Adjustable line thickness

I absolutely agree with the extra mode in your last paragraph. Most simple to implement too.
And maybe also no fade-to-black.

Or something else instead of a black background that doesn’t shock the eye…it really is very distracting and uncomfortable, particularly when you you’re sitting at a desk and/or using stationary guardian boundary (which works in any light conditions, unlike the room boundary which is light sensitive). When you stand, you navigate differently, as you rightly pointed out earlier. Thank you!

If there is any other app you should copy your movement controls from, it’s the google earth pcvr app.
Of all the different apps & implementations I’ve tried, theirs felt the smoothest & most natural. With options that cater to jelly legs and iron stomachs alike without tons of config options.

  1. Line thicknesses continued: I noticed that lines drawn on transparent or semi-transparent objects inherit the transparency levels. While from previous discussions I understand why this may be the case, is there any way of this not happening, please?

  2. Object outlines: it would be good to have another line visibility mode: object outline where the outline of curved objects (where lines between triangulated surfaces are invisible) is made visible. In certain light conditions, and particularly when textures are not applied in SketchUp, objects can look undefined. I believe SketchUp has this line setting somewhere.

  3. teleporting and dimensions in online mode: I noticed that in online mode only (not also in locally connected mode), both the teleporting and the dimensioning tools appear to bounce off objects present in invisible layers. While one can still navigate by dragging oneself around the space, the dimensioning tool is less forgiving. This can be quite frustrating at times, particularly for guests or the less initiated ones.

  4. Interactive components in online mode: as an extension of the problem at 2 above, interacting with dynamic components suffers too as other invisible objects get in the way.

Thanks for looking into this soon, there may be a quick fix to it.

@Christopher_Overbeck: personally, I ended up sick after half an hour the first time I used Google Earth. After some more experimentation I now know what I can safely use and what I should avoid in Google Earth. But that’s the kind of experience we want to avoid as much as possible with VR Sketch.

  1. Edges of semi-transparent faces: oops, indeed they should not be transparent. Fixed.

  2. Outlines: this is problematic with VR. I tried to play with it some time ago but the issue is that in the scene, there are often faces that are visible from one eye but not from the other. Consider a cube where your gaze is grazing a side face. The edge between the side face and the front face is the problem. For one eye, the side face is just not visible and so the edge is drawn in black. For the other eye, both faces are visible and no edge is drawn. This makes the view highly inconsistent between the two eyes. Of course, the side face itself is only drawn for one eye, but our brains expect this kind of inconsistency. But having a strong black edge in one eye only when both eyes clearly see the edge in question, this is really not OK. I don’t know how to fix this problem…

  3. Oops. Thank you for the report. Yes, when viewing a cloud-stored model, the invisible layers are present (but hidden); this is different from the other modes, where the invisible layers are not present at all in VR Sketch (they are requested from Sketchup later if that layer is made visible). I’ll fix this to properly ignore invisible layers.

  4. Same underlying problem as 3, I guess.

I’ll fix and send an updated Quest version here, also with an attempt at making a single slider for the width of the edges.

I’ll also soon start to post beta versions of VR Sketch 18, improving performance a lot. If you want to help us (even more than you already did with all these discussions), it would be great if you could test it! I’ll make a new post in this forum after testing some more myself. I needed to change many internal things for VR Sketch 18, and so it’s likely I failed to test everything!

Here is an updated Quest build with many of the things we discussed here: https://vrsketch.eu/download/baroquesoftware_vrsketch_v17.0.10PRE3.apk

Like before, installation instructions: https://vrsketch.eu/docs-getting-started-quest.html, “Installation without using the App Lab”.

I can’t imagine how much it must suck to be a VR dev when you can’t engage with half the content. The only thing that’s ever been to much for me is spiderman VR.

I take the point of view that VR technology is not ready yet. The very old CRT monitors could cause dizziness for various reasons; with time, it was fixed. The problem with VR is more software than hardware, but otherwise it is similar. I’m sure that technological improvements will help, but most importantly, tools and APIs and knowledge will develop over time until applications are safe by default. For now, anyway, a large portion of VR content is either cinematic or games and I’m not feeling particularly bad about not being able to enjoy half of them.

Thanks for this.
navigation seems to be working fine online,
transparent lines are now solid- looks great.
looking left and right is definitely much much better.
Dimensions unfortunately don’t seem to work properly as invisible layers are still in the way.
Also, is there a way of adjusting section line thickness?
Will keep eye open for more.
Many thanks!

Interaction with dynamic components also works a treat now- thanks. It would be good if they didn’t revert status in editing mode but we’ve been there before :slight_smile:

Not tested for shared status of interractive objects (all viewers see the same thing) when the model is in collaboration mode (whether online or local connection)

Also not tested to see whether all viewers in collaboration mode see the same light setting. This would be most welcome, please.

Also, can we have X-ray mode, pretty please?! And, just like with the fog settings, can we please adjust the opacity so that there is only as deep as you can see through a wall as you allow it. This would be immensely useful in identifying structures behind walls without making layers invisible. Not sure I saw this done anywhere else in a cad program so that could be your own lovely little innovation :slight_smile:

Finally, when in section mode, not sure how difficult it would be to have clipped shadows as an additional view setting. the shadow setting as you have it now is still very useful as it can give you a truer sense of how deep direct sunlight penetrates a building but for presentation purposes, clipped shadows would be better. I can tell tho’ that the section mode is pretty task intensive… so this may add to the challenge of keeping the rendering ‘light’ .
Thank you!

Oops, found why the tape measure tool still didn’t ignore invisible layers. Fixed.

The section line is rendered like a normal edge so far, but maybe it could be a bit thicker.

X-ray mode is very hard to do if we want to keep a minimum of performance. The problem is that in that mode, all faces everywhere are visible and must be drawn. This can take many, many times longer than a regular scene, because there, even though all the faces are theoretically drawn too, most of them are behind other faces and the GPU is extremely good at figuring this out and dropping them without further processing. And I mean extremely good: it takes longer for the GPU to load the 3 positions defining each triangle from its main memory than doing all the math needed to figure out the triangle is behind something else. So far, I don’t see a way to do it. Even if you want only one level of see-through, i.e. see through the first face but stop at the second one, then it might double the time taken.

Shadows in section mode: do you mean that the problem is that the geometry that is hidden by the section plane no longer cast a shadow? And you would like the option to keep the original shadow of the complete, un-cut model? Yes, this would be easy to add.

Dimensions: super, thx!

Thicker section lines would be great, please - thx.

X-ray: I thought that would be the case :frowning: Maybe for the next gen headsets :slight_smile:
Shadows: not a problem, just nice option to have- thx.

Thicker section lines: done!

Shadows of geometry hidden by the section plane: done too. It’s an option in the section plane’s context menu. You can set these shadows to On, Off, or Half-strength… For the last one I found a simple way to implement it, so why not. It would be much harder to implement proper partial-strength (or even colored!) shadows when the light goes through a half-transparent material. But it turns out that in this simple case, we can use dithering: half the pixels opaque, half the pixels fully transparent. With the smoothing we do at the end anyway, it makes a half-strength shadow.

X-ray: if the goal is to see behind walls, I can think of some other approaches that could be implemented efficiently, but they might be too gimmicky. For example, you could have a tool like a laser pointer, but the faces it points to are drawn transparently, up to some depth (configurable?). It’s a lot less costly to find faces that touch a line and render 5 or 10 of them transparent, than it is to render everything in the whole model transparent.

Well, very interesting! The same reason why I mentioned the fog in my earlier message, something that allows you to set a depth. Actually, I also do not see how being able to see the whole model in XRay in VR (unlike in SketchUp) could be anything else than disorienting and so, totally useless.

But being to locally explore under the skin, like Superman, that would be super cool. You can even call it Krypton mode :slight_smile: … In the same way you scale the model with the two controllers, you could open up a surface of the model to peer in behind it by controlling a ‘soft-surfaced’ sphere or cube with adjustable diameter/diagonal, perhaps … ? Only the surfaces within the sphere/cube would be rendered in X-ray, so very local?

Or a circle at the end of a laser pointer as you suggest projected on the nearest visible surfaces which then defines an ellipsoid which you control its diameter with the joystick left-right and it’s depth ( ie transparency level) with push/pull….

Just played with it; it’s simple enough to make faces inside a given sphere invisible. The sphere is (for now) a constant size at the end of the Tape Measure laser, when Tape Measure is set in a 5th new mode. The result is quite funny, and also possibly useful! We might actually want to add it even though it feels “not serious” :slight_smile:

I think you’re quite enjoying yourself now :slight_smile:
Sounds intriguing.