Posts by tonesh

    With this version the file generated by MAKE VRTOUR droplet the tour_testingserver_macos does not open.
    It gives the error: "the file is corrupted".
    If I copy the file from krpano folder (templates>html) and paste into the generated vrtour folder it works.
    This was not with version 1.20, same mac, same OS (13.4)

    This is a project made first in swf in 2011.

    Then I update it for html5 with version krpano 1.18 and everithing was working, on desktop, tablet and mobile.
    To make the movement of the objects it use a “crop animation”, single column cycling process.

    With the first version,http://</a> the images that were processed were made with only one columm, with this dimensions:

    sphere (objects/sfera.png) = 90 x 10.800 px
    cube (objects/cubo.png) = 202 x 24.240 px
    equi (objects/quadro.jpg) = 170 x 12.000 px
    cross (objects/croce.png) = 90 x 14.400 px

    everything ok, desktop, tablet, mobile.

    Then I updated to version 1.20.6 see
    same xml, it works for all images below 15.500 px length (I made several tests):
    on desktop all ok, except for cube that is > 15.000 px
    tablet and mobile are not working at all.
    And there are strange issues on dragging the pano, as there is an invisible hotspot...

    Then with version 1.20.6 see I updated the cycling process using a multi-column one (do crop animation,…anihotspots.xml )
    so to reduce the lenght of the images that are now :

    sphere = 600 x 4500 px (4 columns)
    cube = 808 x 6060 px (4 columns)
    equi = 1020 x 8000 px (3 columns)
    cross = 720 x 7200 px (4 columns)

    desktop is ok for all, tablet and mobile not.
    (Same strange behavior as it have a hidden hotspot)

    So, what are the dimensions limits to import images with new Krpano version?
    (I notice that 1.19 version too had same problem…)

    Thanks, Toni

    Hi Jeff, nice to hear you after a lot of time!
    So, after some tests (on Mac) with different browsers I got that with Firefox I was able to see both, html5 and flash regularly.
    In Safari I see the html5, but flash returns : ERROR: hotspot[video1] loading error: %SWFPATH%/plugins/videoplayer.js
    In Chrome both ok.
    Maybe you resolved the problem last days...

    Little advice: if you update the swf, you should update all plugins too.

    BTW nice video & pano

    Cheers, Toni

    Can't change the "flag" in webVR.
    Passing from a scene to another in weber, is
    always a mix of BLEND and KEEPVIEW: horizontal look at (panning) is set to
    0 and vertical look at (tilt) is get from the previous scene.

    Setting or changing flag in skin_settings does not work on webvr.

    Is this the standard flag for webvr?

    Reported on iOS with krpano 1.19-pr14

    I can't change the flag for hotspots operating in web-vr.
    If I set something like this, in for example "onentervr",

    set(style[skin_hotspotstyle].loadscene_flags, KEEPVIEW);

    the flag still remains in MERGE setting.
    Maybe I'm doing something wrong, but trying for example "scale",

    set(style[skin_hotspotstyle].scale, 0.5);

    is working well.

    (modified some days later)

    here I've put here a very simple project that can explain the problem:
    all the "flags" are set to KEEPVIEW, in all versions are working well except
    the webvr where the passing from a scene to the other is in MERGE flag.
    In vtourskin.xml the code in "skin_webvr_setup" action was modified like this:

    <action name="skin_webvr_setup">	if(webvr.isenabled,	copy(skin_settings.loadscene_flags_backup, skin_settings.loadscene_flags);	set(skin_settings.loadscene_flags, KEEPVIEW);

    Error checked with krpano 1.19-pr14 on iOS, with chrome, safari, firefox apps.

    Thank you, great improvement! *thumbup* *thumbsup*
    Specially the ICC color profile support, but can't understand how it works: in the .config files that comes with the new release I did not find any reference to profile,
    do I have to add it manually, and where?
    Thanks again

    Ok, I solved the problem:
    on a previous thread, Klaus said:
    "it depends on the 'converttocubelimit' config setting if a partial-sphere will be automatically converted to cube:…verttocubelimit
    The default setting is 360x45 in the most config files."

    I just opened the multires.config file and changed the default setting (converttocubelimit=360x45) to 360x30.
    That works.
    Have to remember that HTML5 is only compatible with CUBE panorama.

    I output with MAKE PANO (1.18.5) some cylindrical panos, and some of them have the html5 setting (in the html page) to "never" others to "prefer".
    Can't understand why...
    For example a 5556 × 737px pano is ok for html5, a 5505 × 600px one not.
    And there is no way to set the html5 to "prefer": it just don't work.

    This is the report of Make VTour:

    processing - 5505×600
    - input: 5505×600.tif (cylinder, size=5505x600, fov=360x38)
    - output: scene, preview, thumb, tiles
    - multires: tilesize=512 levels=1 - 5506x600
    - making images...

    processing - 5556×737
    - converting cylinder to cube format...
    - input: *5556×737*.kro (converted cube, size=1770x1770)
    - output: scene, preview, thumb, tiles
    - multires: tilesize=512 levels=2 - 1770x1770 886x886
    - making images...


    2013 tour: 35mm zeiss + d800e for the biggest one (some focus stitching needed) , 16 nikon fisheye + d800e for the rest. The front (outside) is made with 105mm + d800e
    2009 tour: 16mm + d700
    2000 tour: Coolpix 950 + wide adaptor

    Here the new virtual tour of the Galleria Nazionale d’Arte Moderna in Rome, one of the biggest museum in Italy.
    Along with the new version, there is the 2009 version and the 2000 tour, so it’s a 13 years work *wink*
    New version displays 120 panoramas, with 38 of them that are bigger than 1 gigapixel, the rest about 1/2 gigapixel.
    Very basic interface, but with all artworks linked (600 links) to the museum database.
    Some movements and videos around.…am-visita-a-360
    (Sorry there is no english version, hope they can make it soon…)
    Thanks Klaus for the new multi resolution, here is really needed!

    Toni Garbasso

    maybe an action like this, triggered by the hotspot "onhover"

    <action name="mouse_link">
       set(layer[my_layer].x, get(mouse.x));
       set(layer[my_layer].y, get(mouse.y));

    or you can use "tween" instead of "set"