gsf wrote:

Hi All,

I'm a newbie to the topic but have been using AutoPano for a couple of years so I'm "somewhat" familiar with it.

Background

I created a robotic mount for my iPhone from a couple of RC Servos and an Arduino board.  I actuate the camera shutter of the iPhone with a hacked headset - the iPhone allows you to actuate the shutter with the "volume up" control which is really useful.  This also works with my iPad Mini.  I also have a Nexus 4 and I found a guy to create a simple App that simulates the "volume up" on the Nexus so I can do the same thing there and will connect to the Android using bluetooth.

I get pretty good results and until now have used three rows of 10 images with 2 at 90 degrees to each other for the zenith.

Problem

Now to get the images to stitch.   When shooting with blue sky outside or white ceilings inside there isn't enough detail to create the control points and ugliness ensues.  I added full rows at the nadir and zenith and tried to use the import function but this didn't work.  It struck me that my problems weren't much different than what happens with DSLRs on robotic mounts so here I am.  My guess is that I need to create an XML file but before doing that I thought I'd ask.

I did look into PW but my problem isn't controlling the camera it's stitching the output.

Can someone point me in the right direction to save me time?  It would be appreciated.

Gordon

ps... I can post some images if it will help.  Also, I'm looking at turning the mount into an open source project with 3D printable components if anyone's interested.....  but first things first.

you can use PW in a"dummy" mode to create XML files without having a Merlin connected. So all you Need is: have a look at PW and create such an XML file which is reflecting your Shooting pattern of your panorma. very easy is : beginning top left->right->one image/row down->left->etc until bottom last image...) you just Need to know the rows and columns you used and the overlap used between your Images. focal/seonso etc data should be possible to guess or find with Google...

Georg

Last edited by gkaefer (2013-03-16 17:08:48)

gsf wrote:

HI Georg,

I looks like I would have to calculate the values as didn't find much on the Web.  By running PW in a "dummy" mode I assume you mean with nothing connected? 

Gordon

Yes, that's correct.

Just set the Plugins to Simulation mode.

gsf wrote:

I have been able to calculate all of the data needed to populate PW however it seems to fall down because of the sensor geometry not conforming to the available options.  I was able to generate the xml file however PW crashed so I probably need to run it again.

make sure you're using the latest development Version and not the last "stable" Version... (I run v2.1.22)

Ah!  Thanks, will install it.

in Win XP, Win7 and now on Win8 papywizard never crashed when in Simulation mode. no experience on Mac...
Georg

MacOS ports are more complicated, because of Apple philosophy concerning free software... And there system changing all the time, braking many things of the previous version is really annoying.

I strongly suggest you try with a linux or windows system, using virtual machine.

OK, installed Ubuntu, installed PapyWizard 2.1, shows as a terminal application(?) so i ran it in the terminal.  Everything seemed to run and left it in Simulation mode.  Loaded configuration and ...... ?

I think I'm closing in on the settings but I'm confused with two settings:

Optical Multiplier coefficient - I have no idea what this is.  The documentation suggests it might be the Crop Factor.

If it is, then what's the Sensor Coefficient?  This was my guess for the Crop Factor.

Google wasn't much help in understanding either of these metrics.

BTW, reading the released documentation it looks like the session XML file is created by the configuration process (duh!) and since I specify where it's located.....

Really sorry about the newbie stuff, but hopefully it will only be once and I'll put it all together in case someone else wants to use a smartphone for capturing spherical panoramas.

gsf wrote:

I think I'm closing in on the settings but I'm confused with two settings:

Optical Multiplier coefficient - I have no idea what this is.  The documentation suggests it might be the Crop Factor.

No, it's not crop factor. It's for situations where you might add a tele or wide optical convertor onto your lens, say a  1.4x tele convertor or something.

If it is, then what's the Sensor Coefficient?  This was my guess for the Crop Factor.

Correct, this is where you enter the crop factor.

BTW, reading the released documentation it looks like the session XML file is created by the configuration process (duh!) and since I specify where it's located.....

No. The XML session file is created when the shoot session completes.

Yoy can specify where the XML format data file recording shooting co-ordinates is stored using the Configure/Data tab.

.................

If you want to create an XML format data file for use with the APP/APG Papywizard Import filter/wizard without actually performing a real shoot using a Panogear/Merlin mount, then configure Papywizard plugins to Simulation, the use Configure to specifiy settings that match the camera and lens you used to shoot your real images, and overlap, camera orientation, and so on... the Connect and perform a 'shoot' with Data file enabled so you can create the XML data file.

But because you are not shooting a regular grid/matrix pattern you will need to create a custom preset to match your shooting pattern and 'shoot' using that preset in Preset mode.

See: http://www.papywizard.org/wiki/UserGuideSvn#Presetmode

Really sorry about the newbie stuff, but hopefully it will only be once and I'll put it all together in case someone else wants to use a smartphone for capturing spherical panoramas.

Hang on a minute.....

It's difficult to create spherical panorams with a pano FOV of 360x 180 without a proper pano head, still harder I imagine using a Smartphone rather than a conventional camera.

If you must try and do it with a Smartphone, may be ditch the robotic head idea and instead see this?:
http://www.fromparis.com/technical.html

A problem:

I have calculated the horizontal and vertical FoV for the iPhone's camera and it's been born out by actual use.  Basically it comes down to the fact that I get my best panos by capturing three rows of 10 images and then capturing 3 each for the nadir and zenith.

The question:

For Autopano, can I just go into the XML file and edit it to reflect the angles that I already know will work?   My goal is simply to get Autopano to automatically put the blue sky and white ceilings in zenith before stitching.

Thanks for you patience....

Here we go again.  Just send me home

OK, opened the XML file to edit it and it looks like it uses a time stamp to tell Autopano where everything belongs.

Can Autopano just look at the mosaic description values and order images were taken and ignore the absolute timestamp?

Or can I, do I have to, go into the XML file and give it the time stamp of the first and last images so it knows what to use?  I can tell I need to give some careful thought to workflow here.


Again, if you can just point me in the right direction, I'll learn the stuff eventually.