After, instead of using subdivision you can also use multires modifier, that allows you to go top subdivision and keep the details, while you can use your low res for animation
Yes! This is a great workflow. And if you use the Multires modifier, you can then use that to bake all that detail into either a normal or displacement map. I cover those steps in another video: czcams.com/video/R2q7oZLs0GE/video.html
@@JamieDunbar true that getting a discplacement or normal map out of it can be useful when exporting your project to an other engine, but if you plan to stay on blender would you use the map or keep the model with the multires modifier ?
If you have a really detailed model the multires modifier can make your file size really large. So I’d usually bake out the maps. That said, Eevee can’t render displacement maps, so you may want to keep the multires.
@@JamieDunbar Ah true ! hope they'll do something about that soon... Also, I hope they'll upgrade it so we can have bouncing lights. There was the SSGI add-on but it mess with the shaders...
I haven't watched all of your videos, but I did go through your channel, and I believe what you are doing here was the real intention behind creating internet: to spread knowledge and awareness of different niches. I personally want to learn 3D character designing, but never had enough resources to learn. I hope that I will get to learn more from this channel, thank you so much.
The retopology tools in blender that come by default are very powerful. A combination of the regular modeling tools along with the F2 addon, Polybuild tool, Bsurfaces, loop tools addon, and Grid fill. These let me do pretty much anything quickly. The only thing that really bothers me is how the smoothing brush from sculpt mode doesn't work with snapping. So you have to smooth, then clone and apply the shrink wrap to fix the base level. Back face culling not hiding the wireframe is also problematic, but the Draw Xray addon fixes that. I tend to like making edge loops first, then using Grid Fill to take care of large areas for me. Same thing with Bsurfaces, where instead of extruding edges, you can sketch out a patch with the grease pencil real quick.
Have you checked out the Speed Retopo addon? It basically takes all of those tools and packages them up in a nice UI. It also has a one click button to help get around reapplying the shrinkwrap modifier after smoothing.
I watched this over a month ago. Then had the opportunity recently to start practicing making things in Blender for the first time a couple weeks ago for the first time. But I didn't remember all the details of your tutorial at the time, so I ended up encountering problems and either solving them differently, or ignoring them. Two things I wish I had remembered, was how to use the Polybuild tool fully, and how to get the snapping to work. Something I discovered about the subsurface modifier, is that if you switch the mode to Simple, it doesn't round the edges. This let me work with the subsurface modifier visible, though I do keep it to only about 1or 2 subdivisions because that's about the poly count I'm targeting anyway. The vast majority of my retopology mesh is quads. But I also have a triangulate modifier on (set to Fixed), before the subdivision modifier. Which results in a smoother surface while still keeping to the framework of the grids layed out by the underlying quads. I do strategic place tris to relieve strain in shapes that don't suit quads very well, such as spheroids. But it's generally so that I can increase or decrease the density of the rows of quads without a weird spike in the number of polys in that region. I haven't textured a mesh yet, so I might be shooting myself in the foot by using any tris. But I'm also doing this as an experiment to create my own workflow. The model I'm working on is terrifying, unintentionally horror film worthy anyway. No big deal if I mess up.
Interesting workflow. I'm curious how big a difference it makes. Firstly, some tris are acceptable. You want them in strategic areas so subidv modifiers don't give strange results, but most models will have at least a couple hidden in armpits or on the bottom of feet. Perfectly normal. Using a subdiv with simple smoothing - does that really make much difference? The whole point of smoothing is to...well smooth. And using simple doesn't actually smooth, it just doubles your polygons. I'm curious what this is added to your workflow? But I'm really curious about adding the triangulate modifier. Doesn't that make it harder to see how your quads are laid out? I'm assuming you delete this once you've finished retopo?
If the sculpted mesh under the retopology mesh is sufficiently smooth, the Shrinkwrap modifier will make the subdivisions from the Simple mode on the Subdivision modifier smooth, despite them being generated by the Simple mode. Although since I wrote the comment your replying to, I've started testing a new workflow. However, the idea behind this one, was to bang out a basic sculpt from a heavily subdivided sphere. Then create a _very_ low resolution retopology of it, and delete the initial sculpt. (I'm trying to work within tight storage space constraints. So the fewer polys the better.) Then I would iterate by creating a subdivided copy of the retopology, and sculpt that more to refine the shape. Then I "retopologize" it using the mesh that generated it in the first place as a starting point, by fitting it on top with Shrinkwrap etc. I then make changes to the topology to alleviate mesh density problems, and to reflect changes made during sculpting. Then I create a new subdivided copy (shaped like the sculpt because of the Shrinkwrap) for sculpting, where I do touch ups that I couldn't do before because of insufficient mesh density. After that, the old mesh gets refitted on top of the new sculpt, and I add in more detail to the topology if it needs it. Each new iteration should be faster, as the changes in the sculpting step will get smaller, requiring less editing in the retopology step. I will always use a mirror modifier for the topology mesh. Not only does it do the obvious of ensuring symmetry and cutting workload down by at least half. It also allows you to turn X-Ray mode on to see the topology mesh ontop on the sculpted mesh, despite any clipping. And you can use the circle select tool in mesh editing mode, in combination with magnetic snapping, and the move tool, to quickly select swathes of vertexes and snap them to the surface of the sculpted mesh without having to worry about accidentally selecting vertex on the opposite side, because if you position the camera right, there's only one layer of vertex in the area you want to select. If I'm repeating your own advice back to you, I am very sorry. My excuse is that this has all been a mad blur of 3D modeling at any opportunity I'm given. Triangulate doesn't really hurt my ability to see my quads. The human mind helps by recognizing patterns. I think I also poked some buttons on the modifier boxes, at the top, that affect how they show up on the mesh I'm editing. I think it was an icon resembling a quad in edit mode, or at least, what it would look like in Inkscape, due to the different stylization of vertex/nodes. Basically what it does, is not show the resulting wireframe of the Triangulate modifier. I sometimes hide its effect as well. It there more to help me visualize the end result of the final mesh. I might have also changed some viewport settings to lower the opacity of some kind of wireframes while leaving the rest at their normal opacity. I'm doing all this on a borrowed computer, to develope my skills in preparation for getting my own computer. (My previous computer died years ago, and is too outdated be worth the effort to revive.) When I do get my own computer, I intend to earn money making both commissioned custom game avatars, and customizable game avatars bases of unique species of creatures. I've joined a community where it's common for people use unique avatars as a form of self expression, in social VR games. Some people dip their toes in the water by using non-customized, publicly available avatars. Others buy base meshes that are already rigged, and either make their own textures, or commission someone to do if for them according to a ref sheet. Sometimes, people spend in the ballpark of $900 to commission an entirely unique avatar from scratch. I'm not really in it to become rich. It's just one of the first things I can honestly envision myself doing for money and enjoying. If I manage to scrape a living from it, I'll be happy. But right now, I am a newbie to all this. I've passively absorbed a lot of knowledge just from conversation that occurs in the community, and I'm not new to raster graphics or vector graphics. I actually have a modest education in graphic design, which seems to have some overlap in skills and knowledge with 3D modeling. I took it upon myself years ago to learn Inkscape, which has served to help me understand concepts in Blender. Right now, I'm enthusiastically throwing my time and energy into learning character modeling, rigging, texturing, etc. My next research subject is the Multi-Resolution modifier, which looks potentially useful for my multipass-iterative re-re-re-topology workflow prototype. But that's a research topic for the next few days. If it does work the way I hope it does, it should cut down the refitting step to almost nothing, saving a ton of work lining up the old version of the topology to the new version of the sculpt. I only read the Blender Manual page on the Multi-Resolution modifier, and it was a bit unclear to me what exactly it did. I was also so tired at the time that I had involuntary muscle twitches in minor muscles of my face, so that didn't help me any. P.S. The Triangle modifier is mostly there to help me visualize how the final mesh will be affected by the Triangulation modifier. I'm working under the assumption that a mesh made only of tris is less expensive for a GPU to animate in a game. And because I'm exclusively targeting a game with my models, the would be converted to tris at the end. And with the game I'm targeting being a social VR game, framerate is at a premium, as higher framerate equates to less VR sickness. No to mention it's considered somewhat rude to have a poorly optimized avatar when there are a lot of people present. If there's no appreciable difference in performance between tris and quads, I can just leave my avatars with their quads intact. I probably missed a point of two to reply to. But I'm a bit tired at the moment, so if you want to know something, just ask me again. And sorry for the mess of a reply.
@@World_Theory Wow, that was a mammoth of a reply! :P Definitely look into the Multires modifier. It sounds like it could simplify your current workflow a bit. Also, have you looked into the sculpting 'remesh' tool? You do tend to lose a bit of detail with it, but it's kind of a quick and dirty retopology and it's much faster than sculpting with Dyntopo. Again, might help simplify your current workflow.
You are amazing! I've watched a ton of retopo lessons, both on CZcams and paid stuff, but yours is one of the better ones without a doubt. Maybe even the best I've found. Thank you, so much. I'll look into your paid courses for sure
Naw, thanks so much Ann-Louise. Retopology is a difficult idea for beginners to wrap their heads around, I think because you have to learn both *why* you need good topology and *how* to do it at the same time. Which of course also makes it hard to teach! Really glad you got so much out of this video.
@@JamieDunbar Hey Jamie. I don't expect you'll see this, and I'd be rude to expect an answer, but I figure'd I'd ask because I havent gotten any worthwhile responses on Reddit or BlenderArtists. Do you have any advice for topology where the eyes open VERTICALLY? I have a creature sculpted where the eyes basically open and close at a right angle, perpendicular to the mouth. And it didn't dawn on me until I started retopologizing how MUCH it breaks open the typical facial edge flows, and I CANNOT for the life of me figure out how I would do it. Is there any advice you might be able to offer?
Hey Michael. Retopology can be painful until you’ve wrapped how to set up good edge flow. Even a small change like the shape of the eyes can easily throw you off. There’s probably not special secret here though - ultimately you should be able to use the same topology, just scaled in and up. You may need to add or remove a couple of edge loops to adjust for the shape difference. The fun part of course will then be figuring out where these new edge loops have to go 😅
Thanks Murillo 🤩 I’ve really only been making these videos for the past year. It’s slow going, but it’s actually been faster growth than I expected. It all just takes time.
I'm glad I saw someone in another video doing this with the poly build tool, because I was able to look for the name of the tool on their screen, and search for it by name. See, I had guessed that such a thing as retopology was possible, but I had no clue what the name of it was, or if it even had one. Now I just need to go search up what the deal is with quads versus triangles. Retopology looks very similar to tracing a bitmat with 2D vector graphics. Or trying to emulate a none linear color gradient using a multi-stop linear color gradient. I think that some areas deserve a higher density of polys. That was something I discovered when doing gradients like I just mentioned. The way the gradient worked in the program I was using, was the entire gradient had a limited number of color steps, no matter how many stops that gradient had. And the number of steps between stops was kept equal. So doing a gradient became more difficult, the more detailed the gradient was, because the limited number of color steps were being stretched in usage, so that banding was becoming visible. So I discovered that I had to use more stops to emulate finer, tighter curves, and fewer stops where the curves are more spread-out. This increased the efficiency with which I used the color steps enough that the banding in the gradient because smoothed out and imperceptible. It looks like a similar situation here with the subsurface modifier being used together with the shrink-wrap modifier, to get more detail. Except that you didn't allow the higher detail area-the ridges where the plate scales meet-to have more detail. So you got areas that look weird where there is supposed to be more detail, because the areas next to it look normal, because that level of detail was sufficient for their geometry, despite the level of detail given by the subdivisions being roughly the same in both areas.
Haha! I’m always concerned I waffle on too much explaining all that extra detail. It definitely adds an extra 5-10 minutes to each video. But I think beginners can get discouraged and lost when things break - and knowing that’s perfectly normal and how to deal with it has got to be helpful, right?
I've done retopo with Max. works like a charm. I'm new to Blender and some features seemed so promising. But as you said: Might not be the best retopo tool. A lot of setup to do. Anyway: This is a very great tutorial for the overall view! thx!
Haha, it’s been so long since I last used Max it didn’t have retopo tools 😅 Blender does alright with retopo, but if you were doing it a lot I’d definitely be recommending you pick up Retopoflow.
This is honestly awesome and I can’t wait for the next two because I want to know what Retopoflow is like. Some say it’s worth it, others say it has issues that it’s not worth it. I want to see what it’s like so I know it’s worth spending
Retopoflow is still in beta (although I received an email a few days ago saying the full release is close) so it definitely has a few issues right now. But once those bugs are fixed I’d say it’s worth the price. Assuming you’re doing a lot of retopology anyway.
Hey just wanna say thanks for this tutorial! I’ve been meaning to learn these workflows for a long time but your tutorial got right into the stuff I needed and I was working on my mesh in no time! I ended up having to apply the shrink wrap modifier after a bit because the settings were messing up my edge flow in bumpier parts of my mesh but the modifiers and settings are super useful and I’ll definitely use them again!
Thanks Emmy. That's brilliant to hear. And yes, I have the same issue where I need to apply the shrinkwrap periodically. Unfortunately I don't think the shrinkwrap was designed with retopo in mind, so although it works it's far from perfect. I believe there's plans to update the retopo workflow in the future - which means I'll have to make a new video at some point!
No worries Ashini. I often find it hard to practice certain skills, because there can be SOOOO much work just getting to that point. I'm happy to help you skip ahead :P
Great video! But 8:06, pretty sure people don't use "in-front" so that they can see it from behind, that's a negative side-effect of in-front usage during retopo. The reason why people use "in-front" is so that they don't have to use the 0.01m displacement, which can be more difficult for very bumpy meshes since if you make the displacement too high it'll be become warped (Like you mentioned). For 10:05, the limitation of not using in-front is that sometimes it forces you to put edge loops in even when you didn't, which doesn't work if you want a really low poly model (Esp. for videogames when you could easily burn that into a normal map and all that matters is that the outline is round)
So is it actually more common for some one to make a sculpt of a character and then retopology them? That sounds far easier to come up with a character design than trying to model them in polygons first..
These days it probably is more common, but by no means is it the only way. I started learning before sculpting really became a thing, so we had to poly model everything by hand. You can really see the difference if you check out my earlier works on my DeviantArt: www.deviantart.com/archdragon/gallery/24079483/old-3d Broadly speaking (because there isn't really a "right way") - if you know what the character already looks like, poly modelling can be faster. Check out my Lapras sculpting video to see what I mean. But if you're still making up the character, sculpting is so much easier to suddenly change a design on the fly. Like if you randomly wanted to add a tail - sculpting is much easier. Sculpting is also great because it separates the creative side of designing a character from the technical side of getting the topology right. Personally I struggle to be both artistic and technical at the same time, so I really like this workflow.
Have have retopoflow and quadremesher TBH I gave up on retopoflow it's quite slow and cumbersome - but i will look again as I seeit has updates, Quadremesher is good, and currently id say better the retopoflow, I hope they work on some of the redirection guides for animated characters - but for static meshes it's quite impressive.
You mentioned that you don't have the mirror modifier active because it doesn't play nice with the Shrinkwrap. That's been my experience tonight as well (just starting to learn how to retopo). Do you leave a gap by the axis then stitch it up in the end? I'm curious what the strategy is.
Usually yes. You’d definitely apply the shrinkwrap. You may or may not apply the subdiv modifier. Depends if it’s for film or games. And for films, you might delete the subsurf and instead add a multires modifer.
Thanks Retro. I really wanted this first one to be as beginner friendly as possible, coz the next two will add a good chunk of complexity. It's actually my birthday this week, so I'm heading away for a few days. Hoping to get the next video (Speed Retopo) done the week after that. Retopoflow will likely be another two weeks after that, so apologies, you'll have to wait a bit. The plus side is, I just got an announcement from Retopoflow and they're about ready to put out their new 2.9 update. Hopefully by the time I do the video the final, full version will be available!
@@JamieDunbar Thanks for the reply! :D And happy pre-birthday 🎂 haha and I totally understand, hope you have a good time, cause with this video you earned it haha, I just was really curious on how long I'd have to wait for the next ones, it's quite a bit, but not that much, I just subscribed to your channel, so I think in the meantime I'll watch some other of your videos :D
Hi there. When you say "apply modifiers" I assume you're applying the 'subdivision' modifier? The whole purpose of this modifier is to add more vertices, so applying will always add more. I'd recommend not applying the subdivision modifier, turning it off in the viewport, but on for rendering. This will give you a fast viewport, but more detail at render time. I'd also look into either displacement maps or normal maps if you want to capture that extra detail. I have tutorials for both. Here's the one for normal maps: czcams.com/video/HPfqI8lB5Q0/video.html
Thank you for this great video! Anyone have any tips for grabbing the vertices when using the PolyBuild Tool? The blue edge often jumps out and I get an extrude when I just want to move the point. Thanks!
That would be pretty cool, but Retopoflow isn’t owned by the Blender Foundation, so I don’t see it being merged any time soon. Most of the problems I have with the default tools comes down to how the snapping in Blender works. If the math behind those could be tweaked, I suspect you could solve most of these problems without the need for a massive overhaul.
Heya Comander5. Someone else in the comments had a similar problem. His name was Sandborn if you want to find the comment. But he fixed the problem by enabling his gizmo. Maybe yours is the same issue? Here's what he said: "I had to enable the gizmo since I had it disabled because I don't need it. Turned out the Poly Build tool also requires the Gizmo to be turned on to function which is interesting. All I had to do is turn it back on."
@@kingwalz Did you figure out the Polybuild tool? Other people have had that problem. I'm pretty sure all they needed to do was enable overlays to see those blue lines.
Hey Elvan. Whether or not it matters will depend on your project. Obviously bigger polygons will mean less polygons - so that might be ideal for a video game model. For film characters you might add more. Generally though, you want to use as few polygons as possible. And it’s relatively easy to add more edge loops, or even subdivide the entire model. So try and have as few polygons as possible - just make sure they’re small enough to capture the detail you need.
@@_Shinique Unfortunately I think that's just bad luck. As far as I know there's no easy way to assign different random colors. You can however rename your object, or duplicate it to generate a new random color. Maybe this post will help? blender.stackexchange.com/questions/143444/is-there-a-way-to-reset-the-seed-of-random-material-in-blender-2-8-viewport
@@JamieDunbar and the thread helped as well. Someone said it may be dependent on numbers. I realize that I like to name my collections with numbers and I name everything under them starting with the collection number, so that must be why I see the same color all the time.
Hey Rizzlord. Check out the second video in the series. At around 24 minutes I retopo the legs. It’s a different tool, but the principles are the same.
@@JamieDunbar haha,i man '' need creat a plane'',I don't understand the meaning of such a design. Blender is always like this. It is clear that only a small change is needed to make its own tools better
@@jochenfong3056 Ah yes, I see what you mean. There's no option to just click and start creating vertices. Not sure how you'd get around that in Blender, as you need to have an object to enter edit mode...
After, instead of using subdivision you can also use multires modifier, that allows you to go top subdivision and keep the details, while you can use your low res for animation
Yes! This is a great workflow. And if you use the Multires modifier, you can then use that to bake all that detail into either a normal or displacement map. I cover those steps in another video:
czcams.com/video/R2q7oZLs0GE/video.html
@@JamieDunbar true that getting a discplacement or normal map out of it can be useful when exporting your project to an other engine, but if you plan to stay on blender would you use the map or keep the model with the multires modifier ?
If you have a really detailed model the multires modifier can make your file size really large. So I’d usually bake out the maps.
That said, Eevee can’t render displacement maps, so you may want to keep the multires.
@@JamieDunbar Ah true ! hope they'll do something about that soon... Also, I hope they'll upgrade it so we can have bouncing lights.
There was the SSGI add-on but it mess with the shaders...
@@sh4dow2a I'd be so ecstatic if they found a way to add something like Unreal's Lumen to Eevee.
Best retopo lesson I had so far lately.
Yay! Success!! 😁
The auto merge feature just changed my world
Kind of breaks your brain the first time you realise it's there - Where has this been my whole life!!??
Great coming from Max, thats funny how retopo seem easy here
It's been quite a few years since I last used Max. For some reason I thought Max had half decent retopo tools. Is this not the case?
@@JamieDunbar you can retopo but its not so smooth
I haven't watched all of your videos, but I did go through your channel, and I believe what you are doing here was the real intention behind creating internet: to spread knowledge and awareness of different niches. I personally want to learn 3D character designing, but never had enough resources to learn. I hope that I will get to learn more from this channel, thank you so much.
❤️❤️❤️
The retopology tools in blender that come by default are very powerful. A combination of the regular modeling tools along with the F2 addon, Polybuild tool, Bsurfaces, loop tools addon, and Grid fill. These let me do pretty much anything quickly. The only thing that really bothers me is how the smoothing brush from sculpt mode doesn't work with snapping. So you have to smooth, then clone and apply the shrink wrap to fix the base level. Back face culling not hiding the wireframe is also problematic, but the Draw Xray addon fixes that. I tend to like making edge loops first, then using Grid Fill to take care of large areas for me. Same thing with Bsurfaces, where instead of extruding edges, you can sketch out a patch with the grease pencil real quick.
Have you checked out the Speed Retopo addon? It basically takes all of those tools and packages them up in a nice UI. It also has a one click button to help get around reapplying the shrinkwrap modifier after smoothing.
Very informative video! Easy to comprehend as well! Great work!
Thanks Plabs. A lot of retopology is hard to understand, so I'm glad you're able to comprehend it easily!
@@JamieDunbar That's only because you explained so well!
Great tutorial with some good hints and tips that are sometimes overlooked by or done differently than other retopology tutorials. Thank you!
I watched this over a month ago. Then had the opportunity recently to start practicing making things in Blender for the first time a couple weeks ago for the first time. But I didn't remember all the details of your tutorial at the time, so I ended up encountering problems and either solving them differently, or ignoring them.
Two things I wish I had remembered, was how to use the Polybuild tool fully, and how to get the snapping to work.
Something I discovered about the subsurface modifier, is that if you switch the mode to Simple, it doesn't round the edges. This let me work with the subsurface modifier visible, though I do keep it to only about 1or 2 subdivisions because that's about the poly count I'm targeting anyway.
The vast majority of my retopology mesh is quads. But I also have a triangulate modifier on (set to Fixed), before the subdivision modifier. Which results in a smoother surface while still keeping to the framework of the grids layed out by the underlying quads. I do strategic place tris to relieve strain in shapes that don't suit quads very well, such as spheroids. But it's generally so that I can increase or decrease the density of the rows of quads without a weird spike in the number of polys in that region.
I haven't textured a mesh yet, so I might be shooting myself in the foot by using any tris. But I'm also doing this as an experiment to create my own workflow. The model I'm working on is terrifying, unintentionally horror film worthy anyway. No big deal if I mess up.
Interesting workflow. I'm curious how big a difference it makes.
Firstly, some tris are acceptable. You want them in strategic areas so subidv modifiers don't give strange results, but most models will have at least a couple hidden in armpits or on the bottom of feet. Perfectly normal.
Using a subdiv with simple smoothing - does that really make much difference? The whole point of smoothing is to...well smooth. And using simple doesn't actually smooth, it just doubles your polygons. I'm curious what this is added to your workflow?
But I'm really curious about adding the triangulate modifier. Doesn't that make it harder to see how your quads are laid out? I'm assuming you delete this once you've finished retopo?
If the sculpted mesh under the retopology mesh is sufficiently smooth, the Shrinkwrap modifier will make the subdivisions from the Simple mode on the Subdivision modifier smooth, despite them being generated by the Simple mode. Although since I wrote the comment your replying to, I've started testing a new workflow. However, the idea behind this one, was to bang out a basic sculpt from a heavily subdivided sphere. Then create a _very_ low resolution retopology of it, and delete the initial sculpt. (I'm trying to work within tight storage space constraints. So the fewer polys the better.) Then I would iterate by creating a subdivided copy of the retopology, and sculpt that more to refine the shape. Then I "retopologize" it using the mesh that generated it in the first place as a starting point, by fitting it on top with Shrinkwrap etc. I then make changes to the topology to alleviate mesh density problems, and to reflect changes made during sculpting. Then I create a new subdivided copy (shaped like the sculpt because of the Shrinkwrap) for sculpting, where I do touch ups that I couldn't do before because of insufficient mesh density. After that, the old mesh gets refitted on top of the new sculpt, and I add in more detail to the topology if it needs it. Each new iteration should be faster, as the changes in the sculpting step will get smaller, requiring less editing in the retopology step.
I will always use a mirror modifier for the topology mesh. Not only does it do the obvious of ensuring symmetry and cutting workload down by at least half. It also allows you to turn X-Ray mode on to see the topology mesh ontop on the sculpted mesh, despite any clipping. And you can use the circle select tool in mesh editing mode, in combination with magnetic snapping, and the move tool, to quickly select swathes of vertexes and snap them to the surface of the sculpted mesh without having to worry about accidentally selecting vertex on the opposite side, because if you position the camera right, there's only one layer of vertex in the area you want to select.
If I'm repeating your own advice back to you, I am very sorry. My excuse is that this has all been a mad blur of 3D modeling at any opportunity I'm given.
Triangulate doesn't really hurt my ability to see my quads. The human mind helps by recognizing patterns. I think I also poked some buttons on the modifier boxes, at the top, that affect how they show up on the mesh I'm editing. I think it was an icon resembling a quad in edit mode, or at least, what it would look like in Inkscape, due to the different stylization of vertex/nodes. Basically what it does, is not show the resulting wireframe of the Triangulate modifier. I sometimes hide its effect as well. It there more to help me visualize the end result of the final mesh. I might have also changed some viewport settings to lower the opacity of some kind of wireframes while leaving the rest at their normal opacity.
I'm doing all this on a borrowed computer, to develope my skills in preparation for getting my own computer. (My previous computer died years ago, and is too outdated be worth the effort to revive.)
When I do get my own computer, I intend to earn money making both commissioned custom game avatars, and customizable game avatars bases of unique species of creatures. I've joined a community where it's common for people use unique avatars as a form of self expression, in social VR games. Some people dip their toes in the water by using non-customized, publicly available avatars. Others buy base meshes that are already rigged, and either make their own textures, or commission someone to do if for them according to a ref sheet. Sometimes, people spend in the ballpark of $900 to commission an entirely unique avatar from scratch.
I'm not really in it to become rich. It's just one of the first things I can honestly envision myself doing for money and enjoying. If I manage to scrape a living from it, I'll be happy.
But right now, I am a newbie to all this. I've passively absorbed a lot of knowledge just from conversation that occurs in the community, and I'm not new to raster graphics or vector graphics. I actually have a modest education in graphic design, which seems to have some overlap in skills and knowledge with 3D modeling. I took it upon myself years ago to learn Inkscape, which has served to help me understand concepts in Blender.
Right now, I'm enthusiastically throwing my time and energy into learning character modeling, rigging, texturing, etc.
My next research subject is the Multi-Resolution modifier, which looks potentially useful for my multipass-iterative re-re-re-topology workflow prototype. But that's a research topic for the next few days. If it does work the way I hope it does, it should cut down the refitting step to almost nothing, saving a ton of work lining up the old version of the topology to the new version of the sculpt. I only read the Blender Manual page on the Multi-Resolution modifier, and it was a bit unclear to me what exactly it did. I was also so tired at the time that I had involuntary muscle twitches in minor muscles of my face, so that didn't help me any.
P.S. The Triangle modifier is mostly there to help me visualize how the final mesh will be affected by the Triangulation modifier. I'm working under the assumption that a mesh made only of tris is less expensive for a GPU to animate in a game. And because I'm exclusively targeting a game with my models, the would be converted to tris at the end. And with the game I'm targeting being a social VR game, framerate is at a premium, as higher framerate equates to less VR sickness. No to mention it's considered somewhat rude to have a poorly optimized avatar when there are a lot of people present.
If there's no appreciable difference in performance between tris and quads, I can just leave my avatars with their quads intact.
I probably missed a point of two to reply to. But I'm a bit tired at the moment, so if you want to know something, just ask me again. And sorry for the mess of a reply.
@@World_Theory Wow, that was a mammoth of a reply! :P
Definitely look into the Multires modifier. It sounds like it could simplify your current workflow a bit. Also, have you looked into the sculpting 'remesh' tool? You do tend to lose a bit of detail with it, but it's kind of a quick and dirty retopology and it's much faster than sculpting with Dyntopo. Again, might help simplify your current workflow.
You are amazing! I've watched a ton of retopo lessons, both on CZcams and paid stuff, but yours is one of the better ones without a doubt. Maybe even the best I've found. Thank you, so much. I'll look into your paid courses for sure
Naw, thanks so much Ann-Louise. Retopology is a difficult idea for beginners to wrap their heads around, I think because you have to learn both *why* you need good topology and *how* to do it at the same time. Which of course also makes it hard to teach!
Really glad you got so much out of this video.
@@JamieDunbar Hey Jamie. I don't expect you'll see this, and I'd be rude to expect an answer, but I figure'd I'd ask because I havent gotten any worthwhile responses on Reddit or BlenderArtists.
Do you have any advice for topology where the eyes open VERTICALLY? I have a creature sculpted where the eyes basically open and close at a right angle, perpendicular to the mouth. And it didn't dawn on me until I started retopologizing how MUCH it breaks open the typical facial edge flows, and I CANNOT for the life of me figure out how I would do it. Is there any advice you might be able to offer?
Hey Michael. Retopology can be painful until you’ve wrapped how to set up good edge flow. Even a small change like the shape of the eyes can easily throw you off.
There’s probably not special secret here though - ultimately you should be able to use the same topology, just scaled in and up. You may need to add or remove a couple of edge loops to adjust for the shape difference.
The fun part of course will then be figuring out where these new edge loops have to go 😅
Incredible way of teaching, surely your channel needs more attention
Thanks Murillo 🤩
I’ve really only been making these videos for the past year. It’s slow going, but it’s actually been faster growth than I expected. It all just takes time.
Broooo... i didn't even realise, that you've been helping me on FB, while i already watched some of your tutorials..... Thx for the great help
Heya Benny! Happy to have helped :D
Great video! Thanks for this
I'm glad I saw someone in another video doing this with the poly build tool, because I was able to look for the name of the tool on their screen, and search for it by name. See, I had guessed that such a thing as retopology was possible, but I had no clue what the name of it was, or if it even had one.
Now I just need to go search up what the deal is with quads versus triangles.
Retopology looks very similar to tracing a bitmat with 2D vector graphics. Or trying to emulate a none linear color gradient using a multi-stop linear color gradient. I think that some areas deserve a higher density of polys. That was something I discovered when doing gradients like I just mentioned.
The way the gradient worked in the program I was using, was the entire gradient had a limited number of color steps, no matter how many stops that gradient had. And the number of steps between stops was kept equal. So doing a gradient became more difficult, the more detailed the gradient was, because the limited number of color steps were being stretched in usage, so that banding was becoming visible. So I discovered that I had to use more stops to emulate finer, tighter curves, and fewer stops where the curves are more spread-out. This increased the efficiency with which I used the color steps enough that the banding in the gradient because smoothed out and imperceptible.
It looks like a similar situation here with the subsurface modifier being used together with the shrink-wrap modifier, to get more detail. Except that you didn't allow the higher detail area-the ridges where the plate scales meet-to have more detail. So you got areas that look weird where there is supposed to be more detail, because the areas next to it look normal, because that level of detail was sufficient for their geometry, despite the level of detail given by the subdivisions being roughly the same in both areas.
You are a literal god, you've explained everything one might have problems with
Haha! I’m always concerned I waffle on too much explaining all that extra detail. It definitely adds an extra 5-10 minutes to each video. But I think beginners can get discouraged and lost when things break - and knowing that’s perfectly normal and how to deal with it has got to be helpful, right?
This is the most helpful tutorial I've seen on retopology. Thanks a lot!
Haha, thanks Spence. I'm glad it helped you ;)
I've done retopo with Max. works like a charm. I'm new to Blender and some features seemed so promising. But as you said: Might not be the best retopo tool. A lot of setup to do. Anyway: This is a very great tutorial for the overall view! thx!
Haha, it’s been so long since I last used Max it didn’t have retopo tools 😅
Blender does alright with retopo, but if you were doing it a lot I’d definitely be recommending you pick up Retopoflow.
This is honestly awesome and I can’t wait for the next two because I want to know what Retopoflow is like. Some say it’s worth it, others say it has issues that it’s not worth it. I want to see what it’s like so I know it’s worth spending
Retopoflow is still in beta (although I received an email a few days ago saying the full release is close) so it definitely has a few issues right now. But once those bugs are fixed I’d say it’s worth the price. Assuming you’re doing a lot of retopology anyway.
@@JamieDunbar I haven’t gone into retopology yet but at least there’s such an add-on that will speed up the workflow. Thanks for the information
It's free atm, you just won't get any customer support if you face issues github.com/CGCookie/retopoflow
Hey just wanna say thanks for this tutorial! I’ve been meaning to learn these workflows for a long time but your tutorial got right into the stuff I needed and I was working on my mesh in no time! I ended up having to apply the shrink wrap modifier after a bit because the settings were messing up my edge flow in bumpier parts of my mesh but the modifiers and settings are super useful and I’ll definitely use them again!
Thanks Emmy. That's brilliant to hear. And yes, I have the same issue where I need to apply the shrinkwrap periodically. Unfortunately I don't think the shrinkwrap was designed with retopo in mind, so although it works it's far from perfect. I believe there's plans to update the retopo workflow in the future - which means I'll have to make a new video at some point!
Thanq so much for giving the link of the model for practise.....
No worries Ashini. I often find it hard to practice certain skills, because there can be SOOOO much work just getting to that point. I'm happy to help you skip ahead :P
Really cool! Thank you Jamie!
Great video! But 8:06, pretty sure people don't use "in-front" so that they can see it from behind, that's a negative side-effect of in-front usage during retopo. The reason why people use "in-front" is so that they don't have to use the 0.01m displacement, which can be more difficult for very bumpy meshes since if you make the displacement too high it'll be become warped (Like you mentioned). For 10:05, the limitation of not using in-front is that sometimes it forces you to put edge loops in even when you didn't, which doesn't work if you want a really low poly model (Esp. for videogames when you could easily burn that into a normal map and all that matters is that the outline is round)
I love you.Thank you for the tutorial.Eng is not my first language but i manage to get the point of what u say 😆.Thx
Naw, love you too Tulian. I'm glad you were able to figure things out.
*THANK YOU SO MUCH*
You just made my life easier, ❤️ here heart for you...
Heart back at you. Glad I could make your life just a little bit easier 😁
The polyquilt addon which free makes retopo super easy in Blender.
its explained super well, im just stupid and still have no idea how to do this
Haha, thanks...I think :P
What part are you still having trouble with?
Thanks for the video!
Thanks for this. Really appreciate it!
Oh my god this is perfect lesson thank you
If you enable backface culling, the negative sides of faces won't show. This way, you can also enable the 'In Front' option in 'Viewport Display'.
Nicely well done, very informative, thank you.
Wow, amazing tutorial set. Thank you
Really nice tutorial, thanks!
thanks, very good to understand video
So is it actually more common for some one to make a sculpt of a character and then retopology them? That sounds far easier to come up with a character design than trying to model them in polygons first..
OMG, watching to the end, it blows my mind how much harder I make it on myself trying to do things by hand
@@FuzzyImages Rofl!
These days it probably is more common, but by no means is it the only way.
I started learning before sculpting really became a thing, so we had to poly model everything by hand. You can really see the difference if you check out my earlier works on my DeviantArt:
www.deviantart.com/archdragon/gallery/24079483/old-3d
Broadly speaking (because there isn't really a "right way") - if you know what the character already looks like, poly modelling can be faster. Check out my Lapras sculpting video to see what I mean.
But if you're still making up the character, sculpting is so much easier to suddenly change a design on the fly. Like if you randomly wanted to add a tail - sculpting is much easier.
Sculpting is also great because it separates the creative side of designing a character from the technical side of getting the topology right. Personally I struggle to be both artistic and technical at the same time, so I really like this workflow.
Have have retopoflow and quadremesher TBH I gave up on retopoflow it's quite slow and cumbersome - but i will look again as I seeit has updates, Quadremesher is good, and currently id say better the retopoflow, I hope they work on some of the redirection guides for animated characters - but for static meshes it's quite impressive.
Which one is the quadremesher? I did a quick search and the only thing I could find was an automatic remesher- is that the one?
You mentioned that you don't have the mirror modifier active because it doesn't play nice with the Shrinkwrap. That's been my experience tonight as well (just starting to learn how to retopo). Do you leave a gap by the axis then stitch it up in the end? I'm curious what the strategy is.
set the subdiv to simple problem solved. otherwise very cool tutorial nobody wants to use polybuild and i had a hard time finding stuff thank you.
HI! but when you finish the model...you apply all the modifiers,no?
Usually yes. You’d definitely apply the shrinkwrap. You may or may not apply the subdiv modifier. Depends if it’s for film or games.
And for films, you might delete the subsurf and instead add a multires modifer.
@@JamieDunbar ed for game? It’s the same?
@@valentinagarrusso9286 Sorry, don't understand??
Amazing video ♥ Really noob friendly :D When are the next two lessons coming out? Cause I'm really curious about Retopoflow :D
Thanks Retro. I really wanted this first one to be as beginner friendly as possible, coz the next two will add a good chunk of complexity.
It's actually my birthday this week, so I'm heading away for a few days. Hoping to get the next video (Speed Retopo) done the week after that.
Retopoflow will likely be another two weeks after that, so apologies, you'll have to wait a bit. The plus side is, I just got an announcement from Retopoflow and they're about ready to put out their new 2.9 update. Hopefully by the time I do the video the final, full version will be available!
@@JamieDunbar Thanks for the reply! :D And happy pre-birthday 🎂 haha and I totally understand, hope you have a good time, cause with this video you earned it haha, I just was really curious on how long I'd have to wait for the next ones, it's quite a bit, but not that much, I just subscribed to your channel, so I think in the meantime I'll watch some other of your videos :D
Yay! Thanks mate. I’ve had a few people ask when the next video will be out, so the pressure is there to get it done ASAP. I’ll do my best 😜
after i apply all modifier the mesh still have many vertexes. how to apply this modifier without having those many vertex
Hi there. When you say "apply modifiers" I assume you're applying the 'subdivision' modifier?
The whole purpose of this modifier is to add more vertices, so applying will always add more.
I'd recommend not applying the subdivision modifier, turning it off in the viewport, but on for rendering. This will give you a fast viewport, but more detail at render time.
I'd also look into either displacement maps or normal maps if you want to capture that extra detail. I have tutorials for both. Here's the one for normal maps:
czcams.com/video/HPfqI8lB5Q0/video.html
Thank you for this great video! Anyone have any tips for grabbing the vertices when using the PolyBuild Tool? The blue edge often jumps out and I get an extrude when I just want to move the point. Thanks!
Background music used here? 11:21
many issues could be solved if retopoflow was merged into blender itself as a separate mode or built in addon
That would be pretty cool, but Retopoflow isn’t owned by the Blender Foundation, so I don’t see it being merged any time soon.
Most of the problems I have with the default tools comes down to how the snapping in Blender works. If the math behind those could be tweaked, I suspect you could solve most of these problems without the need for a massive overhaul.
I can't seem to get this to work, for some reason. My Polybuild tool won't highlight edges at all, and won't extrude whatsoever.
Heya Comander5. Someone else in the comments had a similar problem. His name was Sandborn if you want to find the comment. But he fixed the problem by enabling his gizmo. Maybe yours is the same issue? Here's what he said:
"I had to enable the gizmo since I had it disabled because I don't need it. Turned out the Poly Build tool also requires the Gizmo to be turned on to function which is interesting. All I had to do is turn it back on."
@@JamieDunbar what is a gizmo? i dont see any blue lines with the polybuild tool either
just figured it out, I hate the gizmo so I always have it disabled
@@kingwalz Did you figure out the Polybuild tool? Other people have had that problem. I'm pretty sure all they needed to do was enable overlays to see those blue lines.
Does the size of a polygon matter as long as you are able to retopologize the whole model ?
Hey Elvan. Whether or not it matters will depend on your project. Obviously bigger polygons will mean less polygons - so that might be ideal for a video game model. For film characters you might add more.
Generally though, you want to use as few polygons as possible. And it’s relatively easy to add more edge loops, or even subdivide the entire model. So try and have as few polygons as possible - just make sure they’re small enough to capture the detail you need.
@@JamieDunbar Thank you. That’s really the only thing that confuses me about Retopology.
You should look at MiraTools.
I found gold
media.tenor.com/images/35b07a9e7f2746fce6048419053d435c/tenor.gif
I failed at repo, everything turned into a 4th dimensional polygons, its a complete mess lol
4th dimension? How on earth did you manage to animate them!?
More like Maya Retopo tool.
My "random" is never really random. It's driving me crazy. I mostly see variations of green.
Hi Shinique. My apologies, it's been a little while since I recorded this video. What is the 'random' you're referring to?
@@JamieDunbar Under the viewport shading you mention at 7:25 . Mine are always a similar color or the same color, and are usually all green.
@@_Shinique Unfortunately I think that's just bad luck. As far as I know there's no easy way to assign different random colors.
You can however rename your object, or duplicate it to generate a new random color. Maybe this post will help?
blender.stackexchange.com/questions/143444/is-there-a-way-to-reset-the-seed-of-random-material-in-blender-2-8-viewport
@@JamieDunbar ok I will rename. That helps. Thank you.
@@JamieDunbar and the thread helped as well. Someone said it may be dependent on numbers. I realize that I like to name my collections with numbers and I name everything under them starting with the collection number, so that must be why I see the same color all the time.
that bensound music tho...
Haha, yeah I know. This was one of my earlier videos. I swear I’ve got a better variety of music now! 😅
why these "tutorials" enver show how to connect any limbs with it??ß
Hey Rizzlord. Check out the second video in the series. At around 24 minutes I retopo the legs.
It’s a different tool, but the principles are the same.
The first step is to create patches, which I hate. Maya doesn't need them
What do you mean by 'patches'? I've done a little retopo in Maya, but admittedly not much. I did really like Maya's retopo tools though.
@@JamieDunbar haha,i man '' need creat a plane'',I don't understand the meaning of such a design. Blender is always like this. It is clear that only a small change is needed to make its own tools better
@@jochenfong3056 Ah yes, I see what you mean. There's no option to just click and start creating vertices. Not sure how you'd get around that in Blender, as you need to have an object to enter edit mode...
@@JamieDunbar czcams.com/video/VKLPHh3CYCU/video.html lookthis ,i think best blender retop tool PolyQuilt(ポリキルト)
@@jochenfong3056 github.com/CGCookie/retopoflow
For God's sake stop this 'vertissy' madness. Singular 'vertex', plural 'vertices'. Easy right?
Get it wrong again and there will be trouble!
*Does Igor impression*
“Yes master”.
No
Yes! Wait...what are we talking about?
@@JamieDunbar this is not the way you do retopology. It will take forever this way.
This is a beginners tutorial. It’s not meant to be fast, it’s meant to teach the basics. If you want speed, check out part 2.
@@JamieDunbar I seen it already.