70-200 opinion piece on dpreview perpetuating the equivalence rubbish

It's just another angle FF users like to take to make themselves feel superior. Of course, as soon as I bring up medium or large format, or the fact that the 35mm sensor size was an amateur format in the film days, they all curl up in the corners and pee on themselves. Most of the confusion has to do with confusing F-Stops with T-Stops.

Most of those people arguing aren't really taking many pictures.

Which reminds me, I have about 1000 photos on my E-M1 that I can't transfer to my computer because I've now run out of storage space. Time to buy a new hard drive!

 

Well that isnt quite the reality is it? For instance if you had a Nikon D810 with a 70-200 2.8 and you shot it in APSC mode you would effectively have a 300 FOV @f2.8.

 

I always though that the end goal was to capture the scene as envisioned and obtain the best possible output quality. To that end, ISO is most definitely effected by sensor size.

 

No, the 'equivalence rubbish' as you called it states that the same exposure and same imaging characteristics (noise, DR, DoF) can be obtained with a larger sensor by using a longer lens, a narrower aperture, a higher ISO and the same shutter speed. And by and large with similar generation sensors, that has proven true. A few underperform (Panasonic 10MP 4/3) and a few over perform (Sony 20MP 1") but on balance the numbers DPR cites match up quite closely with the reality.

 

"Best Possible Output Quality" is subjective.

Amazing output has been made with m43, and mediocre output has been made with FF.

It seems that what REALLY determines the best possible output quality, if you want to take that objectively, is the drive of the photographer/editor and the PP knowledge they have to get what they need out of whatever sensor or film they're working with.

 

ISO is meaningless on its own when it comes to final image quality. What matters is signal to noise. And signal to noise is a function of both ISO and sensor size.

I have no interest in FF and find the IQ of my m43 system to be excellent and more than I need, but I understand basic math, exposure and image quality. You are using arbitrary and frankly rather senseless measures simply because they correspond to dials on your camera. What most photographers care about is not numbers, but the final image. And within that image they typically care about:

1. Angle of view
2. Depth of field
3. Motion blur (shutter speed)
4. Noise/grain (a *combination* of ISO and sensor size)

You've decided the number ISO, which on its own is a meaningless number, is some how significant. It is completely useless on its own when it comes to characterizing what matters to a photographer - the final image. In the final image you *must* consider ISO and sensor size together.

 

Do you feel better now that you've let that out?

 

Yup. Basic physics. But what is missing from this discussion is the fact that below some value, noise is negligible . As the ISO is increased, the larger sensor will begin to outperform the smaller sensor simply because it is capturing more photons (signal). As a practical matter, however, most of my (GX7) shooting is at ISOs where noise is negligible for me. So I am not too concerned by my sensor's smaller size. And, of course, many of the larger-sensor cameras out there are older sensor technology. Against those my GX7s might even run neck and neck.

Looked at another way, we might be wanting to open up our shorter-focal-length lenses by a stop compared to larger-sensor cameras just to get equivalent DOF. In that case we will be shooting at a lower ISO so noise differences are reduced as well.

If I shot a lot of moving subjects in dim or dark contexts, I would probably be more concerned about this issue.

 

I am just happy that faster f-stops help with light gathering and compensates for the smaller sensor. In all honesty I could never get a decent focus shot with a FF 1.2 lens, but it works great on m43. Plenty thin DoF for me Plus my nocticron has OIS so that helps it further in light gathering vs a FF lens that is not stabilized and a bit slower. Granted the 42.5 is probably more expensive than such a lens so we may be paying less for camera but more for lenses. Stll, even the nocticron isn't that big and with other lenses I retain size advantage. In my view putting money in lenses rather than cameras is also a good strategy.

 

There is Angle of View equivalence and it will tell you something about "magnification". Aperture equivalence will tell you something about DOF (and noise).
The problem with the "aperture equivalence" is that aperture is relevant for exposure too and so you should declare two apertures, something like:

Oly 12-40 2.8 -> 24-80 aov equiv, 5.6 dof equiv, 2.8 exposure equiv

And this is going to really confuse beginners who are still trying to grasp the exposure triangle.

The other "trick" of the "aperture equivalence" is that it can be also used for noise comparison. It will somehow tell you that the 1600 iso of a FF is not the same 1600 of a m43 because the "equivalent aperture" is different and so you need four times the shutter speed to get the same noise level (approximately). But this also means to completely ignore how shutter speed affects your pictures, and this is where this thing gets too messy for me.

I like the AOV equivalence, it's handy and easy. The dof equivalence is real, good to know, but IMO too confusing and of relative value to be promoted. The 35mm is as good as anything else as a common reference, so be it.

 

The equivalence stuff is only useful if you are used to other systems and need a frame of reference to compare fields of view. The rest is nonsense that is used to compensate. For m4/3 users as my lens I better then it seems or ff guy as I can pretend my photos are better because only the eye is in focus and everything else is a blur.

 

Or a 512GB SD card

 

The key point is that for equal FoV, equal DoF and equal shutter speed you will shoot the larger sensor at a higher ISO. In that sense the larger sensor has bought you nothing when it comes to final image quality. Yes the sensor is bigger, but you had to shoot a smaller aperture for equal DoF and to get the same shutter speed you had to increase the ISO. And as you point out some FF sensors don't perform as well area for area as smaller sensors (though the newer Sony sensors in the D810 and A7r do appear to do so, but the Canon do not).

What FF does is give you some more options once you run out of equivalent adjustments. If shooting in very low light you have the option to go with even shallower DoF on FF thus getting the shutter speed you want with a larger sensor at the same ISO as the smaller sensor. You'll have lower noise in this case, but you are in fact no longer "equivalent" - you have to deal with shallower DoF to get this noise advantage.

Similar case, at base ISO the FF has a noise advantage over the m43 because the sensor is larger and they are at the same ISO. But again, at this point "equivalence" has been broken, you'll have to shoot the FF camera either with different DoF or different shutter speed to realize this noise advantage. And of course only a very small group of people even care about base ISO noise to begin with, but it can be an issue for landscape photographers printing very large where noise in blue skies can be noticeable even at low ISO.

So really the "advantage" of a larger sensor really only comes into play when shot in a "non-equivalent" manner. The larger sensor just gives you more options. When shot truly "equivalently" it doesn't have any particular advantage.

And this is nothing new, has been true for decades with film that follows the exact same scaling laws. Of course there we didn't even have sensor technology differences, the emulsion for a given ISO was the same whether cut to 35mm or 4x5. So "equivalent" was even more "equivalent" than it is now!