Well, I'm already doing the basics, that's why I'm asking. I'm thinking to try a mix of single row with focus stacking and that will take much longer, possible multirow.
Hi,
Indeed, focus stacking will add to the time to acquire each image tile before you can move to the next. Multirow then requires more planning, and a procedure that can be executed relatively quickly. Using a click-stop indexer really helps a lot, because t allows you to concentrate on other things, and the amount of overlap between tiles can be set in advance.
Even with my current technique (I don't have a pano head/rail) it can take longer than I want to when the sun it's just setting and there is a difference in light in just a minute.
That will complicate some things, but you can gradually build the pano rig if you choose a flexible modular platform. For your particular use, a click-stop with clamp, and a No-Parallax-Point (NPP) bar with a camera L-bracket should go a (very) long way towards successful pano-stitching when under time pressure. The click-stop will be your time saver at shooting time, the rest will help to avoid/reduce stitching issues with busy foreground/layered depth clues.
I use the
Manfrotto 300N (with an RRS ArcaSwiss style clamp on top) which was at that time sold at a significant discount, so look around for good prices.
I did some vertical panoramas with autoexposure (by mistake) and the result was good.
Here is an example. It ended up ok but the difference from left to right wasn't as big when I started.
Yet the change in color temperature is (perhaps too) obvious at this small size. At a larger size the color temperature gradient becomes harder to spot because we need more time to take in the whole scene. I typically produce TIFF tiles for stitching, already focus stacked and or HDR tonemapped if needed, and I can already process the Raws to different white-balance settings going left to right, thus leveling the differences a bit. Also exposure differences can be somewhat leveled between joining tiles, but it's not necessary to level the brightness across the entire pano width, unless one shoots 360 degree panos.
This is one of my rare
very wide angle panos, which required both exposure leveling (camera on auto exposure, except for the HDR with bracketing ones shooting into the sun), and White balance leveling:
The angle of view was close to 180 degrees, so the left side was warm and totally front lit with a rising sun, and the right side was cool in the shadows and totally back lit. The benefit of exposure bracketing is that one can also vary the White-balance of the Raw exposure brackets before blending/tonemapping, i.e. warmer shadow brackets if they are too blue-ish, cooler highlight brackets, if needed.
Having so many exposures to take, requires a setup that allows to save time when shooting, and for that a click-stop indexer is paramount IMHO. It also makes it easier to place tiles without features to place control-points on, like featureless sky or water.
Cheers,
Bart
P.S. I used the SmartBlend blending engine to allow for a gradual transition of smooth exposure and WB gradients in the sky. It can also be used as a plug-in from within PTGUI and PTAssembler, and if I'm not mistaken from Hugin, and the '-HiPassLevel ' commandline parameter controls the width of the transition zone, so it can be user adapted to different scenarios.