The truth about SS & sharp images

_vlad wrote:

gardenersassistant wrote:

TomFid wrote:

Guy Parsons wrote:

All I can put it down to is spurious OIS/IBIS/Digital corrections happening during a frame for no obvious reason.

I've seen this too. It was merely a curiosity for a while, but came to a pointy head when I found that I couldn't stack eclipse photos, because each one had the sun a little less than round in a different way.

Oh dear. This is exactly what I was concerned about. I do a lot of stacking. Looks like my replacement lens won't cure the problem.

I think it's basically a mild version of rolling shutter distortion from camera movement. It seems to me that iS should help rather than hurt, but I'm sure it could be complex.

Presumably then you were shooting the eclipse hand-held?

It would be easy to verify this with a tripod, though I haven't bothered - hoping instead for a global shutter before the next eclipse.

The thinking here being that it shouldn't occur if using a tripod? That wouldn't help me unfortunately but it would be an interesting experiment to try. I did a couple of tripod bursts and didn't find anything untoward, but not enough for a decent test. Perhaps I'll try that more thoroughly.

I do not think it is bad even handheld - did a lot of stacked images without a tripod and so far not a big problem. This one I shot right now - first is distortion GIF and second is stacked image (blended). (I hope that gif will show-up correctly i.e animated)

This is animated gif - GH5 ES slowed down to 1/25s, handheld no extra preparations - point and shoot

And this stacked output - certainly there must be some distortions - but can you see them?

I can't see any distortions in the capture gif, nothing like the frame to frame distortion that I posted above. So I wouldn't expect to see any problems with the stack, and I don't.

Adrian Harris wrote:

gardenersassistant wrote:

Adrian Harris wrote:

It is well known that the lightweight Panasonic 14-140mk2 can misbehave when using mechanical shutter, which is easily solved by avoidance.

Is this a reference to shutter shock, or other issues too? The reason I ask is that I have just arranged to get a replacement for a new 14-140 mk2 because of a strange issue with it on my G80 and I'm wondering whether this is a known about "feature" of the lens in general rather than a malfunction with this particular sample of the lens. (So I won't be surprised if it happens with the replacement.)

I first noticed it with a single shot but the easiest way to illustrate it is with successive shots captured using aperture bracketing, with one of the shots having a different geometry from the others. They were captured hand-held but I don't believe such lopsided geometric distortions can arise from camera and/or subject movement, especially when the next shot after the distorted one reverts to the previous geometry.

There are three frames in the first two animations, one before and one after the distorted frame. The first two are shown for 1 second, the third for 2 seconds. The third example is only two frames, a normal frame and then a distorted one. Since these were captured as bracketed images they were captured in quick succession at a rate of several per second.

There are another five in this album at Flickr.

Click on Original size beneath the image to see the animation.

Is this one of the misbehaviours that has been discussed previously?

Hi Nick, I finally got the chance to see the distortion animation on a PC and recognised it immediately - I think !

I have seen this effect many times and with all types of lenses when using electronic shutter (it's especially bad on Panasonic's slow readout electronic shutter).

With electronic shutter the scene is read from the image sensor line by line, so any slight sideways movement by the photographer means that each part of the image is recorded and readout at a different time and hence camera movement introduces aspect distortion of some sort. Even IBIS doesn't stop this and it is far worse with high magnification - such as when shooting macro or when using a long telephoto lens, which is why I believe that OIS is necessary to compliment IBIS on long lenses. Hence I should imagine the reason that even Olympus thought it necessary - despite their fantastic IBIS - to fit OIS on their 300mm Pro level telephoto.

Using mechanical shutter should significantly reduce this type of distortion as it scans the sensor much faster.

However if you were not using E-shutter when you shot these, then I have no idea what caused it in your particular case!

Yes, I was using mechanical shutter.

For reference: OIS = Optical Image stabilisation (fitted on lens), and IBIS + In Body Stabilisation.

Yes. I was using both. Dual IS. IBIS + OIS.

-- hide signature --

Adrian
http://www.artfotografia.co.uk/

alcelc wrote:

I don't think there was nothing wrong with your lens.

Very often when shooting Exposure Bracketing shots handheld we might not realize how big the movement would be between shots. Generally by a good HDR program we can eliminate the issue making us not easy to really know how serious it would be.

There following 3 shots are partial (other 2 had been long deleted) of a sequence of 5 shots of 1/3ev apart shoot under Exposure Bracketing on GX7 and 12-35 f/2.8 that I can still located on my Hard Disk:

-0.66ev

0ev

+0.33ev

These were shot between 1/8000" ~1/4000" and theoretically under that fast shutter speed and on consecutive shooting, it might not expect to have any significant movement among shots. But they did.

I trust if we look hard into the detail of each of the image from the sort of handheld Exposure Bracketing shooting, we might not difficult to see the movement involved not limited to any particular model or lens.

Movement between shots in a hand-held burst is normal. However, I don't think that is what is going on with my examples. For example, this animation shows normal movement between two shots in a burst. Everything moves the same amount in the same direction.

(Note: This animation uses two successive images in a burst. They are overlaid and the opacity of the upper version goes from 0 to 100% during the animation. This is done to illustrate the direction of movement from the first to the second image.)

Click on Original size beneath the image to see the animation

The next animation shows the movement in one of my distortion examples. Different parts of the image move in different ways. The top (as we look at it here) moves just a little to the left. The bottom moves a lot to the right. The image is being skewed.

Click on Original size beneath the image to see the animation

You might think it is because the flower is twisting around in the breeze. However, that would not explain similar skewed movement in the following example where we are looking at objects in fixed positions on the ground.

(Note: This animation just uses two successive images in a burst without illustrating the direction of movement with intermediate opacity versions.)

Click on Original size beneath the image to see the animation

Adrian Harris wrote:

Almost all cameras experience some form of SS under certain circumstances.

I was first made aware of it when I came across a Nikon shooter draped bodily over a long lens mounted on a tripod so sturdy that I could not have lifted it. Of course I asked him why he was shooting like that and he said he was trying to reduce mirror slap when activating the shutter.

I bought a Sony A77 with a fixed translucent mirror and always used the electronic first curtain, yet I have from it some of the worst examples of SS that I have ever seen!

My Panasonic gx7 would produce SS easily with mk1 14-140 (ruined masses of shots). And the flash sync was a lowly 1/160.

Next was my dream camera the GX8. With a seriously fast shutter and a flash sync of 1/250 fabulous and just what I needed.

Like all cameras and tools nothing is ever perfect and it does have some caveats. However I quickly learned how to get the best out of it, which included finding out which lenses performed best with it to ensure I didn't lose any images through SS. .... And guess what, surprisingly when I am forced to use the mechanical shutter due to flickering LED lighting, I can guarantee that 100% of the time when using my Olympus lenses I never get any SS with it at all !

It is well known that the lightweight Panasonic 14-140mk2 can misbehave when using mechanical shutter, which is easily solved by avoidance.

So Whose lenses should we use for performance and sharp shots...

Interestingly I have repeatedly found that for really sharp action images the Olympus 40-150 f2.8 pro lens performs so much better on the GX8 than it does on the Olympus em1-mk2 !

Conversely - and this is getting weird - the Panasonic 100-400 performs better and produces sharper images on the Olympus em1-mk2 than it does on the Panasonic GX8 or Gx7 !

How do I know all this, unfortunately I found out the expensive way by spending the kids inheritance and my pension

A final note: sadly although many suggest Panasonic 'upgraded' the 'faulty' GX8 shutter, for my photography it was a huge downgrade. We lost the fast shutter speed and also the fast flash sync speed, yet strangely so many seem thrilled about that ?

Let's bring back equivalence threads.  I mean what the heck, there are plenty of horses to beat to death.