Відео

I have heard of things like real-time mocap where live rig data from a mocap suit is fed in to drive a mesh, which can then be reskinned using an SD model. And there are a few pipelines that generate textures that then get piped into Blender. I have no idea if the actual mesh scene with face and vertex info could be fed in before rendering to frames, but it seems there are so many tools being produced, who knows, might already exist or be on it's way?

Hi there, I teach a broad range of things, from traditional drawing to 3D graphics, fabrication and interactivity. I make these videos to help with my classes, but I also keep them around on UA-cam in case they are useful to anyone else, or in case I make some dumb mistake, in which case people usually correct me in the comments, and I try to fix my approach 😀

If the pattern is articulated by the geometry of the mesh (not just a texture), then you need to adjust the Freestyle export settings in the View Layer tab (below render and output in your Outliner). Try 'Crease' or 'Suggestive Contour' or 'Ridge and Valley' and see if you get the results you want. If the pattern is just a texture, you can always render an image, then open it in Inkscape and do a bitmap trace to generate vector lines.

Lol. Thanks. I guess it's not really supposed to make sense by itself, it's a documentation reel from an exhibition we did 3 years ago that we usually present in context, either on our website or when describing what certain parts of the exhibition or performance program looked like to people who want to know more about it. I didn't expect many people to watch this out of context, if I did I could have put a voice-over or some text in.

Yeah, I can see that this must have looked very random, I'm amazed that it made it into people's feeds. I've updated the video description to give more context. This short reel is supplementary material for when we want to present some of the artworks and performances that were part of an exhibition and performance series we made 3 years ago. More context on the video is usually given in person when we present it, or when it's embedded in a website.

@@peternelson8856 I think the fact it had little context actually made it more interesting to me personally. I also think it's the reason it reached my feed so its a welcomed surprise.

​@@skrillex544 That's super funny. Glad you enjoyed the out-of-context exhibition documentation!

@@sakharov00in hi there, sorry for the delay! Did you press the play icon in TouchOSC to make sure you are sending data?

omg lol. Thank you! There was a crazy flood in my neighbourhood the day I recorded this, clearly my brain wasn't 100%. Thanks for pointing out the error!

@@peternelson8856 Ouch - I hope the neighborhood is recovering well...... Lol - at least you hard a good reason. Many people just forget...

Hi there, sorry for the late response. I don't know the answer to this question, other than to export .svg frame by frame and then use CSS or something like that for your animation. There are other UA-cam tutorials available for this sort of thing.

You're welcome, happy printing!

You're welcome!

If you go into the 'View Layer' Properties tab, you can change which types of edges that will be rendered as vectors, from visible, to silhouette, creases, borders, and so on. To render all edges, use Edge Mark, but to follow certain geometry features, untick Edge Mark and play around with border, silhouette, crease, etc. Does that solve the problem?

If you go into the 'View Layer' Properties tab, you can change which types of edges that will be rendered as vectors, from visible, to silhouette, creases, borders, and so on. To render all edges, use Edge Mark, but to follow certain geometry features, untick Edge Mark and play around with border, silhouette, crease, etc. Does that solve the problem?

@@peternelson8856 will try.

Amazing. I'm happy to hear that!

There are some settings to play around with the colours of the vector lines, but as far as I'm aware, you can't convert colours from the shader to a vector file directly. But read the documentation here and if you find a way, let me know! docs.blender.org/manual/en/latest/render/freestyle/view_layer/line_set.html#face-marks

Have you ticked all the settings for Freestyle export as well as set an output location for the SVG file?

Can you be a little more specific - which dimensions are changing and in what part of the process?

@@peternelson8856 I created a mesh (100mmx25mm) in blender. When I open from inkspace svg file, it’s like 8 times bigger than what it supposed to be. I still didn’t figure out why it is changing.

@@elifmiami Yes, sorry this is something I need to research a little further and probably need to record another video. I think it might relate to the resolution of your output and the page setup you have in Inkscape. It is discussed in a number of forums: inkscape.org/forums/cutplot/svg-scale-exporting-off/

Yes sorry about the sound. I had the microphone setting wrong for a while. Hopefully I'll find the time to fix it

@@peternelson8856 no worries, don’t mean to discount your efforts(: you helped me figure out exactly what I needed to.

Hi MikeGG, I don't know the answer to your question off the top of my head, but from a quick search, it seems that this forum post might have your answer: blender.stackexchange.com/questions/151299/plot-single-vertices-based-on-xy-coordinates-from-csv-file . I don't know if this would be a super clean solution for GPS data, but I'd love to see the result!

@@peternelson8856 Thanks again, indeed that got me started! If you read the CSV into a list mylist = []; while... mylist.append([float(x), float(y), float(z)]) you can then create the point cloud like this pc = point_cloud("point-cloud", mylist) Unfortunately that opens up a whole can of worms on how exactly to go from a point cloud -> surface.

@@manmustbuild Ever got further with that?

Hi Hadrian, do you mean adding a 360 camera to every GPS co-ordinate within Blender? If that's the case. you would replace the "bpy.ops.mesh" line in the if statement with a line that adds a camera at the float x, y and z coordinates. To work out exactly what this should look like in your Python script, I recommend you change one of your Blender windows to 'Info' and add a camera. The info panel will show you the python command for actions you do in the editor. For example, when you add a camera, the info panel shows "bpy.ops.object.camera_add(enter_editmode=False, align='VIEW', location=(0, 0, 0), rotation=(1.10871, 0.013265, 1.14827), scale=(1, 1, 1))" , and when you change the camera type to panoramic, it will show "bpy.context.object.data.type = 'PANO' " . So if I were you, I'd make a csv with a small number of values and test a script by changing the add cube line to an add camera and set camera to panoramic and equirectangular. If you can get it to work with a small .csv file, then I think you'll be on your way. Have I understood the question correctly?

@@peternelson8856 Hello PN, Thank you for the prompt reply! You certainly did! I will make this attempt and see what sort of results I can get. I find this very exciting and can't wait to experiment. I will share my findings with you and ask if I have any further questions - provided that is alright with you. Thank you so much and have a great week!

​@@hadrianlaing7251 Absolutely. I have no idea what you're making, but would be super keen to see the results! Best of luck!

Sorry for the late reply. Haversine formula converts longitude and latitude to metres. You can copy the formula setup from this spreadsheet: docs.google.com/spreadsheets/d/1p0xs_bRE3p-d9-lKL7cuPmq6sAavJ9LIW1U_pi3PDdw/edit?usp=sharing

yeah, you can tell I'm a real pro at this lol

Check the Freestyle settings, but there could be outlying factors such as clipping settings on your camera?

Check that you've enabled both the FreeStyle SCG Add-on in your Blender Preferences and that you've enabled Freestyle in your Render settings

probably if you reverse normals?

In the View Layer Properties, scroll down to the Freestyle settings and try changing the type from 'visible' to 'hidden'. I think this should give you the hidden side you are after?

don't use the search function, you need to drag the little rectangle to where you want to download. Also, I have noticed that the website doesn't work that well on Safari

It seems that terrain.party is having some problems. Here is an alternative: tangrams.github.io/heightmapper/#12.48762/22.2338/114.2226