[Hattifnatt]

Of all the aspects of sensor quality that you can measure, dynamic range is one of the least size-constrained in my opinion.

For example, Nikon offers 24 megapixel sensors both in 1.5x and FX sizes. The Nikon D750 and D7200 are two of the most recent iterations of the original sensor designs, (both from Sony in one way or another) ...and yet the D7200 is measured to have 0.1 stops better DR than the D750, by DXO.

Even if you like to throw DXO measurements out the window, and/or allow for a real-world fudge-factor, ...this still proves that generally speaking, a crop sensor is still capable of quite a great amount of dynamic range.

Heck, the Pentax 16 MP and 24 MP APS-C sensors have already far surpassed any full-frame Canon sensor for dynamic range, and that's a difference you'll see even in the most casual real-world situations!

In other words, unless there is some major technological breakthrough in sensor technology, (Sony's BSI technology in the A7R II actually HARMED dynamic range, compared to the original A7R sensor, according to DXO) ...you can expect the full-frame Pentax to have roughly similar or only marginally better dynamic range than the existing 16 MP and 24 MP crop sensors, which already have pretty amazing DR.

TLDR; if I were a betting man, I'd put $10 on 14-15 stops of DR.

=Matt=

I see. But then, why do people say the FF will benefit landscape shooters? maybe because of possibility to print bigger size because of increased resolution? I don't print anyway, I'm just asking.

So far I'm definitively satisfied with the DR of K5IIs, the amount of detail you can recover form shadows it's amazing. But better never hurts. I also started to shoot a bit in pubs and places with bad lightning, where I could use the low-noise-at-high-ISO of the FF. But from my very limited experience, having a bit more noise in those photos doesn't hurt THAT much. It's much more important to seize the moment and get the nice shot. Heck sometimes I even add noise in PP for these shots.

Maybe you can say the same about street photography. I was imagining myself getting the FF, the new 70-200 and then hit the streets for some shots. But now that I think of it, I can do it without FF gear. If I get a good shot, probably noise would not matter that much.

[Matthew Saville]

Personally, using real cameras, I did not find the same result as DXO. Even if on DXO the K-5 and K-3 are supposed to have a better DR than the 5DII, it is practically not true. Practically, K-3 has less than 10 stops of DR, not more. The image quality that we get out of APSC now is great. But the image quality from FF of more than five years ago still beats the best APSC. When I see photos from the 5DII and 5DIII, I don't buy DXO scores.

I don't blindly trust DXO, that's for sure. I've been working in wedding post-production for 5+ years now, and as a wedding photographer / educator for ~12 years. So I've post-produced literally a half-million images from Canon 5-series bodies alone, and maybe another half-million images from various other cameras from Nikon to Pentax, Sony, etc.

My personal experience is that Canon dynamic range is abysmal. Compared to any Sony sensor, whether Pentax or Nikon, the shadow banding on Canon is just a total show-stopper. Yes, Canon has a bit more highlight headroom than some, but it's not that much more, not enough to make up for the sorely lacking shadow recovery. I understand people have their own personal standards for shadow noise or highlight.....funkyness, but whether you "draw the line" at 10 stops, or 13/14 stops, I see Canon coming in noticeably behind most other sensors on the market today.

TLDR; as a landscape photographer shooting at lower ISOs, I'd take a K-3 or K-3 II ANY DAY over any Canon DSLR, APS-C or full-frame.

---------- Post added 09-21-15 at 03:32 PM ----------

I see. But then, why do people say the FF will benefit landscape shooters? maybe because of possibility to print bigger size because of increased resolution? I don't print anyway, I'm just asking.

So far I'm definitively satisfied with the DR of K5IIs, the amount of detail you can recover form shadows it's amazing. But better never hurts. I also started to shoot a bit in pubs and places with bad lightning, where I could use the low-noise-at-high-ISO of the FF. But from my very limited experience, having a bit more noise in those photos doesn't hurt THAT much. It's much more important to seize the moment and get the nice shot. Heck sometimes I even add noise in PP for these shots.

Maybe you can say the same about street photography. I was imagining myself getting the FF, the new 70-200 and then hit the streets for some shots. But now that I think of it, I can do it without FF gear. If I get a good shot, probably noise would not matter that much.

While reviewing various Nikon cameras, I've had the good fortune of gaining access to sensors with identical megapixel counts, and release dates very close, but different sensor sizes.

In short, here's how I'd answer your question: A 24 megapixel 1.5x crop sensor will have *slightly* less dynamic range, *slightly* more noise overall, and *slightly* less per-pixel acuity. However, none of these slight differences will be really noticeable. The only real big difference will be at ISO noise above 400 or so, when the "faint" amount of greater noise in a crop sensor slowly begins to become more significant.

The main problem is that there's almost always a generation gap, or a sensor processing engine gap, from sensor to sensor. We'll almost never get two sensors of identical megapixel counts, identical on-chip and off-chip processing, etc. And every 6-12 months, a slight improvement is made in one area or another.

However, considering Pentax' 645Z image quality, we could still expect a shocker from their FF sensor. I just wonder where that sensor is going to come from, and how many megapixels it will "get away with"....

[goldenarrow]

How about a FF Brownie Box camera with no automation, no LCD, no IBIS, no "scenes", no "filters", etc. Just a viewfinder and K-mount and a 42+ mpix sensor and card slot for $500 (literally, a black box, I guess there is a bootable ROM and battery in there somewhere). It would be so dreadfully lacking that professionals would gladly go for the $2500+ FF. The two paths would never cross and not cannabalize each other. Yet hobbyists, and especially students, would be able to afford digital FF. And if you can produce a current student ID, give away the 50mm DA since it is already FF compatible. Its pure genius! DxO would probably not even bother to test it, it would be so beneath them!!

[biz-engineer]

n short, here's how I'd answer your question: A 24 megapixel 1.5x crop sensor will have *slightly* less dynamic range, *slightly* more noise overall, and *slightly* less per-pixel acuity. However, none of these slight differences will be really noticeable. The only real big difference will be at ISO noise above 400 or so, when the "faint" amount of greater noise in a crop sensor slowly begins to become more significant.

Between APSc and FF sensor, due to physics of CMOS imagers, there's about one stop difference of performance + minor differences (+ -15%) depending on implementation tradeoffs. ISO noise difference between APSc and FF is considered to be larger above ISO400 because of the eye can perceive it above ISO400, however, from base ISO up to max ISO, there is a constant one stop advantage to the full frame sensor.

However, considering Pentax' 645Z image quality, we could still expect a shocker from their FF sensor. I just wonder where that sensor is going to come from, and how many megapixels it will "get away with"....

BSI enhancement in quantum efficiency of the 42Mp sensor, is so that it approximately matches a 36Mp sensor, so it is only a small improvement. Between 645z sensor and FF sensor, again one stop advantage.

[drougge]

Personally, using real cameras, I did not find the same result as DXO. Even if on DXO the K-5 and K-3 are supposed to have a better DR than the 5DII, it is practically not true. Practically, K-3 has less than 10 stops of DR, not more. The image quality that we get out of APSC now is great. But the image quality from FF of more than five years ago still beats the best APSC. When I see photos from the 5DII and 5DIII, I don't buy DXO scores.

I don't know how the numbers translate to real photos, but I have both a K5 and 5D2, and the K5 beats the 5D2 by a lot if you want to push the shadows.

[Rondec]

I see. But then, why do people say the FF will benefit landscape shooters? maybe because of possibility to print bigger size because of increased resolution? I don't print anyway, I'm just asking.

So far I'm definitively satisfied with the DR of K5IIs, the amount of detail you can recover form shadows it's amazing. But better never hurts. I also started to shoot a bit in pubs and places with bad lightning, where I could use the low-noise-at-high-ISO of the FF. But from my very limited experience, having a bit more noise in those photos doesn't hurt THAT much. It's much more important to seize the moment and get the nice shot. Heck sometimes I even add noise in PP for these shots.

Maybe you can say the same about street photography. I was imagining myself getting the FF, the new 70-200 and then hit the streets for some shots. But now that I think of it, I can do it without FF gear. If I get a good shot, probably noise would not matter that much.

I think there are two things. First of all, any time that you get above base iso, Sony full frame sensors tend to keep dynamic range better than APS-C sensors. If you shoot at base iso, there are few cameras that do better than the K5/K5 II with respect to dynamic range. The most obvious difference between the two formats is SNR. Full frame has less noise at all isos than APS-C does. This could certainly allow you to push your image a little harder without needing any noise reduction. The final thing is obviously that you tend to be able to print bigger with full frame sensors.

If you are getting good results with APS-C cameras, using the same techniques, you will probably be able to get some improvement in results using the Pentax full frame (assuming similar quality lens).

[jatrax]

The most obvious difference between the two formats is SNR. Full frame has less noise at all isos than APS-C does. This could certainly allow you to push your image a little harder without needing any noise reduction.

I see this stated a lot but curious if this is a function of the sensor size or the pixel spacing/pitch? What I mean is, it is often considered 'fact' that a FF sensor will have a better SNR (less noise) at similar ISO than an APS-C sized sensor. But why is that? And at what point does that go away? Does a 50MP FF sensor still beat a 16MP APS-C? 60MP? Just trying to understand this as it seems to me that as the MP count continues to explode some of the supposed FF advantage starts to go away. Or maybe I am completely misunderstanding how this works.

The final thing is obviously that you tend to be able to print bigger with full frame sensors.

Is this just a function of the larger MP or is there some other factor? I always thought print size was determined by the pixel dimensions / DPI of the print resolution.

[Rondec]

I see this stated a lot but curious if this is a function of the sensor size or the pixel spacing/pitch? What I mean is, it is often considered 'fact' that a FF sensor will have a better SNR (less noise) at similar ISO than an APS-C sized sensor. But why is that? And at what point does that go away? Does a 50MP FF sensor still beat a 16MP APS-C? 60MP? Just trying to understand this as it seems to me that as the MP count continues to explode some of the supposed FF advantage starts to go away. Or maybe I am completely misunderstanding how this works.

Is this just a function of the larger MP or is there some other factor? I always thought print size was determined by the pixel dimensions / DPI of the print resolution.

If you look at current generation full frames, they actually are very close to each other with regards to SNR, they differ greatly with regard to how they hold on to dynamic range as iso goes up.

The way I think about APS-C versus full frame is that APS-C has to be magnified more to get to your final print size. So if you look at a pixel level, a D810 will be almost exactly the same as a K5 II with regard to noise and dynamic range, but if you print at the same size, you will see noise on the K5 II print long before you see it on the D810. At low iso, there is little enough noise in either image that you probably won't see the difference, but as you go up in iso, you will certainly see more noise. It only makes sense to compare sensors with the same final print or view size, in which case a full frame camera will have less visible noise (doesn't really matter how many megapixels).

[osv]

Of all the aspects of sensor quality that you can measure, dynamic range is one of the least size-constrained in my opinion.

dr and file latitude are not the same thing... the former can be measured, the latter can't, and it's why god invented ff, and to some extent mf

Heck, the Pentax 16 MP and 24 MP APS-C sensors have already far surpassed any full-frame Canon sensor for dynamic range, and that's a difference you'll see even in the most casual real-world situations!

dr can vary with iso, and downrezzing improves all aspects of pq, which can be seen with the dxo test results.

holding canon sensors up as examples of ff dr is

In other words, unless there is some major technological breakthrough in sensor technology, (Sony's BSI technology in the A7R II actually HARMED dynamic range, compared to the original A7R sensor, according to DXO)

??? "An even greater divergence can be seen at higher ISOs. Dynamic range is slightly lower than its predecessor at base; however, the difference is minimal and, as with color sensitivity, the improvement in noise levels apparent at higher ISO settings translates into an impressive increase in dynamic range at these settings, particularly between ISO 1600 and 6400. The A7R II also achieved close to a half-stop (+0.5 Ev) improvement in our low-light ISO score." Sony A7R II vs. Sony A7R: Lower noise levels bring improvements in Low-Light ISO and Color Depth - DxOMark

---------- Post added 09-22-15 at 12:24 PM ----------

The way I think about APS-C versus full frame is that APS-C has to be magnified more to get to your final print size. So if you look at a pixel level, a D810 will be almost exactly the same as a K5 II with regard to noise and dynamic range, but if you print at the same size, you will see noise on the K5 II print long before you see it on the D810. At low iso, there is little enough noise in either image that you probably won't see the difference, but as you go up in iso, you will certainly see more noise. It only makes sense to compare sensors with the same final print or view size, in which case a full frame camera will have less visible noise (doesn't really matter how many megapixels).

thank goodness, somebody in cropland gets it

nice post.

[Rondec]

dr and file latitude are not the same thing... the former can be measured, the latter can't, and it's why god invented ff, and to some extent mf



dr can vary with iso, and downrezzing improves all aspects of pq, which can be seen with the dxo test results.

holding canon sensors up as examples of ff dr is



??? "An even greater divergence can be seen at higher ISOs. Dynamic range is slightly lower than its predecessor at base; however, the difference is minimal and, as with color sensitivity, the improvement in noise levels apparent at higher ISO settings translates into an impressive increase in dynamic range at these settings, particularly between ISO 1600 and 6400. The A7R II also achieved close to a half-stop (+0.5 Ev) improvement in our low-light ISO score." Sony A7R II vs. Sony A7R: Lower noise levels bring improvements in Low-Light ISO and Color Depth - DxOMark

---------- Post added 09-22-15 at 12:24 PM ----------



thank goodness, somebody in cropland gets it

nice post.

I guess I don't really know what you mean with respect to latitude. I would just say that the difference between the sensors in the D810 and the A7r II isn't huge. The D810 has better low iso (tested at about 50) and 14.8 EV of dynamic range there, while after iso 800, the A7r II has an extra 0.6 EVs of dynamic range (SNR curves for both camera basically lie right on top of each other). I doubt very highly that you could actually find many situations where this small extra dynamic range of the A7r II would allow it shoot and the D810 wouldn't get very similar results. Maybe at iso 12,800 or after? But photos shot at those isos, even on full frame cameras, are sub optimal.

In the end, both sensors are great -- the 810 a little better at low iso, the A7r II a little better at high iso -- the DXO Mark scores of 97 and 98 probably tell the story as well as anything else.

The reason to go with A7r II's sensor seems to be more with regard to video performance and frame rates than anything else (not sure how Pentax is going to manage decent frame rates with that many megapixels).

[jatrax]

The way I think about APS-C versus full frame is that APS-C has to be magnified more to get to your final print size.

I am elderly and sometimes not very quick but sorry I really, really do not understand this. A 24MP APS-C has the same number of pixels as a 24MP FF sensor, right? Or do I just not understand this? So a 4000x6000 image is the same on both, yes? So the magnification is the same? But on the FF the pixels are larger or spaced more or something so the SNR is better? Sorry to be so dense, sometimes I just have a hard time getting a concept.

[osv]

I guess I don't really know what you mean with respect to latitude. I would just say that the difference between the sensors in the D810 and the A7r II isn't huge..

agree they are close, d810 at iso64 and a7rii at iso1600 and up, is how i would look at it... pretty subtle differences.

i think of latitude in terms of video, but with digital stills people think "exposure latitude", which is similar... it is relevant to dr, but not the same thing, because it can't be measured... you can test for it: https://www.hurlbutvisuals.com/blog/2013/01/film-education-testing-your-cameras-latitude/

i like the clarkvision explanation, because latitude depends on what is being shot: "Exposure Latitude: The amount of overexposure or underexposure used in acquiring an image (e.g. with a given type of film or a digital camera) that can still produce acceptable results. Exposure latitude is obviously a subjective thing as different people have different ideas at to what is acceptable. You can examine the results in this article and decide what is acceptable for your imaging. The noise in the resulting images was measured and plotted to give quantitative information about the results, and I will present my derivations of exposure latitude based on the noise analysis.

Exposure latitude is related to but different than dynamic range. Exposure latitude depends on dynamic range. The higher the dynamic range, the greater a scene with a lower dynamic range can be recorded with different exposures without compromising highlights and shadow detail. If a large dynamic range scene just matched the capability of the recording medium, then the exposure latitude would be zero. If the scene is low in contrast, the recording system could record the scene at several different exposures and still produce an acceptable image. Thus exposure latitude is also scene dependent." Clarkvision: Exposure Latitude of a Digital Camera

also this: https://books.google.com/books?id=2mp4AgAAQBAJ&pg=PA69&lpg=PA69&dq=latitude+...nition&f=false

so when people talk about latitude, it's not dr, even tho michael says that, what he's really referring to is latitude... the 645z kills the a7rii when it comes to iso performance: "When it comes to noise at high ISO the Pentax 645Z wins. As for dynamic range, both cameras appear comparable in the shadow areas while the Sony has about a stop more recoverable in the highlight end."

[Rondec]

I am elderly and sometimes not very quick but sorry I really, really do not understand this. A 24MP APS-C has the same number of pixels as a 24MP FF sensor, right? Or do I just not understand this? So a 4000x6000 image is the same on both, yes? So the magnification is the same? But on the FF the pixels are larger or spaced more or something so the SNR is better? Sorry to be so dense, sometimes I just have a hard time getting a concept.

So, if you have two sensor with the same megapixels and the only difference is the overall size of the sensor, the larger sensor will have better SNR. I think you can just say that the larger pixels are "better quality" pixels and that while there are the same number on both sensors, the smaller ones tend to be noisier.

You won't see a difference in low iso situations, but if you push the shadows or shoot higher iso, you will see the noise quicker.

I do think that if you have the same number of megapixels, both shot at low iso, you probably would be hard put to tell the difference, with regard to noise, dynamic range, or maximal printing size.

[IchabodCrane]

I am elderly and sometimes not very quick but sorry I really, really do not understand this. A 24MP APS-C has the same number of pixels as a 24MP FF sensor, right? Or do I just not understand this? So a 4000x6000 image is the same on both, yes? So the magnification is the same? But on the FF the pixels are larger or spaced more or something so the SNR is better? Sorry to be so dense, sometimes I just have a hard time getting a concept.

Rondec may come back with his own answer but my way of thinking of this is the ratio of print size to sensor size in physical dimensions.

Example: an 8" X 12" print is 72 times the size of a FF sensor but is 168 times the size of an APS-C sensor.

[jatrax]

I think you can just say that the larger pixels are "better quality" pixels and that while there are the same number on both sensors, the smaller ones tend to be noisier.

Yes, I understand that. So at what point does a FF have the same size pixels or pixel spacing or pixel pitch or whatever, as an APS-C? 50MP? 60MP? What I am asking is if this FF advantage for SNR is lost when the pixel size becomes the same? Which I think for 24MP APS-C is about a 55MP FF.

Example: an 8" X 12" print is 72 times the size of a FF sensor but is 168 times the size of an APS-C sensor.

OK, I guess that makes sense. Sort of. But, still, it is the same number of pixels, right? Is the printer going to print the FF pixels bigger than the APS-C pixels? I thought a pixel was a pixel, a fixed dot with color information telling the printer what ink to put there.

I don't know, maybe this stuff is just too far over my head. Sorry to be so dense.