[WPRESTO]

It would appear that the possibility of SR jiggling the sensor when a camera is tripod-mounted, and thereby degrading IQ, is one of those based-on-theory ideas that when tested is false.

BUT

The SR system responds to movement detected on the sensor. If the camera is tripod-mounted, and there is no motion by any object in the FOV, presumably SR would remain inactive. That is certainly the situation if you're testing the theory using a test chart or other totally fixed target. That mirror-lockup does improve IQ when using long shutter times even when mounted on a tripod should not be too surprising. The sensor is delicately mounted so the SR system can move it, so vibration induced by mirror-slap is probably a greater problem than in the film era. MFT owners (including myself) know that sensor vibration induced by the first curtain of a mechanical shutter can degrade IQ when using long shutter speeds regardless of the massiveness of tripod and head (it's the delicately mounted sensor that's vibrating, not the camera/lens). Consequently MFT owners when using shutter speeds slower than about 1/30, will or should engage the vibrationless electronic shutter or electronic 1st curtain option, despite the possibility of smearing if something moves. But, if the latter is a risk, there would also be risk of blurring because of subject motion as there always has been when using a long shutter speed*. At long shutter speeds there are multiple ways IQ might be reduced because of movement.

*Two famous examples from way, way back of subject motion causing blurring at longer shutter speeds:
1) A famous studio portrait of Abe Lincoln in which he's sitting in a chair with his legs crossed and the foot that's not on the ground is blurred. This has been taken as evidence of a heart problem associated with Marfan's syndrome, the pumping of blood causing his foot to bounce up and down.
2) A picture of the hanging of the commander of the infamous Confederate prison camp at Andersonville where conditions were so horrible. Taken just after the trap was released, the dangling man is blurred because his neck was not broken by the drop and he was writhing in agony as he strangled to death.

[vonBaloney]

Doesn't it really depend on the user and the conditions? If you are using IR remote, timer, or MLU, it will be off anyway. So the only time it would be on when using tripod is manual shutter release with your hand on the camera or possibly wired remote. And if you are using manual focus lenses or panning/tilting around (maybe on a gimbal head), you may also have a hand on the lens itself. Some simple tests in the field under actual conditions will tell you whether it is better or worse right now in those conditions. I've seen it help, I've seen it hurt, and I've seen it make no difference. Under ideal fixed conditions, yeah it probably makes no difference, but if you are under ideal fixed conditions that would probably include some sort of hands-off trigger where it wouldn't be activated anyway. So it just depends, there are a million variables. But just generally, the longer the focal length, the more likely it will be to make a difference (for better or worse). (Shutter speed also very important, obviously.)

[stevebrot]

The SR system responds to movement detected on the sensor. If the camera is tripod-mounted, and there is no motion by any object in the FOV, presumably SR would remain inactive.

The SR system uses inertial sensors. There is no optical component.

BTW, I like your way-back-when examples.


Steve

[WPRESTO]

The SR system uses inertial sensors. There is no optical component.

BTW, I like your way-back-when examples. Steve

Thanks for the technical correction. I find the second example of way-back-when rather creepy.

[photoptimist]

It would appear that the possibility of SR jiggling the sensor when a camera is tripod-mounted, and thereby degrading IQ, is one of those based-on-theory ideas that when tested is false.

It's not false, just not true for everyone's tripod-lens-camera combination.

Whether a specific tripod-lens-camera combination has the problem depends on the mass distribution of the lens-camera-tripod head, tripod stiffness, and damping of the tripod. In particular, if the tripod-lens-camera has little damping (measurable by how many wobbles the camera takes to stop vibrating after a tap on the body), then the combination will be more prone to problems. And if the natural resonant frequency tripod-lens-camera combination happens to be a harmonic of the SR system control loop, then it can actually delay damping of those vibrations or even make them worse.

[WPRESTO]

It's not false, just not true for everyone's tripod-lens-camera combination.

Whether a specific tripod-lens-camera combination has the problem depends on the mass distribution of the lens-camera-tripod head, tripod stiffness, and damping of the tripod. In particular, if the tripod-lens-camera has little damping (measurable by how many wobbles the camera takes to stop vibrating after a tap on the body), then the combination will be more prone to problems. And if the natural resonant frequency tripod-lens-camera combination happens to be a harmonic of the SR system control loop, then it can actually delay damping of those vibrations or even make them worse.

I basically agree with you BUT, tests of MFT cameras placed on very heavy tripods showed IQ degradation caused by first shutter curtain induced vibration when the camera was using longish shutter speeds.I suspect a shutter curtain induces less vibration that a mirror slap, but I'm not sure whether the shutter curtain of an MFT is a greater source of vibration-inducing impact than a DSLR shutter or mirror, because DSLRs may have a better ratio of mass-of-shutter-curtain or mass-of-mirror to mass-of-camera-body. I don't think it's something one could or should calculate or guess. Determining whether physically bigger cameras have an advantage would require actual testing.


BTW, in theory the vibration problem should disappear at very long shutter speeds. As the shutter remains open longer, the vibrations die out and the duration of them comes to constitute only a small fraction of the entire exposure time, eventually becoming too small to notice. This effect was explained to me by a pro who sometimes took pictures by UV light in an otherwise totally darkened room with exposures that were 15 minutes up to more than an hour long. Aside from closing doors and people walking by, there was a heavy electric fork-lift (called "Uncle Louis") that sometimes drove down the hall shaking the entire floor of the building - impossible to ignore if you were sitting in one of the offices. But those vibrations were so brief relative to the length of the exposure, their effect could not be detected on the negatives or prints.

[photoptimist]

I basically agree with you BUT, tests of MFT cameras placed on very heavy tripods showed IQ degradation caused by first shutter curtain induced vibration when the camera was using longish shutter speeds.I suspect a shutter curtain induces less vibration that a mirror slap, but I'm not sure whether the shutter curtain of an MFT is a greater source of vibration-inducing impact than a DSLR shutter or mirror, because DSLRs may have a better ratio of mass-of-shutter-curtain or mass-of-mirror to mass-of-camera-body. I don't think it's something one could or should calculate or guess. Determining whether physically bigger cameras have an advantage would require actual testing.


BTW, in theory the vibration problem should disappear at very long shutter speeds. As the shutter remains open longer, the vibrations die out and the duration of them comes to constitute only a small fraction of the entire exposure time, eventually becoming too small to notice. This effect was explained to me by a pro who sometimes took pictures by UV light in an otherwise totally darkened room with exposures that were 15 minutes up to more than an hour long. Aside from closing doors and people walking by, there was a heavy electric fork-lift (called "Uncle Louis") that sometimes drove down the hall shaking the entire floor of the building - impossible to ignore if you were sitting in one of the offices. But those vibrations were so brief relative to the length of the exposure, their effect could not be detected on the negatives or prints.

Vibrations only die out if there's nothing that keeps exciting the system either from the outside (e.g., "Uncle Louis") or from the inside (e.g., the SR system twitching the sensor a bit to keep it in place).

These SR-induced vibrations don't always die out. Electronic noise in the SR system sensors causes tiny twitches. Under most conditions, the twitches have no noticeable effect and do die out. But if the natural frequency of the tripod-head-camera-lens system matches the resonant frequency of the SR system control loop, that doesn't happen. In the worst case, the tiniest vibration of sensor noise amplifies and the vibrations get worse and worse over time.

As for calculating the effect, it's exactly what mechanical engineers and control system designers are paid to do. There's various ways to model the tripod-head-camera-lens system to estimate the dynamical behavior of it as a mechanical system (e.g., the natural frequencies of the system in various vibrational directions). And an electrical engineer can calculate the response of the SR system to various frequencies. Or these can be measured in various ways, too.

[stevebrot]

I suspect a shutter curtain induces less vibration that a mirror slap,

It depends on the camera. The mirror has already completed its travel and hopefully settled down by the time the shutter opens and most of its impact (so to speak) is in the form of residual harmonics. The curtains on the other hand, despite being very light, are accelerated very quickly to a fairly high velocity resulting in two pulses in one direction, followed by a single deceleration to zero in the opposite direction, all within the context of when the exposure is actually being made. Shutter actuation is fairly violent.

You Tube

You Tube

Love this video


Steve

[jbinpg]

The acid test for any mirror slap/shutter shock is one or more specular highlights on a dark field. Yes, think astro and a field of stars. Even a very short shock will lead to squiggly stars.

[Rondec]

It probably depends on how steady your tripod is. Certainly it is reasonable to leave it on on a monopod. If you are using the two second self timer option then it is automatically turned off. I don't think if you are using a really long lens and your tripod isn't the steadiest that it would be a problem at all to leave it on. The big question here, I suppose, is how steady your overall tripod/camera structure is. SR may not worsen things, but if it is good enough without it then there is no sense adding that variable into the situation.

[stevebrot]

The big question here, I suppose, is how steady your overall tripod/camera structure is.

That, my friends, is the elephant in the room. The problem is that the problem is not very approachable both in terms of measuring and in terms of solving.


Steve

(...the problem is the problem is the problem is...)

[photoptimist]

That, my friends, is the elephant in the room. The problem is that the problem is not very approachable both in terms of measuring and in terms of solving.


Steve

(...the problem is the problem is the problem is...)

Exactly! And that means that a sturdy tripod might be fine unless the center column is extended. Or it might be fine as long as the thinnest leg segments are retracted. Or it might be fine with a small lens but horrible with a large lens using the lens tripod foot.

As a mechanical structure, a tripod legs + tripod column + head + camera body + lens has very different vibrational properties depending on the exact configuration.

[vonBaloney]

That, my friends, is the elephant in the room. The problem is that the problem is not very approachable both in terms of measuring and in terms of solving.

Yep.

As a mechanical structure, a tripod legs + tripod column + head + camera body + lens has very different vibrational properties depending on the exact configuration.

Yep.

Like I said above, test in the field under real conditions, get to know what happens: what helps, what hurts, what doesn't matter.