Another vote for Panavue. PC only though...and now I have gone over to the Mac world...no regrets for that though...Windows sucks !
Another vote for Panavue. PC only though...and now I have gone over to the Mac world...no regrets for that though...Windows sucks !
Hi Doug
"I have gone over to the Mac world...no regrets for that though...Windows sucks !"
Better late then never;--))
Armin
Other than Photoshop CS 3, which programs offer the option of doing a "perspective" (i.e., normal Renaissance-based perspective) stitch, as well as a cylndrical stitch. I find the perspective stitch to be my preference, but for some scenes where you're going really really wide, the cylndrical stitches are better (pardon my spelling, which has never been a strong suit).
CS3 is pretty damn good. Below is a 2x2 hand held stitch (1DsII + 50mm 1.4) with CS3. After stitching, straightening, and cropping it comes to 7219x7581 or ~54.7 megapixels
A 100% zoomable version of this image can be found here
Here is another hand held CS3 stitch, this one is only 2x1, 5560x5017 or 27.8 megapixels.
100% zoomable version here.
Nice gallery at the Autopano site.
here are a few done in Autopano and PTgui. They range from small 4x4 grids, to larger 5x5 and 7x7 grids (which end up being on the order of 25,000px wide) done with a Canon 1Ds2 with 85mm and 135mm lenses for some DOF experimenting I was doing. (the last one was with a 28mm, 6 vertical shots I believe)
I am confused as to the function and purpose of the various methods offered in CS3 in Photomerge, i.e. Auto, Perspective, Cylindrical and Reposition. Could someone briefly explain, or direct me to a page that does so, the basic advantage of each method and when it might be recommended?
So far I have been very pleased with CS3 in stitching scanned negatives using the Reposition Only option. However, these negatives, which were scanned with close registration, are perhaps less demanding than some other types of stitching, hand-held for example.
Sandy
OK, really briefly:
- Perspective uses one of the images from the set as an anchor around which all the others are then manipulated to fit (aligned, skewed, transformed...) and blended together.
- Cylindrical treates all the images in the set as if they were layed out on an inside of a cylinder which is then unrolled.
Both methods use center image as an anchor.
- Reposition Only arranges images in a linear fashion and does not do any manipulation other than aligning and blending layers. The ideal purpose for this would be stitching a multi-sectional scan of a very large image and you are right when you say it is a simpler case than the others.
- Auto is prety much obvious, as is Interactive Layout.
As a side note, aligning and blending stacked layers are also available separately in PS CS3 as Auto Align and Auto Blend and come in very handy in other types of combining multiple images.
Marko,
Thanks for the details. I just picked up a small book on some of the new features on CS3 (Ben Wilmore) and it briefly describes the various features of auto-align and blend. Your comments, though brief, add to my understanding of Wilmore's explanation, especially as regards the anchoring point and cylindrical setting.
I can see that some people have much more complicated stitching scenarios than me. To this point I have only been interested in straight stitching of scanned negatives, which is very easy assuming they were scanned in reasonable register, and of 6X9cm negatives exposed with a level tripod and level head. These too are stitched very easily by the Reposition method, perfectly in fact if the head and tripod were perfectly level.
BTW, I upgraded to CS3 primarily for the photomerge potential. Previously I had been manually stitching 12X20 negatives and the operation took a long time. With the new photomerge option in CS3 I feel that I just stepped out of a Model-T into a brand new BMW.
Sandy
Bookmarks