pico Posted August 17, 2018 Share #461 Posted August 17, 2018 (edited) Advertisement (gone after registration) [...] Personally, I am seeing resolutions (fine details) from the M10 that outperform what the equation predicts. I can see hairs reproduced one pixel (1/167th mm) wide in some pictures. Some physical physical features, such as a hair can be attributed to our brain's wired-in mechanics. IOW, details not in the image, but made up in the brain as if perceived. Much later I can extend the idea to so-called bokeh. Edited August 17, 2018 by pico Link to post Share on other sites More sharing options...
Advertisement Posted August 17, 2018 Posted August 17, 2018 Hi pico, Take a look here Really “Is it the end of M road”?. I'm sure you'll find what you were looking for!
adan Posted August 17, 2018 Share #462 Posted August 17, 2018 Being pedantic here, please check sensor figures, doubling sensor pixel increases linear resolution by sqrt(2)=1.41. That is accounted for: divide or multiply my micron dimensions by 1.41, and you'll see they are correctly scaled. e.g. 48Mp sensor has pixels of dimension 4.2 microns, and 24Mp sensor has pixels of 6 microns (4.2 x sqrt(2) - allowing for rounding) But - the classic total system linear resolution equation (interaction of lens plus sensor resolutions, plus any other element you care to assign a resolution for (teleconverter)) is not simple multiplication by sqrt(2), it is a polynomial equation. (It is basically a mathematical expression of the idea that you can't add anything to an optical system that does not in some way reduce the final resolution: haze, filters, lenses, teleconverters, film, sensors - all will be "less than perfect resolution" and thus degrade the output at every step between "subject" and "final image.") Below is graphed the total system resolution (y) for a lens of 80 lppm, on sensor resolutions (x, in pixels per mm) from zero to 640 pixels per mm (all 24mm x 36mm sensor resolutions between 0 and 353 megapixels). Not a straight-line function, but a flattening curve eventually approaching but never reaching 80 lppm total system resolution, as sensor resolution approaches ∞. Important point: this graph is strictly for 24 x 36 sensors. A C-L or T or S will need separate calculations and graphing. One could swap in a fixed sensor resolution for x and get a similar graph of TSR, for a range of lens resolutions x from 0 to 200 lppm, or any other range. 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! 3 Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/287168-really-%E2%80%9Cis-it-the-end-of-m-road%E2%80%9D/?do=findComment&comment=3574199'>More sharing options...
pico Posted August 17, 2018 Share #463 Posted August 17, 2018 (edited) Adan is the only very good photographer I have witnessed in my 72 years who also has a mathematical, scientific talent. I used to think the abilities were mutually exclusive. No longer. Fortunately he does not evaluate photos using scientific metrics. Vision rules. Edited August 17, 2018 by pico 3 Link to post Share on other sites More sharing options...
mmradman Posted August 17, 2018 Share #464 Posted August 17, 2018 (edited) That is accounted for: divide or multiply my micron dimensions by 1.41, and you'll see they are correctly scaled. e.g. 48Mp sensor has pixels of dimension 4.2 microns, and 24Mp sensor has pixels of 6 microns (4.2 x sqrt(2) - allowing for rounding) But - the classic total system linear resolution equation (interaction of lens plus sensor resolutions, plus any other element you care to assign a resolution for (teleconverter)) is not simple multiplication by sqrt(2), it is a polynomial equation. (It is basically a mathematical expression of the idea that you can't add anything to an optical system that does not in some way reduce the final resolution: haze, filters, lenses, teleconverters, film, sensors - all will be "less than perfect resolution" and thus degrade the output at every step between "subject" and "final image.") Below is graphed the total system resolution (y) for a lens of 80 lppm, on sensor resolutions (x, in pixels per mm) from zero to 640 pixels per mm (all 24mm x 36mm sensor resolutions between 0 and 353 megapixels). Not a straight-line function, but a flattening curve eventually approaching but never reaching 80 lppm total system resolution, as sensor resolution approaches ∞. Important point: this graph is strictly for 24 x 36 sensors. A C-L or T or S will need separate calculations and graphing. One could swap in a fixed sensor resolution for x and get a similar graph of TSR, for a range of lens resolutions x from 0 to 200 lppm, or any other range. sysresgraph.jpg Adan, before we get carried away maths i suggest we go back to the classic equation and agree that both lens and sensor are measured in same lp/mm. These are my figures... For 24Mp sensor that is 6000 x 4000 pixels or 3000 lp [lp = line pair] or 3000:36 = 83.33 lp/mm or rounded down to 83 lp/mm. For double the size sensor or 48Mp that would be 8460 x 5460 pixels [approx], or 4230 lp or 117 lp/mm. So assuming lens with 80 lp/mm [one you mentioned earlier] or any other figure resultant image resolution { Ri = Rs * Rl / (Rs + Rl)} would be always higher with higher res sensor. For example:- a) For lens & sensor having equal lp//mm Ri = 0.5. For one having higher res than the other Ri>0.5. Note; your polynomial equation is same as 01af's #413 equation, one i wrote in brackets above. Above is just proving the obvious. I agree with those here who assert here that usefulness of higher res sensor above 24Mp for average user is unnecessary, 24Mp is adequate for me too. Edit - note added. Edited August 17, 2018 by mmradman Link to post Share on other sites More sharing options...
lct Posted August 17, 2018 Share #465 Posted August 17, 2018 Interesting that the rumoured new Nikon Z6 is 24MP [...] But the Z7 should be 45MP. Hard to stop progress... Link to post Share on other sites More sharing options...
mmradman Posted August 17, 2018 Share #466 Posted August 17, 2018 But the Z7 should be 45MP. Hard to stop progress... Yes, Z6 and Z7 mirrorless counterparts to D750 [24mp] and D850 [45Mp], both announced next Thursday. Link to post Share on other sites More sharing options...
adan Posted August 17, 2018 Share #467 Posted August 17, 2018 Advertisement (gone after registration) Pico, old friend - Magritte notwithstanding (a picture is not the thing itself) - these are no figment of my imagination, boyo. 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! Anyway: Tailwagger mentioned that Leica's "special microlenses" may impose a limit on how small they can make their pixels. This occurred to me also, but since I have no idea how CMOSIS/STMicro actually manufacture those, I don't know if it is actually a problem. Some microlenses are molded, some are layed down on the sensor as little blocks and then heated until they melt and slump into a lens shape. The latter might be affected by such "size" effects as surface tension strength or similar. For a given sensor technology, there is always a trade-off between number of pixels, and things like dynamic range and noise/ISO. It is nice that Sony has back-lit CMOS construction - but Leica doesn't (yet). Things like patents mean not all manufacturers can simply jump on the newest technology any time they want. One thing that is certain for at least the next 20 years (and perhaps much longer), is that a core design parameter for any Leica camera bearing the "M" designation is that it will work as well as possible with virtually any M lenses of any era. And retain the optical rangefinder and the size and form-factor. No matter how much that compromises performance in other ways. If better performance of any type can be made to work within that restriction, Leica will no doubt do so eventually. But they will not sacrifice that core goal for anything - in fact the reverse will be true. Leica will make other cameras and lenses to take technology and features to their limits, unrestricted by their legacy. But not the M. Don't buy into the M system unless that is acceptable, because you will be disappointed if you don't believe it. Go back and look at their history, and what they kept in the line, and what they eliminated, and what they changed their minds about. The M in virtually unaltered form has been the one Leica product that has always survived. mmradman: my only quibble with your response is that an image sensor is really a digital sampling device. As such, what it really has is a "spatial sampling rate." One sample per 6 microns in the M10 (166.67 samples/mm); one sample per 4.2 microns on 48 Mp (235 samples/mm). But the equation can be run with either the full sampling rate, or the Nyquist Frequency limit (half the sampling rate) and get meaningful results. It just depends on whether one is more interested on the maximum amount of (compromised) detail (which is the "Leica way" - no AA filtering - artifacts are better than "fuzzy") or the stricter minimal-artifacts standard. 1 Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! Anyway: Tailwagger mentioned that Leica's "special microlenses" may impose a limit on how small they can make their pixels. This occurred to me also, but since I have no idea how CMOSIS/STMicro actually manufacture those, I don't know if it is actually a problem. Some microlenses are molded, some are layed down on the sensor as little blocks and then heated until they melt and slump into a lens shape. The latter might be affected by such "size" effects as surface tension strength or similar. For a given sensor technology, there is always a trade-off between number of pixels, and things like dynamic range and noise/ISO. It is nice that Sony has back-lit CMOS construction - but Leica doesn't (yet). Things like patents mean not all manufacturers can simply jump on the newest technology any time they want. One thing that is certain for at least the next 20 years (and perhaps much longer), is that a core design parameter for any Leica camera bearing the "M" designation is that it will work as well as possible with virtually any M lenses of any era. And retain the optical rangefinder and the size and form-factor. No matter how much that compromises performance in other ways. If better performance of any type can be made to work within that restriction, Leica will no doubt do so eventually. But they will not sacrifice that core goal for anything - in fact the reverse will be true. Leica will make other cameras and lenses to take technology and features to their limits, unrestricted by their legacy. But not the M. Don't buy into the M system unless that is acceptable, because you will be disappointed if you don't believe it. Go back and look at their history, and what they kept in the line, and what they eliminated, and what they changed their minds about. The M in virtually unaltered form has been the one Leica product that has always survived. mmradman: my only quibble with your response is that an image sensor is really a digital sampling device. As such, what it really has is a "spatial sampling rate." One sample per 6 microns in the M10 (166.67 samples/mm); one sample per 4.2 microns on 48 Mp (235 samples/mm). But the equation can be run with either the full sampling rate, or the Nyquist Frequency limit (half the sampling rate) and get meaningful results. It just depends on whether one is more interested on the maximum amount of (compromised) detail (which is the "Leica way" - no AA filtering - artifacts are better than "fuzzy") or the stricter minimal-artifacts standard. ' data-webShareUrl='https://www.l-camera-forum.com/topic/287168-really-%E2%80%9Cis-it-the-end-of-m-road%E2%80%9D/?do=findComment&comment=3574217'>More sharing options...
Jeff S Posted August 17, 2018 Share #468 Posted August 17, 2018 I used to think the abilities were mutually exclusive. While before the age of photography, I think da Vinci settled that issue (math, art and more) long ago. But, yes, seemingly rare among mortals. Although an understanding of math (golden ratio, etc) can help with composition and certain types of photography. I can think of some people with brilliant minds in computers, engineering, physics, etc who are decent photographers. Bill Atkinson, for example. Jeff Link to post Share on other sites More sharing options...
mdemeyer Posted August 18, 2018 Share #469 Posted August 18, 2018 One reason I would welcome a higher pixel count sensor would be to reduce the frequency with which I have false color from moire in my M10 images, especially in architecture and fabrics. This would, to me, be a meaningful benefit. Link to post Share on other sites More sharing options...
IkarusJohn Posted August 18, 2018 Share #470 Posted August 18, 2018 (edited) But the Z7 should be 45MP. Hard to stop progress... True, but the point is both make 24MP versions of their cameras. They don’t just chase the MP rabbit. Edited August 18, 2018 by IkarusJohn 1 Link to post Share on other sites More sharing options...
pgk Posted August 18, 2018 Share #471 Posted August 18, 2018 Leica will make other cameras and lenses to take technology and features to their limits, unrestricted by their legacy. But not the M. Don't buy into the M system unless that is acceptable, because you will be disappointed if you don't believe it. Go back and look at their history, and what they kept in the line, and what they eliminated, and what they changed their minds about. The M in virtually unaltered form has been the one Leica product that has always survived. Surely this is the point. The M camera is already outperforming its designers wildest dreams. Its legacy restricts its unrestricted modification. Trying to get it to be an 'ultimate' resolution camera is neither practical nor logical. 1 Link to post Share on other sites More sharing options...
Chaemono Posted August 18, 2018 Share #472 Posted August 18, 2018 (edited) Some more with the gorgeous 1955 Collapsible 50 Summicron. There is no point in doing the comparison at f/4.0 since this would increase contrast and the lens would therefore resolve more details. The advantages of a 42 MPx sensor in terms of detail would become more evident. We want a lens that is a soft as possible to see if MPx count makes a difference to the resolution of the image. A 1955 Summicron does not necessarily fall into that category, at least not in the center, but it's the best M-mount lens I have access to for these comparisons. At f/2.0 but I will redo them because the blur in the Sony picture doesn't look as soft as the rendering of the M10. Also, while the focus in both pictures is right on the same spot, the focus point is not quite in the center for the M10. But let's look at the two pictures and the crops anyway. Equal amount of sharpening and clarity applied in LR to both. Raw files can be made available. Less compressed JPEGs here: https://www.smugmug.com/gallery/n-8pL8zB/ α7R III + Collapsible Leica Summicron 5cm f/2 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! ISO 125 f/2.0 @1/4000 sec. Edited August 18, 2018 by Chaemono Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ISO 125 f/2.0 @1/4000 sec. ' data-webShareUrl='https://www.l-camera-forum.com/topic/287168-really-%E2%80%9Cis-it-the-end-of-m-road%E2%80%9D/?do=findComment&comment=3574609'>More sharing options...
Chaemono Posted August 18, 2018 Share #473 Posted August 18, 2018 Less compressed JPEGs here: https://www.smugmug.com/gallery/n-8pL8zB/ M10 + Collapsible Leica Summicron 5cm f/2 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! ISO 200 f/2.0 @1/4000 sec. Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ISO 200 f/2.0 @1/4000 sec. ' data-webShareUrl='https://www.l-camera-forum.com/topic/287168-really-%E2%80%9Cis-it-the-end-of-m-road%E2%80%9D/?do=findComment&comment=3574611'>More sharing options...
Chaemono Posted August 18, 2018 Share #474 Posted August 18, 2018 And now the crops. Can you see 42 MPx detail? Not yet. Less compressed JPEGs here: https://www.smugmug.com/gallery/n-8pL8zB/ α7R III + Collapsible Leica Summicron 5cm f/2 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! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/287168-really-%E2%80%9Cis-it-the-end-of-m-road%E2%80%9D/?do=findComment&comment=3574612'>More sharing options...
Chaemono Posted August 18, 2018 Share #475 Posted August 18, 2018 Less compressed JPEGs here: https://www.smugmug.com/gallery/n-8pL8zB/ M10 + Collapsible Leica Summicron 5cm f/2 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! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/287168-really-%E2%80%9Cis-it-the-end-of-m-road%E2%80%9D/?do=findComment&comment=3574614'>More sharing options...
Chaemono Posted August 18, 2018 Share #476 Posted August 18, 2018 How about now? Well, yes, sort of. But it's almost like splitting hairs. And we are talking the Summicron here, small, light and amazingly sharp. Less compressed JPEGs here: https://www.smugmug.com/gallery/n-8pL8zB/ α7R III + Collapsible Leica Summicron 5cm f/2 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! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/287168-really-%E2%80%9Cis-it-the-end-of-m-road%E2%80%9D/?do=findComment&comment=3574617'>More sharing options...
Chaemono Posted August 18, 2018 Share #477 Posted August 18, 2018 Less compressed JPEGs here: https://www.smugmug.com/gallery/n-8pL8zB/ M10 + Collapsible Leica Summicron 5cm f/2 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! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/287168-really-%E2%80%9Cis-it-the-end-of-m-road%E2%80%9D/?do=findComment&comment=3574618'>More sharing options...
otto.f Posted August 18, 2018 Share #478 Posted August 18, 2018 Did you focus with EVF on the M10? It seems the focus plane lies at a different point than with the Sony Link to post Share on other sites More sharing options...
Chaemono Posted August 18, 2018 Share #479 Posted August 18, 2018 Yes, I did. They are not ideal but the focus point is right on the same part of the plant. 2 Link to post Share on other sites More sharing options...
jaapv Posted August 19, 2018 Share #480 Posted August 19, 2018 Did you focus with EVF on the M10? It seems the focus plane lies at a different point than with the Sony Agree. Exposure/contrast is different too. Link to post Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now