Panoramas shot in 1:1

It seems to me that for panoramas shot in a 1:1 aspect ratio, it wouldn't matter if they were shot horizontally or vertically, or am I missing something?

Steve Thomas

You're talking about the roll orientation of the camera during the captures. You're right, it won't matter.

Jim,

Thanks.

Steve Thomas

Panorama is basically stitching of multiple shots for a wider angle of view that a single shot of our lens can't produce.

Therefore theoretically using shots of any aspect ratio should be alright.

However, normally we shall leave as much overlaps between shots could be for a better chance of good stitching. If we shoot in an aspect ratio wider on the width side, we can use less shots for the same AoV. We know the smaller number of shots used to stitch could be easier for the job. Hence 16:9, 3:2 or 4:3 could be better than 1:1 IMHO.

e.g. a camera of 6000 x 4000 @3:2. Says if we need to leave 1000 pixels on both edges of the width side for (overlap) stitching, it would leave 4000 pixels on the width side full used. If shoot at 1:1, the shot will become 4000 x 4000 and if we still have to leave similar overlapping, it will have only 2000 pixels on the width side used. Not hard to expect a lot more shot has to be used for the same AoV that other non 1:1. Worse stilt, if some element(s), e.g. pedestrian, move(s) in the scene, if he/she will not walk very fast, the wider the width of the shot the smaller chance for he/she to be captured in the next shot. Hence on stitching we can expect lesser chance to see the same pedestrian appearing in 2 ~ 3 places inside a panorama output.

Therefore I would use the aspect ration having widest on width, so will use the widest angle lens for my panorama shooting.

My 2 cents.

But not necessarily in the horizontal direction. There are vertical panos, and square ones, that, at 2x2, 3x3, 4x4, etc with a 1:1 aspect ratio capture, could have the optimum overlap.

Disagree, In a single row [let's call it a 'classic' panorama, having the camera in portrait orientation gives you the angle of view of the larger sensor dimension, which is often useful.
Most advice I've seen is not to use the widest lens for this, as that will probably lead to more distortion.
Both of these points mean you need more shots, a disadvantage in creating the pano, but to me the above advantages are more important

For the original question. I think Jim described it well.

I'm a little confused here about why you would use 1:1 and not the camera's natural aspect ratio.

I normally use portrait orientation which will give me the most pixels in the final stitch and then crop the final stitch to suit.

What am I missing?

Not sure the exact meaning on your "optimal overlap".

As far as experience told me, that was explained in my original response: more overlapping can ensure a successful stitching easier. I suppose the figures provided therein could reflect this. Of course I would never narrower shot can't do it. Just need more shots such that might have more uncertainty on stitching.

I do vertical stitching too. 9:16, 2:3 or 3:4 still hold the adventage over 1:1. For horizontal stitching, I leave certain margin on left and right for overlapping. For portrait orientation panorama, I have to leave the top and bottom for overlapping. So the same should apply.

Sorry can't get your meaning. I suppose the point here is the adventage of using 1:1 shots (square aspect ratio) vs non 1:1 (rectangular aspect ratio) for stitching.

No matter it is landscape orientation (overlap on left and right) or protrait (overlapp on top and bottom) orientation panorama, the same principal apply.

Not sure how the angle of view of larger sensor dimension will fit in and be related to this topic.

For the wider angle lens I perfer over narrower lens, I do a lot from shots using 24~28mm EFL, and for certain situation would not hesitate to use wide as 20mm EFL for successful stitching.

You might try it. For digital camera basically it is cost free. It is not 100% bad as some might expect.

nothing missed I think. Unless I am mistaken, the OP wants to know what would be the best way to hold the camera (landscape or portrait) if your goal is square output.

In that case, the answer depends on the desired field of view of the output and of the lens/sensor combination and on the stitching pattern chosen. In single-row panos, I often use portrait orientation for the captures, but that's because of the vertical FOV I want.

Horizontal panoramas are short and wide.
Vertical panoramas are tall and skinny.

Here's a six shot panorama I shot in a 1:1 aspect ratio at 50mm.
I was in Aperture Priority mode, so the exposure was not constant throughout, but it only varied slightly.

Steve Thomas
.

This was an old output from my album (had little chance to shoot in recent months), a pano made from 4~5 shots using 24mm focal length in 4:3 (full resolution of my system) by ICE for an around 180° AoV:

Some even older Panos made from a lot more shots mostly from 24-70 or 14-28 zoom lenses for wider than 180° AoV. They were all from non 1:1 aspect ratio shots.

The following were even taken handheld on a sailing boat:

Thank you Robert. The op's question has already been answered.

I was just querying the benefit of shooting the shots to stitch in 1:1 aspect ratio.

I normally use portrait orientation as described earlier. Below is an example of an 8 shot stitched panorama using the camera's natural aspect ratio.

DannoLeftForums,

That 's a wonderful picture.

Steve Thomas

Seems I misunderstood you. Why do you shoot like that?
(as I use raw's for my pano's it's no option for me anyway]

Thank you Steve. I'm glad you like it.

But I didn't use a 1:1 aspect ratio for the pano components shots. I still can't head my head around how shooting the components at 1:1 could have been beneficial.

Any particular reason for shooting in 1:1 aspect ratio?

My idea of shooting in-camera panorama, or stitching multiple photos on a computer, is to get a wider image that is not possible with available lens or to get greater resolution. If so, isn't 1:1 aspect ratio wasting some area of the sensor that could have added more coverage or resolution to the image?

I am just trying to understand the use case, not your question about the camera orientation. Thanks.