There's a thread here on polycount somewhere called How you model dem shapes. It has a lot of shapes and tricks to model dificult shapes so you might want to check it out.
I seems like the shape on the bottom right is unnecessarily high poly. On the shape on the top left you can see your dent is being caused by an imperfect circular loop, I agree how you are reducing your edges, but if I were you I'd do that junction away from your hard edge. See image attached
I seems like the shape on the bottom right is unnecessarily high poly. On the shape on the top left you can see your dent is being caused by an imperfect circular loop, I agree how you are reducing your edges, but if I were you I'd do that junction away from your hard edge. See image attached
HI I did as you said and still getting the same error
To add to the good advice jose.fuentes and wirrexx have provided:
It's often helpful to block out the basic shapes to dimension, add the support loops to the primary feature(s) then adjust the segment count on the secondary features to match the segment count on the primary features. Blocking out the basic shapes will also help highlight where and how the shape geometry intersects. This helps inform what the topology should look like and what modeling strategies can be used to develop the shapes.
The pinching in the transition between the countersink and the hex broach is caused by: mismatched segment counts, edge loops that end inside a complex shape transition and inaccurate shape transition geometry. Jose and wirrexx have covered the first two issues so I'm only going to cover the shape transition geometry between a round countersink and a hex broach.
Here's a comparison of two strategies for developing the shape transitions. Using a flat loop to transition between the round countersink and the hexagonal broach results in each segment of the chamfer having an inconsistent angle. The inconsistent slope angles causes undulations in the shape transition when subdivision smoothing is applied. A more accurate approach to this shape transition is to use a 3D loop that has been cut out of the countersink chamfer and maintains a consistent slope angle.
The advantage to using the flat loop strategy is it's quick and fairly straightforward: Block out the basic geometry, add the support loops to the primary feature, match the secondary feature's segment count to the primary feature, bridge the edge loops and add a support loop above and below the transition area. Although the shape isn't completely accurate it's probably good enough for small fasteners that won't be viewed up close.
Cutting the 3D loop out of the countersink takes a little longer but the result is more accurate. Start by blocking out the shapes (ensuring the segment counts will match when the support loops are added), add a loop half way up the wall of the hex, run a Boolean union, delete the resulting n-gon ring and add support loops to the hex with a chamfer / bevel operation, bridge the loops between the hex and countersink, create the rest of the fastener's head geometry and add support loops around the shape transition.
Blocking out the shapes, matching segment counts and using the optimum amount of geometry to hold the desired level of detail will help inform which strategy is best for your application. Something as small as a regular fastener head isn't something that's likely to be closely inspected by players. For most assets it's going to be a case of close enough is good enough.
Replies
\