giellaleafapmu
Well-known
Yep, I didn't want to say this but I have really the feeling that we got already anything needed as far as image quality concerns, so companies will start a new game. You get the new 100Mpx body, but then you need the new supersharp lens and, yep, last year supersharp lens doesn't work because now we have a new 10 time faster autofocus system, so you need the lens with the hypersonic motor, not the ultrasonic one but, wait, there is a new 200Mpx camera...
I shall be buying all this stuff as everybody else but I still think that D300 (or equivalent) are going to be the pinnacle of "usable" technology as far as image quality (in print) concerns.
GLF
GLF
Darthfeeble
But you can call me Steve
Resolution, Resolution, Resolution..... S
willie_901
Veteran
As sensor (or film) area increases, everything gets better. The reason for this is: you collect more data... which means you have more information. Keith's anecdotal test is simply empirical confirmation of what happens as sensor area increases. The effect is not linear, so the increase in information content as you go up in sensor size is greater than you might assume.
Of course other many factors affect the information content. An older 24x36 sensor, or APS-C sensor might underperform the very best m4/3 sensor. Poor exposure (under exposure) is very bad because less data is collected. Overexposure is even worse for a different reason: you have lost all information about the light hitting a photosite.
Dynamic range is related to the sensor's signal-to-noise ratio because as the noise level decreases less of the available bits are used to digitize noise. This is why dynamic range can decrease as electronic ISO amplification increases. However other aspects of sensor design are critical to dynamic range as well.
I think the each photographer has to decide at what point the image quality for the money they pay is sufficient for them to get the job done. Included the financial end of the decision are the funds already invested in the lenses they happen to own. The emotional tie to their lenses is no less important for some people.
Of course other many factors affect the information content. An older 24x36 sensor, or APS-C sensor might underperform the very best m4/3 sensor. Poor exposure (under exposure) is very bad because less data is collected. Overexposure is even worse for a different reason: you have lost all information about the light hitting a photosite.
Dynamic range is related to the sensor's signal-to-noise ratio because as the noise level decreases less of the available bits are used to digitize noise. This is why dynamic range can decrease as electronic ISO amplification increases. However other aspects of sensor design are critical to dynamic range as well.
I think the each photographer has to decide at what point the image quality for the money they pay is sufficient for them to get the job done. Included the financial end of the decision are the funds already invested in the lenses they happen to own. The emotional tie to their lenses is no less important for some people.
gavinlg
Veteran
The D800 is none too gentle on lenses from what I read, and think its more to do with pixel density.
Yeah it is pixel density - the current 20ish mp aps-c bodies have the same pixel density as the d800 too...
craygc
Well-known
Yeah it is pixel density - the current 20ish mp aps-c bodies have the same pixel density as the d800 too...
You only need 16 MP APS-C to have the same pixel density as the D800
richardhkirkando
Well-known
If I didn't use film, I don't think I'd care about full frame vs crop digital. But I do, and I like the idea of one set of lenses and being able to use them to get the angle of view they're supposed to give.
The image quality of my APS-C NEX-7 is more than acceptable - I just wish my 35mm lens looked like a 35mm lens on it.
The image quality of my APS-C NEX-7 is more than acceptable - I just wish my 35mm lens looked like a 35mm lens on it.
I can honestly say I prefer the look from the x-pro1 to the 5d mkii... So much so that I ended up not using the 5d mkii.
Adanac
Well-known
That is also true, but lenses like the ZM 25 that have the highest center resolution ever measured at over 400 line pairs per mm are not scared by that![]()
But don't underestimate the lens design on the format it was designed for. Many Leica and Zeiss lenses loose their magic when used on a crop sensor.
The ZM25 is an excellent, classic, example of the phenomena Edward speaks of.
On APS-C I don't really like the character of the ZM25. On film I do. I've never shot it on full frame digital but expect I'll like it better there too.
Unfortunately, it is my most used lens on the GXR given it's my stand-in for 35mm equivalent field of view on APS-C. While the GXR does it justice from a resolution perspective, due to the crop, it just isn't the same as 35 on 35.
I thought I'd be patient and wait out the arrival of a full frame solution from a vendor like Ricoh but at this point I'm this close to selling the ZM25 (since I've retired my darkroom and am not shooting film any more) and spending even more for the RX1.
BobYIL
Well-known
Yep, I didn't want to say this but I have really the feeling that we got already anything needed as far as image quality concerns, so companies will start a new game. You get the new 100Mpx body, but then you need the new supersharp lens and, yep, last year supersharp lens doesn't work because now we have a new 10 time faster autofocus system, so you need the lens with the hypersonic motor, not the ultrasonic one but, wait, there is a new 200Mpx camera...I shall be buying all this stuff as everybody else but I still think that D300 (or equivalent) are going to be the pinnacle of "usable" technology as far as image quality (in print) concerns.
GLF
Not true.. with the same analogy we should not have anything above film-MF, let alone 4x5 or 8x10 formats.
I was trying to point to the requirements of our day; on one end the sensors are developing, on the other end to get the most out of these sensors the lens technology should develop too. If the R-lenses were performing not up to my expectations then do not blame the sensor or the resolution race; they were originally designed for film.
Then how do those new lenses designed specifically for sensors perform? Stellar!.. I can easily state that the lowly $219 Nikon 50/1.8G can outperform all 50mm R-lenses produced by Leica and can compete with the 50mm Summicron too at all apertures.
Regarding to your contentment with the D300 as "pinnacle of usable technology".. You are not modest at all...
As a side note I can also state that those who are considering to buy mirrorless cameras with the APS-C or especially FF sensors with the hope of using their M-mount legacy lenses at hand should better concentrate on the latest formulas or better, on the native lenses to come with those bodies. Especially the native lenses.. as they would exhibit not only the most optimum formulas and technology for that specific sensor but also "incorporate" with the inherent corrections programs against vignetting, distortion, color-shifts toward corners, CA, etc. The majority of the M-lenses in our hands are quite inferior from this regard.
craygc
Well-known
As a side note I can also state that those who are considering to buy mirrorless cameras with the APS-C or especially FF sensors with the hope of using their M-mount legacy lenses at hand should better concentrate on the latest formulas or better, on the native lenses to come with those bodies. Especially the native lenses.. as they would exhibit not only the most optimum formulas and technology for that specific sensor but also "incorporate" with the inherent corrections programs against vignetting, distortion, color-shifts toward corners, CA, etc. The majority of the M-lenses in our hands are quite inferior from this regard.
Beyond the issues of retro focused versus symmetrical issues, I would suggest that many of the "native" lenses for recent mirrorless cameras are designed as less than optimal but with their deficiencies "corrected" in firmware - its just cheaper to do it that way. If like the Sony Nex 5n the retro versus symmetrical design issue is addressed I would still expect exceptional performance from Leica lenses on these new cameras.
giellaleafapmu
Well-known
Not true.. with the same analogy we should not have anything above film-MF, let alone 4x5 or 8x10 formats.
I was trying to point to the requirements of our day; on one end the sensors are developing, on the other end to get the most out of these sensors the lens technology should develop too. If the R-lenses were performing not up to my expectations then do not blame the sensor or the resolution race; they were originally designed for film.
Then how do those new lenses designed specifically for sensors perform? Stellar!.. I can easily state that the lowly $219 Nikon 50/1.8G can outperform all 50mm R-lenses produced by Leica and can compete with the 50mm Summicron too at all apertures.
Regarding to your contentment with the D300 as "pinnacle of usable technology".. You are not modest at all...The D300 in FF terms is 28MP!! You will hardly be satisfied with any legacy lens including Leica-R ones.. Even the Leica M has newly been upgraded to 24MP..
As a side note I can also state that those who are considering to buy mirrorless cameras with the APS-C or especially FF sensors with the hope of using their M-mount legacy lenses at hand should better concentrate on the latest formulas or better, on the native lenses to come with those bodies. Especially the native lenses.. as they would exhibit not only the most optimum formulas and technology for that specific sensor but also "incorporate" with the inherent corrections programs against vignetting, distortion, color-shifts toward corners, CA, etc. The majority of the M-lenses in our hands are quite inferior from this regard.
I agrre 100% with everything, just one thing I wanted to state again: I am not saying full frame has no reasons to exists (or 8x10" or whatever) just that the general entusiasm seems to be more for the word "Full Frame" than for the real need, of course some people need it or need even larger format but just that..."Fuji is planning to go FF"...seem to me usually only for the sound of "Full frame", not for the size of prints users get from the camera. But, never mind, I am happy all this is happening, eventually I shall buy one instead of renting when needed and, despite the difference in quality being invisible in most prints, probably clients (and, why not, myself...) will be happy shooting..."Fuuuuuullll Fraaaaaame".
Have fun with anything you happoen to use!
GLF
teleparallel
Established
I made a comparison in tricky lighting between my D700 and OM-D with a series of images from both cameras used alternately for virtually the same shots. Initially they seemed well matched when viewing the results but that soon changed when I started trying to get more information from the Oly files. They fell apart quite quickly in comparison to the Nikon's files ... their workability was nowhere near the full frame files which have amazing recovery capabilities if you under expose ... with minimal loss in quality.
Actually, this happen because the noise threshold is really low, not because because it's bigger. It's also likely related to tone curve of the basic files from the d700.
teleparallel
Established
As sensor (or film) area increases, everything gets better. The reason for this is: you collect more data... which means you have more information. Keith's anecdotal test is simply empirical confirmation of what happens as sensor area increases. The effect is not linear, so the increase in information content as you go up in sensor size is greater than you might assume.
Of course other many factors affect the information content. An older 24x36 sensor, or APS-C sensor might underperform the very best m4/3 sensor. Poor exposure (under exposure) is very bad because less data is collected. Overexposure is even worse for a different reason: you have lost all information about the light hitting a photosite.
Dynamic range is related to the sensor's signal-to-noise ratio because as the noise level decreases less of the available bits are used to digitize noise. This is why dynamic range can decrease as electronic ISO amplification increases. However other aspects of sensor design are critical to dynamic range as well.
I think the each photographer has to decide at what point the image quality for the money they pay is sufficient for them to get the job done. Included the financial end of the decision are the funds already invested in the lenses they happen to own. The emotional tie to their lenses is no less important for some people.
I don't think information is related to size, really. Specially in digital technology. It's correct regarding the noise, and the reason bigger sensor usually performs better is due to the easier construction of low noise photodiodes. But it's now way because it's smaller. Of course there is a physical limit to size. Very small system do have limits, but if a 1/2.3'' sensor can have 12 megapixels and yet performs so well(pentax Q for instance), I don't think this limit is now reached. It's more related to purity of materials and correct crystals growth. Reducing thermal noise is really hard, and increases price heavily, and is probably easier to make things bigger and increase signal amount.
helvetica
Well-known
One huge factor that I haven't seen mentioned has nothing to do with the sensor and everything to do with the mirror in SLR's. Look through an APS-C camera and then look through the viewfinder on a full frame camera - night and day difference. Bigger mirror and (often) pentaprism vs pentamirror.
teleparallel
Established
Finally my point of view.
To buy a bigger sensor for legacy lenses is not really an argument here, since it's not really a technical reason, and is valid only for us, old photography lovers who already have a bunch of old lenses. Also we love FF, because we are used to it. It seems correct. It's an psycho-sociological effect called asymmetrical insight, wicth simply states: "Our way is better because we, a established "familiar" group(for us photographers using 35mm), do it this way. The others are just wrong"(and we have the audacity to point out problems for the other ones, feeling that we just know it all better, and that's why it's asymmetrical!). It's pointless to say that if fells right, or anything.
There's no correct sensor size, or field of view to lens. They are capable of cover a certain image circle, and that's it. Theoretically, any non fisheye lens can cover 180 degrees field of view with an enough big image circle(infinite, to be precise). Perspective does change, but correctable with special designs and post, to look the way we want. Also in the old days we did not complain that our MF lenses did not have the "correct" field of view on 35mm film. We did not use nikon lenses on canon cameras(usually), since it did not make sense.
APS-C has these basics pointed out limitations, but, plenty of good characteristics. Lens designs are easier for small image circles, even in faster designs, so a 23mm APS-C only, is very likely to cost less than an equivalent 35mm FF, in the same lunch date. They require less technique to be built, and controlled(it's harder to make big sensor with live view, for instance). Also, although nonexistent, one can builds(with less effort compared to 35mm format) multi-aspect ratio sensors, for squared, 645 or 16:9 formats.
Today we do have a limitation of lenses of aps-c, specially primes for dSLR's. But its a market problem. Many of us love to shoot with primes over zooms, but not younger first buyers of dslr's cameras, before talked into primes for they're "advantages"(asymmetrical insight again). So manufactures prefers make zoom designs, for small sensor. And they do primes for old users who get full frame(and sometimes charge a lot for them).
Finally, I don't think anyone can tell the difference of FF and APS-C. Not comparatively at least. I mean, using two cameras, one full frame and other APS-C, I take one photo with each, using different lenses, in different situations. If I show you these two different photos, can you tell witch camera I used in each occasion? I know I can tell film MF from film 35mm most of the time, since the quality is blowing away different. But it's so not the case in digital APS-C vs FF.
Camera manufactures move to full-frame for the simple reason: The market asks for it(and is also driven to). We get to keep in mind that there's always are more than one solution to problems such ISO performance and DR. To make the sensor bigger is one, but is it justified? This, I just can't know, or presume.
I Hope many of you read this to the end.
To buy a bigger sensor for legacy lenses is not really an argument here, since it's not really a technical reason, and is valid only for us, old photography lovers who already have a bunch of old lenses. Also we love FF, because we are used to it. It seems correct. It's an psycho-sociological effect called asymmetrical insight, wicth simply states: "Our way is better because we, a established "familiar" group(for us photographers using 35mm), do it this way. The others are just wrong"(and we have the audacity to point out problems for the other ones, feeling that we just know it all better, and that's why it's asymmetrical!). It's pointless to say that if fells right, or anything.
There's no correct sensor size, or field of view to lens. They are capable of cover a certain image circle, and that's it. Theoretically, any non fisheye lens can cover 180 degrees field of view with an enough big image circle(infinite, to be precise). Perspective does change, but correctable with special designs and post, to look the way we want. Also in the old days we did not complain that our MF lenses did not have the "correct" field of view on 35mm film. We did not use nikon lenses on canon cameras(usually), since it did not make sense.
APS-C has these basics pointed out limitations, but, plenty of good characteristics. Lens designs are easier for small image circles, even in faster designs, so a 23mm APS-C only, is very likely to cost less than an equivalent 35mm FF, in the same lunch date. They require less technique to be built, and controlled(it's harder to make big sensor with live view, for instance). Also, although nonexistent, one can builds(with less effort compared to 35mm format) multi-aspect ratio sensors, for squared, 645 or 16:9 formats.
Today we do have a limitation of lenses of aps-c, specially primes for dSLR's. But its a market problem. Many of us love to shoot with primes over zooms, but not younger first buyers of dslr's cameras, before talked into primes for they're "advantages"(asymmetrical insight again). So manufactures prefers make zoom designs, for small sensor. And they do primes for old users who get full frame(and sometimes charge a lot for them).
Finally, I don't think anyone can tell the difference of FF and APS-C. Not comparatively at least. I mean, using two cameras, one full frame and other APS-C, I take one photo with each, using different lenses, in different situations. If I show you these two different photos, can you tell witch camera I used in each occasion? I know I can tell film MF from film 35mm most of the time, since the quality is blowing away different. But it's so not the case in digital APS-C vs FF.
Camera manufactures move to full-frame for the simple reason: The market asks for it(and is also driven to). We get to keep in mind that there's always are more than one solution to problems such ISO performance and DR. To make the sensor bigger is one, but is it justified? This, I just can't know, or presume.
I Hope many of you read this to the end.
Share:
-
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.