Announcement Soon: Canon RF 16mm f/2.8 and Canon RF 70-400mm f/5.6-7.1 IS USM

neuroanatomist

I post too Much on Here!!
CR Pro
Jul 21, 2010
25,084
3,047
…so many people get it wrong…
Indeed.

To say f/2.8 on full frame is "equivalent" to f/4.5 on APS-C *might* be valid if your context is DoF, with all other factors being the same. But the amount of light the lens transmits to the sensor per unit area does not change, so in terms of exposure calculation for low-light scenarios, the lens is f/2.8 for either sensor.
That’s technically correct, but ignores the effect of sensor size on perceived noise, i.e., the concept of equivalence also applies to ISO.

In practice, that means that if using the same exposure settings for a lens on FF and APS-C, the image from the latter will have more noise. Conversely, with a FF sensor one could choose a lower ISO setting for less noise, a faster shutter speed with the same noise, or a narrower aperture with the same noise to match the deeper DoF.

Larger sensor: equivalent…but better.
 
  • Like
Reactions: mpphoto and AlanF

SnowMiku

EOS M6 Mark II
Oct 4, 2020
81
53
To clear up any confusion, this is the aperture range of the Canon RF 100-500mm F4.5-7.1 L IS USM from the-digital-picture.com

Aperture Max by Focal Length
100-150mm = f/4.5
151-253mm = f/5.0
254-362mm = f/5.6
363-471mm = f/6.3
472-500mm = f/7.1
 

neuroanatomist

I post too Much on Here!!
CR Pro
Jul 21, 2010
25,084
3,047
To clear up any confusion, this is the aperture range of the Canon RF 100-500mm F4.5-7.1 L IS USM from the-digital-picture.com

Aperture Max by Focal Length
100-150mm = f/4.5
151-253mm = f/5.0
254-362mm = f/5.6
363-471mm = f/6.3
472-500mm = f/7.1
However, note that if you set the camera to 1/2-stop increments instead of 1/3-stop, the 100-500 at 400mm is recorded at f/5.6, meaning they actual iris diaphragm diameter at 400mm corresponds to an f/number somewhere between f/5.6 and f/6.3…which is exactly where it is when the 100-400 is at 400mm.
 

unfocused

EOS-1D X Mark III
Jul 20, 2010
6,025
3,341
67
Springfield, IL
www.mgordoncommunications.com
...To say f/2.8 on full frame is "equivalent" to f/4.5 on APS-C *might* be valid if your context is DoF, with all other factors being the same. But the amount of light the lens transmits to the sensor per unit area does not change, so in terms of exposure calculation for low-light scenarios, the lens is f/2.8 for either sensor.

Without that context, a blanket statement like "f/2.8 on full frame is equivalent to f/4.5 on crop" is meaningless.

Good luck with this. I've heard these statements for years and you just can't knock any sense into people. By the way, even if your context is DoF, it's not really accurate. As I've explained many times on this forum, DoF is dependent on aperture and distance from the subject, not sensor size. Take any lens and put it on a crop sensor camera and a full frame camera. Take a picture of any subject from the exact same spot at the exact same focal length and exact same f-stop. Once you crop the full frame image to the same size as the crop sensor image, the depth of field, and indeed the framing, etc., is identical. The only difference is that in one case, you are cropping in-camera and in one case you are cropping in post.
 

unfocused

EOS-1D X Mark III
Jul 20, 2010
6,025
3,341
67
Springfield, IL
www.mgordoncommunications.com
...That’s technically correct, but ignores the effect of sensor size on perceived noise, i.e., the concept of equivalence also applies to ISO.

In practice, that means that if using the same exposure settings for a lens on FF and APS-C, the image from the latter will have more noise....

Sort of. Back in the days of the 18mp APS-C sensors this was true. But with modern sensors the differences are becoming less and less apparent in practical use, especially at lower ISOs. Just as the small fraction of light difference at the various focal length ranges of the 100-400 vs. 100-500 has little to no practical effect, the small noise differences between full frame and APS-C sensors today are of little practical impact at many ISOs, especially if you are competent at processing the images.
 
  • Like
Reactions: mdcmdcmdc

neuroanatomist

I post too Much on Here!!
CR Pro
Jul 21, 2010
25,084
3,047
Good luck with this. I've heard these statements for years and you just can't knock any sense into people. By the way, even if your context is DoF, it's not really accurate. As I've explained many times on this forum, DoF is dependent on aperture and distance from the subject, not sensor size. Take any lens and put it on a crop sensor camera and a full frame camera. Take a picture of any subject from the exact same spot at the exact same focal length and exact same f-stop. Once you crop the full frame image to the same size as the crop sensor image, the depth of field, and indeed the framing, etc., is identical. The only difference is that in one case, you are cropping in-camera and in one case you are cropping in post.
Under 'equivalence' the effect regarding aperture is on DoF for the same subject framing, i.e. the distance from the subject is not the same because the crop factor results in a different FoV for the same focal length.

It's more logical that way, for example consider a headshot with an 85mm EF lens – if you pick your distance with the FF camera and frame the shot, then switch to the APS-C camera and stay in the exact same spot, sure if you cropped the FF image to the APS-C framing and compared them the DoF would be the same...but then you'd have an eyes-and-nose shot, not a headshot. So of course you'd back up when switching from APS-C to FF.
 

neuroanatomist

I post too Much on Here!!
CR Pro
Jul 21, 2010
25,084
3,047
Sort of. Back in the days of the 18mp APS-C sensors this was true. But with modern sensors the differences are becoming less and less apparent in practical use, especially at lower ISOs. Just as the small fraction of light difference at the various focal length ranges of the 100-400 vs. 100-500 has little to no practical effect, the small noise differences between full frame and APS-C sensors today are of little practical impact at many ISOs, especially if you are competent at processing the images.
It is objectively true, but practically of no significance at low-to-mid ISO settings. If you're at ISO 6400 it's going to make a difference, regardless of your post-processing.
 

H. Jones

Photojournalist
Aug 1, 2014
695
1,326
However, note that if you set the camera to 1/2-stop increments instead of 1/3-stop, the 100-500 at 400mm is recorded at f/5.6, meaning they actual iris diaphragm diameter at 400mm corresponds to an f/number somewhere between f/5.6 and f/6.3…which is exactly where it is when the 100-400 is at 400mm.
At this point I'm pretty sure we could all scream this from the rooftops and still get the same braindead "hurrdurr Canon ruined the 100-500" despite the fact it's basically the same lens. I appreciate what Canon is doing and all the ways they're breaking the mold and trying something new, but it's frustrating the way the internet totally misses the point because they're stuck in the lens design of the 1990s.

It makes me realize what an uphill camera companies are dealing with when it comes to marketing, and the reason why some companies out there just decide to fudge the numbers on this stuff.

It's crazy how far detached from the reality of actually using the equipment some people are. In college I made perfectly fine photographs out of pinhole cameras made of tissue boxes and aluminum foil, but somehow people will fall on a sword for (not-even) a third of a stop in a highly advanced lens that is even more capable than its predecessor.
 

neuroanatomist

I post too Much on Here!!
CR Pro
Jul 21, 2010
25,084
3,047
It makes me realize what an uphill camera companies are dealing with when it comes to marketing, and the reason why some companies out there just decide to fudge the numbers on this stuff.
Not new. Here's a Panasonic bridge camera from 2012, with a 25-600mm f/2.8 zoom lens!

Screen Shot 2021-06-23 at 11.47.11 PM.png


Leica made the lens, and on the front of the barrel it correctly indicates that it's a 4.5-108mm f/2.8 lens.
 
  • Like
Reactions: H. Jones

AlanF

Stay at home
CR Pro
Aug 16, 2012
8,312
10,456
Not new. Here's a Panasonic bridge camera from 2012, with a 25-600mm f/2.8 zoom lens!

View attachment 198531

Leica made the lens, and on the front of the barrel it correctly indicates that it's a 4.5-108mm f/2.8 lens.
@mpmark who wrote I was mistaken when I wrote that a 500mm f/7.1 lets in as much light as a 400mm f/5.6 would by his reasoning claim that a 4.5-108mm f/2.8 lens lets in more light than a 25-600mm f/4 lens. In fact, it lets in the amount of light of an f/16 25-600mm lens. As anyone who has used one of those cameras with a tiny sensor and apparently wide lens knows, you have to use very low isos before noise becomes intrusive. A 400mm f/5.6 in terms of iso is equivalent to a 500mm f/7.1 or 1000mm f/14 when a crop from the shorter is viewed at the same size as from the larger.
 

unfocused

EOS-1D X Mark III
Jul 20, 2010
6,025
3,341
67
Springfield, IL
www.mgordoncommunications.com
What do you mean by that comment?
The Fact is that exposure is determined by time, aperture and ISO speed. Anyone who has ever used a light meter knows that there is no setting on the meter for the focal length of the lens or the size of the film (sensor today). No matter what focal length lens you use, the meter will give you the same exposure setting. And, if you are using 135mm, 120, 4x5, 8x10 or even 110, the meter gives you the same exposure based on the light in the scene. Same thing with sensors.
 
  • Like
Reactions: melgross

AlanF

Stay at home
CR Pro
Aug 16, 2012
8,312
10,456
The Fact is that exposure is determined by time, aperture and ISO speed. Anyone who has ever used a light meter knows that there is no setting on the meter for the focal length of the lens or the size of the film (sensor today). No matter what focal length lens you use, the meter will give you the same exposure setting. And, if you are using 135mm, 120, 4x5, 8x10 or even 110, the meter gives you the same exposure based on the light in the scene. Same thing with sensors.
It is indeed a fact that exposure is determined by shutter speed, aperture and ISO speed. It is also a fact that the signal to noise of a digital image is determined by the number of photons hitting the sensor, its efficiency of conversion of those photons into electronic current and the noise introduced by the circuitry, and the final noise in the image depends also on the size of enlargement. - your lightmeter doesn't tell you that. A 108 mm f/2.8 lens has the same diameter as a 600mm f/16 lens and so both let in the same number of photons and have the same amount of photon noise. The lightmeter tells you that with the "bright" small lens that you can use a low iso with a particular shutter speed but 4 stops more iso with the f/16 lens. However, when you view the resultant images from the short and long lenses at the same size, they will both have the same signal to noise as the iso noise introduced by a good amplifier circuit will be the same as the noise introduced by enlarging the smaller image from the shorter focal length lens. And, that is what I was getting at - a 500mm f/7.1 is no noisier than a 400mm f/5.6 when you view the image at the same size, despite having to use a higher iso. This is where Canon has been clever - the 800mm f/11 lens is criticised for its high f-number but it can be used at high isos satisfactorily and has the same diameter as a 400mm f/5.6.
 

neuroanatomist

I post too Much on Here!!
CR Pro
Jul 21, 2010
25,084
3,047
The Fact is that exposure is determined by time, aperture and ISO speed. Anyone who has ever used a light meter knows that there is no setting on the meter for the focal length of the lens or the size of the film (sensor today). No matter what focal length lens you use, the meter will give you the same exposure setting. And, if you are using 135mm, 120, 4x5, 8x10 or even 110, the meter gives you the same exposure based on the light in the scene. Same thing with sensors.
The Fact is that you don’t really understand equivalence.
 

unfocused

EOS-1D X Mark III
Jul 20, 2010
6,025
3,341
67
Springfield, IL
www.mgordoncommunications.com
The Fact is that you don’t really understand equivalence.
The fact is these arguments have been going around and around for years on this forum. People bend themselves into pretzels to "prove" convoluted, esoteric "points" (See the previous post by @AlanF).

"Equivalence" is too often synonymous with fuzzy thinking that allows people to confirm their own biases.

What is objectively true is that an exposure of f5.6 at 1/250 sec at ISO 400 is not going to change no matter the focal length of the lens or the size of the recording medium.

Now, if you want to argue that the grain in a 110 film image is going to be greater than the grain using a 4x5 view camera when both have an ISO of 400, well, that's a different matter. Or, if you want to argue that depth of field changes when you change position to get the same view with a smaller sensor that you would get with a larger sensor, that's true as well.

But statements like "a 500mm f/7.1 lets in as much light as a 400mm f/5.6" are plain silly. Okay, maybe he figures that since both lenses have the same diameter at the front end, they are "letting in" an equal amount of light. But that has zero use in the real world. What matters is the amount of light that will actually hit the sensor and an f7.1 lens is going to deliver less light to the sensor than an f5.6 lens, regardless of the focal length of the lens or the size of the sensor.
 
  • Like
Reactions: melgross and SteveC

SteveC

R5
CR Pro
Sep 3, 2019
2,348
2,192
But statements like "a 500mm f/7.1 lets in as much light as a 400mm f/5.6" are plain silly. Okay, maybe he figures that since both lenses have the same diameter at the front end, they are "letting in" an equal amount of light. But that has zero use in the real world. What matters is the amount of light that will actually hit the sensor and an f7.1 lens is going to deliver less light to the sensor than an f5.6 lens, regardless of the focal length of the lens or the size of the sensor.

Or to come at it from another direction, one more directly targeted to the example you debunk, let's talk about two different lenses with, specifically, the same entrance pupil, a 50 mm one (to pick a number out of thin air). If the lens has a 100mm focal length, it will put four times as much light onto the sensor as it would were it 200mm; that's because the first one is f/2.0 and the second is f/4.0. Yes the same amount of light passes the filter threads.

As you rightly point out, so what?

There is a reason lenses are rated by f/ ratio rather than the raw diameter of the entrance pupil. It's because the former can be directly used to determine exposure, while the latter requires you to compute the former before you can determine exposure.

(I had to sit down a couple of years ago and think through why the ratio should matter, once I did I resolved to always write "f/2.8" rather than just "2.8". It for one makes it clear the number is a denominator, so it therefore makes sense for what it represents to get smaller as the bare number itself looks bigger.)
 

privatebydesign

I post too Much on Here!!
CR Pro
Jan 29, 2011
10,306
5,443
iso is sensitivity, or the ability of the sensor to collect light.

I don't understand why people find it hard to accept that a sensor at the same iso/sensitivity but half the size collects half the light.

If you talk about equivalence you cannot do it without accepting an image can only be 'equivalent'/equal/identical if it has the same fov, dof, subject movement, AND noise (number of photons).

I made this for this forum back in 2012....

Screen Shot 2021-06-24 at 23.35.12.png
 
  • Like
Reactions: Maximilian

neuroanatomist

I post too Much on Here!!
CR Pro
Jul 21, 2010
25,084
3,047
The fact is these arguments have been going around and around for years on this forum. People bend themselves into pretzels to "prove" convoluted, esoteric "points" (See the previous post by @AlanF).

"Equivalence" is too often synonymous with fuzzy thinking that allows people to confirm their own biases.
Your use of words like convoluted, fuzzy and silly merely confirms that you don’t comprehend the concepts being discussed.

Sure, people who do not understand a concept can misuse it to support their biases. That doesn’t negate the concept itself, for those who do understand it.

What is objectively true is that an exposure of f5.6 at 1/250 sec at ISO 400 is not going to change no matter the focal length of the lens or the size of the recording medium.
True. Was someone suggesting that is not the case? If so, I missed it. But the fact that the exposure settings don’t change doesn’t mean the resulting images will be identical if focal length or sensor size are changed.

Now, if you want to argue that the grain in a 110 film image is going to be greater than the grain using a 4x5 view camera when both have an ISO of 400, well, that's a different matter. Or, if you want to argue that depth of field changes when you change position to get the same view with a smaller sensor that you would get with a larger sensor, that's true as well.
Yes, and of course we also can stipulate other facts that no one is arguing, like force equals mass times acceleration and the existence of electromagnetic radiation.

But statements like "a 500mm f/7.1 lets in as much light as a 400mm f/5.6" are plain silly. Okay, maybe he figures that since both lenses have the same diameter at the front end, they are "letting in" an equal amount of light. But that has zero use in the real world.
That statement is a fact (given a willingness to accept an approximation). 500mm / 7.1 = 70.4mm, 400mm / 5.6 = 71.4mm. You think it’s ‘plain silly’ because you fail to understand it’s significance.

What matters is the amount of light that will actually hit the sensor and an f7.1 lens is going to deliver less light to the sensor than an f5.6 lens, regardless of the focal length of the lens or the size of the sensor.
The point to which @AlanF was replying was a claim that, “In the end the 100-400 does let in more light, comparing it to the 100-500 is misleading,” and that, “…500 f/7.1 is ridiculously slow and useless for [his] photography.”

The fact is 100-500 at 500mm f/7.1 is letting in about the same amount of light as the 100-400 at 400mm f/5.6. Yes, 500mm f/7.1 delivers less light to the sensor than 400mm f/5.6 (because of the longer focal length, as @SteveC points out, that’s why we use the f/number ratio). So to match shutter speeds you raise the ISO slightly. Alan’s point was that if you crop the 400mm image to the FoV of the 500mm lens, the noise in the resulting picture will be the same as the noise in the picture taken at 500mm f/7.1 when you compare them (that’s another example of equivalence, something you should really try to understand).