[Fogel70]

Yes I mix resolution and sensor size, since they both come into play to produce the final image quality.

But that has very little to do with image stabilization or the specification of the image stabilization. You just have to have to handle cameras differently regarding shutter speeds and resolution, as a higher resolution require holding the camera steadier if you want to fully utilize the higher
resolution.

Good point. The SR of the Pentax Q, claim to be able to stabilize 4 stops, so, as you mean, it's not better than the claimed SR perf of the APS-C line. The reason for this is the smaller sensor of the Q has less inertia , but at the same time, Q pixels are 1.54 microns pitched. So, you've just figured out that if you change the size of pixels and sensor in roughly the same amount, the SR performance spec stay the same.

Now, if you apply this to the move from APSC to FF, to keep the same SR performance on FF, you'd have to increase the size of pixel on the full frame sensor, so that the SR can compensate for the same amount of blur (pixel relative). For example, the A7 24Mpixel FF sensor SR is also 4.5 stops, like the K-3 24Mp APSc SR perf. spec. of 4.5 stops.

I let you think about what SR perf. would be for a Pentax FF with a 24Mpixels sensor, or 36Mpixel sensor, 42Mpixels, and 50Mpixels.

The specification of image stabilization do not care of the resolution used on the camera, it only says how much difference it make on a handheld shot with or without stabilization. That does not change with resolution.

FI Sony A7 II with 24MP and A7R II with 42MP is using the same sensor stabilization hardware and both are specified with same capacity of the image stabilization (4.5 stop). But if looking at images from them at 100% you might find that A7R II have blurred images when A7 II have not, when using same exposure settings on both.

Edit:
BTW I believe one reason for newer camera having higher ratings on image stabilization is because all now use the CIPA standard for specifying image stabilization. It's good that all use the same test method, but this test seems to make the effect of image stabilization more overrated than before, and most systems get rating on 4.5 - 5 stops. But in real use you probably get around 3 stops of image stabilization.

And the CIPA test only tests 2-axis of stabilization, so there is no added rating if offering more axis than that

---

Last edited by Fogel70; 08-07-2015 at 05:06 AM.

[Rondec]

But that has very little to do with image stabilization or the specification of the image stabilization. You just have to have to handle cameras differently regarding shutter speeds and resolution, as a higher resolution require holding the camera steadier if you want to fully utilize the higher
resolution.
The specification of image stabilization do not care of the resolution used on the camera, it only says how much difference it make on a handheld shot with or without stabilization. That does not change with resolution.

FI Sony A7 II with 24MP and A7R II with 42MP is using the same sensor stabilization hardware and both are specified with same capacity of the image stabilization (4.5 stop). But if looking at images from them at 100% you might find that A7R II have blurred images when A7 II have not, when using same exposure settings on both.

Edit:
BTW I believe one reason for newer camera having higher ratings on image stabilization is because all now use the CIPA standard for specifying image stabilization. It's good that all use the same test method, but this test seems to make the effect of image stabilization more overrated than before, and most systems get rating on 4.5 - 5 stops. But in real use you probably get around 3 stops of image stabilization.

And the CIPA test only tests 2-axis of stabilization, so there is no added rating if offering more axis than that

The more pixel dense a sensor, the harder it is to stabilize it.

That said, I don't really buy the estimations people make on how effective SR is. 3 stops, 4 stops -- I just don't believe the numbers they put out. SR is definitely worth something, but a couple of stops is what I generally see for improvement on average.

[biz-engineer]

The specification of image stabilization do not care of the resolution used on the camera, it only says how much difference it make on a handheld shot with or without stabilization. That does not change with resolution.

I hope that the spec is related to reality. Otherwise it is pretty useless. We test what we specify and we reject units that are not at least better than the spec. But, I don't know if Ricoh SR specs are guaranteed and if we can return a camera that does not provide 4.5 stops of SR, with a given amount of blur. If the spec is not related to actual blur amount on sensor, this mean you can dump this spec. If the spec is related to the amount of blur on sensor, given certain circumstances, then you can consider the sensor resolution as well.

FI Sony A7 II with 24MP and A7R II with 42MP is using the same sensor stabilization hardware and both are specified with same capacity of the image stabilization (4.5 stop). But if looking at images from them at 100% you might find that A7R II have blurred images when A7 II have not, when using same exposure settings on both.

Well, in this case, the spec of Sony is rubbish, or... it means that you don't get the resolution of 42Mpixel, but 24Mp equivalent sharpness, when shooting with moving camera.

---------- Post added 07-08-15 at 15:25 ----------

That said, I don't really buy the estimations people make on how effective SR is. 3 stops, 4 stops -- I just don't believe the numbers they put out. SR is definitely worth something, but a couple of stops is what I generally see for improvement on average.

Actually, that's very interesting. With Pentax K- bodies, I've not experienced SR effectiveness in terms of stops only, but found that it works best at some shutter speeds. For example, if I use a 50mm lens and shot with 1/25th of sec. I get more blur than if I shot with 1/10th of sec. It happened to me to shot with the DA300mm and get incredibly sharp images with shutter speed of 1/60th of sec. However, if I shot with a K-3 and 50mm lens at 1/30th of sec., with SR ON, I get 50% of blurred images.

[Glorfindelrb]

I agree with Fogel70, it's just a relative result. You get as much blur as 4.5 stops faster without SR. If your initial situation is blurry with 42MP (even if it would not with 24) you're just looking for the same amount of blur without SR. AS far as you don't change your resolution with or without SR it does not change your result.

[Fogel70]

The more pixel dense a sensor, the harder it is to stabilize it.

Yes, and this is true both with or without images stabilization.
If you use a 50 mm lens on a FF camera without stabilization you might be able to use 1/25s with a 12MP sensor and get pixel sharp images, but with 42MP you might have to use 1/100s to get pixel sharpness. With 3 stops stabilization you then can use 1/3s with the 12MP and 1/12s with 42MP camera.

The images stabilization is only compensating the camera shake, and that do not change with resolution of the sensor. But higher resolution on the sensor will put more demand on how steady the camera is to get pixel sharp images.

[D1N0]

I think the more shaky the photographer is, the harder it is to stabilize. Want pixel perfect sharpness? Don't rely on stabilisation, but use a tripod or a higher shutter time. All stabilisation does is provide acceptable sharpness. Therefore pixel density doesn't matter very much to sr. It is within the margin of error of a stabilized pic. It will make the same correction as on aps-c as on ff when the same lens is used. We cannot say how much more it will weigh because we don't know how Ricoh is going to solve it. Will the simply blow up the entire array? Will the fit a FF sensor on the same size array as the K-3 II? Will it be somewhere in between, or will they totally redesign the system?

[Nicolas06]

Good point. The SR of the Pentax Q, claim to be able to stabilize 4 stops, so, as you mean, it's not better than the claimed SR perf of the APS-C line. The reason for this is the smaller sensor of the Q has less inertia , but at the same time, Q pixels are 1.54 microns pitched. So, you've just figured out that if you change the size of pixels and sensor in roughly the same amount, the SR performance spec stay the same.

The formulas used for camera shake, shutter speed or deph of field are based on the notion of circle of confusion and for a print size of around 29x25cm (12x10") viewed at 25cm (10"). That require a resolution of 8MP.

[Fogel70]

Well, in this case, the spec of Sony is rubbish, or... it means that you don't get the resolution of 42Mpixel, but 24Mp equivalent sharpness, when shooting with moving camera.

Not really, you just need to use faster shutter speed to get full potential of the higher resolution sensor, with or without stabilization.

[Nicolas06]

Yes, and this is true both with or without images stabilization.
If you use a 50 mm lens on a FF camera without stabilization you might be able to use 1/25s with a 12MP sensor and get pixel sharp images, but with 42MP you might have to use 1/100s to get pixel sharpness. With 3 stops stabilization you then can use 1/3s with the 12MP and 1/12s with 42MP camera.

The images stabilization is only compensating the camera shake, and that do not change with resolution of the sensor. But higher resolution on the sensor will put more demand on how steady the camera is to get pixel sharp images.

A 42MP sensor will not be less sharp than a 24MP one in the same shooting conditions. At worst the condition as not good enough to get more than 24MP worth of detail and you'll be limited by the shooting condition (the lense, it's apperture, camera shake, noise).

So say if the picture was not so sharp and is worth for 8MP only, it will look sharp enough for a 10x12" print viewed from 10" away. This will be valid both for the 42MP sensor and the 24MP sensor.

In reality of course, this is not true, the 42MP sensor will be a bit sharper because there you'll loose less in the conversion to the digital image.

But if by any luck the condition are better and allow to get more than your 24MP worth of sharpness, the 42MP sensor will allow you to benefit of it, while the 24MP would not. This will be completely useless in many cases: internet sharing, printing at size smaller than 24x18", consummer grade teles/UWA, shooting at large or very small appertures, high isos settings. In some cases it could proove usefull: iso 100, huge print where people will stare at details, heavy cropping, pro lense used at it's best apperture...

By no way a sensor with more resolution is more demanding. It can provide more in ideal conditions, but doesn"t provide less in other conditions.

[Glorfindelrb]

A 42MP sensor will not be less sharp than a 24MP one in the same shooting conditions. At worst the condition as not good enough to get more than 24MP worth of detail and you'll be limited by the shooting condition (the lense, it's apperture, camera shake, noise).

So say if the picture was not so sharp and is worth for 8MP only, it will look sharp enough for a 10x12" print viewed from 10" away. This will be valid both for the 42MP sensor and the 24MP sensor.

In reality of course, this is not true, the 42MP sensor will be a bit sharper because there you'll loose less in the conversion to the digital image.

But if by any luck the condition are better and allow to get more than your 24MP worth of sharpness, the 42MP sensor will allow you to benefit of it, while the 24MP would not. This will be completely useless in many cases: internet sharing, printing at size smaller than 24x18", consummer grade teles/UWA, shooting at large or very small appertures, high isos settings. In some cases it could proove usefull: iso 100, huge print where people will stare at details, heavy cropping, pro lense used at it's best apperture...

By no way a sensor with more resolution is more demanding. It can provide more in ideal conditions, but doesn"t provide less in other conditions.

That's exactly why he said "pixel sharp" ...

[biz-engineer]

Not really, you just need to use faster shutter speed to get full potential of the higher resolution sensor, with or without stabilization.

What is 4.5 stops of stabilization if you need to use a higher speed for the same result ? Please provide an example, focal length, shutter speed and iso.

[Fogel70]

What is 4.5 stops of stabilization if you need to use a higher speed for the same result ? Please provide an example, focal length, shutter speed and iso.

For the same result you do not need to use faster shutter speed (FI to get 12MP resolution image from both 12MP and 42MP), but if you want to get higher resolution out of a higher resolution sensor you need to use faster shutter speed for that sensor, if image blur on pixel level is limited by shutter speed.

If you do not use image stabilization, do you believe it's harder to get pixel sharp image out of a 12MP sensor than a 42MP sensor?
Do you believe that changes when you add image stabilization?

But there are other things you might have to be more careful about maximizing resolution out of a high resolution sensor.
The lens used might not be good enough, higher demand on focus, the aperture used might limit resolution of the lens, ISO might limit resolution...

[Nicolas06]

That's exactly why he said "pixel sharp" ...

And that's exactly why the whole aspect you need a better SR to get same effectiveness is not relevant for general use.

Pixel sharp mean that you ask more for one sensor than the other but if you take the lower pixel count sensor and upsample the image, it will look as blury or more. While SR can fix the higher MP sensor, it can't fix the lower MP sensor image.

If you need that level of sharpness and the lower end sensor doesn't provide it mean that you have brought a camera that is innadequate to your needs.

[dr100hr]

so ,is there any news for next apc cam ?
i skip k3 (have s2 so resolution vise i'm kinda ok,but it's too small) dont like to buy each body
so i go tic toc k10 buy k20 buy,k7 buy ,k5ii buy and k3 skip (well k3ii is same cam just have some tweeks in firmware)
to finaly buy 500mm lens ,so plan is skip 1 gen and buy some good lens (or lenses) to easy up bank acc
so now my lenses need new body to play with

it there any news for next crop body ?

i dont need ff ,, if i do i will sell all i got from pentax and go nikon way ,and i love my pentax lenses too much to do that so nop for ff ,even pentax ...

[Glorfindelrb]

This thread is based on nothing real, the usual delay between 2 Pentax APS-C is around 16 months so don't expect one so early.
If you feel that you need something more than your K5ii, K3 or K3ii are your current options.