Jump to content

line-weight

Member
  • Posts

    3,708
  • Joined

  • Last visited

Everything posted by line-weight

  1. Thank you - I will have a look at that. I too have come to the conclusion that I am going to have to draw the path as a segmented line. This can be done by changing the degree of the NURBS path (to create more vertices) and/or converting each vertice to a corner type. Unfortunately this then results in the EAP object becoming hollow, which means I can't then subtract from it (for example to make the arches through the viaduct). So I'm now looking into using a loft instead of an EAP.
  2. @Matt Panzer would it be fair to classify this as a bug? Effectively it would seem that there is certain geometry that is unusable/undrawable in Vectorworks. It's not something obscure that I want to draw - just something with a gradual curve. It looks like the only way I can do it is to manually draw it as a series of short, straight segments, that are shorter than the smallest faceting I can get VW to do.
  3. An observation: if I draw a circle and zoom in (in OpenGL) there is a facet approximately every 5.5 degrees (image on left, orange line is the 'true' curve) But if I chop that circle back, so it's an arc with a sweep of only 5.5 degrees, it's faceted into much smaller pieces (to the extent that you have to really zoom in to see it, image on right) So it seems to be related to length of curve as well as radius (at least, as far as a circle is concerned, not sure the same applies to a NURBS path)
  4. I don't think that's true - the facets are the same size regardless of how far in or out I'm zoomed (at least, looking at it in OpenGL in VW2018) edit - fiddling around, I think I see what you mean, regardless of the radius of a curve in units, the 'closer' you look at it the more obvious they are. There is a facet per x angular degrees of curve. As @Kevin McAllister's diagram shows. Strange thing is that pasting one of the objects into a new document seems to halve the facet angle. I don't understand why; I can't seem to find any settings that are different.
  5. Thanks for the answer, which of course is not really the one I wanted to hear. So... VW can't draw large radius curves. OK. I assume the 'large' is relative to drawing units? So it could potentially be improved by changing what my drawing units are?
  6. VWX file with relevant geometry attached: nurbs-v.vwx
  7. I am having similar issues, but can't seem to resolve it by changing any render settings. The images below show a viewpoint above a viaduct that diverges into two branches. Each of the viaducts is generated using a nurbs curve (in an extrude along path). In wireframe everything appears as it should - nice smooth curves, and the point where the two viaduct parapets converge (near the bottom right of the image) is correct - they join in line with each other. However in OpenGL, they are heavily segmented and the meeting point appears well out of alignment as a result. In hidden line, they are a bit better, but the same problems are there. Is there any way I can improve this? Is it because my Nurbs curves have too few vertices (I want to keep them to a minimum to ensure smooth curves and for filesize)
  8. @drelARCH all I can do is offer my sympathy, I am still using 2018 and fighting various dysfunctions in sectioning. Just the other week I quite literally lost a whole day trying to get it to section a site model properly. Disappointing to see that it's still a thing in 2020 and for me another reason not to spend the money upgrading. The ability to draw an accurate and clean section is completely fundamental to architectural drafting, and it's something the software should be getting right all of the time.
  9. I just tried it on a DTM in one of my drawings. Moved it up by 100m. The contour line labels did change - after updating the DTM. (I'm on VW2018) I'm a bit confused though, did you move your DTM up by 1300ft, or did you move the elevation of the design layer it's on?
  10. How are the contours created? I've sometimes got very confused when what I assumed was a 3d polygon all on one plane turns out to have one vertex that is 1000 units higher or lower than the other, could something like that be going on?
  11. It doesn't seem to have been marketed much at all in the release of 2020. There was mention of something about 'horizontal sections' and I was curious about that, but it was not explained at all and I assumed it to be something trivial. I'll reserve final judgement until I actually try out 2020 of course, but what you describe gives me the most positive feeling about a future for Vectorworks that I've had for some time. Finally an indication that somewhere someone making strategic design decisions 'gets it'. I've put off upgrading to 2019 and 2020 so far, partly because of perceived unreliability and general disillusionment but also because none of the developments introduced seemed particularly worth it. Had I understood better the thinking behind the horizontal section advances, I might have been more tempted. I've been banging on for some time about how VW should stop introducing new bits and pieces and focus on sorting out the core functionality. This sounds like an actual move in that direction. Back in 2016 I wrote a rant about what I saw as a dead end for top/plan view. Turns out that some of what I wanted is finally coming true. Pleasing news.
  12. That's another question I had - does doing this make a significant difference to overhead, either when zooming around in OpenGL 3d views or when things like section viewports are generated.
  13. By the way I realised back in 2016 there was something going on with horizontal viewports; I even asked about it on here because it was confusing me - https://forum.vectorworks.net/index.php?/topic/46515-are-there-two-confusingly-slightly-different-types-of-section-viewport/ and never really got an answer, but now it makes more sense. So is there now only one type of section - however you create it - with the difference being controlled with the 'display 2d components' option?
  14. @Matt Panzer thanks very much for taking the time to answer my questions. I now understand. I am still using 2018 which is why I'd not realised these concepts had been introduced into 2019 (I was aware of 2d components being added into 'normal' sections but hadn't appreciated that this was now part of horizontal sections too. Once I get onto 2020 I'll look forward to exploring this a bit more - it sounds like a positive step in the right direction. In fact it means that I can continue with my own horizontal-section approach to floor plans, with VW features converging with it rather than the opposite, which is very encouraging news.
  15. Let's say the object I want to end up with can be made from the following objects, added and subtracted as follows: (A+B) - (C+D+E) It could also be made, from the same basic objects but as follows: (A+B) -C -D -E Or ((A+B) -C) - (D+E) This is an object that I want to be able to edit by going back to edit any of the objects A-E, so it will stay as an addition/subtraction and is not to be converted into a generic solid. Other than the configuration that's most convenient for me in terms of subsequent editing, is one of these better than the others, for any reason? For example the first option involves 3 operations (an addition, another addition, then a subtraction) whilst the second involves 4 operations (an addition and then three subtractions). Does this have any significance for things like memory, or likelihood of the resulting object getting corrupted? In reality the objects I'm wanting to make will have more than just 5 component solids.
  16. @Matt Panzer actually yes, I do want to create the construction documents with pure 3d sectioned geometry, and this is what I now do, because a while ago I decided that top/plan simply doesn't work, at least for the kind of buildings I work on (appreciate this might not apply for everyone). I found that it could not generate floorplans that were of acceptable quality. In fact the issue raised in this thread is an example of the kind of thing that just doesn't work properly Yes, it can be frustrating that my section-generated floorplans can take a while to update. I do use top/plan for editing purposes - especially for things like laying out doors, windows and walls, but the final drawing is generated from a horizontal section. Would it not be preferable to try and speed up the process of updating sections, instead of continuing with the top/plan concept which just seems like a kind of massive workaround to me? Can you clarify exactly what a "horizontal section viewport" is though? When I look at the all my horizontal sections, they don't have a "display 2d components" option - but is this because of the way they are generated? Is this something to do with generating them from the clip cue rather than from an elevation view?
  17. I'm confused... is this "horizontal section" a new thing in VW2020? Is it some hybrid version of top/plan? A real horizontal section would not have this problem, because it would be looking at the actual 3d geometry of the 3d part of the stair object, and would know which bits of the stair were hidden by the slab. How do we end up with a situation where VW is suspending a 2d symbol in 3d space? It doesn't make any sense to me - either you stick with the top/plan concept of 2d symbols layered on each other, or you go proper 3d, and project the 3d geometry that's actually there.
  18. Selecting site modifiers is a pain. One simple thing to watch out for though - make sure you don't have the site model itself selected. If it's selected it's then impossible to select the site modifiers at all. This really confused me for some time. This behaviour is different to everywhere else in VW, where if one object is currently selected, it doesn't stop you from selecting something else instead.
  19. It doesn't look to me like @Danilo is talking about annotations - they talk about line types 'in' the viewport.
  20. Yes - to some extent. It's mixed with a feeling of nervousness that I will break something in the model, and then *all* of my plans, sections and elevations will need fixing, something that doesn't happen when you're doing things in 2d. But, so far, nothing too bad has happened. Absolutely agree about a clean model. It means investing quite a lot of time 'up front' but you get the payback at the end. There are some changes to the design that are ultra easy - adjust something a bit, and all of the drawings update, just like that. Unlike in the old days where a relatively trivial change meant a tedious process of going through all the relevant drawings and updating them one by one. Then there are some changes that aren't as easy. These are where VW's parametric tools can't do the job. For example, on this one, the walls were directly modelled - each layer of the buildup. This is because they had funny geometry that the wall tool simply couldn't cope with, and in any case it can't cope with the complexity of most eg. wall/floor junctions in such a way that I could extract construction details from them. So, moving a wall 100mm means making sure I catch all the relevant components, move them together, and then go round each of its junctions with other walls, floors and ceilings and extend/trim them as necessary (again component by component). This takes quite a lot of time... certainly compared to moving a wall that's been made with the wall tool. Now the project is being built on site ... there are certain changes I've decided it's not worth making in the model. For these, I can get away with annotations and a bit of 'cheating'. It might not matter if there's a slightly inconsistency between two drawings because the change has been discussed with the builder, and what's needed is understood. All that the drawing has to do is perhaps confirm a dimension or two. It's all about trying to make a sensible decision about what's necessary to draw (as has always been the case with draughting).
  21. @Jakerhp you might have seen from the other thread that that fix didn't work on another model. However - another thing to try: make the DTM such that no point along its edge is higher than any point within it. (I did this by drawing a 3d polygon around all of the source data, with elevation 0. Then cropping the model so this edge wasn't visible). This seems to have fixed 99% of my sections. (So far)
  22. ^^ well, I thought I had it, but trying to apply this to a more complex model with the same problem has failed. The only way I can get things to reliably behave is to remove all of the "spoil pile" site modifiers. They clearly corrupt the DTM somehow. So, giving up on this for now. Another few hours of my life wasted. Thanks Vectorworks.
  23. Yes. Since I wrote that post - I've taken one job (architectural, small scale) through to construction stage where none of the drawings have used top/plan. All floorplans are horizontal sections. I also managed to create quite a few of the large scale detail drawings, for construction, directly from the model, with relatively little adjustment in the annotation layer. I've had to accept a few compromises in how my floorplans look, but on the whole have managed to get much closer to what I want, with less fiddling around, than would have been the case had I tried to generate the plans for that job using top/plan mode.
  24. @Alan Woodwell I am not using it for cut & fill calculations in this instance. I am simply using the spoil pile modifier to create areas that are slightly higher than the underlying terrain. In this case, they are 100mm higher, and represent areas of paving with a 100mm kerb around them. This is the only modifier that will raise the site model surface parallel to the underlying terrain, rather than creating a raised area with a flat surface. Maybe you have been confused by my OP where I talk about problems showing 'fill' in sections. To clarify, when I'm talking about 'fill', I'm not talking about cut and fill. I was talking about the solid red fill style that is applied to the cut plane in a section viewport. It was missing in my section viewport - in other words the site model was showing as 'hollow' rather than solid red where it was cut.
×
×
  • Create New...