Posts by Zoom

    I had trouble uploading pictures, so I had to reduce the size of the pictures, which affected its quality.

    I think he is talking about the quality of the models. For example certain things that should be oval (e.g. the basin) look polygonal.


    That said this is a very challenging scene. It has a mirror, that enscape can't render properly, and it also uses artificial lighting, which is probably the second weakest point of Enscape when it comes to rendering quality.


    I'd really like to see more scenes like this (directly from Enscape, without post) to see what the limits of Enscape are under such conditions.

    Sketchup is slow, it is not the fault of your system and adding more RAM wouldn't make any difference.


    A CPU with a higher clock speed could help a bit, and I guess Enscape could utilize an even better GPU (but the 2070 super should be plenty), but what you are experiencing is really Sketchup's fault, not a deficiency of your system.


    Try the following: Keep your materials and outliner panels closed at all times, and open them only when you need them. Sketchup has some bugs which make certain actions slower when those panels are open, especially when working with larger models. Also, keep shadows off, and turn them on only when you need them. Lastly in "View -> Edge Style" have only "Edges" turned on, and turn off "Profiles" and everything else.

    BUT! we as users have to accept that this is much more difficult to achieve for the developers, than simply turning on a feature in the code.

    Enscape is great. Compared to traditional rendering engines you usually give up a little bit of fidelity for far greater speed. This is an excellent trade off in most situations for the target market of Enscape (which I assume it is architects, interior designers etc). However this argument collapses when it comes to mirrors. In this case the result is not acceptable, and it doesn't matter how fast you can produce a bad result.


    Clearly it isn't easy, but my understanding is that the main issue is the time that such renders would take, and not that such thing is technically impossible. IF this is the case, then we should let Enscape know that we understand that such render is impossible to be done in real time, but we would still love this feature (as an option) for when we render the final image, even if this would take 50 times longer to render.


    If my understanding is wrong, and such thing is just impossible with the existing engine, then Enscape still needs a solution. How about having the option of using Cycles (of Blender) which is open source and with a permissive licence that lets it be integrated in commercial software? https://www.cycles-renderer.org/

    This way Enscape can have the best of both worlds: A great real time engine & an alternative traditional rendering engine.

    "Oculus Connect" is the name of the conference. Probably he means the Oculus Link, but that is a wire, not wireless.


    For wireless there are apps like ALVR. I am using it with my Go and it works fine, but there is a bit of additional latency which could potentially cause motion sickness.

    Thank you for letting us know. Now if it comes with 2.8 it will be pleasant surprise ;)


    Would the assets created with Pieter's program continue to work once the official Custom Asset import is implemented?

    There actually is already a way to control tree height instance parameter. If you link Revit family to asset it creates a height instance parameter in your Revit family. You can then use this to control the height of the rendered tree.

    Are you talking about scaling it down? This is not what I mean, because a tree that is 2 years old is not just a scaled down version of a tree which is 20 years old.

    +1 on FBX / Obj imports (Custom Library).


    +1 on different Hedges at different heights (with a few variations)


    + 1 on supporting laubwerk plants (for Sketchup at least, for which they have their own plugin)


    Also:


    - Some younger trees. A lot of the trees are huge, which is great for creating a forest, but for landscaping a new project it would be more realistic to have younger trees (i.e. trees that can grow tall, but while they are young and not very tall yet)


    - Since you are taking requests for specific items, mine would be Cypress trees ( 5-7m tall)

    In the Enscape 2.8 development agenda it is mentioned that "Custom Asset Library Import" is a "Stretch Goal".


    That post was last updated in March, and the ETA for 2.8 is end of June, so I wonder what the status of that feature currently is.


    Personally I think this is the most important feature by far compared to everything else mentioned in the development agenda for 2.8.


    The reason I am asking now is because I am debating whether or not to create a custom library based on Pieter's Tutorial, or if I should just wait for a month or so for the official Custom Asset Library Import.

    OK, many thanks!


    I managed to get it to work by editing the file. It is a bit of manual work, but it works, and the resulting asset is far smaller (10 times smaller with the object I tried) compared to converting an .obj to .skp.


    Now I am trying with another file which is fbx. The resulting gltf is over 300mb, and I can't manage to get it to show at all. I don't know if Enscape has a size limit for the assets or if it is something else.

    EDIT: answered my own question, thank you for the pointers! I just had to really dive into the syntax pieces and make sure I had no mistakes! However, I do think a tutorial on these types of material tweaks would be immensely helpful to many! I just scratched the surface and was able to flip one switch!

    So how did you do it? Is this something you need to do in Blender, or by manually editing the gltf file?


    I tried to manually add the mask images in the gltf file, but still the parts that should be transparent show as black.


    In Blender, when Rendering with Cycles, it was fine.

    Zoom Is it allowed to use the Eevee engine for own products? If I remember me right, people ask to get Eevee as engine for Rhino and the McNeel says, this isn't allowed.

    I gave Blender as an example of having 2 render engines - a real time one, and one for best possible results. I didn't mean that Enscape should use either of those engines, but if it did then it would need Cycles, not Eevee.


    While Blender's licence is GPL which I believe doesn't allow its use in commercial software, the Cycles engine is Apache 2.0, so it could theoretically be used by Enscape as a secondary engine if that was technically possible.

    +1


    I think features such as "collaboration" will be somewhat useful to some users, but a feature to get even better quality final results would be useful to nearly everybody.


    I love how Enscape is real time, but there is a compromise in terms of image quality when compared to something like V-ray, Corona etc, especially in things like mirror reflections and artificial lighting.


    If Enscape could get much closer to the more traditional rendering engines then it would be a no compromise solution. If that is not possible with the same real time engine, then maybe you could have a secondary non-real time engine for this purpose. Something like Eevee and Cycles for Blender.