Well, then. I'll try to do something like wip🙂 I hope this topic will help improve skills for you and me. I will make a model excavator, you will observe, ask questions, comment, advise and correct. I think everyone will benefit so much. In general, "the main thing to get involved in the fight, but there we'll figure it out."It is planned to do:1) High-poly model Hyundai R16-9 Mini Range Excavator for rendering.2) Low-poly model for the game engine UE4.3) Textures.4) Rigging.5) Animation.The plans are great, I hope that everything will be done.HUGE request: if you see that something done by me can be done better / faster / more optimal / cleaner, etc., please do not keep knowledge in yourself, share. I also want to be able to do better / faster / better / cleaner.Part 1. High-poly model.Something about preparations. First of all, I created a project folder in which I started creating logical folders:
Separately for photo-ref, scenes of different polygonality, etc. After that, the project folder was synchronized with the cloud storage with auto-update function, which means automatic backup of all project files to the Internet.Next, I collected the initial reference, these are pictures of the object from different sides, close plans of details, interior (perhaps I will do it). You need a lot of photos and in good resolution. The more details you can see, the better. In the course of the matter, if anything, you can search for more ref. Naturally found something like a drawing for the studio. Also, it should be noted that I found a convenient scheme, in which the main details of the excavator are indicated - a convenient thing for the correct name of the objects of the scene and the fastest search of photos of those objects, if necessary.
In the max, I set the units of measurement cm for the subsequent work with the engine UE4. I always use layers, it's very convenient, I have layers also available from the toolbar. On a separate layer created a studio with drawings, froze, so as not to move. Another layer made a blockout - a quick 3d sketch in order to determine the size of the object.For today, everything, tomorrow I'll start modeling.
We must already share something here)He made the frame. He made all the primitives, giving them a shape by dragging points, cutting geometry, polygon modeling tools extrude, inset, chamfer, etc. It is quite standard.The cabin was made of primitives, giving the rounded shape with the help of chambers and manual edits. Naturally, the geometry is made by individual elements by discontinuities on a real sample. The lower part was more convenient to do without the chamfer, but give two iterations of smoothing to a normal box, adjust the roundings with additional geometry, translate into a single poly and, of course, adjust manually🙂 I thought, are not I too big to leave gaps between the elements, I do not think that I will do a separate geometry in loupoli, too "expensive" will do. Most likely, the joints will be edited later.The problems were mainly with the fact that the drawings do not beat each other (these are not drawings, after all) and, especially, do not beat with samples on the reference. Basically, the proportions and radii of curvatures. The priority over the drawings was an outside view and a visual comparison with the reference. What to do 🙄I started to make a cabin, it will be with a cabin, like here:Made of primitive, but this is not the most successful solution. The reference shows that in fact there is the same pipe frame as in the open model. Hence, it will be more convenient to make the frame splines, then add walls, as in the real model.So far, I do not make any specific details on the terrain and details, I will first get a general picture, then I'll have to figure out how and how to model to better transfer to loupoli.
This project I started doing in the 15th version of Max, in which the developers tried to make an alternative to the support edge loop - TurboSmooth - another link "crease edge loop - OpenSubdive". To whom it is interesting, look at youtube about opensubdive. I, in turn, tried to introduce this into my work. The result of the joint use of both ligaments:
You can notice that on the grill I do not have any supporting ribs, only form-forming ribs. Yes, sometimes crease does not cope and you need to add a reference edge (adequate chamfers are obtained for some reason only at values below 0.3), but the opensubdives mod works exactly the same as turbosmooth in this respect. Particularly clearly the advantages of this tool are manifested in places where there is a lot of geometry that requires support ribs, as I do with the grille. To cut it is not a problem, but my radius would look awful and I would have to manually edit the geometry.From the minuses I can still note that the sets of edges with values fly off with changes in geometry, not even directly related to these edges. Remember the selection, too, for some reason did not work, although the modifier remained in place all the time. It was necessary to make all appointments at the slightest changes anew. But let's hope that this bug will soon fix 👍.P.S. The Crease function was previously available, but the values set for the edges did not work in other programs. Now, developers argue that the transfer can be done without any problems, even a video somewhere on this topic. Ok, so you can implement it 🙂
While there is work and there is nothing to spread, since I want to finish the whole cabin, I decided to share a little advice or a couple of tips that came to mind during the work. For someone, they may seem obvious, but for someone they will be a revelation😁- if any tool does not work correctly - 99%, that you have an incorrect topology;- In the process of work, check the polygons for the number of vertices. This menu is located on the ribbon in the Selection tab on the right and allows you to select polygons for the entire object by the number of vertices. It is necessary to be on the sub-level of polygons so that it is active.Here is a small example:The swift loop does not work, although, in appearance, there are 4-angles everywhere.First of all, we check for open edges where they should not be, that is, gaps. To do this, switch to the border sublayer (key 3) and select all (ctrl + A). In this case, my red ribs are not highlighted, so there are no gaps.Next, you need to switch to the sublevel of polygons (4), go to the selection tab in the selection tab and set 4 (vertices) in By numeric, then select all polygons with, respectively, the <and> and the Select buttons, which, respectively, have more or less than 4 vertices . Here is my result:Found 4 pentagons (2 on the reverse side), formed due to uncrossed vertices in the corner.In general, I regularly use this check, which I advise you. If someone is interested to learn more about the function of the tape tools, here are two excellent tutors:Http://www.digitaltutors.com/tutorial/127-Getting-Started-with-the-Freeform-Modeling-Tools-in-3ds-MaxHttp://www.digitaltutors.com/tutorial/128-Getting-Started-with-the-Graphite-Modeling-Tools-in-3ds-MaxI also figured out how to prevent the settings meeting in CreaseSet, so as not to reconfigure everything again. After configuring the sets and modifying the modifier before modifying the modifier, the modifier collapses, which will wrap the crease settings in the EditPoly Edge sublayer, after changing the new modifier (make sure new edges assign a crease, if necessary) to CreaseSet and in it, click the auto-generate tab in the options tab. You can not touch), after which you will generate sets depending on the size of the crease. Quick and painless 👍
Welding seams I'm doing fairly uncomplicated. I have two options:- Create geometry from the edges using Create shape from selection. Ribs are selected where the suture should pass. If there are not enough ribs, you can create them temporarily. I would like to add that in order to get the most successful welding seams, you need to use Create shape from selection with the number of iterations of smoothing, which will be the final model and model in zbrush.- Use the Welder script, which allows you to create geometry at the intersection of two objects. Also with anti-aliasing.Having created all the seams, I combine them into one object. Next, the geometry of the excavator goes to zbrush (I did not send all the geometry, but only one that connects to welding to make the scene easier), seams are exported as a separate file. Excavator and seams should be separate files, so that in zbrush they were separate meshes.What to do in zbrush:1. Import both files in turn;2. Save as a tool (this is already a zbrush object);3. Load both objects into the scene;4. Through Append in SubTool we let one object to another;5. The second object should appear in the SubTool columnDone. Now you need to prepare the seams, because they are created from the edges and the geometry is uneven, some polygons are short, others are long:To fix this, you need to select seams and press DynaMesh in Geometry / DynaMesh. Of the parameters, I changed only Resolution to 1024. There is a uniform distribution of polygons:After that, I made the Divide seams (analog of the smoothing iterations) to level 3.I do not plan for large render plans or very high texture resolution, so I did not work hard on the welding seams, I quickly walked on them with a ClayBuildup brush with a negative effect (Alt-clamped). It turned out something like this:Note in the upper left corner that the seams have 700k polygons. This is quite a lot and someone such a polygon can slow down the work with the stage. In zbrush there is a good optimizer. He is in Zplugin / Decimation Master. There are not many settings, on the offsite there is a detailed instruction. Get 70k polygons on the seams without losing quality - it's easy. You can do less.But, as promised, the model in zbrush:
Before moving on to baking maps, especially the normal map, I would like to tell you a bit about the preparation of the scene and objects.1) Save the scene to a new folder with a new name, with a new number _001. This is a new stage.2) Clean the scene, leaving only low-poly and high-poly models. Throw them on different layers. I made one layer for all the high-poly parts, one for the low-poly model and one for the welded seams.3) Assign low-level anti-aliasing to the low-poly group. Normal map does not like angles of 90 degrees or more (more than 90 degrees is generally better not to do). If at you with such angles on the model one group of smoothing will be assigned, on normal it will appear unnecessary gradients and, naturally, it will not work correctly. Therefore, in such places we share the smoothing groups. Here it is important to note that wherever the boundary of the smoothing groups passes, there must be a seam on the texture scan. Therefore, you can assign anti-aliasing groups before creating a sweep, if someone so is more convenient. There will be one group of anti-aliasing for high-poly.4) Exclude the intersection of the projection modifier cage for the various elements. There are several methods. One of them is explode:Http://cgi.tutsplus.com/tutorials/how-to-bake-a-flawless-normal-map-in-3ds-max--cg-925I use a different method, when different elements are assigned to those elements that should not intersect. For convenience, the object is assigned a subObject mat with multicet material. Naturally, it is necessary to assign the same ID to high-poly elements as to low-poly elements, to which the detailing will be transferred. Both models are superimposed on each other. In general, the more low-poly corresponds to hi-poly, the better the result of baking. But we take into account the poly-account😁 I've decided that it's better to sing the most insidious places first, and then to optimize.5) Select low-poly, enable Default Scanline Renderer, disable Gamma / LUT Correction, open render to texture (0) via the rendering tab.Settings:-padding I delivered 4. Depends on the requirements of the project. If there are no requirements, then you can put more, 8-16.-projection mapping-pick-select all baked high-poly.-options:Here it is necessary to include Hit only matching material ID, if ID is used.Ray miss check will show the red in the render window those places on the texture where the information is not baked.Three main reasons:1) cage does not block high-poly fully2) there is a hole in high-poly that is not available on low-poly3) there is generally no high-poly element that would correspond to the low-poly element, for example, if the IDNormal map space. Green switch to Up, if the model with the normal goes to the game engine, like mine. In general, as far as I know, it depends on the render engine (?), DirectX or OpenGl (xs, did not go deep, I'll be glad if someone will explain or give a good link to read). As a result, in one software, the green channel should be Up, in the other Down. In Max, the default is Down, so my map to be displayed in the viewport will be inverted on the green channel.Global Supersampler. To be honest, you can bake without it, but for more good results, many include Global Supersampler Hammerslay 1.0. This significantly increases the render time, so it's worth only to include the final render.6) In the output, click add and select NormalsMap. Below enter the name, select the folder to save and the resolution of the textures.7) Exit settings, low-poly will already have the projection modifier. In it, include Shade for better visibility of the cage, reset the curved cell by the reset button.8) Now we need to get the maximum correspondence of the cell models. The bottom line is that the cell must completely cover both models, but they should be as tight as possible. Means stupidly to increase the value will not work 🙄I do this:-I throw the cage-make a small push / amount. This evenly increases the cell in volume.-select the element, switch the selection to the vertices (this is mandatory), I make the push / amount more, if necessary- I go through all the elements.9) Again open the render to texture, press the render button.In the window, baked cards are not displayed. Here we see an ordinary render, only in a two-dimensional sweep. With the checker on, you can save this image, throw it on the loupoli as a texture and see on the model where exactly the information is lost, that is, exactly where the red areas are.For today, everyone 🙂I think, tomorrow bake the cabin, then you can still disassemble the main artifacts and their elimination.
Fixed in high-poly:- in places the radius of facets is increased, especially on small details. This gives better readability of the corners formed by the normal, when the camera is distant. One of these details is on the screen. I drew a side view of the part and the direction of the projection cage. It is seen that each face is at an angle to the projection, there are no parallel faces.- almost all the ledges and depressions added bevel. Angles of 90 degrees are poorly considered when removing the normals. This is especially true for those parts that are not supported by the low-poly geometry- very small details are removed, because even in the resolution of 4k they were not drawn.- Removed all the holes that are not on low-poly.Fixed low-poly:- Geometry on rounding is added. Not everywhere, but where the angularity was striking- the welds are newly made on the basis of low-poly to better fit the shape. He dropped low-poly in zbrush, applied dynamesh with a resolution of more than 1k. Dynamesh is needed in order to literally weld individual elements, a large resolution so that the object does not lose shape. I went through the claybuildup brush where I needed the seams, disguised, inverted the mask, removed the excess, exported the seams back to max.Crane recommend doing similar actions using a tablet to save the hand🙂🙂- made a duplicate low-poly and made the following changes only on a duplicate. What is it for? Duplicate can be seriously modified and reworked, which will get better results when baking normal maps. And since the scan and the geometry of the duplicate and the original are completely identical, the map of normals obtained from the duplicate can be safely applied to the original. The most important condition is to keep the sweep as large as possible.
Made on a duplicate:- The geometry of splines and radii is added here and there. The normal is baked more cleanly.- Geometry is added where distorted details that are not supported by geometry. This is due to the fact that cage at an increase stretches at the corners, as it were up and away. This is the average direction for two surfaces adjacent to the corner. As a rule, it is enough to start on the low poly edge on the place where the curved part is on the normal.- Stitched elements, where the welds are clearly visible. This operation most spoils the sweep. Some details had to be re-deployed. In order to restore order, in the scan window I put the pre-fabricated texture of the scan with the original low poly (done through Tools / Render UVW Remplates), and on it, as in the model, everything corrected. This is a decent piece of work, but the result is on the face. Here are the screenshots of how seams looked before and after editing:- added even more geometry to the radii and fillets. Because, although some surfaces look good, unpleasant shoals emerge:1) Illusion of the gap under the welded seam. According to the geometry, the weld seams are clearly low-poly, but slightly above the high poly, hence the visibility of the gap. It is corrected either by filling the gap between the seam and hi-poly, or by adding the geometry to the low-poly radius so that it coincides with the hi-poly radius.2) Creepy artifacts. It is corrected by the maximum approximation of the low-poly radius to the hi-poly radius.Fixed on sweep:- In the main, seam rips were made where the corners were almost 90 degrees and black lines appeared in the seam areas.- just want to demonstrate what is fraught with non-compliance with the rule, which states that the elements of the sweep should be placed vertically or horizontally. Here such jambs arise:In this case, from the ladder, I could not get rid of the specifics of the details, it's just very uneven. To arrange vertically or horizontally one could only one side. The same sides (ribs), which are laid out at an angle, reproduce this artifact. I think it's because of the pixel pitch. You can fix it manually with a smudge brush in Photoshop. But I will not do it, because I do not have the function of turning the canvas ... it seems because of the video card, and without it I can not imagine how to level the line at an angle. I'll try another smoothing with baking, about which I wrote earlier. It just takes a lot of rendering time 🙄 If it works out, I'll unsubscribe.That's all I could remember. If I remember anything else, I'll add 🙂Good luck in the hard business 👍Now I'm going to start texturing and rigging.
Thanks for the help and the link.I found the scene, because I have hdri 360 and several backgrounds from different angles. And obviously they are made separately, and not just cut out. Distorted perspective, apparently just crept into the photo. But, I think, this is not a problem. That's better?And here comes the first rough render.The picture is still very, very raw. The materials are primitive, there is only color and uniform reflections. Total 16 materials:1. yellow paint2. the paint is dark blue3. plastic for rear lights4. plastic for headlights5. iron without painting6. Mirror7. Glass8. Solid black color (for holes and slits)9. rubber for the case10. rubber for trucks11. Chrome12. White plastic13. Black plastic14. gray plastic15. Red plastic16. skinWhat else is planned to do:- Work on materials. Add a variety of maps and effects for color (color heterogeneity, external effects), reflections (heterogeneity of reflection due to external influences), bump (texture and irregularities). Work on the pollution of the tracks. As this will be practically the only place with noticeable dirt (not through the air it flew into the yard😁). Work on the glass, now the salon is practically invisible. By the way, the question is:Behind the transparent glass, instead of the background, black can be seen nothing. Where is it corrected?- add welded seams. They already exist, but so far they have not added it to the scene.Stickers, etc.- adjust the lighting from the hdri card to make the shadows better and, if possible, the shade of the materials becomes more correct. Now somehow the model does not fit into the general background.And another question:Does anyone know of a good way to procedurally change the fading texture? I have an assumption that it is necessary to mix it somehow with a noise or something like that, but so far the result is unsatisfactory.After the materials are finally ready, I will make a description of the most, in my opinion, interesting samples with macro-renderer.
Oleg, welcome! 😁I see it turned out to adjust the glass. What was the problem?I think it turned out too transparent. Well this is my subjective opinion. My perception, not supported by anything. And all that I will say below is also subjective.Light is not enough on the model, or rather it seems that this light is low-contrast.Was going to dirty caterpillars - did not. Immediately the excavator flew into the yard and landed very neatly😁😁😁I do not like this metal on the ladle and where it exists on the model. It looks more like plastic, and look at the place where the elements from this material are applied (bucket + scraper) - somehow there is a depth that is not felt .... They seem to be on the same line, and only the prospect makes it clear that This is a different depth of the picture. Certainly, rust, abrasions, abrasions, scratches should be added to this material - at least to the scoop and scraper (they also worked, they should not be new). Also with glass at the bottom of the cabin - drops, mud splashes .... And with Metal pistons. On good everything must be dirty. Especially the scan did, it is already easier to be with the cards.Well I want to say - well done! I would not do that. You have something to be proud of in just a year of studying the program. It is necessary to do post-processing - to correct the white balance, from contrast, to deduce accents to the foreground and vice versa to take away the distant ... But I think you still have it ahead of you.So, while writing already forgot what she wanted to help .....😁Concerning the sub-materials:Go to the Global DMC tab in the V-Ray tab (in the Virax settings) for version 3.0 (in the versions below it is located in the Settings tab). There you will see the Global subdivs mult parameter. There is value = 1. With this, work like this: If all the subdivs in the scene = 8, then in this tab, let's say 2, then all the sub-materials in the scene are multiplied by two. If some material has 8, and another has 16, then both multiply and it turns out 16 and 32, respectively. Actually, you can also lower the overall level of subdivs in the scene by putting Global subdivs mult. = 0,5 (exactly half) .... But, as you know, this is a lazy and clumsy approach. On good it is necessary to test each material and assign to it as many subordinates, how many it demands in a concrete scene under those or other problems. (In the material itself, in the materials editor).
Thanks for the tips)If I remember correctly, then for the transparency of the glass it was necessary to remove the daw with the reflaction / refraction environment in the settings of the virai. There was a black color.I like the transparency of windows. The more obscure, together with the reflection, completely concealed the interior.Light is not enough, I'll try to adjust hdri.Excavator brand new, they have not yet dug any pit. I can not look at the exact reference, how his caterpillars looked after a trip on dry concrete, but I do not think that they are very dirty. Perhaps a bit more dusty, but I'll see what I can do with the dust, then, perhaps, and the caterpillars will correct. Why without dirt and wear? Because it will all be done on low-poly in another program. I do not see the point of wasting time on the same thing.On the ladle is not metal, it's paint, and the metal under the paint is not visible. Paint by the properties of reflection a bit like plastic, so for the basis I took the plastic and adjusted to the eye on the reference.Does the yellow color look ok? It's the same paint, only of a different color🙂. Here is one of the pictures I was guided by:I did the sweep on low-poly, not high-poly. I think that deploying a model with many millions of polygons is a waste of time. Here I managed uvw mapping.I think that after repeating everything from collecting the reference to visualization, if desired and having time, anyone will do something similar. After all, in fact it is a set of modified primitives and splines🙂 Yes, I did not describe in detail every step, but I do not pretend to be 100% correct approach to the matter, so others can get even better.
The metal paint is yellow.This is what the tree of nodes looks like for this material:If someone seems too sophisticated for such a simple material, nothing terrible, in fact, not everything is so difficult. Now I will explain.1. The main material that connects to the Multi / Sub material. This is a blend. It is necessary to mix two paint materials with different reflective properties, which together with a blending map give an opportunity to get complex glare closer to the real ones.2. Two paint materials and a Falloff card for mixing. The falloff map gives me, by adjusting the curve in its settings, the ability to make the glare dependent on the angle of view. More blurred with a direct look and more glossy when viewed at a sharp angle. The materials, first and second, essentially differ in the setting of the flare. That it had a heterogeneous structure, more pronounced in the center and more diffuse farther from the center. When adjusting the reflection, I turned off reflections on the fresnel, and threw a falloff map (3 columns) into the reflections. This allowed me to customizeFresnel reflections by hand using a curve, based on the wavelength for the material. Data on the materials are taken here:Http://refractiveindex.info/3. Here I have rendered everything that I put in the paint materials, like cards. Basically it is composite materials. They are just needed to mix several cards or materials. Very convenient, you can work as in Photoshop, using the transparency settings, the blending mode, the mask for each material.4. Nothing special, the output map for some textures. I do not remember why I did not make adjustments in the output scroll inside the maps themselves🙂 Apparently, for flexibility.5. Here the textures, raster and procedural.6. Separate text for a larger modification of the noise card Bercon-noise. Not standard, but free, look on the Internet, if needed.7. Separately rendered composite maps, which are a layer in the previous composite maps. This is a layer of labels and inscriptions.8. Raster and procedural textures of labels and inscriptions.9. Masks for stickers and inscriptions. I draw your attention that all this is mixed through the composite material with the base color, and the textures do not have tiling. This means that nothing is mixed in the base color beyond the boundaries of the texture, and nothing extra is mixed into it inside the texture, because there is a mask.And another clue about the exact and fast location of the labels. For each new sticker I put a new map channel. Then I applied a new uvw map modifier on the object to the plannar mode, in which I put the corresponding channel. Adjusted position and size with a gizmo. One minus - the image is displayed the same way and on the other side, but I have not yet dealt with this, because it does not interfere. If you need to render the reverse side, you can simply turn off the uwv map layer for a while.Here such here a paint. Next time will be rubber, slightly stained, trucks.
Hello everyone, I have not written anything here for a long time.I had nothing to write, I was versed in texturing, I was trying to texturize🙂 Now the result is, although, of course, I did not realize everything and not always the way I wanted to. At the end of the scene, because of the amount and complexity of the materials, it was terribly slow and the program often crashed, which could not but affect the result. 🙄 Although for the first serious acquaintance with Substance Designer the result satisfied me completely.In general, what is this I'm telling everything, here are the pictures:I did not make much sense as I did what I did, as I worked mostly as I learned from free official video lessons that can be found here:Https://www.youtube.com/user/AllegorithmicFrom myself I can, perhaps, describe the work with materials, as I organized it in my project:In the first field I have folders with objects, materials, textures. The materials folder is broken for convenience on a lot of other folders. Again, for the sake of convenience, I created the materials layer by layer and each separately. First he created a clean material, then he did damage to it, then he combined all the materials into the one on the site, and then he polluted the whole thing. Naturally, for each part of the excavator, which had its sweep, I did separate folders with materials that are worn and polluted. For pure materials, this is not necessary. But worn or contaminated materials often have in their composition specific for each object textures.I combined the materials in the second field, the tree looks huge, but everything is done on the same basis. Namely, two materials are mixed using a mask, which is generated from a color map (id map), in which each material on the scan has its own color. It looks like this:I did not create it entirely manually, I baked a diffuse map with highpoles, to which I assigned all these colors via id, and manually only brought a little to mind in photoshop.There is another way to use the color map, without such a huge tree, but I had some sort of jambs with mixing materials, apparently because of the incorrect setting of treshhold, which constantly tried to mix something that should not be mixed🙂Well, now there was the final touch - rigging and, perhaps, simple animation.
It should be, as promised, to show what I screwed up with texel🙂 So, in the heat of work I forgot that I have several different sweeps for different objects, for which I am going to use the same basic materials, and , Then the density of pixels per unit square of each of the objects should be, if not identical, then close to that. Why should we worry about this? Because, for example, I have a bucket and ... a front knife (or whatever it's called) for which I will have two different textures, but I want to take one material of dirt and rust and assign it to both objects. My situation with texel was originally this:I think it's clear that if the squares look different in size, then the texture of the same dirt or rust will look differently on these elements.What to do? If you have a customer, you will most likely have to redo the sweep and bring these details to a single texel. But you do not have time to make textures, since the scan is checked before the beginning of texturing😁 If there are no such strict requirements, then you can do this as follows:1. Correct the scale of the invoice for a particular object. In the sub-designer, I added a deformation node and scaled all the materials that contained the texture. At me it turned out not super precisely, but it is obviously better than it was originally🙂2. Reduce the size of the texture, which got a more dense texel. In my case it's a front knife. What for? At least in order for you to have textures of the same clarity. Initially, I have a very small checker on the caterpillars, but I do not need such a pixel density. If next to the metal parts with a texture of 4k I can only see lumps of dirt, then on rubber caterpillars with a texture of 4k, I can see every grain of sand. This difference immediately catches the eye. So, on caterpillars I'll use a texture of about 512.That's all. Rigging I will not do in the near future ... because I do not know how 🙄 I wanted to do everything cool so that not only the details would move with the reverse kinematics, but the wires would move realistically. But something it looks like a lot of work on learning rigging and animation, which is not yet part of my plans.Thank you all for participating 👍