Plausibility of a 100-400mm f/4 L IS USM?

[jrista]

I'm a big fan of my 100-400mm f/4.5-5.6 L IS USM lens. Its served me quite well, and better than the 400mm f/5.6 lenses I've tried from friends or other fellow photographers met out in the field. The only real detractor, however, is the aperture range...f/4.5 to f/5.6 is a bit limiting, from an AF standpoint, and from the standpoint of using a teleconverter. I'd love to have a relatively affordable telephoto zoom lens that I could slap a 1.4x TC on to get 140-560mm f/5.6, something cheaper than the new 200-400mm f/4 L IS USM.

Is a 100-400m f/4 a plausibility, for a "reasonable" cost around $3500 or so? Would it be a lens anyone else even cared about, or is that just a focal range and price that doesn't serve anyone's needs? Personally, I'd find 560/5.6 to be far more useful for general bird photography, without having to lug around a heavy 500mm or 600m L.

Thoughts?

 

[Fleetie]

So you're asking for a lens with twice the zoom ratio of the 200-400L at the same aperture, for a lower price?

Ok, I seem to remember the 200-400L is the one with a built-in 1.4x teleconverter, isn't it?

But apart from the lack of that built in*, why would your proposed lens be much cheaper?

Maybe you're suggesting that you'd live with slightly lower IQ in your lens, to reduce cost?

Unfortunately, both lenses are 400 f/4 at the long end, so both need at least a 100mm front element, and similar barrel sizes, so not much potential for cost saving there, I'd have thought.

And a 4x zoom ratio is asking a lot if you want good IQ throughout - which is going to cost a lot.

(I might be thinking of the wrong lens when I'm talking about the 200-400 with 1.4 built in, though....)


* But you've said you need this 1.4x converter anyway so there's another £450 you'd have to spend on that, too, which for your requirements detracts from any cost advantage of your proposed lens even further.

 

[lol]

Never mind the zoom part, a 400mm f/4 is 400mm f/4 regardless, and not going to be cheap. Especially with the upcoming availability of the 200-400/4 I doubt they'll do another zoom that close in the ball park.

I think the nearest affordable option is the Sigma 120-300/2.8 OS which seems great value for money. I'd love Canon to do their own one too, but you know it wont be so affordable any more!

 

[smirkypants]

The excellent Nikon 200-400/f4 costs about $7000 street and weighs twice what the 100-400 weighs. Adding another 100mm on the short end would significantly increase the weigh and price. Sorry to burst your bubble, but I'm afraid you are asking for too much.

 

[KeithR]

Is a 100-400m f/4 a plausibility, for a "reasonable" cost around $3500 or so?

If you do what I did - buy a Sigma 120-300mm f/2.8 OS and slap a 1.4x on it...

 

[Cosk]

both lenses are 400 f/4 at the long end, so both need at least a 100mm front element

Just to build on that... aperture diameter is focal length / f*stop, and the front element has to be at least as large as the aperture... usually more.

So a 400mm f/4 requires a 100mm aperture...
a 400mm f/5.6 is 72mm

Physically, each stop (f4>5.6) requires 2x the surface area of glass of every element.
Generally, each stop means ~2x the weight and ~2x the cost.

So, assuming similar build and quality, features, and nothing fancy like Diffractive Optics, a 100-400 f/4 will be in the same ballpark from a weight and cost standpoint as 200-400 f/4 or a 300-400 f/4.

Personally, I'd love to see an anything-400 f/5.6 that's less than half the weight and less than half the cost as the 200-400 f/4... with the latest generation of IS, a twist zoom, and compacts to the length of the 70-300L (and those specs are physically possible to achieve).

 

[jrista]

Well, I guess I figured the integrated 1.4x TC added a fair bit to the cost, but perhaps not. I'm also not sure I believe the difference between a 71.4mm aperture (400/5.6) and a 100mm aperture would literally cost thousands of dollars for the front element...were talking a difference of roughly 28mm in size. I'm sure it adds to the cost, but if we take the current 100-400 f/4.5-5.6 which runs about $1800, to the 200-400 f/4 which is supposed to run around $7000...were talking a $5200 difference in price. I don't think the entire difference is allocated to glass...there is the integrated TC, R&D costs to design the thing that need to be recouped, etc.

A normal 1.4x TC costs some $500, and if we assume the built-in TC for the 200-400 is explicitly tuned for that particular lens, it probably cost more than that. Knock off $800-$1000 of the price of the 200-400 for the integrated TC, and we have a price difference of around $4300. I would NOT necessarily expect the kind of IQ from the 200-400 out of a 100-400...I'd expect it to be about as good as or maybe slightly better than the current 100-400, and there wouldn't be an integrated TC.

So, perhaps $3500 is overly hopeful...lets say $4500? Is that any more reasonable, assuming current 100-400 f/4.5-5.6 IQ, with a modern IS system? I'll certainly add that the short end really doesn't matter a wit to me...it could be 200-400, or even 250-400, if that would make it cheaper, as I rarely ever use the 100-200mm range on my current one, as I mostly photograph birds and wildlife with it.

Again, I'm just probing the pluasibility of such a lens, and currently, it seems particularly implausible at around $3500...

 

[jrista]

Is a 100-400m f/4 a plausibility, for a "reasonable" cost around $3500 or so?

If you do what I did - buy a Sigma 120-300mm f/2.8 OS and slap a 1.4x on it...

The sigma+1.4xTC would only get me 420 on the long end. It would get me 600mm on the long end with a 2x TC, but I'm curious how much that would affect IQ...more than I'd care for, I think. I want 560 on the long end, and I would gladly take 200 or 250 on the short end without any TC, as I don't really shoot shorter than that anyway.

 

[jrista]

both lenses are 400 f/4 at the long end, so both need at least a 100mm front element

Just to build on that... aperture diameter is focal length / f*stop, and the front element has to be at least as large as the aperture... usually more.

Technically speaking, the aperture is the entrance pupil (physical aperture as viewed through the front element)...so the front element has to be large enough to make the aperture appear to be 100mm. Physically, the aperture would be much smaller, given the magnification factor from the diaphragm to the front element. I'm not really sure how that translates into front element surface area though...

So a 400mm f/4 requires a 100mm aperture...
a 400mm f/5.6 is 72mm

Physically, each stop (f4>5.6) requires 2x the surface area of glass of every element.
Generally, each stop means ~2x the weight and ~2x the cost.

I know it would require roughly that much more surface area for the front element...but I'm skeptical it would need that much for every element...the point is to magnify the entrance pupil to appear to be 100mm when observed through the front of the lens, not actually make the physical diameter of the aperture 100mm.

So, assuming similar build and quality, features, and nothing fancy like Diffractive Optics, a 100-400 f/4 will be in the same ballpark from a weight and cost standpoint as 200-400 f/4 or a 300-400 f/4.

Personally, I'd love to see an anything-400 f/5.6 that's less than half the weight and less than half the cost as the 200-400 f/4... with the latest generation of IS, a twist zoom, and compacts to the length of the 70-300L (and those specs are physically possible to achieve).

Totally agreed, I don't actually need 100mm on the short end. I could do with 200, even 250mm, without a built-in TC. The desire is to have a lens that can go from 400mm on the long end to 560mm on the long end with a 1.4x TC, improving its versatility without pushing cost beyond reach. Its too bad all of Canon's pro bodies, including the 7D, don't support AF at f/8. I'd gladly take a 250-400 f/5.6 w/o integrated TC, slap on a 1.4x, and use it from 350-560 f/8 for birding...if it cost around $3500-$4000.

 

[KeithR]

It would get me 600mm on the long end with a 2x TC, but I'm curious how much that would affect IQ...more than I'd care for, I think.

That's not what I'm finding.

What I'm seeing is more like this and this.

 

[jwong]

Is a 100-400m f/4 a plausibility, for a "reasonable" cost around $3500 or so? Would it be a lens anyone else even cared about, or is that just a focal range and price that doesn't serve anyone's needs? Personally, I'd find 560/5.6 to be far more useful for general bird photography, without having to lug around a heavy 500mm or 600m L.

Thoughts?

I don't think that it's possible for 3.5k. For larger lenses, weight does tend to double for 1 stop increase in speed. For example, the 200mm f/2.8 is 27 oz while the 70-200II f/2.8 is 52 oz and the 200mm f/2.0 is 91 oz. At 300mm, the f/4 weighs 42 oz and the f/2.8 weighs 83 oz. Therefore, it wouldn't be unexpected that a f/4 zoom to 400 would weigh roughly twice the weight of the 100-400 f/4.5-5.6. Given that the existing 100-400 weighs 48 oz, it would be surprising that the f/4 zoom weigh close to 100 oz. The 300mm f/2.8 weighs 83 oz, which would be lighter than a 100-400mm f/4 zoom, and that thing already $7300. So, I'd have to agree with others that it would not be possible to build it for 3.5k.

 

[jasonsim]

No way they will make a 100-400mm f4L IS lens. No reason for them to since the 200-400 f/4L with 1.4x is on the way.

They should make a version II of the 100-400mm f/4.5-5.6L IS USM: give it latest IS, and twist zoom, and latest flourite coatings.

They should also think about coming out with a new 400mm f/4L...the one with DO is...well...outdated and the 400mm f/5.6L has no IS. Not everyone wants a 400mm f/2.8L beast to lug around.

 

[lol]

Technically speaking, the aperture is the entrance pupil (physical aperture as viewed through the front element)...so the front element has to be large enough to make the aperture appear to be 100mm. Physically, the aperture would be much smaller, given the magnification factor from the diaphragm to the front element. I'm not really sure how that translates into front element surface area though...

For longer focal lengths as discussed in this thread, you can take it as given the lens front size is same as the entrance pupil. You only tend to see a smaller entrance pupil than the lens front for wide angles.

 

[Mt Spokane Photography]

The 100-400mm when upgraded will sell for over $3000, a f/4 version about $7500, maybe more. Just chack the price of a 400mm f/4 DO right now, and imagine a lot more.

 

[jrista]

Technically speaking, the aperture is the entrance pupil (physical aperture as viewed through the front element)...so the front element has to be large enough to make the aperture appear to be 100mm. Physically, the aperture would be much smaller, given the magnification factor from the diaphragm to the front element. I'm not really sure how that translates into front element surface area though...

For longer focal lengths as discussed in this thread, you can take it as given the lens front size is same as the entrance pupil. You only tend to see a smaller entrance pupil than the lens front for wide angles.

Sorry, but its physically impossible for the entrance pupil to be exactly as large as the front element...your looking down a tube, your going to see the first half of the tube that leads up to the diaphragm regardless, so that guarantees the entrance pupil is smaller than the front element. For a visual example of how large the entire diaphragm is compared to the front element, take a look at this image from Canon's patent on the unreleased EF 100-400mm f/4-5.6 L: http://egami.blog.so-net.ne.jp/_images/blog/_f82/egami/2011_215218_fig04-9b046.png. The diaphragm and aperture are smaller than the first three elements.

My point was not that the front lens element would be smaller than the entrance pupil...it has to be BIGGER! Otherwise you can't magnify the aperture to appear as large as it would need to be. Anyway, my point was, its not a gargantuan difference between an f/5.6 aperture and an f/4...not like a move to f/2.8 would be (in which case, yes, I'd expect the lens to be monstrously HUGE and extremely expensive...something along the lines of Sigma's Bigma.)

Anyway, it was wishful thinking to start with, and the simple fact is that Canon is already releasing a 200-400 f/4 L w/ 1.4x TC. I'll just have to wait for that sucker to go on a nice sale and pick one of those up, as its essentially what I want, just a couple thousand dollars more than I want to spend. I guess we'll see where street price ends up after a few months on the market.

 

[lol]

If we're going down to this level of nit picking, let me clarify what I said. Observing from infinity, the apparent size of the entrance pupil will tend towards that of the lens front opening for longer focal length lenses. That is the measure that is of practical significance, regardless where or how big the actual physical stop is. I think we were saying the same thing but from totally different directions.

And I think most people would say the difference of 1 stop is significant enough. Approx. 1.4x linear, 2x in area and possibly more in volume terms since individual lenses elements aren't paper thin. For the same focal length and a comparable quality of design, there's no escaping the increase in glass required as aperture increases.

Side note: Bigma commonly refers to the 50-500. I'm guessing the earlier reference was more towards the 200-500/2.8 which certainly is a monster! You can hand hold that...

But on that note, and not something Canon would ever do... there is the third party trick of making the lenses f/6.3 on the long end which would give a small saving on size, weight and cost. Not a huge amount for sure, but when you're getting to longer focal lengths it could help a bit.

Personally I'm debating between the new 200-400 and the Sigma 120-300 with a 2x strapped on it. I'm pretty much waiting confirmation the 200-400 will cost more than my car before I get the Sigma!

And on one final thought, if anyone is still reading at this point. Would you buy a lens that was intentionally under-designed if it made a significant reduction to the cost? I know we all want the latest L lens, not so much the price that goes with it. But look at the lessons from the mirrorless cameras. Allow the easier to fix flaws of a lens to remain to help make them smaller and cheaper. Lateral CA and distortion are both easy to correct in post processing. If you allow them to remain in the lens design, how much simpler could the lens be? Maybe require less exotic elements too.

 

[jrista]

If we're going down to this level of nit picking, let me clarify what I said. Observing from infinity, the apparent size of the entrance pupil will tend towards that of the lens front opening for longer focal length lenses. That is the measure that is of practical significance, regardless where or how big the actual physical stop is. I think we were saying the same thing but from totally different directions.

And I think most people would say the difference of 1 stop is significant enough. Approx. 1.4x linear, 2x in area and possibly more in volume terms since individual lenses elements aren't paper thin. For the same focal length and a comparable quality of design, there's no escaping the increase in glass required as aperture increases.

True, comparing my 100-400 vs. my 16-35, I see what you mean by entrance pupil size (that telephotos tend towards the size of the front element, where as wide angles don't even come close.)

Side note: Bigma commonly refers to the 50-500. I'm guessing the earlier reference was more towards the 200-500/2.8 which certainly is a monster! You can hand hold that...

Sorry, it is the 200-500/2.8 that I was referring to. I've seen that picture before...its hilarious, and not something I would venture to do myself. I'd slap that puppy onto one of Gitzo's 5000 series uberpods with a gimbal.

But on that note, and not something Canon would ever do... there is the third party trick of making the lenses f/6.3 on the long end which would give a small saving on size, weight and cost. Not a huge amount for sure, but when you're getting to longer focal lengths it could help a bit.

I'd be willing to go for an f/6.3 aperture so long as it still autofocused. I'm never sure what the limit is on non-1D bodies: AF for less than f/8, or AF up to f/5.6? A third of a stop won't affect DOF that much, and I often shoot f/7.1 or f/8 anyway.

Personally I'm debating between the new 200-400 and the Sigma 120-300 with a 2x strapped on it. I'm pretty much waiting confirmation the 200-400 will cost more than my car before I get the Sigma!

I'll have to take a look at the Sigma. I know Canon embeds a lot of AF related functionality into the microprocessors of their lenses, as well as their teleconverters. I've always worried that if I used a non-Canon lens that AF would not work as well or as efficiently. Particularly with one of their TC III's.

And on one final thought, if anyone is still reading at this point. Would you buy a lens that was intentionally under-designed if it made a significant reduction to the cost? I know we all want the latest L lens, not so much the price that goes with it. But look at the lessons from the mirrorless cameras. Allow the easier to fix flaws of a lens to remain to help make them smaller and cheaper. Lateral CA and distortion are both easy to correct in post processing. If you allow them to remain in the lens design, how much simpler could the lens be? Maybe require less exotic elements too.

You can correct both in post processing, however only to a degree without affecting IQ. Most CA corrections fix the discoloration and try to rebalance tone, but the IQ impact is always there, to one degree or another. Correcting distortion in post requires shifting parts of the image, which causes interpolation (sometimes a lot, depending on how bad the distortion is). So, if those kinds of things don't matter, then sure, you could probably get away with buying much cheaper uncorrected lenses. If they do matter, then your just kind of stuck... :\

 

[Caps18]

If only I had won the lottery...

The 200-400 f/4 looks nice. I would have liked it to have the tripod ring collar that my 300 f/4 has...

If they mad the 300 f/4 with a built in 1.4x extender, that would be nice. I'm not sure if they could make one that is f/4 at both focal lengths though.

 

[DavidRiesenberg]

Taking into consideration Canon's latest lens pricing policies, I find it hard to believe the 200-400 TC will cost less than 10K. And it wouldn't surprise me if they'll also release sometime in the future a version without the TC but that will also be very expensive. Probably in the 7K region.

 

[lol]

You can correct both in post processing, however only to a degree without affecting IQ. Most CA corrections fix the discoloration and try to rebalance tone, but the IQ impact is always there, to one degree or another. Correcting distortion in post requires shifting parts of the image, which causes interpolation (sometimes a lot, depending on how bad the distortion is). So, if those kinds of things don't matter, then sure, you could probably get away with buying much cheaper uncorrected lenses. If they do matter, then your just kind of stuck... :\

With lateral CAs, it can be dealt with relatively easily by scaling the colour channels appropriately. Yes, this and any distortion correction will cost resolution. But the point is, how much cheaper could a lens be if they did this? It might or might not be significant. Put it this way, say you had two choices:
1: best ever, very expensive
2: good enough, affordable
If you can't afford 1, wouldn't you be looking at #2?