Jump to content

Time to face the sensor problem


lburn

Recommended Posts

2 minutes ago, Jipster said:

Well said. Reading some of these comments about the shutter sound makes me feel some of us should become musicians. Only professional violinists are that picky about being in tune (tongue in cheek).

I'm guilty of the shutter criticism. But I also don't think the colour digital M cameras offer anything in terms of image quality that outperforms any other modern full frame digital. I currently have three M cameras and another on the way, I obviously love them, but I'm not kidding myself here. It's not about the end result. 

So I'm not in it for the image quality, I'm in it for the experience of taking the photograph with a rangefinder camera and using the manual lenses. I enjoy the process of taking a photograph with the M.      Anything that takes away from how pleasant that experience is, is a step backwards IMO. 

Don't get me wrong I want the best image quality I can get out of the M system, hence why I'm trying out the M11... but I don't want any of the extra image quality to come at the expense of a pleasing user experience. How the camera 'feels' is completely subjective I know, so time will tell whether I keep it or not. The M10 is a perfectly good camera if the M11 feels wrong. 

  • Like 1
  • Thanks 1
Link to post
Share on other sites

3 hours ago, lburn said:

Presumably everyone who buys the M11 will want to use it at full resolution (60MP) at least some of the time.  And they will do so because they want the best possible detail in their image.  So they will not want the image downgraded by camera shake.  After all what is the point of a high resolution photo which isn’t sharp? So, given that Leica, contrary to every other high resolution 35mm camera manufacturer in the current market, has chosen to omit putting a stabilised sensor in the camera, how can this problem be avoided?  Well, you can downgrade the resolution as Leica has helpfully enabled you to do.  But then there is no high res image anymore.  You can raise the ISO until it is so high that the shutter speed is very fast. But then the image is downgraded. You can stop using fast or longer lenses.  But that is one of the strengths and great pleasures of the M camera. You can buy a sturdy tripod.  That will solve the problem. Use the M11 on a good tripod.  But then the whole idea of the M camera, fast and compact, rather goes out of the window. So is the M11 a flawed concept in its current form? Do all the owners of Sony, Nikon, Cannon etc high res 35 m cameras all go round with image stabilisation turned off?  No.  Why not? Because they need it to get consistently sharp high res images.  I suggest so does the M11.  Surprisingly, at least to me, two of leading reviews out so far, Jono Slack and Sean Reid, do not even mention this topic.  

Blah, blah, blah… 

That’s all I have to say. Obviously not a camera for you, so just stop waisting time. Seriously why anybody would take the time to write a bunch of nonsense like this is beyond me. 

Link to post
Share on other sites

5 minutes ago, Artin said:

I really do not understand the comment, how has Leica moved away from a rangefinder only camera ?  What is the definition of a rangefinder camera ? 
as far as I know every Leica M camera is a rangefinder camera. Your M262 has exactly the same capability of my Mp-240 both great cameras. The M10R is a more capable image making tool and the M11 is even better it’s just plain science.  Better ISO .. better dynamic range .. better rangefinder optically .. better metering.. faster buffer.  Better resolution.. evolution  just like any other brand. d1x . D2x  . d3 . D4. D5. D6. Z9. 

Sorry,

Just to clarify the M262 just like my Minolta x-700 has everything i need and nothing i do not need.

Any camera that has stuff i do not require like auto-focus.live view,EVF ,touch screen does not appeal to me.

Centre weighted metering is the best system for me ,i have plenty of resolution with my m262 and have incredibly detailed 24 by 16 prints on my wall although i have lots of beautiful softish glowy  images of my grand kids so lets not over rate the god of sharpness.

I just buy a camera that i like and when it dies i get a new one ,same with cars but that is just me.

I look at leica for a model i can get when my 2 M262 bodies die but they are drifting away from my idea of a good camera.

My cameras are minolta x-700 1983-2012 ,olympus ep-5 pen 2012-2015 ,M262 2015- now .

Did own a mamiya c330 TLR which actually was a medium format camera in the 90's.

Sometimes great dynamic range produces those silly images with the sun shining and perfect detail in the foreground like the over used ND grad filter so i am not bothered much about that.

We are all different you know.

Link to post
Share on other sites

4 hours ago, setuporg said:

It’s about time this myth of blurry high MP images is debunked.  Kind of M10-R usage in the wild simply does not support it.  Ask Steven who was swearing on his M10-P he’ll never succumb to the blur of R, then one day he sings Osanna to his M10-RBP, and no mention of the blur.

if we can do it at 40, we can do it at 60.  60 is the new 40!:)

Wrong. There is blur with M10R below 1/90 if you aren’t super mindful. 
 

my M11 arrives tomorrow but ive heard from numerous sources the same issue doesn’t exist. 

Link to post
Share on other sites

I like the M11 concept and will receive mine tomorrow so looking forward to figuring out if I like it. I’m keeping my M10R for now. 
 

I will point out criticism where it is due even thought I’m a loyal customer. The M is just as big if not bigger then a Sony body and they have IBIS so i don’t get all of the arguments for why Leica can’t due to in the body size. Simply a bad argument. 

Hopefully this is a non issue. It’s certainly an issue with the R, but I still live the R. First or Second best M digital ever made, TBD. 

  • Like 1
  • Thanks 1
Link to post
Share on other sites

Will one eventually resolve movement blurs on higher and higher megapixel cameras, that will not be visible on lower megapixel cameras?

Yes, of course. A single-point/pixel detail that moves during the exposure will eventually register across two (or more) pixels as the pixels get smaller (and the number gets higher).

A sharp point (•) will eventually become •• or ••• . Which is what movement blur looks like on pixels (regardless of whether it is shutter shock, camera shake, or subject movement.)

................

The real question is - when and how often does that happen?

It happens when the point's (•) movement is large enough to cross a border with a neighboring pixel, and register on two or more pixels ••. If the blur-trail fits into one pixel - it can't be detected. So what are the odds that a blur is large enough to do that - we need statistics.

Take a pixel from the 18-Mpixel M9.

6.8 microns across (9.6 microns diagonal). It will require a "point detail" to blur to a streak at least 6.9 microns long to guarantee that it will register onto two separate pixels (and 9.7 pixels long if it is on a diagonal orientation. / ) to show as blur.

Using "naive" statistics - a blur 3 microns long might only register on more than one pixel 50% of the time (someone with more time and more experience with statistics can correct that and the other "naive" assumptions below, for me ;) )

Now, take a sensor (same 24 x 36 total dimensions) - with pixels half the dimensions of the M9 pixel - 3.4 microns x 3.4 microns. And a diagonal dimension of 4.8 microns.

(BTW - that will be a 72-Mpixel sensor (twice as many pixels wide AND twice as many pixels tall - people always forget that part). The M11 is junior-league compared to that ;))

A 6.8-micron streak is virtually guaranteed to cross a pixel border and register as a blur ( •• or / ) The M11 will show it every time, the M9 almost every time (maybe 90%?)

A 3-micron blur will probably (again, naive statistics) cross a border maybe 75% of the time on the smaller pixel, but again only 40% of the time on the M9's larger pixels.

Additionally, the size of the blur streaks can be presumed to be proportional to the exposure time - 1/30th sec. will produce, on average, blurs twice as long as 1/60th sec.

And so will magnification (i.e. focal length) - thus the rule of thumb for shutter speed = 1/focal length or 1/2x focal length, etc.

Additionally (again), remember that there is no such thing as detail "inside" a pixel - the whole pixel only has one brightness value. If any part of a "streak" is only 2% of a pixel's area, it will only contribute 2% to the brightness of that pixel, raising or lowering, for example, a pixel's brightness of 100 to 102, or lowering it to 98 and a bit.

The image below illustrates the general principle - orange blur streaks are not large enough to significantly register on more than one pixel (i.e. undetectable) - black blur streaks (at least 30% of the blur is not on one pixel) probably will be recorded by more than one pixel. The image is not a random distribution of blur streaks (since a human made them), however. Not for statistical counting, just to demonstrate the idea.

So - yes it is probable the number of detectable blurs will increase with the number (and smallness) of pixels. But it is probably on the order of 2x the blur lengths - at some critical speed (say, 1 shutter step = 1/30th to 1/60th = 1 f/stop = double the ISO) based on focal length and photographer steadiness, shutter force (newtons) and subject movement.

And that is for 18 vs. 72 Mpixels.

24 or 40 compared to 60 Mpixels will produce a lot less than one shutter speed/ISO/aperture difference. ;)

In short, one could be worried about moving from 24 or 40 Mpixels to 240 or 400 Mpixels. Probably.

But moving from 40 to 60 Mpixels will need more like 1.2x the shutter speed/ISO/aperture. Moving from 24 to 60 will be 1.6x (ISO 200 to ISO 320; 1/125th > 1/200th sec.)

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 by adan
  • Like 9
  • Thanks 6
Link to post
Share on other sites

You said it better then me. Had I not noticed it at regular viewing size I would have never known. I’m not a zoom in and inspect the image kind of guy. But, If I expect the image to be fully sharper and in focus it’s readily apparent after hundreds of thousands of images taken with Leica cameras. There is no argument to be made here. The R below 1/90 or maybe even 1/125 requires extra attention in direct contrast to the M10P at 1/30. Easily seen at normal viewing. 

Edited by dkmoore
Link to post
Share on other sites

8 minutes ago, setuporg said:

I dunno I haven’t seen it.  Perhaps I’m mindful.  Perhaps I don’t look at it at huge magnification.  Anyhow I can see motion blur but not any specific high-res blur.

Quite honestly I don’t know what even causes it or what it is but it’s there when it’s not at 24 Mp, Leica versus Leica. I’ve taken enought images to know when it’s me versus something inherent with the camera. Once I knew it, it wasn’t super difficult to overcome.  I still really enjoy the M10R. 
 

Link to post
Share on other sites

3 hours ago, Stevejack said:

Just on this point, mine is off most of the time on my Sony A1. It is a great feature, but I'm usually either shooting above 1/4000sec or I'm shooting below 1/60 sec and panning with my subject. I don't really take many photos where the subject is still, or where I want my image to have no sense of movement at all. I also think a little shake makes for a great image in my book, but I understand that it's not the case for everyone.

I would argue that IBIS is particularly useful in an M camera just because a lot of people are shooting in that 1/15 to 1/125 range pretty frequently. I don't care that it's not on the M11, there are plenty of workarounds both in-camera and in post if I want an absolutely sharp image...  but I feel bad for those who really wanted it in the M11 - Leica should absolutely add it to the M as soon as they can. 

From the FAQ sent to Leica dealers (apparently):

An in-body image stabilization is a hardware function that moves the image sensor. Unfortunately, there is not enough space for this feature in the body of the M11.
 

  • Like 1
  • Thanks 1
Link to post
Share on other sites

BTW - I was in the business and tried out the Nikon D800E (36 Mpixels) when the shop got one in used 7-8 years ago.

And yes, it showed some little 2-4-pixel-long streaks hand-held at 1/250 with a 60mm Micro-Nikkor and a motionless subject (parking-lot gravel). Clear at 100% viewing.

But that was a classic SLR with a big mirror and and aperture stop-down lever thumping around inside. And it was a 3x jump in Mpixels from the D700 (12 Mpixels).

I tried a Sony A7r when one came in - it also had a rep for shake (again, an over-powerful shutter). I doubled my shutter speeds (1 stop) to 1/2x focal length for that, and did not note any blur streaks.

Leica's been specializing in "low-impact" shutters for 97 years now. With a few detours (e.g. the original M8 shutter went to 1/8000th - changed for the M8.2).

I kinda suspect they know what 60 Mp needs.... ;)

 

  • Like 3
Link to post
Share on other sites

4 hours ago, ynp said:

There is a Medium format digital world where the resolution above 60 mpx has been a common thing for many many  years. And before the GFX100 there were no IBIS. And working ISO was usually 100-320. Those cameras were not always being used on tripod in studio, a lot of people shot outside. Millions of pictures were shot without any stabilization. It’s insane that  people suddenly started to think that a 60 mpx Leica is unusable without stabilization. The design of the M body allows for it. It’s not a coincidence that Texan Leicas ( Fuji rangefinders) were so popular in the film era. 

I've shot with medium format digital since a Phase One P30 on a Hasselblad 501CM. The first thing I discovered was that the 30mp digital back picked up vibrations which were "buried" in the grain of film (color or B&W).  To make truly sharp images required a shutter speed of 3x focal length if hand held and using mirror lock up on a tripod...all the time.  I sold the P30 for a CFV50 which was even more sensitive. Currently I have a Leaf Aptus 75s (33mp) and Hasselblad 907x with CFV50c ii, I use all of them both in the studio and mostly in the field but vibration remains an issue, so yes, 60 mp in an M body with no IBIS will be a challenge in the real world.  Not sayin' it will be impossible to hand hold but know there will be A LOT of fuzzy images over the learning curve. Being mirrorless the M11 will be better than a HB500 series or the S(006) but it will take some practice.

 

Edited by Sailronin
  • Thanks 1
Link to post
Share on other sites

BTW - it should be remembered that in linear resolution (pickets in the fence) the M11 is only 21% sharper than the M10-R. Not all that big a leap.

9528 pixels/7864 pixels = 1.2115

...........................

For Sailronin: True, but a 500-series Hasselblad firing is like a collision between two semis. ;)

I never use anything but 1/500th hand-held, even on gritty Tri-X. And even with the Distagon 40 wide-angle (which I swapped for an SWC - no mirror slap).

  • Like 2
Link to post
Share on other sites

Just now, adan said:

BTW - it should be remembered that in linear resolution (pickets in the fence) the M11 is only 21% sharper than the M10-R. Not all that big a leap.

9528 pixels/7864 pixels = 1.2115

...........................

For Sailronin: True, but a 500-series Hasselblad firing is like a collision between two semis. ;)

I never use anything but 1/500th hand-held, even on gritty Tri-X. And even with the Distagon 40 wide-angle (which I swapped for an SWC - no mirror slap).

The S(006) is also very sensitive but again, there is a mirror.  The Hasselblad 907x (mirrorless with leaf shutter) is much better but even so to get optimum sharpness at least 1/2x focal length so I would imagine the M11 will be similar.  I agree 1/500 on the HB500 series, the SWC is much better, I've hand held down to 1/30th on occasion but pretty careful and try to use a monopod at least.

Link to post
Share on other sites

4 hours ago, Stevejack said:

I'm guilty of the shutter criticism. But I also don't think the colour digital M cameras offer anything in terms of image quality that outperforms any other modern full frame digital. I currently have three M cameras and another on the way, I obviously love them, but I'm not kidding myself here. It's not about the end result. 

So I'm not in it for the image quality, I'm in it for the experience of taking the photograph with a rangefinder camera and using the manual lenses. I enjoy the process of taking a photograph with the M.      Anything that takes away from how pleasant that experience is, is a step backwards IMO. 

Don't get me wrong I want the best image quality I can get out of the M system, hence why I'm trying out the M11... but I don't want any of the extra image quality to come at the expense of a pleasing user experience. How the camera 'feels' is completely subjective I know, so time will tell whether I keep it or not. The M10 is a perfectly good camera if the M11 feels wrong. 

Since my first Leica, an M8, I always felt there was something special in files produced by Leica camera and lenses. The reason why I am shooting Leica is the end result.

  • Like 1
  • Thanks 1
Link to post
Share on other sites

7 hours ago, lburn said:

Presumably everyone who buys the M11 will want to use it at full resolution (60MP) at least some of the time.  And they will do so because they want the best possible detail in their image.  So they will not want the image downgraded by camera shake.  After all what is the point of a high resolution photo which isn’t sharp? So, given that Leica, contrary to every other high resolution 35mm camera manufacturer in the current market, has chosen to omit putting a stabilised sensor in the camera, how can this problem be avoided?  Well, you can downgrade the resolution as Leica has helpfully enabled you to do.  But then there is no high res image anymore.  You can raise the ISO until it is so high that the shutter speed is very fast. But then the image is downgraded. You can stop using fast or longer lenses.  But that is one of the strengths and great pleasures of the M camera. You can buy a sturdy tripod.  That will solve the problem. Use the M11 on a good tripod.  But then the whole idea of the M camera, fast and compact, rather goes out of the window. So is the M11 a flawed concept in its current form? Do all the owners of Sony, Nikon, Cannon etc high res 35 m cameras all go round with image stabilisation turned off?  No.  Why not? Because they need it to get consistently sharp high res images.  I suggest so does the M11.  Surprisingly, at least to me, two of leading reviews out so far, Jono Slack and Sean Reid, do not even mention this topic.  

http://phaseoneiq4.com/yes-you-can-hand-hold-150-megapixels/

  • Thanks 1
Link to post
Share on other sites

3 hours ago, dkmoore said:

Wrong. There is blur with M10R below 1/90 if you aren’t super mindful.

That hasn't been my experience. I've been using the R the same way I used my previous M digital cameras, and film cameras before that. I typically find 1/15 with a 35mm lens to be my floor as long as I do my part.

There is some truth to blur with large MP sensors, but I haven't really seen it in my M10-R. However, I have seen it while shooting pro sports with a Canon R5 and Sony A9II. I usually use a Canon R5 with a 400 f2.8 for NFL football and MLB baseball. Recently I had to borrow a Sony with a 400 f2.8 for football and I found that I wasn't stopping motion at 1/1000, which has traditionally been the minimum for sports. I found that I had to shoot at a minimum of 1/2000 or I lost a lot a frames to lack of sharpness. It was more noticeable with the Sony than it was with the Canon R5. Both cameras have built in IBIS.

Apples and oranges I know. But I'm not ready to condemn the M11 until I try it.

  • Like 3
Link to post
Share on other sites

2 hours ago, Sailronin said:

I've shot with medium format digital since a Phase One P30 on a Hasselblad 501CM. The first thing I discovered was that the 30mp digital back picked up vibrations which were "buried" in the grain of film (color or B&W).  To make truly sharp images required a shutter speed of 3x focal length if hand held and using mirror lock up on a tripod...all the time.  I sold the P30 for a CFV50 which was even more sensitive. Currently I have a Leaf Aptus 75s (33mp) and Hasselblad 907x with CFV50c ii, I use all of them both in the studio and mostly in the field but vibration remains an issue, so yes, 60 mp in an M body with no IBIS will be a challenge in the real world.  Not sayin' it will be impossible to hand hold but know there will be A LOT of fuzzy images over the learning curve. Being mirrorless the M11 will be better than a HB500 series or the S(006) but it will take some practice.

 

Concur with the above. I agree that a high megapixel camera needs a good technique to operate. Proper shutter speed is a part of it.  I am just saying that one can operate a high megapixel camera without stabilization. 

  • Like 2
Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...