A New Constant f/4 Aperture RF-S Zoom Coming

[zapper]

Sorry, but no. You can argue with physics, but you will lose. Every. Single. Time.

There is no 'unused diameter' to remove. With telephoto lens designs, the limiting factor is the entrance pupil diameter and that is coincident with the front element. A 400mm f/4 lens will need a 100mm front element (slightly less, because really a lens called a 400/4 would be something like a 392mm f/4.13 and thus could have a 95mm front element). A smaller sensor won't change that.

DO will make the lens shorter, not lighter or smaller in diameter.

Try an empirical comparison. The OM 150-400mm f/4.5 is 115mm in diameter and weighs 1.9 kg. The Canon EF 400mm f/4 DO is 128mm in diameter and weighs 2.1 kg. The OM lens is for m4/3 sensors with a 2x crop factor, yet it's pretty much the same diameter and weight as the FF lens from Canon (the differences are because the OM lens is 1/3-stop slower).

Another FF vs APS-C comparison, between lenses I own:

Nikon Z 24-70 f/4 S: ø: 77.5 x L: 88.5 mm; 500g
Sigma RF-S 18-50 f/2.8: ø: 69.2 x L: 74.5 mm; 300 g

Nikon Z 14-30mm f/4 S: ø: 89 x L: 85 mm; 485g
Sigma RF-S 10-18m f/2.8: ø: 71.1 x L: 61 mm; 270g

Not a strict apples to apples comparison, but two pairs of approximately equivalent lenses in terms of focal range and aperture. There is a significant reduction in size and weight for the APS-C lenses. That is partly due to differences in construction and materials (the Nikon S lenses are a higher grade of lens than the Sigma Contemporary), but the reduction in lens size due to the smaller sensor is a major factor.

The R7 and the Sigma f/2.8 lenses will accompany me to Spain this week, and the Z7 and the Nikon f/4 lenses will stay home, because I'm packing light.

 

[zapper]

Benefits of collecting more total light, and not just at base ISO. At lower ISOs, the DR advantage is what is most noticeable - presumably what you mean by 'advantage in IQ'. At the other end of the ISO scale, the ~1.3 stops more light collected by the FF camera benefits not only DR but also image noise directly. I routinely use my R1 at ISO 25,600, a setting I wouldn't even consider with an APS-C camera.

View attachment 228774
(Showing the R3 because DPR's comparator doesn't have the R1, but I also routinely used the R3 at ISO 25,600.)

These examples definitely show the advantage of a larger sensor, but they aren't taken under the conditions of equivalency, where the R10 would have an ISO of 25600 / 1.6 = 16,000. Under those conditions the two images should look similar. But in real world low light conditions, you would likely shoot the two cameras at equal exposure (maximum aperture, as low a shutter speed as possible, and as high an ISO as you can tolerate). Then the images won't be equivalent. The R3 will have lower noise, and the R10 greater depth of field.

 

[neuroanatomist]

Another FF vs APS-C comparison, between lenses I own...

Not a strict apples to apples comparison, but two pairs of approximately equivalent lenses in terms of focal range and aperture. There is a significant reduction in size and weight for the APS-C lenses. That is partly due to differences in construction and materials (the Nikon S lenses are a higher grade of lens than the Sigma Contemporary), but the reduction in lens size due to the smaller sensor is a major factor.

The R7 and the Sigma f/2.8 lenses will accompany me to Spain this week, and the Z7 and the Nikon f/4 lenses will stay home, because I'm packing light.

Indeed, there is a size/weight advantage when designing lenses for a smaller sensor for focal lengths in the ultrawide, wide, normal and short/mid telephoto ranges. But at longer focal lengths, there is no advantage to designing lenses for a smaller sensor.

The main advantages of crop over FF are that the former can yield a system that is smaller, lighter and cheaper. Those are definitely tangible and significant benefits. Higher pixel density can be another one, for some use cases. But FF systems generally offer better image quality in many settings, and more control over DoF when wanted (try finding an APS-C lens to match the framing and DoF of an 85mm f/1.2 lens on FF).

 

[neuroanatomist]

These examples definitely show the advantage of a larger sensor, but they aren't taken under the conditions of equivalency, where the R10 would have an ISO of 25600 / 1.6 = 16,000. Under those conditions the two images should look similar. But in real world low light conditions, you would likely shoot the two cameras at equal exposure (maximum aperture, as low a shutter speed as possible, and as high an ISO as you can tolerate). Then the images won't be equivalent. The R3 will have lower noise, and the R10 greater depth of field.

My point was not about equivalence nor to show that, but merely to show the reason that I would never set an APS-C camera to ISO 25,600 while I routinely use that setting on modern FF cameras.

 

[AlanF]

Another FF vs APS-C comparison, between lenses I own:

Nikon Z 24-70 f/4 S: ø: 77.5 x L: 88.5 mm; 500g
Sigma RF-S 18-50 f/2.8: ø: 69.2 x L: 74.5 mm; 300 g

Nikon Z 14-30mm f/4 S: ø: 89 x L: 85 mm; 485g
Sigma RF-S 10-18m f/2.8: ø: 71.1 x L: 61 mm; 270g

Not a strict apples to apples comparison, but two pairs of approximately equivalent lenses in terms of focal range and aperture. There is a significant reduction in size and weight for the APS-C lenses. That is partly due to differences in construction and materials (the Nikon S lenses are a higher grade of lens than the Sigma Contemporary), but the reduction in lens size due to the smaller sensor is a major factor.

The R7 and the Sigma f/2.8 lenses will accompany me to Spain this week, and the Z7 and the Nikon f/4 lenses will stay home, because I'm packing light.

The crucial point in these discussions is as follows. Above a certain focal length, 50-100mm or so, the natural image diameter of a simple lens is larger, or as the focal length increases, many times larger than the diameter of the circle required to cover a full frame sensor. So, a lens for sensors smaller than full frame will also cover full frame, and the lens cannot be made smaller just to have a smaller diameter for crop, M4/3 etc. Conversely, for shorter focal length lenses, the natural image circle of a simple lens becomes smaller than required to cover the full frame and so clever optics with complex elements is required to expand effectively the image circle. The smaller the sensor, the easier it is to make a lens to cover its image circle. The comparison is not a "Not a strict apple to apples..." but is a quite different situation.

 

[Andreasb]

FINALLY!!! Why has it taken so long for Canon to realize that there is a real need for an all around 24mm equivalent zoom for apc-c, after all there are 3 FF lenses that has this 24mm starting point. This will by my travel and walk around combo, small light hopefully F4 through out.

 

[zapper]

Indeed, there is a size/weight advantage when designing lenses for a smaller sensor for focal lengths in the ultrawide, wide, normal and short/mid telephoto ranges. But at longer focal lengths, there is no advantage to designing lenses for a smaller sensor.

Right, if D = f/N, (D=entrance pupil, f=focal length, N=f number) and f and N both are scaled by the same crop factor for an equivalent focal length and aperture, then D is the same for a FF and equivalent APS-C lens. And the diameter of the lenses is the biggest factor in weight. However, since f does scale down for APS-C, that equivalent lens will be shorter, which should yield some saving in the length dimension and weight too.

The main advantages of crop over FF are that the former can yield a system that is smaller, lighter and cheaper. Those are definitely tangible and significant benefits. Higher pixel density can be another one, for some use cases. But FF systems generally offer better image quality in many settings, and more control over DoF when wanted (try finding an APS-C lens to match the framing and DoF of an 85mm f/1.2 lens on FF).

I agree. I have a FF Nikon (Z7), which is what I use when I want top image quality. Especially when I can use a tripod and use the base ISO of 80, the IQ is top notch. I don't have any f/1.2 lenses though - I'm not really into super short DOF. The R7 makes a great birding camera, due to high pixels density, as you mention, and a great lightweight travel camera with limited financial exposure should some low life steal it.

 

[mb66energy]

15-70 f/4 would be a great video lens for me - especially when it is equipped with motorized zoom (PZ). Could "replace" the cinema lens with 18-80 range for simple applications.
I always wanted a 15-60 f/4 with high correction - if this lens has it, might be a replacement for my RF-S 24-105.

 

[zapper]

The crucial point in these discussions is as follows. Above a certain focal length, 50-100mm or so, the natural image diameter of a simple lens is larger, or as the focal length increases, many times larger than the diameter of the circle required to cover a full frame sensor. So, a lens for sensors smaller than full frame will also cover full frame, and the lens cannot be made smaller just to have a smaller diameter for crop, M4/3 etc. Conversely, for shorter focal length lenses, the natural image circle of a simple lens becomes smaller than required to cover the full frame and so clever optics with complex elements is required to expand effectively the image circle. The smaller the sensor, the easier it is to make a lens to cover its image circle. The comparison is not a "Not a strict apple to apples..." but is a quite different situation.

But no commercial lens is a simple lens. In a complex lens, the size of the rear and intermediate elements can scale with sensor size, as these elements determine the image circle, not the front elements that set the entrance pupil.

 

[phree_ed]

How I would have liked a replacement to the 15-85 for my R7 but sorry Canon - too late. I got tired of waiting and just three days ago picked up a Sigma 17-40 f1.8.

Yeah I know it's only 40mm at the long end but I've got it and I won't be stuffing around selling it and purchasing another lens.

But never say never I guess

 

[neuroanatomist]

Right, if D = f/N, (D=entrance pupil, f=focal length, N=f number) and f and N both are scaled by the same crop factor for an equivalent focal length and aperture, then D is the same for a FF and equivalent APS-C lens. And the diameter of the lenses is the biggest factor in weight.

It doesn’t scale like that, as @AlanF stated. For example, my RF 100-400mm f/8 has a front element of ~50 mm in diameter. 400 mm / 8 = 50 mm. So far so good.

If that relationship scaled across all focal lengths, my RF 10-20mm f/4 would have a front element of ~5 mm in diameter. In fact, its front element is ~60 mm in diameter, larger than that of the 100-400/8.

Lens design matters. The takeaway is that the entrance pupil is at the front element for long telephoto designs, so that becomes the limiting factor in lens diameter. At shorter focal lengths, other factors are limiting.

However, since f does scale down for APS-C, that equivalent lens will be shorter, which should yield some saving in the length dimension and weight too.

For equivalent framing. Focal length is an intrinsic property of a lens, independent of the sensor behind it.

 

[Aldherrian]

This supposed RF-s 15-70 F4 is a lens I would have gladly bought to upgrade from the EF-s 18-135 and update the EF-s 15-85.

Both are variable apertures of 3.5 - 5.6 so a constant aperture at 4 is nice.
I don't find realistic to expect such a wide range of focal length to offer F2.8 unless it's a L lens and totally not the same package of size and price (and ultimately audience)

The lack of such a lens made me upgrade to FF format but I'm glad some stuff is finally happening to the APSC side

 

[Tom R5R7]

How I would have liked a replacement to the 15-85 for my R7 but sorry Canon - too late. I got tired of waiting and just three days ago picked up a Sigma 17-40 f1.8.

Yeah I know it's only 40mm at the long end but I've got it and I won't be stuffing around selling it and purchasing another lens.

But never say never I guess

I used the Canon 17-40 f/4L on my 10D and 30D cameras back in the day. At the time, it was about the sharpest ultra-wide Canon had, with some very rich colors and great contrast. 40 is a little bit short, but that Sigma f/1.8 has the potential to be a real gem.

 

[InternalRevenue]

How I would have liked a replacement to the 15-85 for my R7 but sorry Canon - too late. I got tired of waiting and just three days ago picked up a Sigma 17-40 f1.8.

Yeah I know it's only 40mm at the long end but I've got it and I won't be stuffing around selling it and purchasing another lens.

But never say never I guess

I somewhat feel the same but 17mm ends up just a little too tight often enough for my video projects that I still kept and use the 15-85 after buying the 17-40, despite the EF-S's many, many quirks and deficiencies for video use. The closeup IQ and MFD isn't that great on the Sigma either, and if Canon manages to get as good a MFD/magnification ratio on this thing like they did with the RF-S 18-150, I might just spring for this 15-70 (or buy it bundled with the R7ii if offered).

 

[Tom R5R7]

I somewhat feel the same but 17mm ends up just a little too tight often enough for my video projects that I still kept and use the 15-85 after buying the 17-40, despite the EF-S's many, many quirks and deficiencies for video use. The closeup IQ and MFD isn't that great on the Sigma either, and if Canon manages to get as good a MFD/magnification ratio on this thing like they did with the RF-S 18-150, I might just spring for this 15-70 (or buy it bundled with the R7ii if offered).

Yeah, even on in the 10D says, 17 was tight. But the standard "kit" lens 18 was even tighter. And for all the concerns about deficiencies, its f/4 aperture was pretty useful. F/2.8 would have been better, but the 16-35 f/2.8 at the time was really not quite as good a lens. it had some really crazy flare issues.

 

[dwkdnvr]

*IF* this lens has any degree of weather sealing it would solve the lens dilemma I've been struggling with and will be an insta-buy for me. But this is RF-S so I'm fully expecting that it will continue the budget build quality tradition. It will definitely cause me to wait and see whether it pans out before going with another option, though.

Coming from the M43 world I didn't entirely realize how spoiled they are for hiking/travel lenses. All sorts of 24-xx (equivalent) options many with robust build and weather sealing at a range of aperture/size/cost. I'm trying to find 'that lens' for my R7. Maybe I'm over-valuing weather resistance, but being in the dusty SW and having been caught in weather when traveling far too frequently I'm far more comfortable having it. (we got drenched several times on a trip to NZ and the E-M1.2/12-100 came through with aplomb)

The only RF-S lens with any advertisement of splash/dust resistance is the Sigma 17-40 which is probably the most logical choice, but I'm unreasonably annoyed that it's 17 rather than 15 on the wide end. And general hiking/travel really doesn't need F/1.8 so there's a size penalty. I find myself very seriously considering the RF 14-35/4 as it seems the best fit - 22-56 equivalent (maybe even a bit wider if you manually tweak the corrections), L build quality, reportedly great optically even with the corrections (and APS-C avoids the worst of the corrections in the first place); just evaluating whether paying the premium for an FF ultra-wide (albeit a fairly reasonably priced one) is worth it.

 

[DocInfoSci]

With the understanding and respect that people who shoot both FF and CF cameras care about equivalency, I think many of the people who purchase a camera will own one at a time. And if a CF edition is their camera, then this constant aperture zoom (among some of the others available for RF CF) is a rather excellent option as compared to the EF era. I think that this lens will make a lot of CF shooters happy if it can be priced less than $1,500. And obviously (but worth stating anyhow) it would pair well with bookend lenses such as the 14-30 and 55-210. That coverage would make an excellent kit for a new shooter, school teacher, junior park warden, soccer mom, limited income retiree, etc.

If Canon could release a similarly inexpensive CF UWA constant f/4 zoom (same build quality as the 15-70) then coupled with a 70-200 f/4 FF lens Canon will basically have said, relative to the EF era, they're taking these customers seriously. It's certainly better than just the 17-55, which was an almost-nice thought that didn't line up with anything equally serious.

That stated, the lesson I learned was buy as little CF gear as possible if there's the slightest chance of an upgrade path to FF. Fortunately, I had some smart elders clue me in early into my lens journey.

 

[neuroanatomist]

With the understanding and respect that people who shoot both FF and CF cameras care about equivalency, I think many of the people who purchase a camera will own one at a time.

I'm not sure that many people really care about equivalence. I certainly don't take a picture with my R1 then try to set up a completely equivalent shot with my M6II.

It's a useful concept to understand, if only to avoid sounding foolish when making comparative statements about different cameras. I've seen many people buy into the idea that there is a free lunch, and that a smaller sensor delivers a smaller, cheaper system with no tradeoffs. For some reason, people stop that line of thought at APS-C (or m4/3 on other forums), but no one seems to carry it forward and believe that an ILC with a P&S-sized sensor would be even better than their APS-C camera body, if their logic was reasonable.

On the flip side, I've seen some FF users treat equivalence as a bludgeon to claim FF cameras are superior, when of course there are good reasons for both formats (and for medium format and P&S cameras, too). As I mentioned above, a system that's smaller, lighter and/or cheaper has meaningful advantages. When packing a carryon for an overnight business trip, it's far easier to pack the M6II, M11-22 and M18-150 in the suitcase than the packing lenses with the corresponding coverage along with the R8. But the shallow DoF you can get with a fast lens, or the wider FoV with a lens like the RF 10-20/4, can't be replicated on a Canon APS-C MILC (at least, I don't think there is a rectilinear 6mm lens for RF-S).

And if a CF edition is their camera, then this constant aperture zoom (among some of the others available for RF CF) is a rather excellent option as compared to the EF era. I think that this lens will make a lot of CF shooters happy if it can be priced less than $1,500. And obviously (but worth stating anyhow) it would pair well with bookend lenses such as the 14-30 and 55-210. That coverage would make an excellent kit for a new shooter, school teacher, junior park warden, soccer mom, limited income retiree, etc.

Absolutely. I think more choice is good. But I can also see some reasons why it might not happen.

If Canon could release a similarly inexpensive CF UWA constant f/4 zoom (same build quality as the 15-70) then coupled with a 70-200 f/4 FF lens Canon will basically have said, relative to the EF era, they're taking these customers seriously. It's certainly better than just the 17-55, which was an almost-nice thought that didn't line up with anything equally serious.

That may be true. Certainly things trended the other way in DSLR days. The EF-S 17-55/2.8 was a great lens for crop cameras, and never updated. The EF-S 10-22/3.5-4.5 was another excellent lens that was downgraded to the slower EF-S 10-18mm. The 90D was essentially the successor to the 7DII. All of those show that APS-C DSLRs were trending downmarket.

The MILC market is trending up overall, but I'm not sure that fully applies to the APS-C segment. By the numbers, that segment is shrinking pretty substantially. From around 90% in the heyday of DSLRs, last year APS-C MILCs were 63% of the market. So Canon may prefer to avoid making 'too good' a range of lenses and bodies with APS-C sensors, to further drive sales of FF MILCs.

That stated, the lesson I learned was buy as little CF gear as possible if there's the slightest chance of an upgrade path to FF. Fortunately, I had some smart elders clue me in early into my lens journey.

My first DSLR was a Rebel T1i/500D, that I replaced with a 7D. I subsequently added a 5DII, and used both side by side until the 1D X delivered both FF and speed, and since then my only APS-C cameras have been M-series for the portability.

The lesson I remembered from shooting film was glass >> body. I skipped the EF-S 18-55mm kit and 50/1.8 that were commonly recommended, and got just the T1i body only. I bought the EF-S 17-55/2.8 and EF 85/1.8 as my starter lenses. The only other EF-S lens that I bought was the EF-S 10-22. I kept both EF-S lenses until I sold the 7D.

 

[zapper]

It doesn’t scale like that, as @AlanF stated. For example, my RF 100-400mm f/8 has a front element of ~50 mm in diameter. 400 mm / 8 = 50 mm. So far so good.

If that relationship scaled across all focal lengths, my RF 10-20mm f/4 would have a front element of ~5 mm in diameter. In fact, its front element is ~60 mm in diameter, larger than that of the 100-400/8.

Lens design matters. The takeaway is that the entrance pupil is at the front element for long telephoto designs, so that becomes the limiting factor in lens diameter. At shorter focal lengths, other factors are limiting.

Yes. For complex lenses, the entrance pupil may be very different than the front element size. I'm fuzzy on the relationship; this is what Google tells me: "The entrance pupil in a complex lens is determined by imaging the physical aperture stop through all preceding optical elements into object space. It is the virtual or real image of the diaphragm as seen from the front of the lens, defining the cone of light that enters the system."

For equivalent framing. Focal length is an intrinsic property of a lens, independent of the sensor behind it.

Of course. And field of view depends on the focal length and the sensor size.

 

[zapper]

I'm not sure that many people really care about equivalence. I certainly don't take a picture with my R1 then try to set up a completely equivalent shot with my M6II.

It's a useful concept to understand, if only to avoid sounding foolish when making comparative statements about different cameras. I've seen many people buy into the idea that there is a free lunch, and that a smaller sensor delivers a smaller, cheaper system with no tradeoffs. For some reason, people stop that line of thought at APS-C (or m4/3 on other forums), but no one seems to carry it forward and believe that an ILC with a P&S-sized sensor would be even better than their APS-C camera body, if their logic was reasonable.

On the flip side, I've seen some FF users treat equivalence as a bludgeon to claim FF cameras are superior, when of course there are good reasons for both formats (and for medium format and P&S cameras, too). As I mentioned above, a system that's smaller, lighter and/or cheaper has meaningful advantages. When packing a carryon for an overnight business trip, it's far easier to pack the M6II, M11-22 and M18-150 in the suitcase than the packing lenses with the corresponding coverage along with the R8. But the shallow DoF you can get with a fast lens, or the wider FoV with a lens like the RF 10-20/4, can't be replicated on a Canon APS-C MILC (at least, I don't think there is a rectilinear 6mm lens for RF-S).

I think the value of equivalence is to understand the range of capabilities of a given set of equipment, and the relative compromises one makes.

My first FF DSLR was the 5Dii, and I spent a lot of money on a f2.8 trinity. I quickly discovered that unless photography was my one and only objective for the day, I hated carrying it. I much preferred to carry the APS-C DSLRs that I had before and after the 5Dii for everything other than totally dedicated photography. But what's the compromise? Well, a f2.8 lens on an APS-C body gives you about the range of capabilities as shooting no faster than f4.5 on the FF body. You give up that first f-stop and a third for a kit that's easier to carry. That's the type of tradeoff that equivalence helps to clarify.

I've since had a FF mirrorless system with f4 lenses, and don't hate it. That's about my happy place for size vs capability.

In the case of the lens that's the topic of this thread, you give up another stop of light. If you can imagine shooting no faster than f6.3 on FF, it will be about the same as that.