[ilya80]

Sometimes I want a larger aperture because I'm in low light, but want increased DOF. In such situations an APS-C sensor would have an advantage over a 36x24mm sensor.

ND filters to the rescue!

[RMabo]

24x36 doesn't automatically have more advanced sensor technology, it can be the other way around - a smaller format having more advanced sensors, because it is more challenging to keep noise down and dynamic range high on a smaller sensor so this can require more technology.

Sony said about their highly advanced Exmor "back lit sensor" technology that it was designed for smaller sensors to expand the light sensitive area and lower the noise, because in a bigger sensor there was no need to use this technology since larger sensors already had a larger light sensitive area and lower noise...

[junyo]

This thing about more megapixels could be questioned.
You can make more Mp on a 24x36 yes, because you have the surface area to do it - this doesn't automatically mean that the sensors makers will make as many Mp's in as possible

Not long time ago, high end APS-C had more Mp's than 24x36.
Now 24x36 leads, but it can change fast.
So amount of Mp is not an "absolute advantage" in my interpretation of the word, since it depends on the decision made by the sensor manufacturer and not an automatic effect of the 24x36 format.

The number of pixels doesn't depend on the size of the sensor. There are 14Mp APS-C sensors and there are 12Mp 36x24mm sensors.

Sensor technology isn't the only factor when considering image qualities related to the size of a sensor. Pixel pitch size also counts, yet it varies with the pixel count (and pixel count is an implementation choice, not related to sensor technology or sensor size).

Which is why I said "with a given sensor technology". Sure you can construct a bazillion MP cellphone sensor, but there's an objective performance tradeoff. Assuming relatively common construction/subcomponent technology so that the performance is roughly similar, the greater sensor size would allow you to either a) pack more photosites on the sensor at a given density or b) reduce the density while maintaining the same pixel count and thus likely decreasing noise. Unless you're telling me that there is such a thing as a free lunch, and using smaller imaging media extracts no penalty? I'm prepared to live with the trade, but I'm not going to say it doesn't exist.

Last, but not least, less DOF can be an advantage or a disadvantage, depending on your goals. Sometimes I want a larger aperture because I'm in low light, but want increased DOF. In such situations an APS-C sensor would have an advantage over a 36x24mm sensor.

I take your point however, you can usually decrease the aperture of a lens and thus gain DOF, it's much more difficult to increase max aperture. I hear of people complaining about not being able to separate with an f2.8 lens much more than I hear about people wishing that they had lenses that went to f48.

Again, I don't have a dog in this fight, just playing devil's advocate trying to be objective.

[WMBP]

I'm agnostic as to FF (I'd probably think about it if one where available, but I don't think it's crucial). But objectively, there are some absolute advantages to FF:
- DOF at a given FOV and F stop
- Lower noise with a given sensor technology
- More pixels to work with so cropping leaves a better quality image

How crucial/big of an advantage those points represent is up to a particular user, but they do exist.

These are advantages, but not absolute advantages. I'm trying to use the word in a very simple sense, really I am. Not sure why this distinction is hard to grasp.

If I show you a video and tell you it was taken on a camera, you can be sure that it was NOT taken on a K20D, because the K20D doesn't do video. This is (I hope) a clearly absolute difference, and if you regard the ability to do video as an advantage, then it's an absolute advantage. This camera can do it, that one simply can't, can't do it even badly. It's a simple, unambiguous distinction. No advantage of 36x24 is this clear cut, this black and white. You simply CANNOT show me a photo taken with a 36x24 camera that simply could not have been taken by an APS-C camera. You can show me 36x24 photos that would be hard to take with APS-C. You can show me photos that, if taken on APS-C, might have a little more noise. But these are matters of degree, not absolutes.


The depth of field advantage?

Let's say that, first, I show you the photo below. This is a very small plant in a very small vase. The mouth of the vase is not much more than an inch in diameter. In the photo, the tip of the nearest leaf of the plant is in focus, and so is the tip of another leaf that seems to be on the same focal plane; but otherwise, the rest of the plant is out of focus, most of it in fact very significantly blurred. Depth of field in this shot is a tiny fraction of an inch.

Now, looking at this photo, you could, with some confidence, say that the photo probably wasn't taken with a point-and-shoot/fixed-lens camera, because none of them are capable of extremely narrow depth of field. But it would only be a safe bet; photographer might have done something to surprise you. I guess I'd say this is an almost absolute advantage. Certainly a compelling advantage of the larger sensors over the (much smaller) point-and-shoot sensors.

On the other hand, if the question was, "Was this picture taken with a K20D or a Canon 5D MkII?", I contend that you simply wouldn't be able to tell from the depth of field alone. Because if you go to a little bit of trouble, you can get very narrow depth of field from a K20D as well. After all, I did it here. (K20D with Sigma 105 macro at f/2.8, with a tip-of-lens to tip-of-leap distance of about 6 inches.)

You might be able to get even narrower depth of field with a 36x24 camera, true. But at some point, depth of field is so narrow that, well, you can't SEE it any more.

Depth-of-Field Master online tells me that, a Pentax K20D with a 100mm lens, at f/2.8, shooting from 3 ft, has a depth of field of 0.03 ft; while a Canon 5D MkII, with a 150mm lens, at f/2.8, shooting from 3 ft, has a depth of field of 0.02 ft. That 0.01 ft difference is, well, not enough to justify the extra expense.

I would note also that the depth of field issue, to some extent, cuts both ways. It is certainly the case that the easy, deep depth of field you get from a point-and-shoot is an advantage for the majority of ordinary users.


The megapixel advantage?

Megapixels are also not an absolute advantage, unless you work in some field where some really high minimum resolution is required, for some reason. There seems to be a reason for medium-format digital cameras to exist. I don't really know what it is, but I acknowledge that it's there.

Still, for the work that most photographers do, it's quite clear that file resolutions in the range of 12-16 MP are more than adequate for most purposes. I have 20"x30" pro-lab prints of my photos that are, well, breathtakingly sharp. Ben Kanarek observed in this forum years ago that his photos (taken with the 10MP K10D) were able to be blown up VERY large indeed for use on kiosks.

The idea that you can, occasionally, crop tighter on a file with more megapixels, is a an accidental and non-absolute advantage. It's nice, when you need it. But I would not pay thousands of dollars for this "feature." If I needed to "get closer" to my subjects, I'd buy a longer telephoto lens.


The wide-angle advantage?

Which brings me to the alleged wide-angle advantage of 36x24 cameras. I don't feel the need to say much about this at all, because we now have very good wide-angle lenses for our Pentax, APS-C cameras. I love my Sigma 10-20, which is a wonderful lens even at 10mm. Pentax's 15mm prime seems to be a terrific lens. Anybody who routinely needs to shoot wider than these lenses can go, might have a good reason to go to 36x24. But for most of us, this just isn't an issue.

I think it IS an issue for the 4/3 and micro-4/3 cameras, although I don't know a great deal what lenses they have. But I'm not defending the 4/3 or micro-4/3 systems.

And again, the "crop-factor" or field-of-view equivalence issue cuts both ways. If what you need is telephoto reach, then APS-C has an advantage.


The high ISO performance advantage?

OF the alleged advantages of 36x24 digital cameras, better high ISO performance seems to me the most compelling. But it's not an absolute advantage either. Indeed, this is an area where it's pretty clear APS-C will continue to improve.

Perhaps 36x24 will improve, too, but to be honest, I don't need a LOT of improvement here. I'm already able to shoot in dark churches, at ISO 1600 or 2000, with good to acceptable results. Yes, of course, I wish that my ISO 2000 shots were reliably GREAT, virtually noise-free, etc. If I were charging $3000 per wedding, I probably would be shooting with a 36x24. But my clients are very happy with the results I get. I make the most of what I've got by shooting with fast primes—I shoot at f/2 a fair bit. I've shot candlelight ceremonies and other very poorly-lit events. So far, nobody has actually asked me to shoot a wedding in total darkness.


No absolute advantages

Bottom line: no absolute advantages. If APS-C can do the job more than adequately, which it clearly already can in the majority of circumstances, then WHY would I want to spend more for 36x24? If I truly needed it, I'd be there already.


Is 36x24 inevitable?

SO let me address the inevitability argument, which I think may be the one that 36x24 advocates are really most persuaded by.

I don't know if it's inevitable. Nobody else does either, including nobody at any of the top camera makers. Somebody at Pentax knows right now if Pentax is releasing a 36x24 camera this year. But whether they do or don't, doesn't tell us whether that camera will be a hit or a miss, and it certainly doesn't tell us whether 36x24 will ever have more than 5% of the interchangeable lens digital camera market.

As I suggested earlier, the market is being pulled in two directions. There clearly is some kind of market for 36x24, as there is some kind of market for medium format. At the same time, there's an enormous market for smaller, more compact and better cameras. 90% of the market wants to take better pictures without having to learn anything about photography. They want to upload their photos to Facebook without effort, too. They can do this now—on their smart phones.

Will 36x24 bodies get cheaper? The components will get cheaper, PERHAPS. That's assuming we don't have a global melt-down of some sort, like a world war, or some other major hit to the economy. And that's an iffy assumption right now. I don't go to bed scared, yet, but I do say my prayers at night.

But even if the components get cheaper, the cameras themselves might not. Nobody is going to be selling 36x24 camera bodies in my lifetime for, oh, $500. Entry level DSLRS have been selling for $600-$800 for the last 4-5 years. My wild guess is that $2000 is the floor for 36x24 bodies. The higher-end bodies always have a higher profit margin. Remember this is a camera that very few people ACTUALLY NEED. Those that think they DO need it, will pay for it.


*


Well, we'll see what Pentax does this year. But as I said, it almost doesn't matter to the points I'm making. If they do release a 36x24, I'll take a look at it. But almost the only thing that would drive me out of APS-C right now, would be the feeling that the engineers had stopped trying to improve APS-C. And if that happens, I won't be buying Pentax anymore.

Will

[wolfier]

The number of pixels doesn't depend on the size of the sensor. There are 14Mp APS-C sensors and there are 12Mp 36x24mm sensors.

Sensor technology isn't the only factor when considering image qualities related to the size of a sensor. Pixel pitch size also counts, yet it varies with the pixel count (and pixel count is an implementation choice, not related to sensor technology or sensor size).

Last, but not least, less DOF can be an advantage or a disadvantage, depending on your goals. Sometimes I want a larger aperture because I'm in low light, but want increased DOF. In such situations an APS-C sensor would have an advantage over a 36x24mm sensor.

Cheers,

36x24 has the advantage of DOF control - not just simply "DOF will be shallower". If you use a lens with the same aperture, same shutter time, and shoot at the same distance and lenses that give you the same FOV, then yes the DOF will be shallower.

e.g. 50/2.8 on APS-C vs 77/2.8 on 24x36, same shutter time, distance, both at ISO 400, for example.

If you desire a DOF the same as what you get on APS-C, it can be achieved on the 24x36 by stopping down by 1.3 stops and turn up the ISO respectively.

e.g. 50/2.8 on APS-C vs 77/4.4 on 24x36, same shutter time, distance, but while remaining at ISO 400 on the APS-C, turn up ISO to 1000 on the 24x36.

You'll get the same DOF. So, until you use up the aperture stops all the way to f/16 or f/22 on most lenses, you can achieve the whatever APS-C DOF you want on a 24x36.

On the other hand, if you want shallower, you have no recourse on APS-C except to pay for a faster lens - but in this case, 24x36 will benefit, too - not to mention a faster lens usually makes the APS-C system larger and heavier.

[WMBP]

36x24 has the advantage of DOF control - not just simply "DOF will be shallower".

Yes, of course. But what everybody wants—well, what the pros and enthusiasts want—is control at the shallow end.

[Pål Jensen]

36x24 has the advantage of DOF control - not just simply "DOF will be shallower". .

You just get shallower DOF thats it. The larger the format the shallower the DOF at same angle of view. Thats why LF cameras have tilt function; they would have been useless otherwise . Shallow DOF does not equal control over DOF. The APS is the sweetspot if control over DOF is considered. It give mm thin DOF if needed and deep DOF that can only be rivaled by LF; 35mm cannot do the latter.

[Pål Jensen]

Yes, of course. But what everybody wants—well, what the pros and enthusiasts want—is control at the shallow end.

No they don't. Hardly any sucessful image ever shot depends of so shallow DOF that it cannot be achieved with any DSLR format.
Photographers need enough DOF; thats whats make an image sucessful. The idea that professional photography is about paper thin DOF is ridiculous....

[slocant]

ND filters to the rescue!

On the contrary, I want the largest aperture possible (for available light photography) while retaining adequate DOF (say, for group shots).

Cheers,

[slocant]

I take your point however, you can usually decrease the aperture of a lens and thus gain DOF, it's much more difficult to increase max aperture. I hear of people complaining about not being able to separate with an f2.8 lens much more than I hear about people wishing that they had lenses that went to f48.

Again, I don't have a dog in this fight, just playing devil's advocate trying to be objective.

You can't usually decrease the aperture if you are constrained by the available light. Because you'd have to decrease shutter speed, or increase ISO, or use artificial lighting.

If I had to choose between shooting at f/2 with APS-C (and its DOF) or at f/2 with a 36x24mm (with less DOF), I'd have to choose APS-C in order to not lose DOF. Considering the ISO and shutter constraints of (my) low light photography.

That is my "complaint" regarding stills, and is also a contradiction in DSLR motion picture production. I've read (and it does make sense to me) that 36x24mm sensors aren't the best choice for motion picture production, since the DOF is so thin (unless that is your goal). Because generally more DOF is required or desired for motion pictures. But if you decrease aperture you have to increase artificial lighting, which adds cost, size and weight to DSLR motion picture production, counteracting its cost, size and weight benefits.

For motion picture (micro) four thirds may be the ideal format, that balances shallow DOF (and excellent image quality), but not so much as to require artificial lighting. I really don't have any experience with different formats for motion picture.

Cheers,

[wolfier]

You just get shallower DOF thats it. The larger the format the shallower the DOF at same angle of view. Thats why LF cameras have tilt function; they would have been useless otherwise . Shallow DOF does not equal control over DOF. The APS is the sweetspot if control over DOF is considered. It give mm thin DOF if needed and deep DOF that can only be rivaled by LF; 35mm cannot do the latter.

It is the ability to go shallower and you're not forced to - if you want as deep, just stop down by 1.3 stops and up the ISO by 1.3 stops.

Unless you constantly shoot at f/16 at 3200 ISO, there is always a setting on a 24x36 to match whatever DOF you can achieve on APS-C. This, is what I call "control".

[WMBP]

You just get shallower DOF thats it. The larger the format the shallower the DOF at same angle of view. Thats why LF cameras have tilt function; they would have been useless otherwise . Shallow DOF does not equal control over DOF. The APS is the sweetspot if control over DOF is considered. It give mm thin DOF if needed and deep DOF that can only be rivaled by LF; 35mm cannot do the latter.

If you are shooting 36x24, and you want MORE depth of field, can't you just stop down? Or step back further? The problem with depth of field (unless I've been missing something for years, which is entirely possible) is that you eventually run into a problem at the SHALLOW end of the depth of field pool. When you get as close to the subject as your lens will allow, and open your lens up as wide as it will go, you've done all you can do.

You just get shallower DOF thats it. The larger the format the shallower the DOF at same angle of view.

.... AND (you forgot to add) the same aperture.

No they don't. Hardly any sucessful image ever shot depends of so shallow DOF that it cannot be achieved with any DSLR format.
Photographers need enough DOF; thats whats make an image sucessful. The idea that professional photography is about paper thin DOF is ridiculous....

Pål, it sounds like you and I may be on the same side here, so let's not throw stones at one another. Let's get together and throw stones together at the other guys. :-)

I NEVER SAID that photography is about razor-thin depth of field. And I DID say exactly what you are saying: the good enough is good enough. You and I are exactly in agreement. If you think I said otherwise, please go back and reread what I said.

*

To reiterate: The depth of field differences are not nothing, but they aren't anything LIKE an order of magnitude, either. To take a simple, classic example. Say you're doing a portrait. You might shoot with your Pentax 70 on your K20D or K-7; if you're using a 5D MkII, you might use a 105mm lens. Now, let's assume you do want shallow depth of field for the portrait—you want a classic portrait with a blurred background.

So you set the 5D MkII to f/2.8 and you position the camera 8 ft (2.4 m) from the subject (focal plane). You've got yourself 0.29ft of depth of field—about 3.5 inches or almost 9 cm.

Or you pick up the Pentax K20D with the 70mm lens on it, stand the same distance, and open up to f/2.8. Now you've got 0.44 ft (13.4 cm) of depth of field.

Difference in depth of field? 1.8 inches.

It's not nothing. I have never said otherwise. But is it worth $2000K? Does it really matter to the client all that much? I doubt it. That 1.8 inches is less than the distance from the tip of the subject's nose to her ears.

(And of course if I open the 70 all the way up to f/2.4 .... )


Will

[WMBP]

Unless you constantly shoot at f/16 at 3200 ISO, there is always a setting on a 24x36 to match whatever DOF you can achieve on APS-C. This, is what I call "control".

Yes, good point, and I forgot that there actually IS a limit to how deep you can go, at some point. But I forgot for a reason: because I have never come close to that point.

On the other hand, the photographers in Group f/64 (Adams, Weston, et al.) DID work at that limiting point, with their large format cameras. But that was THEIR problem, not mine. Adams seems to have figured things out somehow. :-)

Will

[RMabo]

Pål, it sounds like you and I may be on the same side here, so let's not throw stones at one another. Let's get together and throw stones together at the other guys. :-)

Jehova! (who throwed that stone?)

[wolfier]

Yes, good point, and I forgot that there actually IS a limit to how deep you can go, at some point. But I forgot for a reason: because I have never come close to that point.

On the other hand, the photographers in Group f/64 (Adams, Weston, et al.) DID work at that limiting point, with their large format cameras. But that was THEIR problem, not mine. Adams seems to have figured things out somehow. :-)

Will

Agreed. No great shot in particular is limited by not having a razor-thin DOF either. So the DOF control on. The larger sensor is not important/noticeable to most people.

I do need to say that however, "DOF control" might be a misnomer, because when photographers shoot wide open for a "thin DOF", what they're usually, actually after is "creamier bokeh". (so the 1 inch vs 1.5 inch argument can fall apart - even the DOF is too close to be noticeable, the background blurring often behaves otherwise). In fact, they usually want both eyes to be in focus, thus creating a self-contradicting situation.

You may achieve the same with APS-C, but must pay more attention in choosing your background. 24x36 shooters need to pay more attention to keep things in focus. Some find one easier to achieve than the other.