Notices

Actually, this would be a really cool method for making "Image Sets" as well. These are a series of four images that apply to the four panoramic desktops in Maemo 5.

Tim

The stitching is all done in software (or at least it could be, no idea what those apps make you do), though I agree that viewing the panorama may not be ideal on a small screen. But who am I to question someone else's dream?

Maybe if you had cam app assistance it would be easier... As in, choose a panorama mode, and then show 1/4 of the previous image overlaid in the cam app and remember to use the same cam settings as previously (exposure, ISO, etc). It probably would be bad at high-res photos, but for wallpaper/web class stuff it might be more than enough and it requires little to no postprocessing.

http://store.ovi.com/content/23154

It seems someone did create a program for shooting panoramic images. (Only a) Trial is available already, but this will be a commercial app.

Just tried the trial version, a cool app. The picture quality sucks big time but with steady hands can it be improved?

With steady hands you could fly a plane in a snowstorm, or walk a rope while wearing a blindfold. With steady hands (translation: tripod stand and alidad) you could take an ideal panorama, and the application wouln't have to compare them to determine their relative positions - it would have only to stitch them accurately.
The fact is, Nokia Panorama application has excellent interface, and not-ideal-but-almost-perfect stitching. However, today as I tried to take a panorama, it crashed during stitching (four photographs, nothing complex) and rebooted the whole operating system. The four photographs and warps.txt are still here. I am attempting to stitch the four photographs with desktop-Hugin, but it's frustrating. Who knows HFOV of Nokia N900 camera? Or its lens type?
The idea: create script which would check whether Panorama is running; if not, then check if there are any files in /home/user/.Panoramatmp ; if yes, then check that the latest photographs (beginning from P0.jpg; number of them could be determined from warps.txt, I expect) don't have a corresponding already-stitched Panorama in /home/user/MyDocs/Panorama ; if yes, then stitch the photographs using the warps.txt and add it to /home/user/MyDocs/Panorama .
This script could be run either manually, if you don't mind having to remember to click it each time Panorama crashes during stitching, or be typed into Panorama shortcut to run after Panorama closes - just in case Panorama crashed. It should not be run before Panorama opens - it is quite uncomfortable to wait for lengthy process of stitching when you just wanted to take a new Panorama.
So, who knows the meaning of numbers in warps.txt? Who would be willing to create a command-line application to stitch the photographs? It should be much easier than porting whole Hugin.
Best wishes.
_________________
Per aspera ad astra...

Today I took the highly scientific method of measuring it with a ruler. At an 11 cm distance from the target, I measured how much of the ruler fitted in the field of view. The result: 11.9 cm, in both 4:3 and 16:9 mode. That means the horizontal field of view (HFOV) is 56.8° (let's say ±0.5° for my precision error).

Does anyone have the official figure to confirm my measurement?

BTW, using the HFOV and the focal length (5.2 mm, according to EXIF in the pictures taken by the phone), the photo sensor horizontal size comes out as 5.6 mm. Which sounds about right

Indeed it does : https://en.wikipedia.org/wiki/Image_sensor_format

Sensor is 1/2.5" with width of 5.76 mm so you were 0.16mm off the mark. Not bad for a... what did you say.... "pano-newbie"?

Thanks, handaxe. Obviously there are a number of rounding errors in all calculations. My sensor width came out exactly at 5.625 mm, I rounded it to one decimal. That puts me only 0.135 mm off The 1/2.5" may also be rounded. My distance from target may have a mm or two error. The focal length may be rounded. And finally, my figure gives the size of the active section, the 1/2.5" may be the absolute size.

Having said all that, I would still like to see the official figure