 Felly ydych yn dda? Am gweithio? Felly dda. Dill yw'n mynd i'ch gwaith. Felly rŵn i'ch gweithio y Cymru 2017. Rwyf yn dda'n gwneud fan hynny'n ceisio'r pwysig, ac mae'r gwirioneddau'r cyffredinol yw hynny'n fwy o'r hyd yn gwneud. Rwy'n gweithio'r cyffredinol sy'n eich ddod, a'r hynny'n gwneud gyda chi'n ddod, mae'r gwrthwyng yn ôl. Rwyf yn gweithio'r gwrthwyng yn llwydd. yn gilydd, mae'r peth eijust yn meddwl i'n meddwl eu bod yn hynny, mae'n meddwl i'n meddwl i'n meddwl i'r goll honodaethau, mae'n meddwl i'n meddwl i wneud dyna i'r niwylo, mae yna'r Rys MacFall, yn mynd i amdano yn unig ar yr anymgau i unig. A wedi gondol, mae'r ardi iddo yn unig, rwyf yn ymddangos, rwyf wedi gdudio, Mae'n gweithio'r ysgolig. Roeddwn i'n gweithio'r llai yn gweithio, mae'r ffordd yn fawr. Mae'r gweithio'r ysgolig yn fawr yn ei wneud y gwaith ar y cyfrifio'r ysgolig o'r fawr. Mae'r fawr yn gweithio'r ysgolig o'r fawr. Mae'r gweithio'r fawr yn ymddangos, i gael y gallu ffasol Lundin, ac ysgolig o'r Ffranc Banc. Mae'n gweithio'r fawr, fel y gallai'r arig ynghylch ar gweithio, digwydd i'n meddwl yw'i palm yn ymweld o'i gweithio amser a'i eich cyflogion iawn i'n dod yw'r galiad, fyddwn i ddim yn gweithio i galiad wrth eu gwirionedd, ac mae ar y ddwy'n meddwl yw ddwy am y cyflogion iawn eich bod hefyd ar gyfer i gael didnig. Felly, mae'n ni fyddon ni. Yn i gwaith, mae'n fyddech chi'n gweithio i'n dda chi'n gweithio ar y pryd yn ddim yn cael ei bod yn gweithio. this is the final piece that I created for said attraction. It's about 2000 frames of meandring under aquatic life and basically, the prerequisite for this job was that it loops and therein lies a great deal of issues because when you're working with something like this you want to simulate is the first protocol is let's simulate this except there's only one very small section yma y gallwn yn cyffredinol yma, y llwyddoedd yn gyfnodol. Yn gyfnodol yn cymhwyl o'r ffordd. Daniel Matinezlaro, y dyma, yn y ffordd, yn ychydig i'r ffordd, a'i gwybod, mae'n gwneud yma, mae'n gweithio'n gweithio'r ffordd a'i gweithio'r ffordd. Mae'n gweithio'n gweithio'n gweithio'n gweithio. Daniel wedi'u gweithio'n gweithio'n gweithio, oherwydd mae'n gweithio'n gweithio'n gweithio. ac wrth gwrs, mae'n dda, i'n edrych arno i'n meddwl eu gwled이. Mae'n idea i ddweud yn hollu'r rhaglaid. Mae'n cre할io iawn yn iawn dyfodol o'r clyweddau. A dyna'n edrych arwys i gael yr ysgrifennu. Mae ydy'r boedd yna pethau ymgyrch yn ghael. Mae'n golygu i'r effeithio cyffredinol ar y rhaid. Mae'n meddwl i'n edrych ar ager, ac mae'n gael yn ganfaenwynau arall. Mae'r hyn i日oeddydd a llu'n ddweud ac wedi fynd i'n meddwl gyda'n ddweud arfod roedd yn y gallu gwirionedd. Rwy'n ddaeth i'n meddwl a'r cyfathor. Rwy'n ddweud yn ymlaes yn gwybod. Mae'r ysgol yn gwybod yn ymlaes yn gwybod. Mae'n ddweud arall, ond, mae'n gwybod yn amlwg ar y bwrdd. Mae'n ddweud arall yn gwybod, ond mae'n gwybod i'n gwybod yn gwybod. Mae'n gwybod i'n gwybod i'r gweithio ar yr awdurdod. Mae'n amlwg. Mae'n gwybod i'r cyffredagol. i roi gweld, er mwyn gweld byddwch yn siaraduaeth o'ch sorg. Roi'n bydwch, ac mae wedi gweithio'r gwybod o'r rhagwch, ac mae'n dweud o bwysig arall. Mae'r rhai gweld, mae'r rhagwch i'w gallu rhoi gwaith gennym'i meddyl a'r rhai. Mae'r rhagwch yn oed o'u gwirio arall. Mae'r rhagwch yn oed i gwasanaeth o'r rhaeg yn unigdag yn cael ei wneud. Mae'r rhagwch yn oed o rhaeg i ddenner sy'n gweithio eich oedol a'i altru mewn mawr o amser, Ie dweud y gwirionedd yn gwneud yma, maen nhw'n mynd i'n gweithio yn glamorous. Mae'r gweithio yn y blenda, rwy'n gobeithio'r gweithio. Mae'r ddiwrnod yng Nghymru, mae'r ddiwrnod yn mynd i gael y port o gyfnod y gwirionedd yw'i gweithio, ond mae'n rhaid i'n gwneud. Mae'r gweithio'r cymryd yma, mae'n rhaid i'n gweithio ar â'r apte. Mae mae'n gobeithio'r gweithio arherwydd i ddiwrnod a oed yn dweithio ar dechau. Mae'n rhaid i'r car, rhaid i'n rhaid i'n rhaid i'n mynd i ddim yn gwybod, dwi'n ddiogel i'n garaed, rhaid i'n bwysig i'n gwneud i chi'n gweld. A y gallwch yn ymdillwyd, mae'n gwybod am ysgrifennu amlaen, o'r amddangos ar gyfer amser o'r ysgrifennu, yn mynd i'n rhaid i'r amser. O'r amddangos, mae'n cael rhaid i'n gwybod, mae'n ddweud yn ymdillel, ond mae'n gwybod. Cymru'r axaton gweld fynd yn gweithio. Cwrwch ei gweld i gyd i chi wedi'i cael ei amserol. Mae'r newydd ei gyfledd o'r unrhyw gwrthoedd yn maen i chi gymryd yn gweithio. Fydden nhw'n gweithio'n gweithio'n gweithio'n llyfr� ac mae'n llyfras fel digwydd arall. Mae'r bobl yn gweithio'r holl. Mae'r bobl wedi gweld i'n gweld i chi'n gweld i chi? Mae'n gweld i chi'n gweld i chi na bai ddigwydd, driver. I can show you the driver. It's probably not of interest. It's so basic. Graph editor drivers. And there it is. It's three letters, VAR, and it's relating to just picking out the master empty and the custom property parenting. So essentially, all that does is every item in the scene has that, and then I dial that up and suddenly we've got our animation in situ. I want to continue working on it. I want to continue fixing the rig, and I want all my axes aligned. I want to be able to work with, you know, orthographic cameras very easily. I can do that for every item in my scene with one control. So that's the first tip. Really important because organisation is everything. From there, I'm going to go straight into just creating the rig, the bones, the bare bones of it. It's very simple. Again, as I said, the easy rigging principle, don't create stuff that you don't need because it's an economy of, you know, of CPU power and so on. So we've got our root bone here, and all we're going to do is apparently have a nervous fit and start selecting the wrong things, editing, and I'm just going to extrude out a bone, and we'll call that bone, yikes. I'm way more nervous than I thought. This is really good fun. So we'll call it bell peak. From a bell peak, we'll extrude another bone and another bone. We'll call this one upper, and because I'm a creative genius, we'll call this one lower. And it's really hard to blend when you've got a shaky hand. That's not right. Cool. We'll call this one control. I'm going to move through this as quickly as possible because obviously it's really boring watching someone build a rig. I could go all blue Peter and go, here's one I made earlier. That's just for the UK people. So we've got these bones. We'll select the bone, select the root, control P. God, that's like trying to play the bass. Lots of finger stretching. Keep offset. Same for the second control bone. Keep offset. In fact, we'll do that with the bell peak as well. Keep offset. And why not upper to bell peak? Keep offset. So now we've got this chain of bones. What does it do? Nothing. Absolutely nothing. And that's OK. Oh, g. That's it. Forget my shortcuts. So now we've got to actually make these bones do stuff. So we'll add a bone constraint, a damp track. The damp track will point towards our first control bone. And the next one, honestly, but whole talk's not this dry. It's just, I figure if I don't actually show you how the bones go together, you might feel cheated and ask for your money back, but you haven't paid for this or whatever. Am I making any sense? I don't know if I am. So we've got this damp track working. And now, if we do this, we've got this nice little filly jobby. But the key part about jellyfish is how it pushes and flows. So we want this very particular motion. And now if we drag the bell peak, you'll see that it actually starts to look like something. Now, it would be a pretty shoddy rig if that's where we ended. So I'm just going to extrude out another bone from bell peak. There we go. And another one. And I'm just going to scale that out so it's easier to select. And these two bones I'm going to make bendy bones. It's riveting stuff, right? So we'll take our bendy bone. We'll give it another constraint. We'll make it a stretch too. And come on, fella. You know you want to. Oh, it's because I'm in the wrong mode. And that's not even the bone constraints, is it? It's like amateur hour over here. So stretch to armature. And we'll have our control one bone. And then we'll do the same with this. Stretch to armature. And I've done the wrong one. That one, haven't I? No, that was the right one. I actually didn't screw something up. Who knew? So now we've got something looking a little bit more like this. And we can use this, these curvy bendy bones, to actually deform the geometry which I'll show you in a short while. So they maintain the nice flow of the curve. And it's really just a super simple bone structure. Now, there was something I was going to say there. It's gone. It doesn't matter. Couldn't have been that important. So select all. And we'll just reset that. So there are a few more elements to the finished rig where I've got another bone just here which will animate rotation in the very tips of the jellyfish's bell. Yeah, every part of the jellyfish's anatomy is rudely named, I do apologise. It's bells, gonads, and tendrils. I mean, you couldn't make it up. So those are the bones. That's the basic bone system. And I'll show you how that actually applies to our geometry. So here's the bells and gonads. I just love saying it, you know. And that's really just, as you can see, it was quite simple geometry, just nicely laid out and split and uvied and projected so that you can do all your texture painting later. We've got our rig here. This is a slightly more advanced version with, you know, just tidied up to look nice. And then we've got this peculiar piece of geometry here. All it is, is a line of edges and vertices traced around the side of our geometry. Again, it's economy of, you know, economy of geometry this time just because we really need the playback to be 25 frames a second so that we can animate it smoothly. And what we can do with that is just give it its armature modifier. I've already done the weight painting, so I don't have to bore you with that. I can see the palpable relief in your faces. And what we'll do is I'll just show you how that looks. As you can see, this line of geometry is now following our armature, which allows us to do all sorts of fun stuff. There's that rotation bone I was saying about. Selector, GSR. Make sure it's all back to normal. Now, that's all well and good, but I don't, you know, you're probably wondering how that helps us to deform this geometry. Well, screw it. Quite literally screw it. There's a screw modifier and what we do is use it to spin the geometry all the way around. Now, because the screw modifier follows the armature, it actually allows us to do something like this. And oh, yes. Oh, mama. So, essentially, we can now animate the entirety of the geometry with a single sort of set, a single sort of branch of armature. And then if we want to break the entire project, what have I done there? This is a problem with touch screen. You don't give flick for screen and everything goes mad. I've actually lost everything. Pardon? How did I even manage to turn that on? I told you it's armature, I didn't know. Right. So, what we can do is select our Bell and Goonag geometry. It's never getting old. And I'm just going to turn on a mesh deform modifier. Now, I've already baked this. Normally this takes a little time to bind. It'll take, you know, 10, 20 minutes if you've really got heavy geometry. This is typically quite quick. But what that does, it binds our geometry to this mesh deform geometry, which with the screw modifier on it, and it will deform along with it. So, now we can move this. And actually, I'm just going to hide that. I'm going to hide that. We don't need to see that. And you can see that we're actually now quite happily deforming our jelly. So, let's give it some animation just so I can show you how quickly this can be done. I'm so used to working with a Cintiq and a Wacom and just sort of, this is all, ooh, a mouse, how very 1990s. So, yeah, we can do something like that. And then pull it back. This is the shoddiest jellyfish animation I've ever seen. Go on, Chris, work the bell. So, yeah, as we can see, it is now animating. And that was relatively painless to do. Now, the thing about jellyfish is that they undulate and they move and they don't have a uniform deformation across the entire surface. We've got this very rudimentary control and I don't really want to have to make controls all the way around so that I can add detail to this deformation. But because we used the child of modifier, it's actually centred in our scene and aligned along specific axes, we can use the wave modifiers. The wave modifiers are very particular modifiers. They want everything to be aligned a certain way. You can't just go, now, do it along that plane, do it along this plane. It has to be done in a certain fashion. So, those wave modifiers, it's very subtle because you don't want to overbake it. I'll just crank it up a little bit so we can see. But these wave modifiers create an uneven sort of deformation along the, am I even doing it right? That's not saying you want to hear in a presentation. So, you can clear keyframes. That's probably screwing it up. You can see this will create some unevenness in the mesh. Something that you can actually do from here, once you've actually got an animation you're really pleased with, you can actually bake that out to an MDD cache and save that animation and add further deformation should you need to. I found I actually didn't have to do that because I got what I needed in principle straight off the bat. Consult my notes. Oh, I missed a stage, actually, because apparently I'm a real pro. Before you start any project, you really should plan it. You should look at your reference. You should look at pictures of what you're trying to create and figure out how you want it to look. So, obviously, I went into lots of imagery looking at Lion's main jellyfish, the particular genus of jellyfish, which the Aquarium wanted. Something which I did which I found really useful in Blender was just a little gift because, you know, I can't screw it up that way, right? I took footage of a jellyfish and I took the grease pencil and I traced around the outline, just frame by frame going around it and just sort of seeing how it felt. There's a couple of reasons to do this. One, when you're playing back video footage in Blender, a couple of nasty things can happen. One, it might slow down your computer. It might slow down your playback and prevent you actually getting that really good frame rate so you can do nice, smooth animation. Two, there's a bug of some description and it might be down to a video codec or corrupt frames where occasionally Blender just won't load and you have to append your entire scene into another file. That's a real pain in the arse. With the grease pencil outline that you've created from your reference footage, you've got a sound which will play back without any issues. Secondly, you can parent it to an empty and therefore you can move it around your scene and realign it with your geometry so you can just check your reference again. Thirdly, and this is the most important part, you familiarise yourself with what is an entirely alien motion and you're familiarising yourself with that motion frame by frame and getting more of an understanding of the flick and the weight and how it moves. That's something just really, really worth doing with any projects you start with. Start with the grease pencil. Start with the most basic way you can. So, yeah, let's continue. So, I've done the grease pencil, workbench, bones, bendy bones. Just a quick tip with bendy bones is there is a time and a place to use them. I like to use them in basically everything, but I recently completed a project for this medical brace and it was a really cool project, really good fun and I wanted it to animate in and this rigid piece of plastic bends in and forms around the leg. Let me just see if I've got that here for you. I originally did it with bendy bones and bendy bones normally work beautifully with everything, but with particularly dense geometry and a particularly dense blender user, you can actually see here that the geometry is getting these ridges where the bone segments are meeting one another and you get this odd deformation and breaking up here. What you can actually do, I probably don't have time to do this today and show you in full but if you replace, if you put with the bendy bones a curve with hooks and attach those to the bones, then you can actually get a much smoother animation and much smoother deformation. As you can see here, it's really a lot better, there's none of that jagginess. That's just something to bear in mind when you are using bendy bones with dense geometry, sometimes it might upset your assets. What was the next bit? I should always refer to the notes because otherwise I'll miss stuff and wind up apologising a lot. We've done some screwing, we've done some bones, tentacles. This is another interesting part about this. With the tendrils of a jellyfish, the first place you'd want to go and the first place I went is hair and cloth simulation, is trying it out, working it, and I spent quite a long time working with simulations and trying to get realistic collisions. Now, the other problem with, as I said, the stipulation in the brief was that this loops, it has to be a loop. When you've got a simulation, you can actually create a loop by creating a shape key of your first frame and then merging it together. That's not going to work so well with hair, so it's going to be a real nightmare, and then you've got, again, you've got collisions, which you can get it to work in blender, but when you've got 2,000 frames, hundreds and hundreds and hundreds of these tendrils, and your realism is compromised by going through other parts of geometry, it's just not practical. You'd be wrangling that for the rest of your life. So, the system I came up with was this. Essentially, we've got a little tentacle collar. This, again, has the parenting child of modifier, which we can snap to our geometry at any time. Then we've got, essentially, just some lines, again, of verts and edges. So, if you were to render this, there's no faces. It won't render at this point, and I've just sort of, like, sculpted it out to be vaguely sort of tendrally in hairy. Hairy tendrils, yet another rude part of a jellyfish. So, I'm sure you can all guess what I'm going to do. I'm just going to slap a curve modifier on it, and we've got this basier curve here, and if I were to, let's just turn off that, if I were to move it, then you can see it deforms quite nicely this geometry. Then, in order to be able to animate that in a reasonable fashion, we've got hooks. So, we can move these hooks around, and it will move with it. Oh, brain fart. Don't just love it. So, obviously, it's not going to look particularly detailed or crazy. It's a rigid mesh, which is just sort of following a curve. It's not going to do a great deal. We need it to look alive. So, the solution to that I found was using a displaced modifier. Now, a displaced modifier, we add a texture to it. We give it global coordinates, so this texture is fixed in space. As we drag, I'm just going to bring up our parent. I shouldn't have deleted the camera from this scene, should I? So, as we drag our stuff through space, you can see, oh, one, one. I keep making really weird noises, don't I? You can see that the displacement modifier actually acts like turbulence, and it starts to bring this stuff to life. Through tweaking the settings, just enough, these branches of vertices will sort of intersect each other and actually really react quite a lot like tendrils. We've also got some wave modifiers over here, which do some more subtle emotion. So, that's pretty cool. Again, this is not... I'm aware, as soon as I sort of proposed to do this talk, I realised this was probably a terrible idea, because when it comes to rigging in a room, in a conference like this, I'm probably not even in the top 10 best riggers in this room. There's probably a lot smarter people out here, but these are very simple solutions to quite a complex problem. So, that's why I wanted to share them with you today. So, if you're just going, but that's really obvious, I do apologise. So, the next thing which I did just in tidying up these tendrils, because obviously they still weren't render, is I've added a subsurface modifier, you know, the ubiquitous subsurface modifier, gives it a bit more smoothing, and then we add a screw modifier like 0.4 of a degree. So, it just brings out the geometry to give it some faces, add a solidify modifier, and suddenly we've got something quite renderable. We've got this geometry which will then render and it works. But again, the importance of this, the reason you start with those verts and edges and not faces, is because if you add the displacements onto that, you'll get, in fact I can show you, if I just move my displace modifier down a bit, come on a little fella, there we go, and I'll crank it up, go away toolbar, taskbar, whatever you're called. If I crank this up, you'll see that it's no longer deforming the path, it's deforming the actual geometry around it, and you get these sort of crinkly sort of edges which are less than ideal. It can be quite a good look actually, but it's not what I was going for. The displacement modifier is merely altering the trajectory of a path and not the actual geometry around it. So, it's all about stacking stuff in the correct order and getting what you need. So, as you can see, we've got that. And we've got this rig now where anytime it sort of intersects with the bell, you can make a real-time correction so that it's no longer intersecting, and you've got a relatively controllable system. So, basically then you need to repeat that about 12 times around the collar, and you get a dense mane of tendrils which you can then manually animate painstakingly whilst crying into your coffee. So, I've done it again. Again, another tip with this is with these hooks you can actually parent them to a variety of bones in the thing, and that will allow you to sort of give them an overall sort of swim around. Now, what time is it? I've already ran out of time, haven't I? That's fantastic. So, I would have liked to show you a little bit about the cloth sim. This is the only part which is using a simulation. It's fairly dense geometry attached to a parent, to the parent empty. And as it's dragged around the scene and stopping and starting, it undulates and sort of catches up with itself and then stretches back out. That's then baked to an MDD cache, and then what you'll find that the issue there with the MDD cache is that the origin will start slipping away from the geometry. So, what you do is you create an inverse parent on that MDD cached geometry, and the inverse parent is basically you're time-zing the motion on the z axis by minus one, and it will create an inverse parent. Then you can MDD cache that back out, and then you can just parent that to the geometry. I'm out of time. I would have loved to have talked about... Sorry. I just want to say one last thing. There's a couple of other things I would like to have talked about about defamation with eyes, but there is actually a tutorial I did for CG Cookie, which is already out there. You can go on CG Cookie. Become a citizen. There's so much great stuff to learn no matter what level you're at. This project wouldn't have been possible without the Blender Network first connecting me to the client without Blender itself, and without Render Street, who helped me so much with actually getting before noise reduction, getting this final piece out. I'll just play it one more time just for those who didn't enter the room, and the final piece. Thank you very much for your attention. Have a great conference. Peace, y'all. I don't know why I did that.