on the first try.

 

So... it seems to be tempermental or something.

 

Anyway! Thanks again for taking a look at it and helping to track down the issue... and enjoy your Twinkies!

[Rodney]

I'm late to the party but... speaking of spline layout, your modeling of this object reminded me of an old tutorial by Rodger Reynolds:

 

http://www.hash.com/r_reynolds/surf_tute/surf_tute.htm

[MikeV]

I'm late to the party but... speaking of spline layout, your modeling of this object reminded me of an old tutorial by Rodger Reynolds:

 

http://www.hash.com/r_reynolds/surf_tute/surf_tute.htm

 

Wow. Yet another awesome resource for A:M.

 

It continues to amaze me how much helpful information there is scattered about these forums and the website in general, in terms of videos, articles and people sharing their knowledge/expertise.

 

I need to set up a special Bookmarks folder specifically for this stuff, as well as download all the helpful videos I've been referred to and otherwise found. This way I can be a bit more "self-helping" and won't have to ask others to use their own time helping me with something that there's already a perfect resource available for. As much as I appreciate the help I receive here, I'm also eager to "take off the training wheels" and start solving and working through things on my own, or being able to recognize issues that other newer users encounter. Will be a great moment when I realize I've reached that point. Right now I'm in that phase of not always knowing what questions I should be asking, or what I should be looking for help with.

 

But anyway, I'm rambling.

 

Thanks for sharing that!

[largento]

It's this cp that is causing the issue. If I break either of these two splines and reattach, I can create a 5 point patch. When Robert deleted the spline, he fixed the problem.

 

Hmm... I re-downloaded the broken project, deleted that CP and rebuilt it. The 5 point patch button lit up, but no patch was made when I clicked it.

 

So I opened the file again, with the issue still there, deleted that CP again and rebuilt it, and the second time, it worked on the first try.

 

So... it seems to be tempermental or something.

 

Anyway! Thanks again for taking a look at it and helping to track down the issue... and enjoy your Twinkies!

 

Are your normals facing the right direction?

[NancyGormezano]

See, at first it didn't work because I had a "broken spline". It wasn't dead-ended, but it ended at one CP and a new one started from the same CP (if that makes sense). I detached and deleted the points, then reattached them to make it one continuous spline. And then it worked.

 

I then did an extrude operation on the "inset" part to bring it in so I could create the flat portion of it. After I did that extrude is when that 5-point patch disappeared and all the trouble started. Oddly enough, the opposite 5PP on the bottom of that section wasn't affected.

 

Glad to see that the phantom CP has been identified, but I am really curious as to how you are getting these.

 

There seems to be some sequence of steps you are doing (not your fault...it's probably a bug). It would be good to identify what causes this - as I remember, you had some other weird model thing going on that could only be solved by copy/paste into a new model (or at least I think it was you, maybe I'm thinking of someone else). So I'm suspecting it's something in your process.

 

(EDIT: yes, it was you. See this post about the funny coffee cup.)

 

I looked at your project. I wonder why you have splines that are not continuous when you are "extruding"? Are you peaking everything? (why?). Are you connecting splines using the shift key? Are you messing with biases (maybe too early in the process)?

 

I am not a mechanical modeler, but these are very odd problems, and perhaps you are using some keying/modeling sequence that might fit for blender, but perhaps is something that showcases a bug in A:M?

 

And don't beat yourself up about asking questions here...WE LOVE IT! It helps everyone to learn something. Even those who have been here since the last Ice Age. It helps us all to shake out bugs, learn new methods. And please try to remember - those TAOAM exercises were constructed pre-Ice Age, and the software has evolved, changed, and what worked then (for the most part, sorta), might not still be true now (less sorta).

Edited July 30, 2012 by NancyGormezano

[MikeV]

Glad to see that the phantom CP has been identified, but I am really curious as to how you are getting these.

 

That seems to remain the mystery lol. It's the first and only time I've gotten that particular problem. If it happens again, though, then I'll try and keep better track of the process I used up to that point.

 

There seems to be some sequence of steps you are doing (not your fault...it's probably a bug). It would be good to identify what causes this - as I remember, you had some other weird model thing going on that could only be solved by copy/paste into a new model (or at least I think it was you, maybe I'm thinking of someone else). So I'm suspecting it's something in your process.

 

(EDIT: yes, it was you. See this post about the funny coffee cup.)

 

Ahh yes, the coffee cup with the wacked-out handle that copy and paste "cured". That was another odd one lol.

 

As for my process, I don't remember all the steps 'cause I was just working along 'til that 5-point patch issue happened.

 

I can give you a basic 10,000 foot overview.

 

There were two shapes, or cross-sections to be more accurate, to start with. One being the solid piece, which I'll call "Base Shape", and the part with the recess, which I'll refer to as "Recessed Shape".

 

I extruded out the Base Shape up to the point where it would connect to the Recessed Shape.

 

Then I split the right-most middle spline Base Shape and lined the CPs up with where they'd need to be to line up with the Recessed Shape. I then lined up the Recessed Shape with the Base Shape and connected those. Then I extruded from that point on out to the end of it. This is all in context of working left to right, using the "Right" viewport (3 on the keypad).

 

At this point, I had all the CPs peaked as I was originally going to have hard edges. You can see the result of that in my first post with the 3 different spline layouts.

 

After that, I decided I wanted it to be more curved, since I'm working with patches and not polygons, I figured I could make it look a lot nicer, less "chunky" and a bit more elegant, by bringing curves into it.

 

It had also been pointed out that I had a dead-end spline running into the recessed area. So I reworked it to try and get bevels into it, which was frustrating at first, but I eventually got to work, more or less. I still have to do another beveled section for the inside of the recessed area.

 

Everything was great up to the point where I extruded the spline for the recessed area in to give it depth. That's when the 5 point patch disappeared and that whole weirdness started.

 

There were certainly smaller in-between steps in there, but I couldn't tell you what they were. Hopefully that wasn't too confusing lol.

 

I looked at your project. I wonder why you have splines that are not continuous when you are "extruding"? Are you peaking everything? (why?). Are you connecting splines using the shift key? Are you messing with biases (maybe too early in the process)?

 

I am not a mechanical modeler, but these are very odd problems, and perhaps you are using some keying/modeling sequence that might fit for blender, but perhaps is something that showcases a bug in A:M?

 

Well, I peaked everything because initially it was going to be a harder-edged shape. Then, when I remembered I wasn't working with polys, I decided it would look nicer to have it more rounded. So, I started with peaked edges, and then went back and adjusted the CP handles to get them more round. I think someone earlier in the thread suggested it that way, actually, rather than just pressing "o" and letting A:M do it automatically. I did find using the CP handles gave me much more direct control over it. Though, I'm guessing that might not be the best way to go about it?

 

 

And don't beat yourself up about asking questions here...WE LOVE IT! It helps everyone to learn something. Even those who have been here since the last Ice Age. It helps us all to shake out bugs, learn new methods. And please try to remember - those TAOAM exercises were constructed pre-Ice Age, and the software has evolved, changed, and what worked then (for the most part, sorta), might not still be true now (less sorta).

 

 

Well great! 'cause I can tend to be full of them! And I enjoy talking to people about anything that's a shared interest ( in case you haven't noticed :-p). You should hear me when I get going on games or game design.

 

It's good to know about the TAoAM exercises... though now it makes me wonder. Am I well served by doing those actual exercises, then? Should I maybe be taking the information provided (regarding animation and such) and apply it to other tutorials outside that manual? I mean, if it's possible that what's in the book is no longer the correct or best way to go about doing something, perhaps I shouldn't be going about it the way it does?

[MikeV]

It's this cp that is causing the issue. If I break either of these two splines and reattach, I can create a 5 point patch. When Robert deleted the spline, he fixed the problem.

 

Hmm... I re-downloaded the broken project, deleted that CP and rebuilt it. The 5 point patch button lit up, but no patch was made when I clicked it.

 

So I opened the file again, with the issue still there, deleted that CP again and rebuilt it, and the second time, it worked on the first try.

 

So... it seems to be tempermental or something.

 

Anyway! Thanks again for taking a look at it and helping to track down the issue... and enjoy your Twinkies!

 

Are your normals facing the right direction?

 

I checked and, at least on the fixed model, they are.

 

I'd have to check to see what the case was with the "broken" one.

[robcat2075]

It's good to know about the TAoAM exercises... though now it makes me wonder. Am I well served by doing those actual exercises, then?

 

the modeling ones are all good and fundamental. They are a good start.

 

The animation ones are a bit basic and do not exploit the timeline as I would like to see, but no harm is done by doing the ones that are there.

 

My own hope is to make a better explanation of rigging than "Show some backbone" but it does show one method to get a rig in a character.

 

the ones on materials, cloth, particles are niche interests but they are decent brief introductions to huge topics.

[NancyGormezano]

It's good to know about the TAoAM exercises... though now it makes me wonder. Am I well served by doing those actual exercises, then? Should I maybe be taking the information provided (regarding animation and such) and apply it to other tutorials outside that manual? I mean, if it's possible that what's in the book is no longer the correct or best way to go about doing something, perhaps I shouldn't be going about it the way it does?

 

I think the TAOAM is a good start, but they are not the endall. They introduce new user to one of many ways to do something, but not necessarily the only way, nor are they very deep. Just don't take them as gospel, never to be questioned. Some of the interface(s) has/have changed. Some of the models are faulty. New users run into the quirks, get frustrated. But the exercises are worth looking at, trying, and then asking questions, and moving on to other tutorials, examples. And yes it's hard to find the info here on forum doing a search. That's why questions are always good, at whatever level you're at.

 

 

There were two shapes, or cross-sections to be more accurate, to start with. One being the solid piece, which I'll call "Base Shape", and the part with the recess, which I'll refer to as "Recessed Shape".

 

I extruded out the Base Shape up to the point where it would connect to the Recessed Shape.

 

Then I split the right-most middle spline Base Shape and lined the CPs up with where they'd need to be to line up with the Recessed Shape. I then lined up the Recessed Shape with the Base Shape and connected those. Then I extruded from that point on out to the end of it. This is all in context of working left to right, using the "Right" viewport (3 on the keypad).

 

At this point, I had all the CPs peaked as I was originally going to have hard edges. You can see the result of that in my first post with the 3 different spline layouts.

 

After that, I decided I wanted it to be more curved, since I'm working with patches and not polygons, I figured I could make it look a lot nicer, less "chunky" and a bit more elegant, by bringing curves into it.

 

It had also been pointed out that I had a dead-end spline running into the recessed area. So I reworked it to try and get bevels into it, which was frustrating at first, but I eventually got to work, more or less. I still have to do another beveled section for the inside of the recessed area.

 

Everything was great up to the point where I extruded the spline for the recessed area in to give it depth. That's when the 5 point patch disappeared and that whole weirdness started.

 

There were certainly smaller in-between steps in there, but I couldn't tell you what they were. Hopefully that wasn't too confusing lol.

 

Well, I peaked everything because initially it was going to be a harder-edged shape. Then, when I remembered I wasn't working with polys, I decided it would look nicer to have it more rounded. So, I started with peaked edges, and then went back and adjusted the CP handles to get them more round. I think someone earlier in the thread suggested it that way, actually, rather than just pressing "o" and letting A:M do it automatically. I did find using the CP handles gave me much more direct control over it. Though, I'm guessing that might not be the best way to go about it?

 

Hmmm...I don't have a real answer other than I too have noticed quirky behavior when switching between peaked and not peaked cps, and with extrusion. As I said before, I'm not a mechanical modeler. I do squishy characters, and am not that picky about exactness. I traditionally work only with continuous curves (not peaked). If I need something to look sharper or rounder, smoother, I will add splines, or work with bias handles (less usual).

 

When I took your model and then hit the "curve switch" I get a very funny looking splines. When I hit the peaked switch again, I get another funny model, different from your original. Me no likee.

More importantly, I also notice that if I take your left most spline that you used for extrusion, that it is NOT a continuous spline - so I will guess you peaked that original spline from which you extruded - not a good idea, I don't think.

 

As an alternative, or something to add to toolbox, I find I have more luck with getting "protrusions" or "indents" on/in a surface when I think of the surface as a grid (after extruding with continuous spline). Outline the cross section or shape to be protruded on the grid/surface with a continuous spline ring, thus creating 3 and 5 point areas. Then select all of the inner area and move it in 3D space out or in from the surface. To refine the indented shape, or protrusion, stitch in more spline rings. Then select all those 5 point areas (when everything is happy), and green donut your way to nirvana.

malomethod.mdl

Edited July 30, 2012 by NancyGormezano

[MikeV]

It's good to know about the TAoAM exercises... though now it makes me wonder. Am I well served by doing those actual exercises, then? Should I maybe be taking the information provided (regarding animation and such) and apply it to other tutorials outside that manual? I mean, if it's possible that what's in the book is no longer the correct or best way to go about doing something, perhaps I shouldn't be going about it the way it does?

 

I think the TAOAM is a good start, but they are not the endall. They introduce new user to one of many ways to do something, but not necessarily the only way, nor are they very deep. Just don't take them as gospel, never to be questioned. Some of the interface(s) has/have changed. Some of the models are faulty. New users run into the quirks, get frustrated. But the exercises are worth looking at, trying, and then asking questions, and moving on to other tutorials, examples. And yes it's hard to find the info here on forum doing a search. That's why questions are always good, at whatever level you're at.

 

 

There were two shapes, or cross-sections to be more accurate, to start with. One being the solid piece, which I'll call "Base Shape", and the part with the recess, which I'll refer to as "Recessed Shape".

 

I extruded out the Base Shape up to the point where it would connect to the Recessed Shape.

 

Then I split the right-most middle spline Base Shape and lined the CPs up with where they'd need to be to line up with the Recessed Shape. I then lined up the Recessed Shape with the Base Shape and connected those. Then I extruded from that point on out to the end of it. This is all in context of working left to right, using the "Right" viewport (3 on the keypad).

 

At this point, I had all the CPs peaked as I was originally going to have hard edges. You can see the result of that in my first post with the 3 different spline layouts.

 

After that, I decided I wanted it to be more curved, since I'm working with patches and not polygons, I figured I could make it look a lot nicer, less "chunky" and a bit more elegant, by bringing curves into it.

 

It had also been pointed out that I had a dead-end spline running into the recessed area. So I reworked it to try and get bevels into it, which was frustrating at first, but I eventually got to work, more or less. I still have to do another beveled section for the inside of the recessed area.

 

Everything was great up to the point where I extruded the spline for the recessed area in to give it depth. That's when the 5 point patch disappeared and that whole weirdness started.

 

There were certainly smaller in-between steps in there, but I couldn't tell you what they were. Hopefully that wasn't too confusing lol.

 

Well, I peaked everything because initially it was going to be a harder-edged shape. Then, when I remembered I wasn't working with polys, I decided it would look nicer to have it more rounded. So, I started with peaked edges, and then went back and adjusted the CP handles to get them more round. I think someone earlier in the thread suggested it that way, actually, rather than just pressing "o" and letting A:M do it automatically. I did find using the CP handles gave me much more direct control over it. Though, I'm guessing that might not be the best way to go about it?

 

Hmmm...I don't have a real answer other than I too have noticed quirky behavior when switching between peaked and not peaked cps, and with extrusion. As I said before, I'm not a mechanical modeler. I do squishy characters, and am not that picky about exactness. I traditionally work only with continuous curves (not peaked). If I need something to look sharper or rounder, smoother, I will add splines, or work with bias handles (less usual).

 

When I took your model and then hit the "curve switch" I get a very funny looking splines. When I hit the peaked switch again, I get another funny model, different from your original. Me no likee.

More importantly, I also notice that if I take your left most spline that you used for extrusion, that it is NOT a continuous spline - so I will guess you peaked that original spline from which you extruded - not a good idea, I don't think.

 

As an alternative, or something to add to toolbox, I find I have more luck with getting "protrusions" or "indents" on/in a surface when I think of the surface as a grid (after extruding with continuous spline). Outline the cross section or shape to be protruded on the grid/surface with a continuous spline ring, thus creating 3 and 5 point areas. Then select all of the inner area and move it in 3D space out or in from the surface. To refine the indented shape, or protrusion, stitch in more spline rings. Then select all those 5 point areas (when everything is happy), and green donut your way to nirvana.

 

Yeah, I think I'm gonna have to just start over with that thing, if I continue to work on it at all.

I hate to say this, but I'm beginning to really feel that maybe A:M and patch modeling is just beyond me or something and I'm barking up the wrong tree trying to learn it. I'm going on a week at least now since I started learning and I feel like I haven't learned or absorbed anything. I figured recreating that market stand would be within my reach, and a good self-assigned project to put what I've learned to practice, to test myself and create something that doesn't come specifically from a pre-written tutorial or exercise. I figured, that beam took me maybe 10 minutes to model and refine in Blender. How difficult could it possibly be with splines and patches? Well, I think I've gotten my answer lol.

 

Apparently I haven't learned anything, 'cause darned if I'm not making mistakes left and right.. most of which I'm not even aware of. All the issues people are pointing out to me about it is really putting it in stark contrast that I really have no clue what I'm doing.

 

Is it *supposed* to be this difficult to learn to use A:M and patches? Am I expecting too much of myself, or am I just not wired for it?

[Rodney]

darned if I'm not making mistakes left and right

 

As the folks at PIXAR would say (Peter Doctor in particular), "Make mistakes faster!"

 

You are doing just fine.

 

It should be noted that you can get away with a lot of splineage that isn't optimal.

What you are doing now however is learning 'optimal'.

This will pay off big dividends in the end.

[itsjustme]

Is it *supposed* to be this difficult to learn to use A:M and patches? Am I expecting too much of myself, or am I just not wired for it?

 

You just have to stop trying to model like it is polygons, Mike...it's a tough transition for some. It takes a little more thinking...I approach mechanical modeling and organic modeling totally differently.

 

For mechanical modeling, Rodger Reynold's tutorials explained it best for me. The biggest eye-opener would be the Modeling Edges and Corners tutorial.

 

Hope that helps.

[Vertexspline]

hey David --thanks for those links -----I struggle with modeling and I have not seen these two ---so thanks-----(you see when you post a helpful hint for someone in the thread --you never know how many folks you may actually help.

[Vertexspline]

MikeV-----do not get down on yourself --------you really are doing great for a weeks go at it. Modeling with patches has some hurdles and obstacles that takes some time to work out - I got frustrated for awhile myself and decided to give it a rest and took some months off but in hindsight I think i wasted some time so ----its better to just keep plowing along -----practicing and practicing......I understand the frustration and have felt like I can't do this as well...but based on what I have seen you do in a week........You Can DO This !!

[MikeV]

Is it *supposed* to be this difficult to learn to use A:M and patches? Am I expecting too much of myself, or am I just not wired for it?

 

You just have to stop trying to model like it is polygons, Mike...it's a tough transition for some. It takes a little more thinking...I approach mechanical modeling and organic modeling totally differently.

 

For mechanical modeling, Rodger Reynold's tutorials explained it best for me. The biggest eye-opener would be the Modeling Edges and Corners tutorial.

 

Hope that helps.

 

Appreciate those links

 

I don't think it's that I'm trying to model like polygons. I'm aware of the differences between poly and patch modeling. Polygons can be modeled a bit more "arbitrarily", patches/splines require more planning and care in their layout. That part I'm fine with and I realize I have to work on.

 

What's getting to me is that there seems to be all these sorta secondary and tertiary rules that are just as important, but are, seemingly, extremely situational and not at all obvious or "tell-tale" in any way.. at least to me.

 

Nancy ran down a long list of questions earlier in the thread that really made that hit home to me. I also notice that in different tutorials, discussions and such, it seems to make all the difference in how you create a new spline, how you connect a new spline, how you continue an existing spline, how you add a new CP to an existing spline, do you press Shift when connecting or not, how you delete or detach a new spline, what direction is the spline going in, how is it flowing through the CPs, when to use peaked, when to use curved, when to set CPs to peaked before adjusting their curves with the handles, when to use the CP handles, when not to use CP handles, when is something mandatory, when is it optional... etc. etc. All these "little things" that seem to have potentially big impacts on the resulting models.

 

It's just so much to take in. Everything seems to have several additional "sub-rules" associated with it, and it seems that if you're not completely versed and understanding of all of them, and all situations that each of them applies in, that you're bound to make mistakes that may not even show up 'til several steps after you've made them - such as with my beam project here.

 

I'm getting feedback about mistakes I'm making that I never even knew you *could* make to begin with.

 

I'm honestly and without sarcasm amazed and all the more impressed at how people are able to grasp and master it all as well as they do. That's what makes me wonder if it's just beyond me. I mean, I consider myself a fairly intelligent person, capable of learning new things... but this spline/patch thing making my head spin.

[robcat2075]

Is it *supposed* to be this difficult to learn to use A:M and patches? Am I expecting too much of myself, or am I just not wired for it?

 

 

 

 

 

Well, I peaked everything because initially it was going to be a harder-edged shape. Then, when I remembered I wasn't working with polys...

 

There's the core misstep at the outset... approaching splining as if it is variation of polygon modeling. A very common misstep among people who have been using polygon modelers but it's still a misstep. If you start out wrong it WILL BE HARD to get right again.

 

No, A:M is NOT intended as the fastest modeler of angular mechanical models, although fine work is done in that area.

 

A:M IS intended as a great platform for modeling, rigging and animating characters. That's where the thin meshes, the splines and patches become advantageous. That said, I feel comfortable making any mechanical thing I need also.

 

Good splining IS much easier if you think through the shape of your object at the start and use strategy in placing your splines. That pre-thinking WILL be harder for some people.

 

For me, i see the splines i need as soon as I see the object. That comes with practice.

 

The sign post is not a bad first project. I'm sure it is taking longer than Blender but you're not a novice in Blender. How long did the first model in Blender take? And will you ever be able to model and rig a character in Blender?

 

Really the only serious bump in the sign project was the 5-pointer that wouldn't close and we solved that one. And many lessons learned along the way.

[MikeV]

Is it *supposed* to be this difficult to learn to use A:M and patches? Am I expecting too much of myself, or am I just not wired for it?

 

 

 

 

 

Well, I peaked everything because initially it was going to be a harder-edged shape. Then, when I remembered I wasn't working with polys...

 

There's the core misstep at the outset... approaching splining as if it is variation of polygon modeling. A very common misstep among people who have been using polygon modelers but it's still a misstep. If you start out wrong it WILL BE HARD to get right again.

 

I may have phrased that bit poorly.

 

I don't mean I realized I started out approaching it as though it were a polygon model and then realized "oops, I'm using patches" and then changed my approach.

 

What I meant is that I was making it with sharp edges because I was trying to recreate it as it looks in Blender. I created it that way in Blender because it was intended to be a low-polygon model to be used in a game engine, and so I had to do it that way to keep performance up when the final model was placed in a real-time environment setting among a lot of other geometry.

 

Once I remembered that I wasn't doing it for a "low poly" or "low resolution" application, and didn't have to worry about a final polygon count, that I could go ahead and let A:M do what it does well, which is handling curved surfaces.

 

The statement had to do with the final look I was going for, not the approach I was taking to it.

 

I just didn't phrase that bit very well

[jakerupert]

Hi MikeV,

 

 

Jts all about spline continuity. You can always check edgeloops hiting the spline and the comma key.

If something really won'T work, don't break your head finding out why, simply break splines

And delete cps til your reach your goal.

You can live well without ever touching the biashandles.

(there was a time, when they had some disadvantages so i never used them ever since)

 

AM is sometimes a bit sturdy , the 5 pointers always have been, but you get used to it.

(peaking is always last step ,because it will make it harder to judge splinecontinuity.)

[MikeV]

Hi MikeV,

 

 

Jts all about spline continuity. You can always check edgeloops hiting the spline and the comma key.

If something really won'T work, don't break your head finding out why, simply break splines

And delete cps til your reach your goal.

You can live well without ever touching the biashandles.

(there was a time, when they had some disadvantages so i never used them ever since)

 

AM is sometimes a bit sturdy , the 5 pointers always have been, but you get used to it.

(peaking is always last step ,because it will make it harder to judge splinecontinuity.)

 

Hey there. Thanks for the feedback/info.

 

You say it's all about spline continuity, and that was my initial perspective. But that doesn't really seem to be the case when you get down to it, because there are so many little individual things that can alter a splines continuity without someone (especially a new user) even realizing it 'til much later.

 

With the model in this thread, I had been checking spline continuity and such to make sure nothing was dead-ending, etc. And nothing was. At one point, even Robcat was stumped because, I presume, he couldn't see any reason why it wouldn't work. It turns out there was a "rogue CP" that he found that seemed to be the problem. But then there was a CP that another user noticed that seemed to be the problem. So, in that case, even having proper spline continuity didn't guarantee anything.

 

Also the thing with peaking the CPs. That part kind of confused me, because at one point - maybe in this thread, perhaps elsewhere - someone suggested peaking the CPs and then using the bias handles to get the curve I wanted. But then later, I think it was Nancy, questioned why I was using the bias handles so early in the modeling. That brings to light another "secondary or tertiary rule" that I seem to have slammed into face-first without even knowing it existed. I had no idea there was a "right" or "wrong" time to use bias handles. To me it was a matter of tweaking the splines at a given CP to get the shape you want. Apparently it's more involved than that.

 

And that, on balance, is what's really throwing me off with A:M. All those hidden rules that seem to go along with, seemingly, everything you can possibly do in the program. I realize for someone more experienced, all of those things are second nature and completely instinctual. I'm not saying "A:M sucks. It's all the program's fault". Again, I've seen what the program is capable of and what it can do. I'm just saying that, to someone just learning the program and trying to come to grips with the absolute basics, it's more than a little overwhelming. And, apparently, the basic tutorials I've been following have not been doing a thorough job of pointing all that out.

 

Speaking of tutorials, another thing that's struck me is how someone earlier in the thread noted that the tutorials in the TaoAM are a bit outdated and not really up-to-date with how the program works now, or with the best practices of how to approach different specific tasks. But then on the other hand, what is the first thing people are told to do when starting out with A:M? To go through the TaoAM book. I mean, think about that.

 

I don't know. I think A:M is a solid and very cool program that can do some amazing stuff. I think it's the structure of the tutorials and such that need to be be revisited, perhaps. Especially if much of it is outdated, using approaches that are no longer valid or ideal in newer versions of the software.

 

Just my two cents.

 

I think I'm gonna take some time away from this program and maybe come back with a fresh brain. We'll see how I do.

 

Thanks again for all the feedback and help. Despite my frustration with learning A:M at the moment, it is appreciated.

[Fuchur]

The problem is, that it can not be said which way of modelling you like and which is "the best".

We all use slightly different approaches, which in the end can result in very nice work. Which one suits you best is your own decission... and a hard one .

 

If I am going for mechanical models I have learned that with mechanical models it is better for me to peak the CPs and after that use the bias-handlers to get the curve right (mechanical models often need very straight parts and they can be very frustrating to create without peaking the CPs used, especially if you try to move the CPs after the Bias-Adjustment.

 

For organical modelling I would as good as never use peaked CPs. That can result in problems when trying to create very nice curves.

 

It is my opinion so... other people love to use non-peaked CPs for mechanical modelling. It is all about what you like to do.

This is something that you have to find out for yourself and I absolutley agree that this is a hard process and that you will sometime be desperate about it.

 

For beginners it is said, that using no bias-handlers is better at the beginning, because they can be hard to overview and handle.

This is totally fine for organic models (since A:M is quite aimed at character-animation this is okay too) but can be problematic for mechanical once.

 

The best way to handle that is: Choose one and start with it. You will go a long way before you really run into disadvantages of the different methodes.

 

See you

*Fuchur*

[Gerry]

Hey MikeV, you're doing fine and asking good questions! The projects in TaoAM may not reflect a lot of newer functionality in the program, but they are still good practice and all will benefit the patient student. I still keep my v8 manual handy for reference and I use it all the time, even though the whole interface (and much else) has been overhauled since then, because the basic rules and tools are unchanged.

[robcat2075]

Shorter answer to your question: No, it's not supposed to be harder, we don't want it to be harder, but it is, a little bit, for this task.

 

 

We might ask the same if we were rigging a face with Blender. One of my Facebook acquaintances has been posting his progress on rigging a face he modeled. He's been going at it for weeks and can just get the mouth to open a bit satisfactorily. I can get that same thing going on an A:M face in an hour maybe:

 

mouth.mov

 

 

Blender is a lot harder for this task.

[Vertexspline]

I think that Mike's experience is a common one. Struggling with AM modeling is almost a given for all new folks. I struggle and feel his frustrations. He obviously wants to learn and make stuff . He obviously has talent but is being frustrated by those patch works rules and practices . And Mike brought up many good points of all the trickiness to getting it all to come out right.

 

And its not like it cannot be done --obviously great modeling is done by the works you can see here.

 

Besides the TAOAM book tutorials ....what really is needed are a point by point modeling tutorials by some of the most experienced modelers with AM .......going over all the planning , and decisions one needs to take. All the continuity strategies used making the models . All the break spline tricks, Add points , etc etc. Every step taken and why. and maybe all the bad habits new folks do ie. peaking to early or late etc ----an a few different models each with different approaches to being successful with etc. I almost think having some do not do this ...type of tutorials would even help many.

 

Heck ---I would even think many folks would pay for some good current tutorials.

 

However- I cannot stress how important --how amazingly valuable this kind of thing would be to growing the user base of AM users. And more important keeping them active and renewing year after year. And sure ---you can't make everyone a great modeler by having some great tutorials but-----you will not lose so many good modelers to frustration. And many less talented will become more talented.

 

Robert has made a few short tutorials of the likes needed ------but many more of a similar vein would be oh so helpful.

 

Robert made some good points how vs blender you can rig super fast with AM but modeling might take longer. But if folks never get to make that model right , never make one character they are happy with --they may never get to see the elegance and speed of AM's rigging.

 

I know resources are tight with HASH at this point in the life of AM --but if I owned this AP -----Working on new current and detailed tutorials would be very high on the to do list......That and t- shirts and hoodies !!!!!

 

Anyhow --just my two cents.

[MikeV]

Shorter answer to your question: No, it's not supposed to be harder, we don't want it to be harder, but it is, a little bit, for this task.

 

 

We might ask the same if we were rigging a face with Blender. One of my Facebook acquaintances has been posting his progress on rigging a face he modeled. He's been going at it for weeks and can just get the mouth to open a bit satisfactorily. I can get that same thing going on an A:M face in an hour maybe:

 

mouth.mov

 

 

Blender is a lot harder for this task.

 

Fair enough, though pointing to face rigging in Blender has nothing to do with modeling a simple beam in Animation Master. It certainly has nothing to do with the issues I'm encountering. I'm not trying to rig a face. I'm trying to model a fairly simple and straight-forward static object.

 

If you want to do a comparison with Blender, then at least keep the comparisons 1-to-1. As in, "modeling the same, simple object in each program".

 

And it's not "because polygons are superior to patches". So please don't take it as me implying that. I think they're both equally useful. Each has its benefits and its drawbacks. It's just that patches are more complicated in their use with more underlying rules to be aware of at every step of the process. Polygons are more free-form and forgiving and have less "baggage" involved in their usage.

 

The simple explanation for that would be that polygon modeling is a lot more.. actually much more straight forward than spline modeling. What you see is quite literally what you get. There are no hidden secondary and tertiary rules governing every single edge and vertex on the surface of a polygon mesh. You don't have to worry about which direction your edges are flowing, or if more than 2 edges meet at a given vertex. There's no such thing as "dead-ending". If you select two faces and then extrude/inset them into another shape, you're not automatically breaking any rules of "proper edge flow". You don't have to worry about internal faces because they don't occur unless you specifically and intentionally create them. The greatest concern you have in terms of the structure of a polygon mesh is "are the normals all facing the right way?", and a fix for that is typically a button or key press away.

 

So, to answer your question earlier in the thread about what I would have been able to do when I first started with Blender... Yes, I would have been able to model something like that - and more. Within the first few hours. No question. With any poly modeler, Blender or otherwise. Because vertices, edges and faces don't carry all the caveats and conditions that splines and patches do. It's constructed with three of the first and most basic tools you learn.. extrude, transform and scale. We're talking the Day One poly-modeling equivalent of creating an S out of a spline and lathing it into a vase in A:M.

 

I don't have to worry about having some mistake I made unknowingly suddenly coming to light an hour or two later when I'm well beyond that point and doing something on another part of the model.

[MikeV]

I think that Mike's experience is a common one. Struggling with AM modeling is almost a given for all new folks. I struggle and feel his frustrations. He obviously wants to learn and make stuff . He obviously has talent but is being frustrated by those patch works rules and practices . And Mike brought up many good points of all the trickiness to getting it all to come out right.

 

And its not like it cannot be done --obviously great modeling is done by the works you can see here.

 

Besides the TAOAM book tutorials ....what really is needed are a point by point modeling tutorials by some of the most experienced modelers with AM .......going over all the planning , and decisions one needs to take. All the continuity strategies used making the models . All the break spline tricks, Add points , etc etc. Every step taken and why. and maybe all the bad habits new folks do ie. peaking to early or late etc ----an a few different models each with different approaches to being successful with etc. I almost think having some do not do this ...type of tutorials would even help many.

 

Heck ---I would even think many folks would pay for some good current tutorials.

 

However- I cannot stress how important --how amazingly valuable this kind of thing would be to growing the user base of AM users. And more important keeping them active and renewing year after year. And sure ---you can't make everyone a great modeler by having some great tutorials but-----you will not lose so many good modelers to frustration. And many less talented will become more talented.

 

Robert has made a few short tutorials of the likes needed ------but many more of a similar vein would be oh so helpful.

 

Robert made some good points how vs blender you can rig super fast with AM but modeling might take longer. But if folks never get to make that model right , never make one character they are happy with --they may never get to see the elegance and speed of AM's rigging.

 

I know resources are tight with HASH at this point in the life of AM --but if I owned this AP -----Working on new current and detailed tutorials would be very high on the to do list......That and t- shirts and hoodies !!!!!

 

Anyhow --just my two cents.

 

Excellent post, Vertexspline. I think you sum it up perfectly.

 

A:M is fine. The way it works is fine. The results it can produce are awesome. None of this is debatable, in my opinion.

 

The thoroughness of the tutorials/excercises and the comprehensiveness of all the things that *really* go into spline/patch modeling, however, need an overhaul.

 

When someone, however new, is running into issues that they didn't even know could be issues to begin with, and it's resulting in the loss of hours, or even days of work... there's something very wrong.

[Gerry]

Besides the TAOAM book tutorials ....what really is needed are a point by point modeling tutorials by some of the most experienced modelers with AM .......going over all the planning , and decisions one needs to take. All the continuity strategies used making the models . All the break spline tricks, Add points , etc etc. Every step taken and why. and maybe all the bad habits new folks do ie. peaking to early or late etc ----an a few different models each with different approaches to being successful with etc. I almost think having some do not do this ...type of tutorials would even help many.

These exist already, in the form of Barry Zundel's tutorial series. You can buy them one at a time, for like five bucks. So you can stop *wishing* they exist. They already do.

[NancyGormezano]

Barry Zundel's tutorial series. You can buy them one at a time, for like five bucks.

here's the link

I haven't seen them, but many people here seem to feel these are quite good.

[Malo]

Hi,

I often read an opposition between the polygon's world and patches. Probably related to the policy of Martin Hash: "Say No to polygons".

Out of habit, we confuse polygons and polygons&Catmull-Clark subdivision.

This combination seems simple to use.

I say "seems" because I recommend reading this post:

http://blenderartists.org/forum/showthread...Poles-and-Loops

 

What I remember from this post have similarities with the patches:

Facesloop and loops are very similar to splines.

The Npoles are inside the 3-sided patches.

The Epoles are inside the 5-sided patches.

The triangles are banished (the patches contain only 4gones).

There is a preference for vertices that are connected to four edges, which is also reflected in the crossing of splines.

 

To return to the starting problem.

Old tutorials use the CP by Cp method (patch by patch) which has the désaventage to get errors such as those in this post.

I prefer the method "box modeling" (starting from a closed volum) ... or even better a mix between the two. Or to use polygonal modeler and Doosabin subdivision ... Clearly there is no one way to treat a model in AM.

 

What could have been informative and helpful to the forum, it would be to create posts with a very simple model of rotoscope (organic, props or Mechanical) in early post, where every persons, it deposits its tutorials to treat the rotoscope model.

[MikeV]

Besides the TAOAM book tutorials ....what really is needed are a point by point modeling tutorials by some of the most experienced modelers with AM .......going over all the planning , and decisions one needs to take. All the continuity strategies used making the models . All the break spline tricks, Add points , etc etc. Every step taken and why. and maybe all the bad habits new folks do ie. peaking to early or late etc ----an a few different models each with different approaches to being successful with etc. I almost think having some do not do this ...type of tutorials would even help many.

These exist already, in the form of Barry Zundel's tutorial series. You can buy them one at a time, for like five bucks. So you can stop *wishing* they exist. They already do.

 

It's great that they exist, and I've seen that video set mentioned with high regard a number of times.

 

My one problem with offering that as a solution is this...

 

If what he's demonstrating in those videos is critical to developing a comprehensive early understanding of patch modeling, of the behaviors and quirks of CPs and splines, and provides instruction/advice on how to properly approach modeling something, to start you off on the right foot, with those characteristics in mind, then they, or something equivalent, should be included with the software itself.

 

We're not talking specialized topics like how to model a character, or how to compose and render a scene well. We're talking nuts-and-bolts, "this is information that will carry over into everything you create with our software" basic information.

 

People buying your software, and then being told "now go spend more money on some videos that teach you the ins-and-outs of patches, splines and CPs and how to plan and work with them properly" seems more than a little unreasonable to me. I would fully, and reasonably, expect to be provided that information with the price of the software itself, especially if such info is as critical to proper modeling technique as it clearly is in this case.