Here is the Canon RF 100-400mm f/5.6-8 IS USM

justaCanonuser

Grab your camera, go out and shoot!
Feb 12, 2014
1,025
923
Frankfurt, Germany
TC compatibility, I wonder if that’s limited like the 100-500. Regardless…1-2 stops lost from f/8 is dark.
A 2x TC e.g. with that lens would work well on sun's surface. Brings me to astro (and particle) physics: they are striving so hard to solve the riddle of Dark Matter, but they only have to check Canon's new RF lens list now... ;)
 
  • Like
Reactions: 1 user
Upvote 0
The *total* light, integrated across the area of the FF sensor at f/8, is indeed greater than the crop sensor at f/5.6, in terms of exposure, f/8 is still f/8. If you shoot the same scene at the same time and the same ISO with both cameras set to f/8, they will both require the same shutter speed.

What you're saying is equivalent to saying "The sun is brighter at my neighbor's house because he has a bigger yard."

Yes... and no. If all you care about is getting equivalent brightness and you ignore all other aspects of the image then sure, f/8 is f/8 is f/8.

(Strap yourselves in, kids, class is in session.)

But 1) not all cameras produce the same brightness at the same ISO setting, and 2) smaller sensors produce more noise at the same ISO (assuming the cameras are of a similar age and resolution), so you're not ending up with equal quality.

Even if we ignore #1 for now, equalising #2 requires you to get more light onto the smaller sensor so a lower ISO can be used. (Or less light onto the larger sensor, to use a higher ISO.) We do this by changing the aperture, since changing the shutter is usually more destructive and causes other differences in the resulting images, while changing the aperture also allows us to equalise the depth of field at the same time. The difference in both depth of field and noise between 35mm and APS-C tends to be roughly one stop and a quarter, so it's an easy compensation to make by just changing the aperture by one stop. (And for 4/3rds cameras to 35mm, it's two stops.)

In this sense, f/8 is no longer f/8 when moving to a different sensor. 1/100 f/8 ISO 800 on a 35mm sensor and 1/100 f/8 ISO 800 on an APS-C sensor may give you the same brightness and motion blur/freezing, but that APS-C shot will be noiser and have a deeper depth of field. If we change the APS-C camera to 1/100 f/5.6 ISO 400, now we've got the same brightness, (roughly) equivalent depth of field, and (roughly) equivalent noise.

So in a total equivalent sense, f/8 is not f/8 when you're talking about different sensor formats. Aperture effectively changes just as much as focal length does. ("f/8", after all, does literally mean 'focal length divided by 8', and if you paid attention in algebra class you know that you can't change 'f' without the rest of the maths changing, too; 'equivalents' and 'crop factor' are applied to the whole equation, not just the focal length.)


To use your garden analogy, yes, having a bigger garden does not mean the sun itself is brighter, but it does mean that garden recieved more sunlight. On average throughout the year, in one minute and per square meter, the amount of sunlight that hits the ground is 6kWh/m². If someone's garden is four square meters then in one minute their garden recieved (or saw, caught, or experienced, or whatever other term you care to use) 24kWh; if their neighbour's garden is eight square meters then that garden, in the same time, recieved 48kWh of sunlight. It's double the area so double the sunlight landed on it, even though the sun itself was putting out light at the same rate.


A garden or a sensor, it's the same thing. It's not that the light is brighter, but that that more of the light is able to be captured. The smaller sensor captures less of the light in the first place (and is less light-efficient at the same resolution), so the signal has to be boosted more and hence you end up with more noise. Open up the aperture, let more light in, less boosting (ISO) is needed, and now everything is nice and even.

But then there's point #1 earlier, which causes big problems when comparing different brands, since every manufacturer rates ISOs differently, and in particular Canon and Fuji cameras are sometimes actually shooting at as little as half the stated sensitivity. So, in some ways, the biggest problem with getting equal exposure between cameras isn't the aperture, but the fact that ISO 100 simply is not always ISO 100.
 
Upvote 0
f/8 is beyond the diffraction limit for a 45MP sensor!
... yepp, and f/11 primes even more! That's why the R6 is the smarter choice for such lenses.

Kind of, but not really. The differences in diffraction limitations only apply when looking at files at their full sizes, comparing each pixel 1:1. If you downsample the larger files size (45mp, in this case) to match the smaller size it is being compared to (20mp) then the diffraction limitation also scales down and equalises. It's the exact same as with noise.

Also, some lenses showcase diffraction worse than others. Lenses which produce higher contrast can typically get away with pushing past the diffraction limits better than a low contrast lens; a lot of what we perceive as 'sharpness' is actually contrast rather than resolution. In the case of the 600 and 800 f/11s their contrast is good enough that even without downsampling, the R5 files show slightly more clearly-defined detail than the R6's. Downsample the R5 to match the R6 and the results are noticeably better. That the R5 is more than double the resolution of the R6 means you can crop in by about 1.2x, downsample that cropped file, and still match the R6's whole file in regards to diffraction specifically.

The R6 is a great camera, but there's no way around the physics; maths is maths.
 
  • Like
Reactions: 1 user
Upvote 0
Jul 21, 2010
31,088
12,851
No, that's not how it works.

The *total* light, integrated across the area of the FF sensor at f/8, is indeed greater than the crop sensor at f/5.6. But in terms of exposure, f/8 is still f/8. If you shoot the same scene at the same time and the same ISO with both cameras set to f/8, they will both require the same shutter speed.

What you're saying is equivalent to saying "The sun is brighter at my neighbor's house because he has a bigger yard."
Sigh. Why is the concept of equivalence so hard to understand? The short version of @aceflibble‘s class is that’s exactly how it works, because noise is proportional to total light integrated across the sensor. So on FF, you can use a 1.3-stop higher ISO for the same noise.

Another way to look at that is you can shoot at f/8 on both sensors, but for the same noise level you can use a higher ISO on FF to set a 1.3-stop faster shutter speed.
 
  • Like
Reactions: 1 users
Upvote 0

mdcmdcmdc

EOS R7, M5, 100 (film), Sony α6400
CR Pro
Sep 4, 2020
318
442
Sigh. Why is the concept of equivalence so hard to understand? The short version of @aceflibble‘s class is that’s exactly how it works, because noise is proportional to total light integrated across the sensor. So on FF, you can use a 1.3-stop higher ISO for the same noise.

Another way to look at that is you can shoot at f/8 on both sensors, but for the same noise level you can use a higher ISO on FF to set a 1.3-stop faster shutter speed.

Sigh. Why is the concept of reading so hard to understand?

The original statement was:

f8 full frame is brighter than f5.6 cropped.

That statement is incorrect. The total amount of light gathered by the FF sensor may be greater, but that's not meaningful for photography. Individual photosites (pixels) gather the light that reaches their own area. That doesn't change if the total sensor size is APS-C, FF, 8x10, or the size of an American Football field. It is the light reaching the individual photosite that matters for exposure.

Other statements about noise and DOF are besides the point because that's not what @JohnOnTheNet said.

If you want to turn this into a "why I hate crop" discussion, have fun. I won't be drawn in.
 
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,342
22,517
Fair enough then, but given my experience with the 100-500 on TCs being weaker than the 800 f11, I'm going to put this down to copy variation and say you can't gurantee results. It could be that you got an unusually good 100-500 or TC, or an unusually poor 800, or I hand my hands on an unusually good 800 or an unusually bad 100-500. Unless someone can test a sample group of at least ten copies of both lenses and TCs, we can't chalk this up as any kind of consistent trend.
I have analysed carefully two copies of the RF 800 f/11 on the R5. Both had the same IQ, and both better than my 400mm DO + 2xTC @f/8 on my R5 and 5DSR, and the digital picture has the 800 f/11 similarly sharper than the 400mm DO + 2xTC @f/8 on a 5DSR, so we can rule out that I had a bad copy of the 800mm f/11. I used the same RF 1.4x and 2xTCs on the 800mm and the 100-500mm, so the results are in parallel and should rule out any effect of a bad TC - in any case the TCs work really well. So that leaves my having an unusually good copy of the 100-500mm or you having a bad copy.
If you say we can't chalk this up until we have sampled 10 copies of both lenses and TCs, then you should practice what you preach and withdraw:

That, plus I've never seen a zoom lens on even just a 1.4x TC which is as good optically as the two f/11 primes are bare, except for the 200-400 which has a custom 1.4x built-in and optimised for that specific optic. Even then, by the time you either add another 1.4x on the back of it, or skip the built-in extender and use a 2x TC, it ''only'' matches the f/11 primes in optical quality, and the focus gets slower.
until you have documentary evidence on 10 copies of each.
 
Upvote 0
Jul 21, 2010
31,088
12,851
The total amount of light gathered by the FF sensor may be greater, but that's not meaningful for photography.
Lol. Did you really suggest that ‘total light gathered is not meaningful for photography’?!? That statement deserves an award for oxymoron of the month.

Sorry, but if that’s your level understanding then you need to engage in some self-education before trying to have a meaningful discussion on the concepts under consideration.

Obviously, you’re correct about light per unit area determining the exposure. But exposure is not the sum total of photography.

If you want to turn this into a "why I hate crop" discussion, have fun. I won't be drawn in.
It’s not about hating crop sensors. That would be pretty silly for me since I have more cameras with APS-C sensors than with FF sensors. It’s about understanding the implications of sensor size for photography, which evidently you don’t.
 
  • Like
Reactions: 1 user
Upvote 0
zoom range and IS are very rarely of any use; what are you photographing at 400mm that you can use a shutter below 1/500th with, and when are you not going to want the maximum focal length? And if you're photographing anything which might move (which is 99% of super-tele use) then your panning can screw up stabilisation systems anyway, even in the panning modes, hence why you'll find many pros just keep the IS off entirely.
You talk a lot of sense but I'd just like to sound a note of dissent here. The reason I didn't get on with the 400L f/5.6 was precisely because it lacked IS; I found it made a big difference. Admittedly that was on the 50D, so you have the crop magnifying camera shake at a given shutter speed, its high ISO capabilities were fairly limiting, and my technique was still developing back then.
But I've shot (perched) birds at shutter speeds well below 1/500, and IS can be invaluable. Of course I never used a monopod, and rarely a tripod. Different styles but IS can absolutely be of value for bird photography. PS I've never found it interfered in the way you describe, maybe I've been lucky.
 
Upvote 0
I like this lens in principle, hopefully the image quality is adequate. When I started out in long lenses on a budget, I mounted the EF 70-200 f/4 with a 2x TC, and it was handy; although my standards are more stringent now, 400 f/8 doesn't worry me.

I'd just like to repeat something I've said on a number of threads where people have asserted f/7.1-8-11 are too dark for bird photography in anything but bright sunshine: that is not true. My usual setup in the last few years has been 1000mm f/10, and except on very dull overcast winter days, at dusk, or in deep shade it works well - and this is on bodies whose noise characteristics was poorer than today's models. Sure, f/4 or f/2.8 will give cleaner images, and especially less cluttered backgrounds, but you absolutely don't need them to photograph birds in Britain or in woodland or cloudy weather.
 
  • Like
Reactions: 1 user
Upvote 0
The point being that aperture is a ratio, not a size. Canon APSC is less than half the area of full frame, so one stop smaller aperture on full frame allows more than twice the light onto the sensor and consequently halves the effective ISO (noise.) My main complaint with crop sensor cameras is the lack of dedicated crop sensor lenses for them. If I attach my EF100-400 lens on my M6, I'm using less than half of the lens and it matters little how light the camera is because the lens is heavy. My EFS 55-250 STM is brilliant on that camera, but there was no EFS 100-400 or similar. A crop sensor EOS R would be a waste of time without good crop sensor R lenses. This is why Canon id making full-frame lenses with smaller apertures. All things being equal, a f8 full frame will give a similar result to a f5.6 crop sensor. A lot of people want lighter and cheaper lenses and will like this. Other people want heavier and more expensive lenses and they will buy something else. Canon will be only too happy to oblige. Heavy and expensive lenses have a lot going for them - except that they are heavy and expensive.
 
Upvote 0
If you say we can't chalk this up until we have sampled 10 copies of both lenses and TCs, then you should practice what you preach and withdraw:
No, because I rather thought it was common sense that anyone commenting here would be doing so based on their subjective and anecdotal experience, and trying to 'out-lab' each other to prove... I'm not sure what point you're even going after... is childish. The whole point I was bringing up about testing a large number of copies was to highlight that nobody here, no matter how much they may insist they have the one definitive test of a lens, actually does get hands-on with enough units for what they say to ever be anything more than subjectiove interpretation and copy variation, which, as I said, should simply be common sense anyway and shouldn't need to be explained to you multiple times.
 
Upvote 0

H. Jones

Photojournalist
Aug 1, 2014
803
1,637
The point being that aperture is a ratio, not a size. Canon APSC is less than half the area of full frame, so one stop smaller aperture on full frame allows more than twice the light onto the sensor and consequently halves the effective ISO (noise.) My main complaint with crop sensor cameras is the lack of dedicated crop sensor lenses for them. If I attach my EF100-400 lens on my M6, I'm using less than half of the lens and it matters little how light the camera is because the lens is heavy. My EFS 55-250 STM is brilliant on that camera, but there was no EFS 100-400 or similar. A crop sensor EOS R would be a waste of time without good crop sensor R lenses. This is why Canon id making full-frame lenses with smaller apertures. All things being equal, a f8 full frame will give a similar result to a f5.6 crop sensor. A lot of people want lighter and cheaper lenses and will like this. Other people want heavier and more expensive lenses and they will buy something else. Canon will be only too happy to oblige. Heavy and expensive lenses have a lot going for them - except that they are heavy and expensive.

All great points here, and exactly why this lens exists.

My daily go-to for wildlife is 560mm F/8 from my 100-400 + 1.4x teleconverter, and it hasn't bothered me enough to actually purchase bigger glass yet. I borrow the 600mm F/4L from Canon time to time, but when you use those huge lenses, it really increases the labor involved in a photo outing.

Not a big deal when it's an important subject and you have a purpose in mind, but if I'm taking a hike to cool off after a long week at work, it's a lot more enjoyable to just grab the 100-400 and teleconverter, throw it in a smaller camera bag, and snap some images. I also find that my reaction time is a lot better with the smaller set-up, and that often gets me images I wouldn't have normally gotten with the bigger glass.
 
  • Like
Reactions: 1 user
Upvote 0

justaCanonuser

Grab your camera, go out and shoot!
Feb 12, 2014
1,025
923
Frankfurt, Germany
Kind of, but not really. The differences in diffraction limitations only apply when looking at files at their full sizes, comparing each pixel 1:1. If you downsample the larger files size (45mp, in this case) to match the smaller size it is being compared to (20mp) then the diffraction limitation also scales down and equalises. It's the exact same as with noise.
I know, I am a physicist, I know about diffraction and Airy discs. That's why I would recommend to select an R6 if you use such slow lenses. Otherwise you'd get only bigger files with no plus of useful information you then can downsample. Not smart if you can have camera that does this job anyway.
 
Upvote 0

AlanF

Desperately seeking birds
CR Pro
Aug 16, 2012
12,342
22,517
I know, I am a physicist, I know about diffraction and Airy discs. That's why I would recommend to select an R6 if you use such slow lenses. Otherwise you'd get only bigger files with no plus of useful information you then can downsample. Not smart if you can have camera that does this job anyway.
As you are a physicist, you might be interested in that I calculated the MTF values of the different slow lenses based on the Airy diffraction and the sensor MTFs and posted here: https://www.canonrumors.com/forum/t...of-f-5-6-f-7-1-and-f-11-lenses-and-tcs.39118/ (I didn't allow for lens aberrations or the Bayer filter).
It's true that you see more advantage with lower resolution sensors of increasing focal length at expense of f-number. I found from actual measurements of putting the 2xTC on the RF 100-500mm and R5 that the resolution increased by ~40% on going from 500mm f/7.1 to 1000mm f/14.
 
Last edited:
  • Like
Reactions: 1 users
Upvote 0

stevelee

FT-QL
CR Pro
Jul 6, 2017
2,383
1,064
Davidson, NC
Nice. Too bad all my Canon gear is EF mount. Eventually I suppose I'll be switching to Fuji or Sony. Sad to see Canon essentially abandon long time EF users as the EF lens portfolio shrinks.
I have all the EF lenses that I need, so it doesn’t matter to me whether they make more. If I bought another lens, it would be the 24mm TS-E, but I really would not use it often enough to buy it. I have rented it before, and may well again if some project presents itself.

I will continue to use those lenses with my DSLR. I have concluded that if I spend $6,000 on a camera, I will go to the 102 MP Fujifilm. I would need to define better how and when I use it before I can decide what lens(es) to buy with it. Probably a good strategy would be for me to shoot more landscapes with my current gear. I might get a better sense of how I might use different focal lengths. I might also discover that what I have now is more than adequate for my purposes. So far I haven’t perceived any unmet needs that R series cameras and lenses would solve for me.
 
Upvote 0

mdcmdcmdc

EOS R7, M5, 100 (film), Sony α6400
CR Pro
Sep 4, 2020
318
442
Let's review. JohnOnTheNet said:

f8 full frame is brighter than f5.6 cropped.

I replied that this was not correct because:

The *total* light, integrated across the area of the FF sensor at f/8, is indeed greater than the crop sensor at f/5.6. But in terms of exposure, f/8 is still f/8. If you shoot the same scene at the same time and the same ISO with both cameras set to f/8, they will both require the same shutter speed.

What you're saying is equivalent to saying "The sun is brighter at my neighbor's house because he has a bigger yard."

Neither one of us said anything about DOF, noise, or "equivalence". The statements were only about the lens being "brighter", i.e, providing greater luminous flux density, when used with one sensor versus another. But the properties of the lens do not change because there is a different size sensor behind it. the amount of light per unit area reaching the sensor is the same (assuming the lens can illuminate the full sensor).

In the context of the original statement, my response was correct and I feel no need to defend it further. Personal attacks on me, @neuroanatomist, won't change that.
 
  • Like
Reactions: 1 users
Upvote 0