[MSL]

I know this topic has been on the Forums before and now I remember why. Some very detailed work done by Falk Lumo (AKA @falconeye) on this topic a few years back which I found when doing a general web search

Falk Lumo: Pentax shake reduction revisited

[Flugelbinder]

Thank you. Will have a read...

[Flugelbinder]

I know this topic has been on the Forums before and now I remember why. Some very detailed work done by Falk Lumo (AKA @falconeye) on this topic a few years back which I found when doing a general web search

Falk Lumo: Pentax shake reduction revisited

Very interesting read, and he comes to the same conclusion as I have; shake-reduction is a good option for low shutter speed values (1/50 and slower).

[kh1234567890]

-CPU: Ops, the user pressed down fully, SR lift up the image sensor and start counteract the movements, and while you do that I will flip up the mirror.

The various patents imply that sensor movement only occurs after mirror up (except when in live view).

[BrianR]

Perhaps before you jump to that conclusion you may wish to read the full report from P. Smith. Read it very thoroughly and carefully you just might gain some more insight. In his testing he mentions some cause due to possible Low Pass Filter, remember neither your K5 lls nor the K3 have one.

You mean this (from Study of the Effectiveness of Shake Reduction in the Pentax K7)

"The small local peak in motion blur is probably a property of the low pass filter in the sensor correction circuitry."

or something else? The above looks like speculation on circuitry that makes the sr ignore higher frequency signals, not the anti-aliasing filter that the k5iis/k3 lack, though I haven't got to the discussion on dpreview yet. Interesting stuff in any case.

[Oldbayrunner]

You mean this (from Study of the Effectiveness of Shake Reduction in the Pentax K7)

"The small local peak in motion blur is probably a property of the low pass filter in the sensor correction circuitry."

or something else? The above looks like speculation on circuitry that makes the sr ignore higher frequency signals, not the anti-aliasing filter that the k5iis/k3 lack, though I haven't got to the discussion on dpreview yet. Interesting stuff in any case.

"It is probable that the sensor correction circuitry has a low pass filter that starts to act at about 1/125 sec and comes into full play at 1/60 sec.From this it appears that there is no harmonic resonance causing sensor vibration as many have speculated. It is simply the effect of a low pass filter without an abrupt corner frequency that starts to reduce motion blur at an approximate shutter speed of 1/80 sec. To clarify this more measurements were made at a shutter speed of 1/80 sec. See illustration 4."


Anti-aliasing filter
From Wikipedia, the free encyclopedia

Optical applications

In the case of optical image sampling, as by image sensors in digital cameras, the anti-aliasing filter is also known as an optical lowpass filter or blur filter or AA filter.

[MSL]

Very interesting read, and he comes to the same conclusion as I have; shake-reduction is a good option for low shutter speed values (1/50 and slower).

Yes, but he also seems to say that it isn't going to be detrimental at fast shutter speeds.

[BrianR]

"It is probable that the sensor correction circuitry has a low pass filter that starts to act at about 1/125 sec and comes into full play at 1/60 sec.From this it appears that there is no harmonic resonance causing sensor vibration as many have speculated. It is simply the effect of a low pass filter without an abrupt corner frequency that starts to reduce motion blur at an approximate shutter speed of 1/80 sec. To clarify this more measurements were made at a shutter speed of 1/80 sec. See illustration 4."


Anti-aliasing filter
From Wikipedia, the free encyclopedia

Optical applications

In the case of optical image sampling, as by image sensors in digital cameras, the anti-aliasing filter is also known as an optical lowpass filter or blur filter or AA filter.

Yes, I know it's also called an optical low pass filter, but I don't think that's he's talking about there... "sensor correction circuitry has a low pass filter". I'm pretty sure he means circuitry (not an optical thingie) that cuts out higher frequency vibrations so the SR can ignore them instead of being confused.

From A study of the effectiveness of K7 shake reduction: Pentax SLR Talk Forum: Digital Photography Review, he says:

"When you eliminate camera movement the movement sensors have nothing to report and as a consequence send no corrective signals to the sensor control motors. As I reported, I believe the movement sensor signals are passed through a low pass filter to eliminate high frequency components in the signal."

[Oldbayrunner]

"sensor correction circuitry has a low pass filter".

There is no such thing!

Infrared Reduction / Low Pass Filters

A low pass filter, as it pertains to a digital SLR camera, is an optical low pass filter that allows certain wavelengths to pass while reducing or disallowing all others. This is used to combat colour aliasing and moiré artifacts in the images. This was accomplished in the past by increasing the birefringent (double refraction) index of the material, usually monocrystalline silica, used to compose the filter. This, unfortunately, had the side effect of blurring the images in these particular areas and usually needed software algorithms to make adjustments. Today advanced proprietary technologies from some of the bigger digital SLR companies are making great advancements in the reduction of distorting artifacts.

What is a Low-Pass Filter?

January 3, 2012 By Nasim Mansurov
A low-pass filter, also known as anti-aliasing or “blur” filter, was designed by camera manufacturers to eliminate the problem of moiré by blurring what actually reaches the sensor. While extreme details are lost in the process, the problem of moiré is completely resolved. Since most cameras are designed to be used for day-to-day photography, where moiré pattern is very common, most cameras on the market today use a low-pass / anti-aliasing filter. While this surely benefits most photographers out there, it is a big blow on landscape photographers that never see moiré and yet end up with blurred details.Because of this problem, some companies on the market started specializing in removing the low-pass / anti-aliasing filter from modern DSLR cameras, specifically targeting landscape photographers. Most digital medium-format and some high-end cameras do not have a low-pass filter, because they want to deliver the best performance from their sensors. While those cameras are affected by moiré, manufacturers leave it up to the photographer to decide on how to avoid it or deal with it in post-processing. Below you will find two examples of low-pass filters used on typical Nikon DSLRs and on the Nikon D800E.

Hence what Pentax did with the K5 lls.

http://www.scribd.com/doc/27443479/Pentax-Shake-Reduction-Patent-20080226276#scribd


Read the patent there is no mention of a Low Pass filters, only High Pass ones.

[BrianR]

You quoted the guy:

"It is probable that the sensor correction circuitry has a low pass filter that starts to act at about 1/125 sec and comes into full play at 1/60 sec.From this it appears that there is no harmonic resonance causing sensor vibration as many have speculated. It is simply the effect of a low pass filter without an abrupt corner frequency that starts to reduce motion blur at an approximate shutter speed of 1/80 sec. To clarify this more measurements were made at a shutter speed of 1/80 sec. See illustration 4."

I pulled out part of his quote:

"sensor correction circuitry has a low pass filter".

There is no such thing!

...and you're now telling me there's no such thing?

You do know the term "low pass filter" can refer to things other than the anti-aliasing filter of a dslr? It's a general term: Low-pass filter - Wikipedia, the free encyclopedia. The SR system is measuring movement, he's speculating there's a low pass filter to cut out high frequency vibrations. This has nothing to do with whether or not there's an optical low pass filter in front of the sensor.

[Oldbayrunner]

Originally posted by BrianR Quote "sensor correction circuitry has a low pass filter".

I believe your misreading his inference and trying to infer he is stating there is a "probable" low pass filter for the shake reduction which there isn't. The only low pass sensor correction filters are called, for one, anti-alias filters. There are two filters one for horizontal and the other for vertical. Which are attributed to image blurring. Which is the way I read into it.

• dichroic mirror - reflects infrared light
• infrared absorbtion glass - absorbs infrared light
• low pass filter 1 - seperates subject horizontally
• phase plate - converts linearly polarized light into circularly polarized light
• low pass filter 2 - seperates subject data vertically

You do know the term "low pass filter" can refer to things other than the anti-aliasing filter of a dslr?

Yep, so what, I posted one from wikipedia that pertains to Optical data of a digital camera for which we are discussing.

The SR system is measuring movement, he's speculating there's a low pass filter to cut out high frequency vibrations. This has nothing to do with whether or not there's an optical low pass filter in front of the sensor.

No your reading into it that manner. I don't read it he's speculating that because I know based on everything I've read there is no high frequency vibration cutting low pass filter in relation to the Pentax SR system , the only sensor correcting low pass filters are the anti-aliasing so that is a more likely the ones he is talking about that cause some indication of blurr.

However...READ the Patent
Pentax Shake Reduction Patent 20080226276

Please, show me where it mentions any Low Pass filters in relation to cutting vibrations. High Pass, you bet but no Low Pass.

[Gimbal]

The various patents imply that sensor movement only occurs after mirror up (except when in live view).

Alright, the sensor is lifted AFTER the mirror is up, but before the shutter opens.
But what do you mean by "except when in live view"? The mirror has to be lifted to enable live view.

---------- Post added 12-31-2014 at 11:42 AM ----------

I believe your misreading his inference and trying to infer he is stating there is a "probable" low pass filter for the shake reduction which there isn't.

There has to be a low pass filter between the gyro sensors and the A/D converters, always. Unless the gyro sensor is limited in its output frequency by default to a frequency below half the sampling rate of the A/D converter. It's all about the Nyquist criterion. Also it makes sense to filter out frequencies that are way over the capabilities of the SR anyway. So even if it isn't mentioned in the patent, I bet there is a low pass filter there.

Edit:
They could of course figure that the camera body in itself functions as a low pass filter as it just takes to much energy to shake the camera in any higher frequencies. But I doubt it.

[Oldbayrunner]

There has to be a low pass filter between the gyro sensors and the A/D converters, always. Unless the gyro sensor is limited in its output frequency by default to a frequency below half the sampling rate of the A/D converter. It's all about the Nyquist criterion. Also it makes sense to filter out frequencies that are way over the capabilities of the SR anyway. So even if it isn't mentioned in the patent, I bet there is a low pass filter there.

[0048]The first high-pass filter circuit 27a reduces the low-frequency component of the signal output from the first angular velocity sensor 26a, because the low-frequency component of the signal output from the first angular velocity sensor 26a includes signal elements that are based on null voltage and panning motion, neither of which are related to hand-shake.

[0049]Similarly, the second high-pass filter circuit 27b reduces the low-frequency component of the signal output from the second angular velocity sensor 26b, because the low-frequency component of the signal output from the second angular velocity sensor 26b includes signal elements that are based on null voltage and panning motion, neither of which are related to hand-shake.

[0050]Likewise, the third high-pass filter circuit 27c reduces the low-frequency component of thesignal output from the third angular velocity sensor 26c, because the low-frequency component of the signal output from the third angular velocity sensor 26c includes signal elements that are based on null voltage and panning motion, neither of which are related to hand-shake.


[0054]The reduction of the low-frequency component is a two-step process. The primary part of the analog high-pass filter processing operation is performed first by the first, second, and third high- pass filter circuits 27a, 27b, and 27c, followed by the secondary part of the digital high-pass filter processing operation that is performed by the CPU 21.


[0055]The cut-off frequency of the secondary part of the digital high-pass filter processing operation is higher than that of the primary part of the analog high-pass filter processing operation.

[0056]In the digital high-pass filter processing operation, the value of time constants (a first high- pass filter time constant hx, a second high-pass filter time constant hy, and a third high-pass filter time constant hθ) can be easily changed

Instead of disputing me, you might try actually reading the patent information I've linked to, you may just gain some insight into how the SR system really works instead of relying on presumption..

Also read; [0063] , [0064] , [0065] & [0184].

Again there is NO low pass filter associated with the SR system according to the patent. Come up with written documented proof in Pentax's SR system there is and I will stand corrected.

[BrianR]

I believe your misreading his inference and trying to infer he is stating there is a "probable" low pass filter for the shake reduction which there isn't.

He's 'labnut' on dpreview,

A study of the effectiveness of K7 shake reduction: Pentax SLR Talk Forum: Digital Photography Review

it seems pretty clear to me he's not talking about the AA-filter. The patent is beside the point (though still interesting), every instance of his use 'low pass filter' is qualified by him stating he's inferring it's existance, and I'd give him credit enough to know the k-7 has an AA-filter.

[Gimbal]

Again there is NO low pass filter associated with the SR system according to the patent. Come up with written documented proof in Pentax's SR system there is and I will stand corrected.

They don't mention anything about a battery either so... But perhaps that is because that isn't necessary in order to describe the SR function, just as every filter or capacitor actually used ins't mentioned.
And according to this https://www.google.se/patents/US20080013939 patent there are no filters at all except the optical ones.