I need to make a hole (with specific diameter) in a cylinder (with specific diameter). I’ve read couple of posts about my problem, watched couple of tutorials but still I encounter the same problem. Even when i have - in my opinion - clean topology the hole still gives me weird shading around its sides. So, am I wrong assuming this is a clean way to do it or is a clean hole in a cylinder without shading problems impossible?
Any rounded surface needs a high enough density that areas that diverge from the rounding do not stretch greatly from a square shape.
In your example, your tube needs at least two more vertical divisions to correct the oddly shaped polygons at the corners.
For the inside of the hole, use a number of evenly spaced loops instead of trying to control the flow with one really tight loop.
The key to maintaining good form is polygon regularity.
Clean topology is not minimal topology, it’s ‘regular’ topology.
It helps to model with subdivision surfaces turned off in edit mode.
Hey, A wire frame overlay would be nice so we could see the topology, but I am going to assume your using sub’d modelling, so topology is very important. Check out these links if you have not already:
Hope my tutorial helps you can just use the ctrl plus r on the inner face and place a loop cut near the flat face on the inside of the hole.
Don’t really know why this works but it does. Regards Terry
@Ben_Morrison thank you for usefull information. To be honest I thought that minimalistic topology is as important as polygon regularity (obviously, i failed in the latter). As i dig trought the problem throughout the last couple of days I always end up with at least four stretched quads. The higher subdivision of the main cylinder the smaller these quads are but the irregularity near the edges remain (but less and less visible of course).
@BlenderTC thank you for your tutorial but it’s about making holes in flat surfaces and i can’t apply your ideas succesfully on a cylinder. Maybe my limited imagination is to blame
@ajcdfin thank you for the link, i’m still reading that
@Roken i recreated your topology but still the distortion near the edges and shading problems remain.
I’m - “happily” - making holes in a cylinder over and over again and trying different things, i’ll let you know when i end up with something better because failure is not an option thank you for your responses once again
Hope this helps you more regards [email protected]
Also, in ortho mode the circle does not look like one. Select your most outer circle, press space, type “to Sphere” and set the value to 100 (or move your mouse cursor). Search with space must be selected in the prefs when using 2.8 and above. The addon Looptools has the same option.
Do this for all circles one after another or use a perfectly round cylinder as base and dont scale or move it’s circle shaping loops. Manually grabbing or moving one the verts that join both cyclinders will give you shading errors in close to every situation. You have to work with hiding, the 3d cursor and orthographic views alot.
Same principles add to your mesh:
Yay! I’m famous
I managed to get quite clean effect by:
- matching the number of vertices of both cylinders (32 main and 16 for the hole worked for my dimensions)
- using boolean modifier, applying it
- merging vertices from the main cylinder closest to the hole cyilnders
- deleting remaining n-gons and dealing with them manually by adding loops and joining vertices
After that I ended with topology similiar to this you presented in the video. It is not perfect but i think it is close enough to perfection.
One thing intrigues me: at the end of the video you mentioned that I could scale the hole to the size matching my dimensions, but scaling will more likely deform the overall curve od the main cylinder in the place where the hole takes place (the hole would scale inward/outward its median point). Is there a way to scale along the surface of the main cylinder?
My nickname is a polish sentence, i didn’t think that trough when it comes to english pronunciation
you deselect the direction going into your model so the scaling is only across two axis and so stays on the surface not perfect but usually good enough.