Posts by tksharpless

I have gotten the desired result by processing the panos with vtour_multires_nocube.config and editing just the <scene> block in tour.xml, without including any other xml. In this case, as Klaus has stated, just limitview="fullrange" is enough.

The tool decided that some panos were cylindrical rather than flat, so I made a modified config file with 'panotype=flat' in place of 'panotype=autodetect,flat', and that one gives all flat panos. It would be good if the tool would automatically set limitview to fullrange in this case; but editing those lines by hand is not so bad.

This time I left 'hfov=360' in the config file. Before, I had removed that line because I thought it would conflict with viewing flat panos. That mistake apparently caused the problem I have been complaining about.

I guess the lesson here is "keep it simple, stupid".

I have <view ... limitview="fullrange"/> in each scene as well as in the setup script.

But I am still seeing each pano scaled to window height. There must be some other adjustment. Bear in mind that that the main skin is the standard vtourskin.xml.

I have tried limitview="fullrange", also "offrange", but neither cures the problem.

When I process a single wide, short flat pano with flat.config, the result shows the full width of the pano.

Now I want to make a tour containing several such panos. I can do this with a modification of vtour_multires.config that specifies panotype=flat and omits hfov=360. However in the tour, the panos are always scaled to fill the screen area vertically, so that it is not possible to see the full width.

I have tried including in tour.xml the script 'flatpano_setup.xml', and removing from the <scene> blocks all code except what is needed to specify the pano and the preview. But the problem persists.

It does. The same .js file that works for the spherical panos works also for the flat ones.

However I consider it a bug that the update tool does not know this.

Using krpano tools 1.21.2, I am trying to update a website containing old Flash-based krpano images. Some are spherical panos, others are partial panos displayed in cylindrical mode (actually orthophotos). The tool replaces the javascript files for all images, after which the spherical panos work properly but the flat ones are black, because the .js files are empty.

The original embed scripts for the flat images are much shorter than the ones for the spherical panos. I tried to attach samples, however your forum s/w rejects the file extension '.js'; so I have renamed them as text files. "right.js.txt" is for a flat pano, "12thPassyunk.js.txt" is for a spherical one.

I hope this problem has a simple solution

12thPassyunk.js.txt

right.js.txt

With 1.21.2 tools, when I build a tour using vtour-vr.config, the resulting tour does not work in VR on either android phone or Meta Quest browsers.

I think this is a longstanding problem, but I have forgotten what to do about it.

It would be better if the defaults worked.

I have been able to get the desired effect by changing the enter VR button's onclick to call a new action that toggles display.anaglyph in desktop mode, otherwise calls enterVR().

The current 'fake VR' mode is nearly useless on desktop platforms. But it could be quite useful for stereo panoramas, if it would automatically display them in anaglyph.

I have tried to get this effect with the XML, but without any luck. The static side-by-side display always appears.

As a further feature request, it would be good to have the option of grayscale anaglyph display.

I don't know of any Quest Pro applications that demonstrate use of eye tracking. Meta/Oculus forums indicate that Meta Pro eye tracking is currently working (badly) only in game engines and not at all on webXR. Meta does intend eventually to support it in the Quest browser.

The currently available VR headsets with eye tracking include the Meta Quest Pro, the HTC Vive XR Elite and Vive Pro Eye, the Pico Neo 3 Eye and the Varjo XR4. They range in price from $1000 (Quest Pro) to nearly $5000 (XR4) and are targeted at the corporate/professional market. That is a market krpano should target too, as there is much need for custom app development.

Do you have any plans to support the eye tracking now available on several VR headsets?

Eye tracking offers several possibilities for big improvements in viewing comfort as well as interactivity.

Dynamic adjustment of the convergence distance, by angular shift between the left and right images, is perhaps the most important application. This is a basic element of cinematic 3D, and should be a feature of all future VR hmds.

Thanks, Klaus.
Clearly, I need to study the documentation on how to access WebXR APIs.
Is it allowed to code any of these directly in JavaScript?
Regards, Tom

Is krpano able to access OpenXR APIs, especially for controller devices? I would like to make virtual tours that play easily on Quest 2 etc.

Two improvements to the vtour editor gui would be welcome.
1) ability to rotate as well as position hotspots
2) ability to re-order the thumbnails, changing the next/previous sequence

In the tools gui, the make vtour droplet can accept image files added one-by-one. the gui shows them as a numbered list, however it always sorts the list by name. It would be convenient if instead it kept them in order of addition.

I just made a multiresolution tour with krpano tools 1.20.9 using vtour-multires.config. In tour.xml I include vtourskin_design_ultra_light.xml.
In other words, everything is stock krpano latest release.

Using krpano testing server, everything looks fine on PC bowsers, including 'fake VR' mode in response to VR button. This is true both on the machine hosting testing server and other PC or Mac on the same LAN; but when I load the tour in a phone or Oculus Go via the LAN there are no zoom, pan, or VR buttons on the control bar. Buttons for fullscreen, thumbs, hide/show bar and next/prev pano are present and work as expected.

What must I do to get those missing buttons on mobile and headset?

I fear that this might have something to do with the lack of https support by testing server, but I hope that is not the case as I intend to host these tours on AWS S3, which also does not support https.

Ah, I understand the distinction. The pano is OK but the depth map/3D model has discontinuities.
I am just now studying how it may be possible to improve depth maps by forcing their edges into alignment with edges detected in the images. I hope to show some results soon.

That is definitely a stitch error.
We should try to make or find some CGI panos with depth maps, those will be geometrically perfect though artificially simple.
I am working with Andrew Hazelden on depth mapping some photographic stereo panos, using available tools that make almost-acceptable maps. Will upload when I have them.

Klaus, I have also been in touch with Andrew Hazelden, who is ready to help by rendering CG views with depth maps. He says he can easily make both omnistereo panoramas and simulated fish-eye photos with depth maps. I have sent you his email, and also asked him to join this forum.