[BrianR]

There is no mention to which degree the lens is mis-focused in order to obtain an accurate reflection of the AF systems speed and responsiveness***.

It's in the italicized text below the charts:

"To minimize the effect of different lens' focusing speed, we test AF-active shutter lag with the lens already set to the correct focal distance."

They are testing how fast the camera can confirm it's already in focus and trip the shutter from different initial camera states and settings, not how fast it can focus on a mis-focused target..

The use of the Nikkor 60mm f/2.8G Macro which is not known for being a AF speed demon**, rather than using a fast 50 is somewhat perplexing.

They use the Sigma 70mm f/2.8 macro when possible, it was supposedly relatively fast at these tests and available for most systems (more under the chart in one of the Canon tests). The focusing ring isn't turning for any of it, so the long focus throw macro lens AF shouldn't be a factor.

For what it's worth, they give info on how they time these tests here Digital camera timing tests at Imaging Resource, and claim to be accurate to 0.001s.

I also see no mention of whether SR is on or not, I wonder if it would make any difference for what they're testing. I'm not convinced SR adds any lag at all to tripping the shutter - doesn't the shutter just fire anyway even if SR isn't ready yet? (This could lead to its own problems)

[Digitalis]

To minimize the effect of different lens' focusing speed, we test AF-active shutter lag with the lens already set to the correct focal distance

This is for pre-focus AF speed test, which incidentally the D5 was still faster than the K-1. I'm referring to the full AF test, if they put the Micro-Nikkor AF-S 60mm f/2.8G ED macro lens at MFD at 1:1, its AF travel would have been significantly further than starting from the MFD of the Pentax DA55mm f/1.4.

They use the Sigma 70mm f/2.8 macro when possible

This statement is irrelevant, IR didn't use that lens for AF testing with either the D5 or the K-1. The lenses they DID use were very different, so making any comparison between the two is slippery ground to be standing upon. Even AF macro lenses have longer focus throws than standard primes.

I also see no mention of whether SR is on or not, I wonder if it would make any difference for what they're testing. I'm not convinced SR adds any lag at all to tripping the shutter - doesn't the shutter just fire anyway even if SR isn't ready yet? (This could lead to its own problems)

The SR system is always active in pentax DSLRS even when it is turned off*, otherwise the sensor wouldn't be in the focal plane to begin with. You know all those complaints about rattling noises from pentax cameras with IBIS? That is because the sensor isn't fixed in position, it is held there by magnets.



*but it no longer corrects for user motion, it merely holds the sensor rigidly in place. I doubt it is always on otherwise pentax cameras would have higher battery drain, I think the mechanism is only active from half-shutter press until the mirror flips back down after exposure.

[BrianR]

This is for pre-focus AF speed test, which incidentally the D5 was still faster than the K-1. I'm referring to the full AF test, if they put the Micro-Nikkor AF-S 60mm f/2.8G ED macro lens at MFD at 1:1, its AF travel would have been significantly further than starting from the MFD of the Pentax DA55mm f/1.4.

They are starting with the lens already set to the correct focus distance, even in the "full AF" test. Compare the Full AF and Manual focus times for the K1, how would it be possible to be only .003 seconds slower if it actually had to spin the focusing ring to acquire focus?

This statement is irrelevant, IR didn't use that lens for AF testing with either the D5 or the K-1. The lenses they DID use were very different, so making any comparison between the two is slippery ground to be standing upon. Even AF macro lenses have longer focus throws than standard primes.

Again, focus throw doesn't matter for any of those tests, the lens does not actually have to focus, just confirm that it's already in focus and fire. I'm just pointing to a possible rationale for using that macro.

Please, see the bit again about the lenses being prefocused. None of these are tests of how fast the lens/camera can spin the barrel and acquire focus.

The SR system is always active in pentax DSLRS even when it is turned off*, otherwise the sensor wouldn't be in the focal plane to begin with. You know all those complaints about rattling noises from pentax cameras with IBIS? That is because the sensor isn't fixed in position, it is held there by magnets.

Sure! My point was those tests don't say anything about SR or not and AF lag. They don't test it on vs off within the K1 and comparing to another camera brand really says nothing about the lag introduced by the SR system. I suppose a masochist could go back and compare a k110d and a k100d. Unfortunately Imaging Resource only did a full test on the k100d.

[leekil]

Usually 3 weeks, but the Pentax men in white coats are hoping to get this down to a day.

For pity's sake folks go out and take pictures. If you want a camera system that will take pictures before you even see the idea go and buy a Canikon.


edit. Apologies to leekil. My frustration should not be aimed at you.

I'm not particularly concerned about the IBIS spool-up lag; I mean, I haven't noticed any issues with taking photos that have been delayed by it, I was just wondering what it was. I wouldn't think it would be very long.

The wakeup time, however, since the K-3, has been a bit longer than I would like.

[Kunzite]

SR - you know...

Is that the one thing slowing Pentax autofocus.

I just think - you elaborate!

Is in camera shake reduction a dead end?

Awaits fire...

By "just thinking" that a negative effect exists - no claim of observing an effect yourself, no detail, nothing at all - this is too similar to flame bait. You just threw the bait, and we are supposed to go all-out making up scenarios. And the "awaits fire"?
So in which ways exactly did you found out that SR is slowing down autofocus? How did you observe the slowdown, conclude it's SR, test and confirm this vague theory of yours?

By the way, what happens on the other forum is because of a single person who doesn't even knows how a lens works but makes up theories after theories (and yeah, most are "Pentax is doomed!"). Don't do that. Don't be Triplet Perar.

[D1N0]

The lenses they DID use were very different,

To minimize the effect of different lens' focusing speed, we test AF-active shutter lag with the lens already set to the correct focal distance.

AF speed of the lens is taken out of the equation, for the KP they used the 35mm F2.8 Macro. It is equally fast as the DA* 55 on the K-1.

[Kunzite]

The SR system is always active in pentax DSLRS even when it is turned off*, otherwise the sensor wouldn't be in the focal plane to begin with. You know all those complaints about rattling noises from pentax cameras with IBIS? That is because the sensor isn't fixed in position, it is held there by magnets.

*but it no longer corrects for user motion, it merely holds the sensor rigidly in place. I doubt it is always on otherwise pentax cameras would have higher battery drain, I think the mechanism is only active from half-shutter press until the mirror flips back down after exposure.

That's a good post, but allow me to add my input.

The sensor, IMO, is always in the focus plane - it's held there not by magnets but by roller balls and pressure. The magnets are moving it within the focus plane - when inactive, the sensor just rests at whatever is "down" at that moment. With SR Off, the SR system raise the sensor to position so the framing will correspond with what you're seeing in the viewfinder, and hold it there.

When and how long it's operating, that depends on the camera. In a quiet room, you can hear it operate (either holding the sensor in place or compensating for movement) - this can be user tested.
With the K-1 (and unlike the K-5 series) SR starts operating right when you half press the shutter (and stops operating shortly after you release it, or as you say right after the exposure). With the K-5 series, SR starts operating after fully pressing the shutter.

[TheOneAndOnlyJH]

That's a good post, but allow me to add my input.

The sensor, IMO, is always in the focus plane - it's held there not by magnets but by roller balls and pressure. The magnets are moving it within the focus plane - when inactive, the sensor just rests at whatever is "down" at that moment. With SR Off, the SR system raise the sensor to position so the framing will correspond with what you're seeing in the viewfinder, and hold it there.

When and how long it's operating, that depends on the camera. In a quiet room, you can hear it operate (either holding the sensor in place or compensating for movement) - this can be user tested.
With the K-1 (and unlike the K-5 series) SR starts operating right when you half press the shutter (and stops operating shortly after you release it, or as you say right after the exposure). With the K-5 series, SR starts operating after fully pressing the shutter.

That's exactly what I was wondering. I have a K-3, so I'll have to play with mine to try to figure out how it does. I always assumed the SR system was active whenever metering was active so it would be ready when you take a picture. I also use back button for focus and have half-shutter press focus disabled. I wonder if pressing the back button wakes up metering or SR like half-pressing the shutter... Questions I'll answer this weekend when I have it in my hands again!

[Digitalis]

The sensor, IMO, is always in the focus plane

This is kind of a Schrodinger cat scenario: the sensor might be in the focal plane, it also might not be. It could also partially be in the focal plane. What we do know is the sensor has to be moved to a position where to focal plane mark on our cameras to have any meaning.

it's held there not by magnets but by roller balls and pressure

From what I have seen of the SR mechanism from pentax literature the entire sensor platform appears to be free floating. Physical contact between the body and the sensor platform would increase friction and reduce the efficiency of the SR mechanism, and cause higher power draw. I don't see any permanent points of contact between the SR mechanism chassis and the sensor platform.

[Gimbal]

From what I have seen of the SR mechanism from pentax literature the entire sensor platform appears to be free floating. Physical contact between the body and the sensor platform would increase friction and reduce the efficiency of the SR mechanism, and cause higher power draw. I don't see any permanent points of contact between the SR mechanism chassis and the sensor platform.

That is how it works though, if I recall correctly there are three pressure points with "ball bearings" holding the sensor in the focal plane at all times. The sensor is free to move left/right, up/down and rotate slightly in the focal plane, the limits being a peg hitting the sides of the rectangular holes you see in the picture.

In this cut away of the K-5 you can actually see two steel balls in the lower part of the right side, with the sensor plate in between.
https://upload.wikimedia.org/wikipedia/commons/2/2e/Flickr_pentax_K5_3_%28CC-BY-SA%29.jpg

[Digitalis]

if I recall correctly there are three pressure points with "ball bearings" holding the sensor in the focal plane at all times.

Here is a clearer image:



I'd say these bearings don't hold the sensor perfectly in the focal plane*, the SR mechanism still needs to get the entire surface of the sensor in position to produce usable images.


* depth of focus at the sensor plane works in reverse to depth of field on the other side of the lens, at infinity the depth of focus at the sensor plane is razor thin - so the SR mechanism still has to be active to hold the sensor in the correct position, otherwise it would be possible to dislodge the sensor if enough force was applied.

[Gimbal]

I'd say these bearings don't hold the sensor perfectly in the focal plane*

Why not? I'll bet they do.
There is a tight fit and that is way some/most cameras need a slight nock on the side before you can here the sensor plate slide inside the body while tilting it from side to side, the famous Pentax "clunk".

[D1N0]

For three axis stabilisation the sensor just stays in the same plane all the time. There is no pitch or yaw. 5 axis only works with lenses that communicate the focussing distance (otherwise the camera cannot know how much to compensate) I would think that in de focus plane is the standard position of the sensor, with some mechanism pitching and yawing it when five axis is used. The fact that the D5 has a faster prefocussed response time has nothing to do with IBIS as focussing distance is already known and IBIS is active because of a half press. It is just faster, that is all.

[Gimbal]

with some mechanism pitching and yawing it when five axis is used.

Actually movement in all 5 "axels" are compensated by moving and rotating in the focal plane. So the senor itself is always in the same plane, only sliding left/right, up/down and rotating a few degrees.

[Kunzite]

Why not? I'll bet they do.
There is a tight fit and that is way some/most cameras need a slight nock on the side before you can here the sensor plate slide inside the body while tilting it from side to side, the famous Pentax "clunk".

Yes, they do - and the magnets cannot control back and forth movement anyway. For that a more complex mechanism would be necessary.
The easiest way is to have the sensor in the focal plane all the time. That's what is happening.
The first Minolta IBIS systems had the sensor gliding on rails, IIRC - that's why they couldn't compensate for rotational movement.

You're absolutely correct regarding compensating for 5-axis camera movements by moving the sensor in the focal plane. Unfortunately, the marketing videos presenting various IBIS systems are misleading, I'm not surprised people believe the sensor can tilt.
And actually the three axis are pitch, yaw and rotation (around the lens' axis). The other two are horizontal and vertical translation (which becomes significant at shorter distances).