GR III and its 3 axis IBIS

thelps · Oct 3, 2018

With all the banter about things missing from the GR III one thing that is included is the 3 axis IBIS.

With the IBIS inclusion I'd like to see a Hi-Res mode if its possible. I know it has its limitations, I have see results from the EM1.2 work and fail. With Ricoh having access to the Pentax developed K1 H-Res mode they could maybe be able to include this, even if it does not get initial firmware release.

I could not see it mentioned in the datasheet nor has it been asked in the videos of reviewers talking the Ricoh execs.

Even with all the caveats and limitations there is just something about having a file and resolution maybe somewhere in the 50 Mpixel range ability in such a small package even if its not used much.

TN Args · Oct 3, 2018

thelps said:
With all the banter about things missing from the GR III one thing that is included is the 3 axis IBIS.

With the IBIS inclusion I'd like to see a Hi-Res mode if its possible. I know it has its limitations, I have see results from the EM1.2 work and fail. With Ricoh having access to the Pentax developed K1 H-Res mode they could maybe be able to include this, even if it does not get initial firmware release.

I could not see it mentioned in the datasheet nor has it been asked in the videos of reviewers talking the Ricoh execs.

Even with all the caveats and limitations there is just something about having a file and resolution maybe somewhere in the 50 Mpixel range ability in such a small package even if its not used much.

3-axis won't do pixel-shift. That requires the second-generation, 5-axis SRII system.

As an aside, I'm not sure where the number of axes was set at 3. It isn't in the Ricoh media release. I'm assuming it was picked up by DPR at a presentation or face-to-face discussion with Ricoh.

So, bad news I guess.

The good news is that I really like the switchable simulated OLPF that will be included in the GR III, enabled by IBIS.

thelps · Oct 3, 2018

TN Args said:
thelps said:

With all the banter about things missing from the GR III one thing that is included is the 3 axis IBIS.

With the IBIS inclusion I'd like to see a Hi-Res mode if its possible. I know it has its limitations, I have see results from the EM1.2 work and fail. With Ricoh having access to the Pentax developed K1 H-Res mode they could maybe be able to include this, even if it does not get initial firmware release.

I could not see it mentioned in the datasheet nor has it been asked in the videos of reviewers talking the Ricoh execs.

Even with all the caveats and limitations there is just something about having a file and resolution maybe somewhere in the 50 Mpixel range ability in such a small package even if its not used much.

Click to expand...

3-axis won't do pixel-shift. That requires the second-generation, 5-axis SRII system.

As an aside, I'm not sure where the number of axes was set at 3. It isn't in the Ricoh media release. I'm assuming it was picked up by DPR at a presentation or face-to-face discussion with Ricoh.

So, bad news I guess.

The good news is that I really like the switchable simulated OLPF that will be included in the GR III, enabled by IBIS.

Thank you.

Yes, I heard 3 axis, so if it needs 5 then perhaps this is why its not been mentioned.

The specs at Ricoh mention 3 axis - http://news.ricoh-imaging.co.jp/rim_info2/2018/20180925_026135.html?_ga=2.44911716.1306194464.1538624411-1053178372.153811872 6

Image StabilizationSensor-shift shake reduction (SR)(3-axis)

TN Args · Oct 3, 2018

thelps said:
TN Args said:

thelps said:

With all the banter about things missing from the GR III one thing that is included is the 3 axis IBIS.

With the IBIS inclusion I'd like to see a Hi-Res mode if its possible. I know it has its limitations, I have see results from the EM1.2 work and fail. With Ricoh having access to the Pentax developed K1 H-Res mode they could maybe be able to include this, even if it does not get initial firmware release.

I could not see it mentioned in the datasheet nor has it been asked in the videos of reviewers talking the Ricoh execs.

Even with all the caveats and limitations there is just something about having a file and resolution maybe somewhere in the 50 Mpixel range ability in such a small package even if its not used much.

Click to expand...

3-axis won't do pixel-shift. That requires the second-generation, 5-axis SRII system.

As an aside, I'm not sure where the number of axes was set at 3. It isn't in the Ricoh media release. I'm assuming it was picked up by DPR at a presentation or face-to-face discussion with Ricoh.

So, bad news I guess.

The good news is that I really like the switchable simulated OLPF that will be included in the GR III, enabled by IBIS.

Click to expand...

Thank you.

Yes, I heard 3 axis, so if it needs 5 then perhaps this is why its not been mentioned.

The specs at Ricoh mention 3 axis - http://news.ricoh-imaging.co.jp/rim_info2/2018/20180925_026135.html?_ga=2.44911716.1306194464.1538624411-1053178372.153811872 6

Image StabilizationSensor-shift shake reduction (SR)(3-axis)

Ah, thanks for the link. It looks like DPR didn't copy the bits in brackets onto the table they published with the announcement, link. I thought they copied the full media release including full Ricoh table, but no.

teemodk · Oct 3, 2018

TN Args said:
thelps said:

With all the banter about things missing from the GR III one thing that is included is the 3 axis IBIS.

With the IBIS inclusion I'd like to see a Hi-Res mode if its possible. I know it has its limitations, I have see results from the EM1.2 work and fail. With Ricoh having access to the Pentax developed K1 H-Res mode they could maybe be able to include this, even if it does not get initial firmware release.

I could not see it mentioned in the datasheet nor has it been asked in the videos of reviewers talking the Ricoh execs.

Even with all the caveats and limitations there is just something about having a file and resolution maybe somewhere in the 50 Mpixel range ability in such a small package even if its not used much.

Click to expand...

3-axis won't do pixel-shift. That requires the second-generation, 5-axis SRII system.

What is the reason behind this? You should think that 3 axes would be enough?

TN Args said:
As an aside, I'm not sure where the number of axes was set at 3. It isn't in the Ricoh media release. I'm assuming it was picked up by DPR at a presentation or face-to-face discussion with Ricoh.

So, bad news I guess.

The good news is that I really like the switchable simulated OLPF that will be included in the GR III, enabled by IBIS.

chankosuto · Oct 3, 2018

Would have been a nice feature for a landscape backpacking camera.

TN Args · Oct 3, 2018

teemodk said:
TN Args said:

thelps said:

With all the banter about things missing from the GR III one thing that is included is the 3 axis IBIS.

With the IBIS inclusion I'd like to see a Hi-Res mode if its possible. I know it has its limitations, I have see results from the EM1.2 work and fail. With Ricoh having access to the Pentax developed K1 H-Res mode they could maybe be able to include this, even if it does not get initial firmware release.

I could not see it mentioned in the datasheet nor has it been asked in the videos of reviewers talking the Ricoh execs.

Even with all the caveats and limitations there is just something about having a file and resolution maybe somewhere in the 50 Mpixel range ability in such a small package even if its not used much.

Click to expand...

3-axis won't do pixel-shift. That requires the second-generation, 5-axis SRII system.

Click to expand...

What is the reason behind this? You should think that 3 axes would be enough?

The extra 2 axes are horizontal and vertical shift, which is how pixel shift works.

Samuel Dilworth · Oct 3, 2018

Well, the 3-axis systems shift the sensor vertically and horizontally to counteract pitching and yawing of the camera body during the exposure. That’s enough to pixel-shift the sensor to simulate a higher-res sensor over multiple exposures.

The sensor can also rotate in plane to compensate for the camera rolling around its lens axis or motion that appears identical to that at infinity focus (such as the right side of the camera going down when the user jabs the shutter button).

The 5-axis systems make exactly the same sensor movements, but they additionally measure the vertical and horizontal shift of the camera and add a compensating response of the sensor to its other movements.

The GR III has no need for shift compensation because its wide lens means camera shift contributes almost nothing to blur, even at quite short focus distances. That is probably why Ricoh didn’t give it the 5-axis system.

It may have the pixel-shift multi-capture method anyway, although the interest in it is probably very small (nil from me, for instance).

TN Args · Oct 4, 2018

chankosuto said:
Would have been a nice feature for a landscape backpacking camera.

Just fire off a burst of 4 or 5 handheld shots, then use Photoshop or equivalent and the "super-resolution" technique. Actually gets better results. And doesn't need a tripod! Win-win. Who needs pixel-shift. (and I'm not biased, as my gear list shows, I have a pixel-shift camera)

chankosuto · Oct 4, 2018

TN Args said:
chankosuto said:

Would have been a nice feature for a landscape backpacking camera.

Click to expand...

Just fire off a burst of 4 or 5 handheld shots, then use Photoshop or equivalent and the "super-resolution" technique. Actually gets better results. And doesn't need a tripod! Win-win. Who needs pixel-shift. (and I'm not biased, as my gear list shows, I have a pixel-shift camera)

I've tried that with a telelens, with a wide angle i don't know how effective would be, been tempted to get a dp Merrill 2 for this sort of thing.

Barry Pearson · Oct 8, 2018

The GR III is claimed to support moiré reduction using the IBIS/SR unit.

If it can do that, doesn't that mean its axes of sensor movement support horizontal and vertical movement?

Isn't that what is needed for pixel-shift?

(I'm asking, not telling!)

teemodk · Oct 8, 2018

Barry Pearson said:
The GR III is claimed to support moiré reduction using the IBIS/SR unit.

If it can do that, doesn't that mean its axes of sensor movement support horizontal and vertical movement?

Isn't that what is needed for pixel-shift?

(I'm asking, not telling!)

If that is correct, then there is a very good reason for delaying the release. I guess this requires lots og fine tuning on both hardware and firmware. That would be a very nice little surprise. Especially because it is on my guessing list lol.

Vitriol · Oct 8, 2018

That's the Shannon rule: aliasing is due to sampling any sort of data of (here photo details) which are of higher frequency than half the sampling frequency (small details which are smaller than the size of one sensor pixel).

So you want to blur these finer details so that they are transformed into a circle of confusion of 1 pixel wide. Certainly that the 3-axis is enough (pitch and yaw together).

And if the blur is just enough (remember:1 pixel wide), you don't lose the useful details (of 1 pixel or more).

preferredfault · Oct 8, 2018

Pixel shift is possible, but it's a software feature in the end. Whether or not they choose to add it hasn't been shown yet I don't think. I'd hope they would add it though.

Barry Pearson · Oct 9, 2018

Vitriol said:
That's the Shannon rule: aliasing is due to sampling any sort of data of (here photo details) which are of higher frequency than half the sampling frequency (small details which are smaller than the size of one sensor pixel).

So you want to blur these finer details so that they are transformed into a circle of confusion of 1 pixel wide.

Indeed. I knew that.

For interest, in February 2013 I bought the Pentax K-5 IIs, which didn't have an anti-aliasing filter (OLPF). There was lots of advice at the time not to buy this version of the K-5-II-series because of the risk of moiré. I chose to ignore that advice in order to get maximum image quality.

Whether or not I was wise to do so is still a matter of debate. But I only detected noticeable moiré in one image! (I still have over 31,000 images taken with that camera in my Lightroom catalogue). Moiré was never the problem in that camera that people originally claimed it would be.

Vitriol said:
Certainly that the 3-axis is enough (pitch and yaw together).

OK, that answers part of my question. Thanks.

Now, what else can be done with a sensor that has pitch and yaw?
(But not simple horizontal and vertical).

Vitriol said:
And if the blur is just enough (remember:1 pixel wide), you don't lose the useful details (of 1 pixel or more).

I think the use of sensor-movement to allow user control of anti-aliasing is a good one. Rightly or wrongly, I'll use "OFF" by default.

Vitriol · Oct 9, 2018

The key point is how often will you take pictures of a material or textile with apparent patterns of about the size of a pixel, and how big must be the area on the picture to be annoying.

I guess nowadays most manufacturers are removing the AA filter because moire circumstances are very close to none.

IMHO the AA-filter simulation should not be advertised as a main feature. And it comes with some limitations : does not work with a flash or at high shutter speed because of the sensor vibration not having enough effect on the picture.

Finally, I am not sure how well AA simulation is compatible with stabilisation on.

Sabud · Oct 9, 2018

Can´t find any information about how many stops you can gain with this 3-axis IBIS.

With the latest 5-axis IBIS (Olympus-Panasonic) you will have 5 stops.

Any ideas?

Barry Pearson · Oct 9, 2018

Vitriol said:
The key point is how often will you take pictures of a material or textile with apparent patterns of about the size of a pixel, and how big must be the area on the picture to be annoying.

Yes, that's important

I was doing a lot of studio work where I could have expected problems from fabrics. The photo below is the example I particularity noticed. (In the grey material in her "chest area"). When I pixel-peep, I think there is a still a trace there.

I'm not ruling out other cases, but I didn't notice them. Or ... sometimes what I notice when I pixel-peep in post-processing "disappears" when I print, or downsize for the web.

Vitriol said:
I guess nowadays most manufacturers are removing the AA filter because moire circumstances are very close to none.

IMHO the AA-filter simulation should not be advertised as a main feature.

Perhaps it is best thought of as as a specialist tool, or "for emergency use only"?

(I don't think I've ever used it. But I might experiment when I get my GR III).

Vitriol said:
And it comes with some limitations : does not work with a flash or at high shutter speed because of the sensor vibration not having enough effect on the picture.

Indeed. Certainly normal flash is ruled out, and in the case of the K-1-series, a shutter speed faster than 1/1000th second isn't long enough.

But ... I use Lightroom Classic CC. Its Graduated Filter and Radial Filter have a "moiré reduction" option.

Vitriol said:
Finally, I am not sure how well AA simulation is compatible with stabilisation on.

As far as I know, it is compatible. I can't find any hint to the contrary in operating manuals. But I haven't tried both together, so I can't say from personal experience.

--

http://www.barrypearson.co.uk/

User:Barry Pearson - Wikipedia

en.wikipedia.org

Barry Pearson · Oct 9, 2018

Sabud said:
Can´t find any information about how many stops you can gain with this 3-axis IBIS.

With the latest 5-axis IBIS (Olympus-Panasonic) you will have 5 stops.

Any ideas?

These are the best descriptions I know of for the GR III, and neither answers your question:

http://news.ricoh-imaging.co.jp/rim_info2/2018/20180925_026135.html

https://shop-uk.ricoh-imaging.eu/ricoh-gr-iii.html

I would welcome more comprehensive descriptions. (As I'm sure you would too!)

The Pentax K-1 (5-axis) has about 5-stops:

http://www.ricoh-imaging.co.jp/english/products/k-1/feature/02.html

--
http://www.barrypearson.co.uk/
http://en.wikipedia.org/wiki/User:Barry_Pearson

Samuel Dilworth · Oct 9, 2018

Sabud said:
Can´t find any information about how many stops you can gain with this 3-axis IBIS.

With the latest 5-axis IBIS (Olympus-Panasonic) you will have 5 stops.

Any ideas?

At infinity, 3-axis and 5-axis systems are identically effective.

At closer distances, the 5-axis systems will pull out a small lead until at 1:1 magnification the benefit will be significant. However, at typical shooting distances with a wide-angle Ricoh GR, the benefit of 5-axis over 3-axis is practically nil.

So it will come down to the technology used and not the number of axes of compensation.

I know this doesn’t answer your question!

GR III and its 3 axis IBIS

Leading Member

Forum Pro

Leading Member

Forum Pro

Senior Member

Well-known member

Forum Pro

Senior Member

Forum Pro

Well-known member

Veteran Member

Senior Member

Active member

Forum Enthusiast

Veteran Member

Active member

Leading Member

Veteran Member

Veteran Member

Senior Member

Similar threads

Keyboard shortcuts