krpano 1.19-pr15 / pr16

  • Thanks, Klaus

    Can one easily set up the new sphere<->cube droplet to create a given cube size?

    The web browser in my LG 3DTV cannot convert an image in either sbs or over/under format to 3D.. With an interlaced 3D image it would not need to do anything, or even know the image is 3D.

    -- Tom

  • Can one easily set up the new sphere<->cube droplet to create a given cube size?

    You could enter a custom size here:

    Or alternatively call from commandline and use the '-outsize=' setting:
    https://krpano.com/tools/ktransform/#top
    E.g.

    Code
    krpanotools sphere2cube cube inputpano.jpg cube.tif -outsize=2048x2048


    The web browser in my LG 3DTV cannot convert an image in either sbs or over/under format to 3D.. With an interlaced 3D image it would not need to do anything, or even know the image is 3D.

    I've already started working on interlaced-stereo support, should be ready in the next release.

    Best regards,
    Klaus

  • Hi, one note about:

    Quote

    The web browser in my LG 3DTV

    I have also a LG 3DTV - it's a 4k TV with passive 3D - but the TV browser there doesn't render webpages in the native 4K resolution, the browser only uses a 1920x1080 resolution - and with that interlaced stereo output is not possible - interlaced stereo would require a 1:1 unscaled rendering at the screens physical resolution...

    But with a PC attached driving the TV in its full 4K resolution interlaced stereo is working...

  • I'm running into situations where, at seemingly random instances, we are getting an "FATAL ERROR: - invalid encryption!" for XML that is being encrypted on-the-fly leveraging the "-out=stdout" parameter with krpanotools -encrypt.
    When I say random, I mean a user will run into this error on a particular scene, reload the tour, go back to the same exact scene, and all is good. If the error was present each time a user went to the scene, then there would be an issue with my XML generation.

    Here is a link to a tour on our "preview" domain with the full-encryption code in-place where I can confirm users have run into this issue.
    https://www.createvtour.com/mammoth-mountain

    I personally have not run into this in my testing, but one user running into this is one too many.

    When I say encrypting on-the-fly, I mean leveraging PHP to handle the encrypting of XML at time of request/load.

    My sequence goes like this:

    1. The xml is generated & loaded into the output-buffer
    2. Krpanotools is registered ( ensuring it is registered ).
    3. If registration is successful we proceed, else we output the buffer and exit, basically rendering unencrypted XML.
    4. If we are successful with putting the output-buffer into a temporary xml file, we proceed, else we output the buffer and exit, basically rendering unencrypted XML.
    5. If we are successful with encryption, the encrypted XML is rendered and we remove the temporary XML file.
    6. The encrypted XML is loaded into the tour.

    I'm assuming that when an encryption error happens, in my case, the unlink ( file delete ) does not occur. That is why I have several XML files in the directory the temporary XML files is placed.
    I've run these files through XML validators ( multiple, to make sure that this is a valid test ) to check their integrity and they come out clean as a whistle.

    I could put an encrypt-error handler in-place at step 5 ( above ) but I don't know what the exact output/return, if any, would be in the case of an error.

    Here is the code-block that handles this:

  • In the previous version I have used "convertdroplets.config" to change the output of Convert CUBE to SPHERE droplet.bat.

    This doesn't work for the new .exe droplet. Is there a way to change the output format (e.g. jpg instead of tiff) of "Convert SPHERE CUBE Droplet.exe"?


    Thanks a mil!

  • Hi,

    When I say random, I mean a user will run into this error on a particular scene, reload the tour, go back to the same exact scene, and all is good. If the error was present each time a user went to the scene, then there would be an issue with my XML generation.

    Can you please describe that case more detailed? Additionally also what browser and browser-version is used.

  • In the previous version I have used "convertdroplets.config" to change the output of Convert CUBE to SPHERE droplet.bat.

    This doesn't work for the new .exe droplet. Is there a way to change the output format (e.g. jpg instead of tiff) of "Convert SPHERE CUBE Droplet.exe"?


    Hi,

    the new droplet doesn't use the convertdroplets.config file. The config file is for direct usage of the transform/sphere2cube command line tool, which the old droplet was using.

    For the new droplet to changing the output format just start the 'krpano Tools' program and change the conversion settings in the 'Convert Tool' tab. All settings there, except the output size, will kept stored and also be used when directly using the droplet (btw - the droplet and the 'Convert Tool' tab in the krpano Tools are the same, just a different window visualization).

    Best regards,
    Klaus

  • Do you have a link to your example?

    Well, it's in examples from krpano package (googlemaps example)
    just replace googlemaps to bingmaps and that's it, we will see wrong radar direction.

    I have tested pr13, pr14, pr16 krpano versions with bingmaps html5 radars, all versions has same bug

    Andrey

  • On my machine with firefox (on win 7-64bit), there is a bug of orientation within bing map(the radar is 180° rotated). It's work fine with chrome.

    Tested also with Firefox and Chrome on W7 and works for me...
    And the radar drawing code is very simple and independent of any browser or system stuff, so not sure what there could go wrong there...

    Well, it's in examples from krpano package (googlemaps example)
    just replace googlemaps to bingmaps and that's it, we will see wrong radar direction.

    Tried too (and example above is basically that) and works for me...

    Strange... I will try further...

  • And the radar drawing code is very simple and independent of any browser or system stuff, so not sure what there could go wrong there...


    Yep, that's what i thought too, and was really wander about this :))

    It's on my Mac

    But here is the test on Windows (Win10 64) system

    http://prntscr.com/j60bkh


    Andrey

  • I was now able to reproduce it on the iPhone (on Desktop browsers I had no change), but even there is appears only randomly every few reloads...

    ...and after some testing and data-logging this is clearly a bug in the Bingmaps API itself - the 'tryPixelToLocation' function there returns wrong (inverted) values in the not working cases...

    After some research I've found another Bingmaps API bug-report about that same problem:
    https://social.msdn.microsoft.com/Forums/en-US/6…?forum=bingmaps

    According to the report above that bug is fixed in the experimental branch of the Bing Maps API (the krpano bingmaps plugin is currently using the 'release branch') - and yes, it seems to be fixed there. But the problem is that this bug-report is already 1,5 years old and the fix is still not in the release branch - and using the 'experimenta branchl' by default could be also a bit risky for other unexpected API changes in Bingmaps that could cause other problems... Will keep on testing...

  • Update on the Bingmaps problem:
    I've found now a simple but effective workaround by testing the Bingmaps API for correctness - I'm requesting the locations of two points, one point a bit below the other - and when their relative position returned by the API is wrong, I know the API doesn't work and can manually correct/invert the values ;-). So everything works and regardless if and when it will be fixed on Microsofts side.

    Here an updated bingmaps.js with that workaround: bingmaps.zip (will be included also in the next krpano release of course).

    Best regards,
    Klaus

  • {how can I delete this unwanted quote?}

    I'm sorry to hear this about the LG browser's video mode, but not surprised. How did you determine it?
    That browser also crashes on panos with large cubeface sizes, so is really not much use.
    But by using the TV as a 4k 3D monitor, we should still be able to show interlaced stereo from web pages.

    I have also a LG 3DTV - it's a 4k TV with passive 3D - but the TV browser there doesn't render webpages in the native 4K resolution, the browser only uses a 1920x1080 resolution - and with that interlaced stereo output is not possible - interlaced stereo would require a 1:1 unscaled rendering at the screens physical resolution...

Participate now!

Don’t have an account yet? Register yourself now and be a part of our community!