Raw & Post Processing, Printing > Digital Image Processing

PTGui vs Autopano Pro vs PTAssembler

<< < (15/20) > >>

Ray:

--- Quote ---If would be really helpful if there is a *set* of *standardized* source images that everyone can use to evaluate and compare this kind of tools. Much like a known test pattern used to evaluate lens focus sharpness.
[a href=\"index.php?act=findpost&pid=113532\"][{POST_SNAPBACK}][/a]
--- End quote ---

It would also be helpful if these trial programs would include a big notice like, 'Click here for maximum quality'. With Autopano (and I presume PTGui) the program first opens with low quality default settings which are good for maximum processing speed, but no good for evaluating what the program can really do in automatic mode. I've wasting a lot of time changing settings by degrees. Eventually, I just moved every slider to maximum and selected every option that appeared to be maximum quality, like Spline 64 instead of bilinear or bicubic interpolation.

It could be that as a result, the stitching process is slower than it need be. However, for an automatic mode, overkill is better than underkill in my view.

Those who opt for CS3 Photomerge will not have these problems, but I have found that stitches that don't work well in the CS3 'auto' layout sometimes work better with the 'cylindrical' layout, which is still automatic.

Since getting all the settings at maximum, I haven't come across a series of images yet that Autopano Pro cannot stitch automatically and seamlessly.  

Christopher:

--- Quote ---It would also be helpful if these trial programs would include a big notice like, 'Click here for maximum quality'. With Autopano (and I presume PTGui) the program first opens with low quality default settings which are good for maximum processing speed, but no good for evaluating what the program can really do in automatic mode. I've wasting a lot of time changing settings by degrees. Eventually, I just moved every slider to maximum and selected every option that appeared to be maximum quality, like Spline 64 instead of bilinear or bicubic interpolation.

It could be that as a result, the stitching process is slower than it need be. However, for an automatic mode, overkill is better than underkill in my view.

Those who opt for CS3 Photomerge will not have these problems, but I have found that stitches that don't work well in the CS3 'auto' layout sometimes work better with the 'cylindrical' layout, which is still automatic.

Since getting all the settings at maximum, I haven't come across a series of images yet that Autopano Pro cannot stitch automatically and seamlessly. 
[a href=\"index.php?act=findpost&pid=113536\"][{POST_SNAPBACK}][/a]
--- End quote ---

Look at the beginning of the topic and yopu will find two :-P

Ray:

--- Quote ---Look at the beginning of the topic and yopu will find two :-P
[a href=\"index.php?act=findpost&pid=113547\"][{POST_SNAPBACK}][/a]
--- End quote ---

I'll rephrase that. I haven't come across any of my images that I haven't been able to stitch seamlessly with Autopano in auto mode, after pulling out all the stops in the settings menu.

Kirk Gittings:
Ray I will try it on the image I showed above and see how Autopano Pro does. Especially because of the 90 degree wrap and the distortion involved from camera position, this image is a serious test of any software.

Ray:

--- Quote ---Ray I will try it on the image I showed above and see how Autopano Pro does. Especially because of the 90 degree wrap and the distortion involved from camera position, this image is a serious test of any software.
[a href=\"index.php?act=findpost&pid=113564\"][{POST_SNAPBACK}][/a]
--- End quote ---

I'd be interested to see the results, but beware of those settings under 'edit'.

Below are the settings which I'm now using for every image. They only have to be set once (most of them). These might not all be appropriate or necessary for each set of images, but they work flawlessly for me so far.

[attachment=2337:attachment]

Navigation

[0] Message Index

[#] Next page

[*] Previous page

Go to full version