M11 rumors : Leica in bed with Sony ?

[rramesh]

Advertisement (gone after registration)

I am happy to have a camera that capture an image the way my eye sees it. With color correction and noise reduction in PP, I think the M9-P or the M10-P is sufficient for my needs. Don't understand this desire for more pixels, unless we want to elicit detail, more than the eye can see in the first look.

 

[Advertisement]

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

[SrMi]

40 minutes ago, rramesh said:

I am happy to have a camera that capture an image the way my eye sees it. With color correction and noise reduction in PP, I think the M9-P or the M10-P is sufficient for my needs. Don't understand this desire for more pixels, unless we want to elicit detail, more than the eye can see in the first look.

 

Higher resolutions have one disadvantage: the file size.

Higher resolutions have several advantages (e.g., M10-R vs. M10-P):

- less false colors (moire)

- larger prints

- more cropping potential

- can get more out of a good lens (according to Sean Reid's tests)

[jaapv]

Further disadvantages:  a tendency for motion blur. The better definition of the plane of focus will show up focusing errors and uneven planes. of focus.  

[Ko.Fe.]

Quote

Leica in bed with Sony ?

Who isn't.

[SrMi]

21 minutes ago, jaapv said:

Further disadvantages:  a tendency for motion blur. The better definition of the plane of focus will show up focusing errors and uneven planes. of focus.  

Only if looking at outputs of different sizes. Same as with noise.

[jaapv]

Somewhat true* -but if you look at the output at the same (smaller) size, the main difference  will be somewhat better acuity.

(*). The motion blur, if small enough to fall within the diameter of a pixel on the lower resolving (hence larger pixel size).sensor, will not be recorded like it would be on the small pixels of the high-res sensor.

[Stevejack]

Advertisement (gone after registration)

For those worried about camera shake, here is a very quick and dirty example to show how easy it is to clean up even very blurry images. This is a panning shot taken at 1/30 second, with the rider going extremely fast. If I was ever going to use this image, I would spend some time actually finessing the result but what you see here is a one-click process.

The image on the left is straight out of camera, the image on the right has been run through Topaz sharpen on the motion blur setting. The forum will compress the hell out of the image but hopefully you can make out the difference. This is about as extreme as it will ever get for most people so any minor camera shake, if you do experience it, is truly a non-issue.

The key here is having enough pixels in your image for the sharpening software to work with. This file is from a 50mp camera, heavily cropped to show just the rider's head.

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 December 30, 2021 by Stevejack

[SrMi]

37 minutes ago, jaapv said:

Somewhat true* -but if you look at the output at the same (smaller) size, the main difference  will be somewhat better acuity.

(*). The motion blur, if small enough to fall within the diameter of a pixel on the lower resolving (hence larger pixel size).sensor, will not be recorded like it would be on the small pixels of the high-res sensor.

That makes sense. I wonder if it makes any difference in practice.

Regardless, I try to get the same acuity at 100%, so I need to shoot a 30% faster shutter speed with M10-R than with M10-P. So your point in post #84 is valid.

For a 60MP sensor, another 21% will be needed, or when compared to M10-P, 60% faster shutter speed is needed (e.g., 1/60sec -> 1/100sec). That is a difference of less than a stop.

[Chaemono]

One reason I decided to keep the M10-P and send back the M10-R is that I often couldn’t get sharp images (viewed at 100%) at the same low shutter speed as with the P.  That meant higher ISO and the higher ISO R files since non-BSI sensor are not that “push-able.” Actually, they kind of suck. For my pixel peeping habit (can’t help it), anything above 40MP FF either needs IBIS or a BSI sensor for high ISO “push-ability” if it’s a color camera.

[hdmesa]

6 hours ago, Stuart Richardson said:

...

When I talk about a sensor "outresolving" a lens, I mean that there is no aperture at which the photo will appear sharp at 100% with that lens and camera combination. So you may gain extra resolution in the center, but the edges will be visibly soft...

 

^ You may be surprised as many were with the GFX 100/100S when several of the GF lenses actually tested better in the corners than they did on the 50S/R. The M11 BSI sensor technology should be better at capturing the steep angles of light from wide angle lenses. In addition to seeing improved corner performance, you may also see less vignetting with some lenses. This is of course assuming a very high-IQ lens to begin with. Vintage lenses on the GFX 100/100S have a strange dichotomy – infinity stopped down can look better than it did on the 50S/R, but wide open shooting really, really suffers if you insist on reviewing images at 1:1 or 2:1 magnification. All those "character lenses" when shot wide open look pretty bad on 100mp at high magnification.

 

3 hours ago, jaapv said:

Further disadvantages:  a tendency for motion blur. The better definition of the plane of focus will show up focusing errors and uneven planes. of focus.  

The motion blur thing is easily resolved with a higher shutter speed or by limiting yourself to viewing your images at 50% instead of 100% or 200%.

A more concerning (to some) issue with viewing images from ultra-high-res sensors at 1:1 or 2:1 is the reduced depth of field you see at those magnifications. On the GFX 100/100S, you have to accept having perfect focus in only one area of the image, or you learn to focus stack.

The biggest issue with going over 50mp is not having IBIS (nor having OIS on the M lenses of course). But I think Leica's tactic here will be the rumored resolution switching of the M11, which will mimic using a lower resolution sensor.

 

Edited December 30, 2021 by hdmesa

[Stevejack]

3 hours ago, Chaemono said:

One reason I decided to keep the M10-P and send back the M10-R is that I often couldn’t get sharp images (viewed at 100%) at the same low shutter speed as with the P.  That meant higher ISO and the higher ISO R files since non-BSI sensor are not that “push-able.” Actually, they kind of suck. For my pixel peeping habit (can’t help it), anything above 40MP FF either needs IBIS or a BSI sensor for high ISO “push-ability” if it’s a color camera.

I haven't had too much trouble with higher ISO myself, and I'm comparing it to my Sony A1 which I often shoot at very high ISO. It's very subjective though and everyone's needs are different.

Another quick and dirty example from the M10R is this one, here at close to 100% crop and this is the file straight out of camera, all noise reduction sliders zeroed out in Lightroom. ISO12,500. 1/90sec f/2.8. Pretty noisy when you pixel peep like this.

 

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 December 30, 2021 by Stevejack

[Stevejack]

Here is the same image with a quick pass through noise reduction. Far from perfect, but if I really needed to I'm sure I could get a decent result out of it. I doubt I would ever crop this heavily on a high ISO image but if I needed to, I would make it work.

 

 

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!

[Chaemono]

vor 6 Minuten schrieb Stevejack:

I haven't had too much trouble with higher ISO myself, and I'm comparing it to my Sony A1 which I often shoot at very high ISO. It's very subjective though and everyone's needs are different.

Another quick and dirty example from the M10R is this one, here at close to 100% crop and this is the file straight out of camera, all noise reduction sliders zeroed out in Lightroom. ISO12,800. 1/90sec f/2.8. Pretty noisy when you pixel peep like this.

 

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!

I didn’t mind the high ISO noise of the M10-R. It’s the “push-ability” of high ISO files that was awful. Try a high contrast scene with the M10-R at ISO 1600, push exposure by two stops and lift shadows to 50. If I wanted to puke, I’d push shadows to 100.

[Stevejack]

6 minutes ago, Chaemono said:

I didn’t mind the high ISO noise of the M10-R. It’s the “push-ability” of high ISO files that was awful. Try a high contrast scene with the M10-R at ISO 1600, push exposure by two stops and lift shadows to 50. If I wanted to puke, I’d push shadows to 100.

Gotchya - yeah I'm probably not pushing my M10R files quite that much in post so haven't really run across it 👍

[Chaemono]

The M10-R is for landscape and architecture photography and for those who print IMO. The camera is complementary to the M10-P. The M11 will likely replace both.

I’m not planning to get an M11 right away. I will wait for the P version and then compare it to the M10-P which I absolutely love up to ISO 640. Above that the Sony FF BSI sensors start to pull ahead. If one has used files from Sony BSI sensors, one can already anticipate where some of the IQ advantages of the M11 over the M10 will lie. I love the noise of the M10-P at ISO 640 in certain pictures and the control of highlights from ISO 200-640. That’s where I will do the comparisons with the M11-P eventually.

[AlanD]

17 hours ago, hdmesa said:

The GFX cover glass is not only too thick for M-sensor-like performance, but it's spaced 9mm from the sensor. This spacing helps defocus dust, but it's hell on M glass. Most M lenses will start to lose sharpness before they even reach the edge of a 35mm crop on the GFX. An additional, if not larger problem, is every single M to GFX adapter does not give perfect hard stop at infinity.

It’s interesting since I find that even my TriElmar 28-35-50 is pretty sharp even in the corners (in 60MP mode).  I have the Contax G45 modified to M mount which is as sharp as a GF45 when both are at f/2.8.  The Nikon-S 50/1.4 from the Y2K special edition is also capable of single pixel eyelashes at f/2.8 on the GFX100.  Both of those will cover the 33x44 which just a bit of hard vignetting at the extreme corners but have sharp foliage to the edge.

IBIS adds a level of added sharpness in my opinion and the 60MP/16-bit pixel density adds a lot of tonality when downsampled.

The thick coverglass adds some negative focus which is why the hard stop at infinity might not really work but the GFX100 BSI sensor seems to be better than anything other than the SL line.

I use a 7artisans M to GFX adapter.  I have found Rayqual adapters to be good at preserving infinity but they don’t have a Leica M to GFX adapter as far as I know.

[Chaemono]

Am 28.12.2021 um 22:03 schrieb hdmesa:

What I see is the same background but the SL2 has a sharper subject, which increases the perceived separation. A high-res BSI sensor should produce the same separation as you're seeing here with the SL2 but with the benefits of the BSI tech.

Sharpness is part of it but not only. I see a clear difference in the OOF areas. BSI sensors with Leica lenses just give more detailed creaminess in the OOF areas IMO. It means, If one gets an M11, one will be compelled to get at least a Summilux to go along with it.😂  In certain situations (depending on distance of the subject from the camera and the background), f/2 will just not cut it in terms of separation. Here the Summilux-SL and the Summicron-SL both on BSI sensors. Roll-off of Summicron-SL is like an f/1.4 but blur is like an f/2. Sharpness of Summilux-SL wide open is like a Summicron. This lens has this warmth and coolness at the same time with the Adobe Standard profile.

Less compressed JPEGs here: https://chaemono.smugmug.com/Folder-2/50-Summilux-SL-vs-APO-50-Summicron-SL-on-BSI-sensor/

50 Summilux-SL on SL2S

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!

 

APO 50 Summicron-SL on S5

 

 

[hdmesa]

2 hours ago, Chaemono said:

Sharpness is part of it but not only. I see a clear difference in the OOF areas. BSI sensors with Leica lenses just give more detailed creaminess in the OOF areas IMO. It means, If one gets an M11, one will be compelled to get at least a Summilux to go along with it.😂  In certain situations (depending on distance of the subject from the camera and the background), f/2 will just not cut it in terms of separation. Here the Summilux-SL and the Summicron-SL both on BSI sensors. Roll-off of Summicron-SL is like an f/1.4 but blur is like an f/2. Sharpness of Summilux-SL wide open is like a Summicron. This lens has this warmth and coolness at the same time with the Adobe Standard profile.

Less compressed JPEGs here: https://chaemono.smugmug.com/Folder-2/50-Summilux-SL-vs-APO-50-Summicron-SL-on-BSI-sensor/

I can see these kinds of minimal changes to bokeh/OOF areas simply by reducing contrast. Since BSI can deliver a wider dynamic range and can reduce vignetting in wide angle lenses, I guess it makes sense it could have some effect on the default image rendering depending on how the profile in C1/LR is set up.

Edited December 31, 2021 by hdmesa

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[RayD28]

On 12/27/2021 at 8:18 PM, stefanusj said:

the race for sensor right now is: high iso/noise controlled and high megapixels. I just dont care about both. if the object is too dark then it's a sign NOT to take a picture. 

Good point but some of us (me) like to use a higher ISO in good light to get a faster shutter speed to avoid motion blur or camera shake.  

 

On 12/28/2021 at 4:01 PM, SrMi said:

Improved light gathering at these sensor sizes is not a benefit for BSI sensors. M43 cameras have smaller pixels than 60MP FF cameras, no BSI sensors and work fine within the constraint of the sensor size.

BSI's advantages are better handling of steep incident angles and faster readouts. The latest Sony sensors also support dual conversion gain. 

Does an M10-R make sense to you? For example, a 60MP M11 needs a 20% faster shutter speed than M10-R to avoid camera shake.

 

 

I have a Nikon D800 with 36MPs.  I pixel peeped and deleted many good shots.  If beard stubble was blurry at 100% crop I deleted it.  It wasn't the MPs that made the shot "bad", it was my belief that a perfect capture can only be determined when viewing at 100%.  

 

On 12/29/2021 at 8:09 PM, jaapv said:

Further disadvantages:  a tendency for motion blur. The better definition of the plane of focus will show up focusing errors and uneven planes. of focus.  

Hence the need for higher ISO performance so faster shutter speed can compensate.  I recall a discussion in another thread about taking photos at a dimly lit school play.  A fast lens can get the exposure but the slow shutter speed can reduce the quality of the shot if actors are moving quickly on the stage.  

On 12/29/2021 at 9:23 PM, Stevejack said:

For those worried about camera shake, here is a very quick and dirty example to show how easy it is to clean up even very blurry images. This is a panning shot taken at 1/30 second, with the rider going extremely fast. If I was ever going to use this image, I would spend some time actually finessing the result but what you see here is a one-click process.

The image on the left is straight out of camera, the image on the right has been run through Topaz sharpen on the motion blur setting. The forum will compress the hell out of the image but hopefully you can make out the difference. This is about as extreme as it will ever get for most people so any minor camera shake, if you do experience it, is truly a non-issue.

The key here is having enough pixels in your image for the sharpening software to work with. This file is from a 50mp camera, heavily cropped to show just the rider's head.

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!

Thanks for posting.  I'll have to try the software.  I may even already have it.  

Edited December 31, 2021 by RayD28

[jaapv]

On 12/30/2021 at 4:23 AM, Stevejack said:

For those worried about camera shake, here is a very quick and dirty example to show how easy it is to clean up even very blurry images. This is a panning shot taken at 1/30 second, with the rider going extremely fast. If I was ever going to use this image, I would spend some time actually finessing the result but what you see here is a one-click process.

The image on the left is straight out of camera, the image on the right has been run through Topaz sharpen on the motion blur setting. The forum will compress the hell out of the image but hopefully you can make out the difference. This is about as extreme as it will ever get for most people so any minor camera shake, if you do experience it, is truly a non-issue.

The key here is having enough pixels in your image for the sharpening software to work with. This file is from a 50mp camera, heavily cropped to show just the rider's head.

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!

What compression? You are allowed 2480 pixels wide….