Posts by robertl


    So when switching the xml or scene before that loading is done, that event can't be fired of course, but onxmlcomplete or onnewscene events would fire of course. When they are missing then there would be either an usage-error or a bug (but currently there is no known one here).

    Thanks for taking the time to reply. I'm sure I'm using it correctly, but I'll try to set up an example and get back to you.

    That was the first place I looked. None of the relevant events work every time.
    I first went with just onnewscene/onnewpano and noticed that if I click fast enough, the event won't fire which is a complete deal breaker for my application.
    Clicking fast is easy to replicate, for example, choosing from the thumbnails, but also just clicking on the hotspots.
    I did get better results adding onnewscene, onblendcomplete and onloadcomplete to fire at the same time. Though it still seems hackish.


    If I click through fast enough the events still sometimes won't fire, but it's much better than just onenewscene.

    What would you have in the background behind the panorama?

    Piotr


    I want to be able to overlay a part of the panorama on top of the original one.
    Say for example, I have a panorama with a table and four chairs, and I would to make each chair a different color when I hover or interact with them.
    Currently I either have to make four different panos and blend them or muck around with distorted image hotspots.
    Being able to use alpha masks with 360 images would make the process a whole lot easier and straightforward and would also open the doors to a whole lot of new possibilities.


    I hope that make sense.

    Hi,


    Okay - there is one special case about the 'custom images' - when the normal image is SMALLER than the custom image size and is of the SAME TYPE, then they are not generated - because in this case they would need to get upsampled and would be only waste of space and resources.

    You can count on me to hit that special edge case! :)
    My input images are 18432x1536.

    Hi,

    I've just tested again and all stereo settings seems to work as they should...

    Maybe try the MAKE VTOUR (VR-OPT) Droplet - it has:
    stereosupport=false
    and
    customimage[vr].stereosupport=true

    and this way it generates a mono-multires image for normal-viewing and stereo-images for VR-viewing.


    I haven't given much thought to stereo multires. I guess I'd like to keep it for future compatibility. I can always generate a new XML.

    So, currently I have enabled the stereosupport for stereo multires and commented out the customimage[vr].stereosupport. That combination produces both a stereo multires and stereo vr.
    If I leave the customimage[vr].stereosupport=true, the vr images aren't generated.
    Here is my full config:


    Hi!
    Converting 12x1 stereo cube strips.

    In my droplet config file I have:
    stereosupport=true

    and
    customimage[vr].stereosupport=true

    The problem is that VR images are not created unless I comment out one of those.
    If I comment out customimage[vr].stereosupport=true, I get both multires in stereo and VR in stereo.
    That seems weird.

    Klaus, would you chime in?

    Hi!
    I've set up a scene with a video hotspot.
    The H264 mp4 files won't play video in Opera browser on Android.
    The webm works fine, but only if I put it as a first URL.

    The H264 plays fine if I open it directly in the browser.

    I tried different encoding settings, levels, bitrates, didn't make a difference.

    Any ideas?

    Pano url:
    https://robertlesac.com/static/pff/
    Video url:
    https://robertlesac.com/static/pff/pff…screen-dark.mp4

    Hi!
    I have all calculated all the heading in the panos so they point north.
    Trying to set up radar, but it seems as every pano needs a different heading for the radar to display correctly.
    I also use a floor plan that is offset 40 deg from north, and include the offset.

    I did some manual testing and calculation, and the calculated heading should work for radar, but it's not matching up.

    How exactly is the radar heading calculated?

    Hi,

    it's always a bit a balance, too much error checking and reporting makes the code big and slow, but too few isn't also okay of course.

    For the next release I have already an advanced checking for invalid variable names (the reason for several strange errors), together with a error-reporting that makes it easier to locate the problem in the xml.


    That sounds great! Keep up the good work.

    Last time I did Krpano was 6-7 years ago and none of the code will run on the current version. It took some time and patience to figure out what has changed in meantime. It took a few very frustrating days, but I seem to have gotten the hang of it.

    Hi,

    even when that might be (or get) be possible, because of the required distortions equirectangular-crops would be only the low-quality solution.

    The better and higher-quality solution would be this:
    Example/Tutorial - Extract a Distorted Hotspot from any kind of Pano Image (with perfect pixel alignment)

    Thanks, found it and I'm successfully using it using it.
    I do have a small issue, but I'll post it in that thread.