Mystery of OIS & 1/50 Solved

[Miltz]

Advertisement (gone after registration)

Since I got my Leica Q3 I’ve been wondering why Leica choose 1/50 of a second to be the Automatic shutter speed. I was keeping sharp images all the way down to 1/15-1/8. 1/50 is also the point where OIS turns on automatically. 

After running multiple tests over the last two weeks I can confirm why Leica chose 1/50. Corner Sharpness. Typically I don’t shoot flat objects and focus in the center, but I have been the last few weeks and well I can confirm when shooting at less than 1/50 corner sharpness plummets and is basically unusable. (When focusing in the center of the frame) This is something you’ll never see if your subject has depth and shoot relatively wide open. If you need sharp corners and are shooting handheld use 1/50, Leica picked it for a reason. Now if you’re shooting something flat and focus on the corner at less than 1/50 the corner will be sharper than the rest of the image. The Q3 lens behaves very differently than my Nikon 35mm 1.8s. Basically with that lens if I focus on the center and shoot at 1/15 with IBIS on the entire frame is sharp. I’m guessing the OIS system in the Leica is more limited due to space constraints. This is only a hypothesis. 

Overall I’m disappointed that shooting at 1/8-1/15 with OIS in isn’t good for flat subjects but it’s better to know and now plan for it. Luckily it’s not something I do often so that’s a plus. 
 

 

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[David Wien]

Hmm. I must investigate this, because if I understand you correctly it could be important when I am photographing archive documents in low light. are you suggesting for focus all over the page that this should be done at 1/50 sec or faster?

David

 

Edited January 30, 2024 by David Wien

[SrMi]

7 hours ago, Miltz said:

Since I got my Leica Q3 I’ve been wondering why Leica choose 1/50 of a second to be the Automatic shutter speed. I was keeping sharp images all the way down to 1/15-1/8. 1/50 is also the point where OIS turns on automatically. 

After running multiple tests over the last two weeks I can confirm why Leica chose 1/50. Corner Sharpness. Typically I don’t shoot flat objects and focus in the center, but I have been the last few weeks and well I can confirm when shooting at less than 1/50 corner sharpness plummets and is basically unusable. (When focusing in the center of the frame) This is something you’ll never see if your subject has depth and shoot relatively wide open. If you need sharp corners and are shooting handheld use 1/50, Leica picked it for a reason. Now if you’re shooting something flat and focus on the corner at less than 1/50 the corner will be sharper than the rest of the image. The Q3 lens behaves very differently than my Nikon 35mm 1.8s. Basically with that lens if I focus on the center and shoot at 1/15 with IBIS on the entire frame is sharp. I’m guessing the OIS system in the Leica is more limited due to space constraints. This is only a hypothesis. 

Overall I’m disappointed that shooting at 1/8-1/15 with OIS in isn’t good for flat subjects but it’s better to know and now plan for it. Luckily it’s not something I do often so that’s a plus. 
 

 

Can you share examples?

[Miltz]

7 hours ago, David Wien said:

Hmm. I must investigate this, because if I understand you correctly it could be important when I am photographing archive documents in low light. are you suggesting for focus all over the page that this should be done at 1/50 sec or faster?

David

 

Hi David, If you are going to photograph a document in low light hand held I would focus in the center and keep your shutter speed at 1/50. No need to go faster than that unless your hands are unstable. F2.8 should be plenty of DOF for a flat subject. 

[Knorp]

Amazing, so shooting with OIS on and a shutter speed slower than 1/50 sec the center of the image is sharp but the corners are soft, even unusable ?

Could it be field curvature ?

[BobsFirstLeica]

5 hours ago, SrMi said:

Can you share examples?

Yes!  OP.  Please supply examples along with "all" shooting parameters.  There may be other factors at play.  Thanks.

Bob

[Miltz]

Advertisement (gone after registration)

9 hours ago, Knorp said:

Amazing, so shooting with OIS on and a shutter speed slower than 1/50 sec the center of the image is sharp but the corners are soft, even unusable ?

Could it be field curvature ?

Only when shooting a flat plane. Normal shooting I’ve even gone down to 1/8 with sharp results. I don’t think it’s field curvature since I was shooting a flat plane. It basically looks like camera shake on all 4 corners. 

[Miltz]

9 hours ago, BobsFirstLeica said:

Yes!  OP.  Please supply examples along with "all" shooting parameters.  There may be other factors at play.  Thanks.

Bob

So when initially started this it wasn’t to test it out or for a forum post. I would just discard the bad images. But since I was doing the same thing over and over for 2 weeks I noticed a pattern and was able to verify it after by doing some tests. I didn’t save these images. The good news is anyone can do the same tests themselves if they don’t believe my findings. The control was the Nikon Z6 with 35mm 1.8s If that image was blurry the entire shot was blurry and if it was sharp the entire shot was sharp. Where the Leica you can have perfect center sharpness and have the corners be blurry. I guess this is also a difference between IBIS and OIS. 
Basically if you want perfect corner to corner sharpness that’s repeatable on a flat plane handheld shoot 1/50 with OIS on. Now you know. 

[hepcat]

12 hours ago, Miltz said:

So when initially started this it wasn’t to test it out or for a forum post. I would just discard the bad images. But since I was doing the same thing over and over for 2 weeks I noticed a pattern and was able to verify it after by doing some tests. I didn’t save these images. The good news is anyone can do the same tests themselves if they don’t believe my findings. The control was the Nikon Z6 with 35mm 1.8s If that image was blurry the entire shot was blurry and if it was sharp the entire shot was sharp. Where the Leica you can have perfect center sharpness and have the corners be blurry. I guess this is also a difference between IBIS and OIS. 
Basically if you want perfect corner to corner sharpness that’s repeatable on a flat plane handheld shoot 1/50 with OIS on. Now you know. 

Forgive me, but I am totally lost here.  Corner sharpness has always been a function of aperture as it relates to relative sharpness over the entire field of view.  Most lenses are soft at the corners wide-open, but begin to sharpen as you stop down, with most lenses' sweet spot somewhere between f/5.6 and f/11.

Blur from camera motion is a function of shutter speed; if the camera moves during a slow exposure, the image is soft, or even blurred.   IBIS and OIS are designed to move the sensor (or lens elements) to keep the sensor stable behind the image that it is capturing by counter-movement and sensing camera shake (or motion.)

Can someone please explain the relationship of shutter speed, OIS, and corner sharpness to me please?  What am I missing here?   Are you not seeing corner sharpness increase with an increase in DOF provided by the aperture as you slow the shutter speed down?

Edited January 31, 2024 by hepcat

[BobsFirstLeica]

13 hours ago, Miltz said:

So when initially started this it wasn’t to test it out or for a forum post. I would just discard the bad images. But since I was doing the same thing over and over for 2 weeks I noticed a pattern and was able to verify it after by doing some tests. I didn’t save these images. The good news is anyone can do the same tests themselves if they don’t believe my findings. The control was the Nikon Z6 with 35mm 1.8s If that image was blurry the entire shot was blurry and if it was sharp the entire shot was sharp. Where the Leica you can have perfect center sharpness and have the corners be blurry. I guess this is also a difference between IBIS and OIS. 
Basically if you want perfect corner to corner sharpness that’s repeatable on a flat plane handheld shoot 1/50 with OIS on. Now you know. 

I have yet to test what you have done but intend to do so. @hepcat is correct with his statements about sharpness, blur, etc.  The weather forecast for my area predicts rain for an extended period, one which I probably will be able to test your "pattern"  out.

Bob

[Knorp]

14 hours ago, Miltz said:

Only when shooting a flat plane. Normal shooting I’ve even gone down to 1/8 with sharp results. I don’t think it’s field curvature since I was shooting a flat plane. It basically looks like camera shake on all 4 corners. 

Exactly. Shooting a flat plane could be prone to field curvature of the lens where the plane of focus is not flat but curved, hence the sharp center and soft corners.

[Miltz]

4 hours ago, Knorp said:

Exactly. Shooting a flat plane could be prone to field curvature of the lens where the plane of focus is not flat but curved, hence the sharp center and soft corners.

It’s not field curvature because the corners get perfectly sharp at or past 1/50. It’s strictly related to the sensor stabilization, or lack of it.

[hepcat]

48 minutes ago, Miltz said:

It’s not field curvature because the corners get perfectly sharp at or past 1/50. It’s strictly related to the sensor stabilization, or lack of it.

Please see my post above.  I still fail to understand how sensor stabilization and shutter speed affect corner sharpness.

 

[Miltz]

4 hours ago, hepcat said:

Please see my post above.  I still fail to understand how sensor stabilization and shutter speed affect corner sharpness.

 

The OIS on the Leica Q3 prioritizes the focus point when it comes to which part of the photo is sharp at slow shutter speeds. This method works for more situations, but doesn’t work for flat subjects if you need corner to corner sharpness and are shooting handheld below 1/50. Remember the Q has a floating lens element. That’s the sound you hear when you shake the camera. 

Edited February 1, 2024 by Miltz

[Ad Dieleman]

8 hours ago, hepcat said:

Please see my post above.  I still fail to understand how sensor stabilization and shutter speed affect corner sharpness.

 

My guess is that the moving element(s) of the optical stabilization system in the lens affects the flatness of the focus plane and/or may induce tilt of the focus plane.

[LocalHero1953]

I don't see how OIS can make one part of the image more in focus than another. Why wouldn't OIS affect all the image equally?
That would imply that the OIS system is analysing the image to decide where the focus point is; I thought it was purely an electro-mechanical process responding to movement sensors, not what's in the image.
I'm happy to be convinced otherwise.

[hepcat]

9 hours ago, Miltz said:

The OIS on the Leica Q3 prioritizes the focus point when it comes to which part of the photo is sharp at slow shutter speeds. This method works for more situations, but doesn’t work for flat subjects if you need corner to corner sharpness and are shooting handheld below 1/50. Remember the Q has a floating lens element. That’s the sound you hear when you shake the camera. 

Picking a focus point in the image using autofocus has been around for a long, long time.   It doesn't affect the corners, just the plane of focus.

From an optical theory standpoint, shutter speed and depth of field are just not related except through aperture.  I still don't understand. The corner sharpness of a lens is determined by its optical formula and the aperture at which it is used.  Shutter speed has no effect on lens corner sharpness.  I have several OIS lenses and two IBIS bodies and you postulation that shutter speed is fixing corner sharpness is akin to suggesting that horsepower affects suspension ride quality in a car.

I'm just not making sense of your assertion.  Unless the Q3 does something no other camera has ever done before, like moving the center of optical element ALL over the sensor so the "sweet spot" is the only thing making the complete image, I'm just not making sense of your assertion.   And that technology would be all over the news instead of the "all at once" sensor technology recently introduced.   The only thing that could affect edge sharpness when you're using a slower shutter speed is that the aperture is smaller (going from f/2.8 at 1/250 to f/8 at 1/50 for example.

Can you please point me to something that supports your assertion so I can figure out what's really happening?  

Thanks.

 

[Daniel C.1975]

Hi, 

 

honestly and being blunt here, while of course in due respect, I call this solved mystery nonsense. From an optical engineering perspective it just makes no sense at all. 

I wanted to write more, but hepcat already described it quite nicely. 

 

I believe the TO just came to wrong conclusions based on a faulty test set-up (and maybe, and I really do not want to overstep, a missing technical background)

It is, simplified, all a function of the plane of focus, depth of field and the field curvature (which can do funny things). That's all. Plus, the corner sharpness of the Q(n) is good, but not great. 

 

Saying this as an engineer. 

 

Cheers

Daniel

 

 

[Advertisement]

Hallo,

Look here: Leica Q

[SrMi]

On 1/29/2024 at 10:10 PM, Miltz said:

Since I got my Leica Q3 I’ve been wondering why Leica choose 1/50 of a second to be the Automatic shutter speed. I was keeping sharp images all the way down to 1/15-1/8. 1/50 is also the point where OIS turns on automatically. 

After running multiple tests over the last two weeks I can confirm why Leica chose 1/50. Corner Sharpness. Typically I don’t shoot flat objects and focus in the center, but I have been the last few weeks and well I can confirm when shooting at less than 1/50 corner sharpness plummets and is basically unusable. (When focusing in the center of the frame) This is something you’ll never see if your subject has depth and shoot relatively wide open. If you need sharp corners and are shooting handheld use 1/50, Leica picked it for a reason. Now if you’re shooting something flat and focus on the corner at less than 1/50 the corner will be sharper than the rest of the image. The Q3 lens behaves very differently than my Nikon 35mm 1.8s. Basically with that lens if I focus on the center and shoot at 1/15 with IBIS on the entire frame is sharp. I’m guessing the OIS system in the Leica is more limited due to space constraints. This is only a hypothesis. 

Overall I’m disappointed that shooting at 1/8-1/15 with OIS in isn’t good for flat subjects but it’s better to know and now plan for it. Luckily it’s not something I do often so that’s a plus. 

To my knowledge, Leica activates OIS at shutter speeds below 1/60 sec. That cutoff is based on the standard double the focal length recommendation (1/(2*28)), but it does not work for everybody as the visible shake depends on the photographer,

Are you saying that the center stays sharp at shutter speeds below 1/50, but the corners get less sharp? Yes, that is possible and is induced by a camera shake when handholding. When you press the shutter button, the camera can rotate slightly. The rotation is much smaller in the center than at the edges; hence, the edges can be blurrier.

The way the camera shake manifests itself in the image can be different.

In my brief tests with Q3, with OIS off and at 1/15 sec  (bookshelf as a target), both the edges and the center lose similar sharpness compared to 1/15 sec with OIS on.

[adan]

39 minutes ago, SrMi said:

Are you saying that the center stays sharp at shutter speeds below 1/50, but the corners get less sharp? Yes, that is possible and is induced by a camera shake when handholding. When you press the shutter button, the camera can rotate slightly. The rotation is much smaller in the center than at the edges; hence, the edges can be blurrier.

FINALLY - someone who understands the physics!

Exaggerated rotational camera shake. M10 (with 28 Summicron to simulate Q), 1/15 second.

The motion blurs get larger towards the edges/corners - even to the extent that some objects take on a cone shape. And certainly to the extent that OIS can't handle them.

Really not any different than "star trail" astronomical time exposures. The star in the center, on the rotational axis, is sharper than the long curving arcs created by the stars progressively farther from the rotational axis.

https://en.wikipedia.org/wiki/Star_trail#/media/File:Star_trail_and_aurora_over_Mount_Wellington,_Tasmania.jpg

Alles ist klar? Ja oder Nein?

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Edited February 1, 2024 by adan