• Content Count

  • Joined

  • Last visited

  • Days Won


Everything posted by javalang

  1. OK, last message, the DS4 Controller is mapped as a BABYLON.GenericPad and has the id: Wireless Controller (STANDARD GAMEPAD Vendor: 054c Product: 09cc). The image below shows the button values. Would be fine if someone creates a class for this widespread controller
  2. It was soo simple -> I didn' realize that chrome also have a develpment-browser version for mobiles (chrome-Dev) , I installed the browser and enabled the chrome://flags: WebVR and Gamepad extensions ->voila Gamepad Tester is running using DS4 wireless on the mobile. BTW, the DS4-controller on the PC (connected via USB) is recognized as BABYLON.GenericPad Now I'm going to figure out which type of controller is attached to the mobile, it's neither a BABYLON.Xbox360Pad nor a BABYLON.GenericPad. There is a controller connected but I have to do some debug now. Thanks for helping me and I hope this info is useful for another people.
  3. @brianzinn thank you for the link, I tried your method today but unfortenately after some debugging my bluetooth connection was corrupted and I gave up, I will give another try. No I don't, but this is probably because the gamepad API is event-driven (so far I've seen) and the controller is already connected (paired) over bluetooth.
  4. @trevordev about html5gamepad: on the android mobile there are no buttons reacting at all! Only the website itself is scrolling if I press the navigation keys. So I have to figure out first this problem first, maybe my controller isn' t supported to use with the html5 spec.
  5. @trevordev Thanks for the reply, I also did some tests with html5gamepad but on the PC side. My interest is on the mobile side and it's very difficult for me to see what's going on. I just started an android app "Gamepad Tester" where I can see that most of the buttons are working well including Joysticks and X/Y buttons. But this app is for sure a native app and probably not using the html5 interfaces. Can you give me a hint how to find out the supported or connected controllers? Is "VRHelper.onControllerMeshLoaded.add(onControllerAttached);" the only way I can log the existence of a controller? And another question: I don't understand why the "onControllerAttached" callback has something to do with a loaded Mesh, what is if I have no Mesh to load? Thanks again for your help
  6. Thanks @Deltakosh, but I'm afraid there is no VRController involved. I think it's just a key emulation like the UP/dn/left/right keys on a pc-keyboard which is active if I pair the contoller with the mobile. Anyway, I will follow the description of "how to use gamepads" to see if any of the methods will be called...
  7. Hello everybody, I'm looking fore some advice using a PS4 (DS4) Contoller for mobiles. So far I've found at least one PLAYGROUND where the camera zoom is responding to my Controller keys. But I have no idea which type of WebVRController is instantiated and how the key-mapping happens (I'm not fit in debugging mobile applications) . Maybe there is one specific WebVRController already implemented or is the Generic type the right one? Does someone have experience with this topic? My hardware conection: Dualshock controller (DS4) paired with an android mobile. Thanks in advance
  8. I wasn't able to load your file into the sandbox, I think there is some mismatch in the obj-file. What I did: importing into blender and exporting it to test.babylon with "Texture location"-option is set to "inline" for enbedding all textures. No special code required to render the file. Hint for loading objects in general: first make sure the import into is correct, after then your programmatically import will also be fine in the attachment you can find my converted .babylon file. Hope this helps. test.babylon
  9. hi, maybe you have a similar issue I had in this thread. I'm pretty sure the materials.hasAlpha flag isn't set to true...
  10. Hmm, I always initialize the GamepadCamera and set the RigMode without initialization problems...playground
  11. For the users of dat.GUI, I've created a new version with three color-controllers, one with fixed height and canvas (no popup) and another handling two colors (i.e for bg/forgroundcolor) and the last with HSV/RGB sliders Available on GitHub. Have fun
  12. @ivan.popelyshev So the short answer is: this.sprite.tileTransform.rotation = props.rotation * Math.PI / 180; and it works , thank you
  13. I understand, I will test it. BTW , better java/kotlin/scala than FORTRAN But, since I changed to javascript, i will never change to java again, I should change the Nic to something like 'javascriptfunction'
  14. @ivan.popelyshev thanks for reply, TilingSprite is not suitable for me because my scale factors are not whole numbers. I added a link of the testcase in the first thread . So you think I should implement the repeat function for example in a filter? Edit: Changed my scenario to TilingSprite, using tileScale and tilePosition works fine now, but how to rotate?
  15. Hello, I'm asking because I'm not sure of my understanding the PIXI.basetexture WRAP_MODES: For example: given one container, one texture (loaded with loader) and one smaller sprite, defining PIXI.WRAP_MODES.REPEAT should fill the whole container, but it doesn't. The following code shows a sprite inbetween the container without repeating, what I'm doing wrong ? see WRAPMODE_TEST_HERE Thanks in advance PS: using PIXI 4.5.5 loader.load((loader, ressources) => { var texture = ressources.rgba.texture; texture.baseTexture.wrapMode = PIXI.WRAP_MODES.REPEAT; var sprite = new PIXI.Sprite(texture); sprite.width = 312 sprite.height = 312 stage.addChild(sprite); ... })
  16. Hmm, all 3 Browsers (FF,chrome,edge) are not reacting properly. Edge sometimes changes from blue to green but sphere wasn't visible, can't reproduce anymore. FF and chrome always green. Changing position of sphere from 1 to 2 causes crash on FF and chrome. ;(
  17. @Mythros thanks for reporting, fix will updated soon
  18. thanks Updated version V1.01: see changelog
  19. Hi, as announced, PBRComposer V1.02 is now ready for boarding. Last changes: 15 studio environment hdrs included, material and environment completed, tests, tests, tests. Project: GitHub Docu: User Interface , Examples , Presets Livedemo including 20 presets: PBRComposer @Deltakosh If you need files for BJS-HPage, let me know or peek/link it from GitHub
  20. I'm sorry, my mistake: besides some understanding problems I had a refresh problem with my glossy (microsurface) parameter. So lets summarize: the HDRCubeTexture, which is used as reflectionTexture is not blurred, but the DefaultSkybox can be blurred. The PBR-parameters are responsible for the blurriness, in metallic model it is the metallic/roughness parameters, in specular model it is the glossiness parameter. The skybox can individually be blurred. So at the moment everything is explainable, question is resolved.
  21. here the same difference without refraction:
  22. Hello all, I'm almost finished with my PBRComposer, but I have a question about reflection of the Skybox: in the glossy model changing the environment-blurriness also changes the blurriness of the material reflection/refraction whereas in the metallic model the blurriness of the environment is not regarded. Can someone confirm that this conforms to the metallic model and is not a bug? IMHO the environment blurriness cannot be skipped away... The following images shows a scene (fingerprints on a glass sphere with some mictosurface and some direct light) with exact the same material parameters (only glossy/metallic texture difference) and with a environment blurriness 0f 0.15. Can you see the difference? thanks in advance PS: Both, the reflection and the refraction does not reflect the environment blurriness in the metalic model
  23. @Deltakosh Sorry, but I can NOT follow you. Yourself wrote an article in Dev Zone if specular is not taken in account from light, how would you explain this playground then? OOhhhh....Maybe I haven't checked that "physic lights" are lights in conjunction with PBR and they are different from "classic lights" (or better: PBR-Shaders handles lights other than traditional Shaders)??? Is there any info about this? I couldn't find something in the Doku "Introduction/Mastering-PBR" about this. When you say does this mean the specular component of a lightsoure (concretely light.specular) is obsolete for a PBRMaterial? If so, then I understood the problematic.