Posts by Nupsi

Hey folks

Maybe I have missed some changes, but why are onover-events are now passed over to all children-elements?

Changing a textlayers css like this:

set(layer[text_01].onover,set(css,font-family:barlow; font-size:16px; color:#ffffff; text-decoration:underline;));

and layer text_02 like this:

set(layer[text_02].parent,text_01);
set(layer[text_02].onover,);

the children-layer text_02 will also trigger the parents onover-event

How do I stop the childrens to trigger their parents onover-events?

THX in advance,
Nupsi

Sorry for pulling up such an old topic, but is it still not possible to create a sphere-grid?

image.reset();
set(preview.type,grid(sphere,32,32,512,0xCCCCCC,0xFFFFFF,0x999999));
loadpanoimage(MERGE,BLEND(0.5));

still gives me a cube-grid

Thx in advance,

Nupsi

Hi and thx for your reply, indexofrefraction. But that´s not what I´m trying to achieve - unfortunately

And of course I know Tuurs Toolbox But placing a distorted hotspot is still an annoying task, especially if you have to place a lot of them. Rotating, scaling, positioning - over and over again until it kinda fits. Placing a polygonal-hs perfectly is such an easy task with just a simple dragging-routine.

There must be a mathematical way of converting the center- and corner-positions of a 4-sided poly-hs into ry,ry,rz-values. Getting the 4 ath/atv-values and the center-position of the poly-hs is an easy task, but figuring out the math is way beyond my skills

I wish we could position the 4 corners of a distorted hotspot just as easily...

Hi folks,

I've busted my head for the last days to convert a polygonal hotspot into a distorted hotspot, but even with the help of AI I'm failing, because the conversion is never even close to be accurate.

Is it even possible to convert a poly-hs into a distorted one? I've found a reference to edge_ath/edge_atv of distorted hotspots, but this feature seems to be only krpano-internal and is not accessible directly, which is a real bummer. Because it would give us direct access to the corners of a distorted hotspot.

Does anyone ever had success in converting a poly into a distorted hs?

Klaus

Is there any way to place a distorted hotspot by its corners instead of rotating it around its center? It is just so damn hard, to place a distorted hotspot for videos etc. correctly. A polygonal hotspot is so easy to place, why I wrote a litte script which places a poly-hs and let me drag the corners around. It would be so awesome to be able to do the same thing with a distorted hotspot Or at least bea able to use the poly-hs as a parent for images/videos or iframes

LG,Nupsi

Yep...seems to be abandoned/dead

Is this plugin still working or abandoned?

Is´nt it possible to use a var?

<events onloadcomplete="if(your_var !== 1,your_action())" />

<action name="your_action">
	do_whatever_you_want;
	set(your_var,1);
</action>

This way the action should be executed only once, until the var is set to 1. That´s at least how I do it.

Hi Index,

like I said, I did try every possibility, even depthmap.align and image.depthmap.align. And all the other settings are working just fine inside the depthmap-line, like cull, axis and subdiv, but not align:
<depthmap url="sphere/earth_rings_satellites.obj" cull="front" background="none" backgroundurl="sphere/space.jpg" subdiv="0" center="0,0,0" axis="+x+y+z"/>

And even prealign works perfect. Only align is not working, no matter if used with MTL or texurl. I just can´t rotate only the texture/image

there is no texurl on your dephmap ... ?
If the texture-url is taken from the material (mtl) there is no need for texurl. It´s only necessary if there is no texture inside the mtl defined. But I did try it with texurl anyway, without luck...

Hi,

I can´t get the image-align to work like it´s described in the documentation:
The align setting rotates the pano-image.
The setting defines the rotation-angles in degrees about these each axis.
Works like the image.prealign setting, but it rotates only the image, not the 3D-geometry.

Where exactly do I have to put the align and how? I did try to put it directly into the depthmap-line, like this:

<image>
	<depthmap url="sphere/earth_rings_satellites.obj" align="90|90|90" cull="front" background="none" backgroundurl="sphere/space.jpg"  subdiv="0" center="0,0,0" axis="+x+y+z"/>
</image>

I also did try it in the view-line:
<view hlookat="0.0" vlookat="0.0" fovtype="MFOV" fov="120" fovmin="70" fovmax="140" limitview="off" align="90|90|90" />

And I even did try to set it in the onstart- and onloadcomplete-event:
set(image.align,0|0|0);

But the texture does not rotate! No matter if I use the MTL or texurl.
set(image.prealign,0|0|0); on the other hand works just fine but it rotates the whole model instead of only the texture

Absolutely stunning, as always, Tuur I wonder if those pylons could work on the surface of OBJs when mixed with the 3d-mouse cursor-function?

Quote from klaus.krpano

Next release

OMG - CAN´T WAIT

EDIT: Oh...and if you could take a look at the cull-function it would be perfect! Right now it´s not working quite right when using cull="twosided" for object-maps with alpha-channel

Quote from kme

I had a play with it and the depth can indeed solve the scaling issue.
But then it also moves the object in 3D space and I want it to be in an exact position. When I then move the hotspt so that the final obj is in the good position, the model is too big and I have to make it smaller again, running into the same quality issue again.

Also, I noticed that the 3D model is not interacting with other 3D models - it is drown over. You can see it in the screenshots below where the person is standing in the hall, but when you move the camera so that the bench is intersecting with the view to the person, the person is drawn over the bench, making it not realistic.

I noticed the same problems. At first it was´nt obvious, because I was just testing with planets, but there is indeed no way of using multiple obj´s right now.
Looking forward to the next release

Quote from kme

I experimented with that approach too, but in my tests, the quality of the generated hotspots was pixelated. If you find a way to improve that, I would be very interested in it.

It´s a scaling-problem! Just use the depth-parameter to set the hotspot in 3d-space instead of scaling.

l.enabled = false;
l.type = "krpano";
l.depth = "500";
l.distorted = "true";
l.width = "2000";
l.height = "2000";
l.scale = "1";
l.tx = "0";
l.ty = "0";
l.tz = "200";

I used oz="10000" and scale="500" for the 3d-model. and it´s sharp.

But unfortunately this method does NOT work in VR
As soon as I switch to VR the hotspot disappears, no matter what I try (renderer="webgl",renderer="css3d",depthbuffer.....).

Quote from lastromantik

There is no way in this version. Only one model at a time is supported. This issue has been long overdue and has been formalized as a wish that many people are waiting for. But we just have to wait for Klaus to realize the possibility to add more than one object/model).

That´s what I thought But it´s always good to ask to be sure

Quote from kme

I saw a post of tuur who did it with a hotspot:
https://krpano.com/forum/wbb/index.php?p…&threadID=19184
In the second example of "hotspot type krpano (no vr):", he loads an .obj as a hotspot.

Wow...thanks for the link. Did´nt see his examples and will check ´em out

That´s exactly what I´m trying to achieve, but there are still quite a few problems to solve.
The biggest problem for me right now is the backface-culling which I just can´t get to work properly

Hi there,

I´m testing right now the possibilities with 3d-objects (OBJ) inside krpano. Loading a 3d-object as the scene itself is pretty easy and straight forward:

<image>
		<depthmap url="model.obj" texurl="texture.jpg" hittest="true" scale="100" />
	</image>

But how can I add a one or more 3d-models above another pano-image? Unfortunately there is no documentation on this topic - or at least I did´nt find any.

Thx in advance,
Nupsi

Hi,

the Quest/Quest 2 and the Rift S are working just fine with the desktop-version of Chrome. But I can´t get the Pico 4 to be recognized by Chrome
SteamVR shows the Pico 4 as the standard device (cable or wifi) and the Pico 4 is working just fine with SteamVR-content.

The SteamVR-web console shows the following error:
Error creating VR renderer.
ComposeLayerProjection: failed to submit view 0: VRCompositorError_SharedTexturesNotSupported
CD3D11VRRenderer::Initialize(): Failed to create input layout for vertex shader shaders/d3d11/distort_geom_uvs_embedded_vs.fxo!

I would like to use the Pico 4 at a trade fair instead of the Quest 2, but for that I need to get the Pico 4 to work with a desktop browser - local browser in the headset is not working because we need to be able to interact from the outside (2nd browser-window).

Hi,

I´m using a main-scroller and on top another scrollarea with a text-layer. But the scrolling on the sub-scroller works really weird and is constantly moving the main-scroller beneath it too
Right now I´m using this to disable the main-scroller while using the sub-scroller, but this does not work on mobiles:

set(layer[sub_scroller].ondown,set(layer[main_scroller].draggable,false);showlog(true));
set(layer[sub_scroller].onup,set(layer[main_scroller].draggable,true);showlog(false));

Why can´t I set draggable to false on mobiles?

Edit: Oh...I´m using krpano 1.21 from 30.04. Tested on iPhone (Safari,Chrome) and Galaxy (Chrome, Samsung Internet)

Hi,

your example works fine on iPhone (Safari & Chrome) and Samsung Galaxy (Chrome & Samsung Internet). But you might set the fullscreen-button to toggle

console.log("myarr[0]",krpano.myarr.getItem(0));
console.log("myarr[0]",krpano.myarr.getItem(0).index);
console.log("myarr[0]",krpano.myarr.getItem(0).value);

Those are exactly the problems I had with the array. Did´nt know that the indices and values are inside another object
Thanks again for clearing that out!

note: getArray() doesn't get you [ "1", "2", "3" ], but only a js array of kprano "base" objects
I´ll keep that in mind

Okay...I think I got where my mistake was. From the array-docu:
array.getItem(name or index)
Get an existing array item.
Returns the array item object or null when the items doesn't exist.

So krpano.my_array.getItem(1) does NOT return the value at index 1 but reads the object from the array which contains the value?! Because:
var my_array = krpano.my_array.getArray();
console.log('Value_1='+my_array[1].value);
returns the correct value at index 1.

This array/object/index/value-thingy is really confusing

So I guess with krpano.my_array.getArray() I can transfer the whole krpano-array from krpano to JS.
And with console.log('MY_ARRAY_VALUE='+krpano.my_array.getItem(1).value); I finally get the value at index 1

I really can´t thank you enough for your help, but anyway...THANKS

NUPSI

P.S.: Sorry, I meant the JS split-function which I used like this: const filter_value = krpano_favs.split(',');. And within krpano I used txtadd to create a var which contains the content from the krpano-array as a workaround

Hi and thanks for the explanation

Unfortunately console.log(krpano.my_array.getItem(0)); just ends up in an error "krpano.my_array not defined"

I really don´t get the difference between the krpano-array and a js-array.
var my_array = krpano.get("my_array");
console.log('MY_ARRAY='+my_array);
shows: MY_ARRAY=[Object object], which seems to me like an array?!

But no matter how I try to access this array:
console.log(my_array[1]);
console.log(my_array.at(1));
the console only reports "undefined".

But if I access the krpano-array directly:
krpano.get("my_array[1].value"); the console returns the correct value at index 1 "17".

There is definitely some misunderstanding on my side

The main reason I want to access the krpano-array directly is that I can use it to filter database-entrys:

const applyFilter = (data, filter) => data.filter(obj =>
	Object.entries(filter).every(([prop, find]) => find.includes(obj[prop]))
	);


	var filter = {[id]:my_array};
						
	result = (applyFilter(data, filter));

Right now I use a workaround with txtsplit, where I loop though the krpano-array and txtadd the values to a var. It´s working but it´s waaay slower and much less elegant

Thanks again, indexofrefraction! I really appreciate your time and help