APS-C lens frustration.

[9VIII]

(I'm sure many of you have read all this before, but being fresh to the hobby I had to get it off my chest.)

So I've been looking for a fast prime lens for my T3 for a while, I could just buy one of the many great prime full frame lenses out there, but I couldn't shake the feeling that I'm not getting the most out of the lens if you go that route.

My understanding is that f-stop numbers on any lens give the same exposure (the amount of light per square area directed to the sensor) regardless of sensor format. The "aperture" or size of the opening in the front of the lens, is directly related to focal length in order to keep the amount of light collected the same across all lenses at the same f-number. There seems to be a huge oversight though.
There are no APS-C lenses that give the same field of view as a full frame lens at the same focal length.

Currently, all APS-C lenses act just like a full frame lens, giving a tighter field of view at the same focal length. To make a 30mm "equivalent" for APS-C bodies, camera companies just make a small 18mm lens and call it a day.
What would happen if you made a 30mm lens that produced the same field of view on APS-C as a full frame lens on a full frame body? Your image would be one stop brighter! Sounds like a win-win situation to me! If you're specifically going after a fast wide lens APS-C can get you an extra stop of light for free. With wide angle lenses generally being sort of slow, but useful indoors, you would think that someone would have made a big deal about that extra stop of light by now. If you're after telephoto, an APS-C equivalent to a FF 400mm lens would either weigh and (theoretically) cost half as much, or give the brighter image at the same size/cost.
Which means that if you went out and bought a 400mm f5.6 to use with an APS-C body because it's the only reasonably priced way to get quality telephoto, you can rest assured that half the value of your lens is utterly wasted on that APS-C body.
If you would rather put money into glass than bodies (considering that APS-C sensors are an order of magnitude cheaper to make), it boggles the mind that Canon (or anyone for that matter) wouldn't be jumping all over the idea of high end APS-C native lenses. They could make them just as expensive, just APS-C would have some advantages, whereas right now it has no advantages.

We're also missing a very valuable tool to gain performance that would otherwise not be possible.
The equivalent of an f1 lens on digital cameras may only be possible with smaller sensors. Take a normal f1.4 lens, adjust it for a smaller sensor and BAM! You have an image brighter than can possibly be achieved on Full Frame.
What really perplexes me is that the sickness (trend) seems to be propagated across the entire industry. The Metabones Speed Booster seems to be the first product ever to acknowledge that getting a brighter image off the same size of lens by using a smaller sensor is even a possibility. Note how the f1 equivalent with that tool does not give more light? I'm assuming that's because the speed booster puts the final element in the lens much closer to the sensor rather than keeping it the same distance as on a full frame mount, so light hits the sensor at more extreme angles. Thus aggravating pixel vignetting.
Done correctly, just imagine what could be achieved with something like the Nikon 1 series, you could get almost three extra stops of light.
Unfortunately, everyone has their reputation to keep. We can't have people walking around with $300 cameras that work better than $3,000 cameras, that just wouldn't do.

Yes, there will be some sacrifice in IQ vs. Full Frame, a sensor with greater surface area will almost always get better IQ (just remember that has little to do with pixel size until you're diffraction limited). As far as the consumer is concerned though, we've had cameras that are "good enough" for just about anything for years now, and given that low light is such a common issue you would think that brighter lenses for crop sensor cameras would be strewn all over the place by now.

"Deep breath"

I'm done now.

[ahab1372]

So are you asking for a 30mm APS-C lens that gives the same framing as a 30mm Full Frame on Full Frame camera?

That's not possible. And why would you want that instead of using a wider lens on the APS-C?


Arnt

[Dave Throgmartin]

There's some stuff out there that is marketed just to be a quality fast lens just for APS-C. The Sigma 30mm f/1.4 comes to mind. It is intended to be a normal lens so the FF equivalent is 48mm f/2.2.

The flange distance used by the Canon EF mount does not allow small size fast primes that are wider angle on APS-C. The Canon 28mm f/1.8 is the widest fast Canon prime. The M mount does have a fast "wide angle-ish" lens in the 22mm f/2. I've heard it is fantastic lens. It won't work on an EF mount though. The M mount has a shorter flange distance. I do not believe it is possible to make a small 18mm prime for APS-C using the EF mount...

You may want to consider the Samyang 14mm f/2.8. It is manual focus only, but given the very wide focal length many find manual focus acceptable. It's a full frame capable lens too. It is said to have some mustache type distortion on full frame, but it is not as bad on APS-C.

The only APS-C telephoto I've seen is the Sigma 50-150 f/2.8. It isn't that dramatically cheaper than a FF capable 70-200 f/2.8 though.

Dave

[neuroanatomist]

There are no APS-C lenses that give the same field of view as a full frame lens at the same focal length.

Ahhh...so, you're asking lens makers to lie. Focal length is the physical distance between the rear nodal point and the image plane, and is an intrinsic property of a lens, regardless of the sensor behind that lens. You are suggesting that an APS-C (EF-S) be labeled with a focal length that is 'adjusted' for the FoV relative to full frame, but doing so would mean the number printed on the lens would not be the actual focal length, i.e. it would be a lie.

If you're after telephoto, an APS-C equivalent to a FF 400mm lens would either weigh and (theoretically) cost half as much, or give the brighter image at the same size/cost.

Actually, it wouldn't. Lens designs differ, but the physical size of the elements in front of the entrance pupil always need to be large enough to fill that entrance pupil with light. For a telephoto design, the entrance pupil is essentially at the front element, so a 400mm f/5.6 lens is going to need at least a 71.4mm diameter front element, regardless of the image circle diameter. So, making a 400/5.6 for APS-C would allow some of the rear elements to be smaller, but not really decrease the overall size or cost.

Now, if you mean a lens that provides the FoV of a 400mm lens on APS-C, well...that's the long end of the EF-S 55-250mm, which is far less than half the size/cost of the 400/5.6.

[9VIII]

I had a quick look at Wikipedia, according to that the size of your projected image is directly dependent on the distance of the lens from the sensor. So EF-M can get boosted lenses but I'm guessing not EF-S.
On the other hand, any home theater projector can zoom the image in and out, giving a varying size of projected image from the same spot. Maybe it's just impractical to put something in that would correct the image size?

[DavidEccleston]

There is probably a good reason for ignoring the APS-C lens market. At my previous job, everybody who had a non point and shoot, had a Rebel XTi, with the OLD 18-55 non-IS kit lens, and was happy. I couldn't convince any of them to even try the 50mm f/1.8. The boss had apparently considered trying the lens a few times, but never did.

I suspect the majority of Rebel buyers don't buy lenses. The camera kit itself was a large enough expense that they're not willing to invest more. Even the 50mm f/1.8 is out of budget. It's not economical to produce fast lenses for the few APS-C users who do buy lenses. The people who are shooting fast lenses are either trying to lower DOF, or lower noise with lower ISOs. That super narrow subset of APS-C shooters aren't a viable market, and this non-viable market can easily put themselves into a viable fullframe market that will give them the thin DOF and low noise they desire at a variety of price points. From $500 used 5DmkI, $1200 used 1DsmkII for servo, $1900 6D, $2700 5DmkIII for servo, or $6000ish 1DX for all the latest bells and whistles.

Because of servo performance needs, a limited budget, I recently went with a used 1Ds2. I *really* like the quality I get, even with an ancient fullframe body. I can't see myself going back to a crop body. I don't shoot enough wildlife to justify lowering the quality of the rest of my photos. My 7D mainly gathers dust along with the T1i. If I *knew* the 6D could handle my servo needs, I'd sell all three bodies and pick up a 6D.

DxOMark show the 5DmkI outperforms your T3, and the semi-pro 7D in image quality and ISO. If you're looking for fast primes, you're looking for the advantages of fullframe. Why not get one? For the price of a lens, you can have everything you want*, just not in lens form.

(* = Assuming you weren't just hoping for something lighter!)

[HDNitehawk]

I don't think your ideas will get much consideration from Canon. Marketing and development may even be moving in the opposite direction. Think about this, if you want your customers to migrate to your higher end models you creat items that do not limit the migration. At present the EF-S line limits upgrades to full frame.

The APS-C line was a compromise when it was introduced. Cost of production and the lack of processing power lead to small sensors. The technology today has brought the digital camera up to what film bodies could do. As full frame sensors become more economical I think you may see the APS-C bodies slowly squeezed out. How many people would buy a crop body over FF if the cost is similar, there is over a 2x cost spread between the two now but the gap is gradually closing.

[9VIII]

There is probably a good reason for ignoring the APS-C lens market. At my previous job, everybody who had a non point and shoot, had a Rebel XTi, with the OLD 18-55 non-IS kit lens, and was happy. I couldn't convince any of them to even try the 50mm f/1.8. The boss had apparently considered trying the lens a few times, but never did.

I suspect the majority of Rebel buyers don't buy lenses. The camera kit itself was a large enough expense that they're not willing to invest more. Even the 50mm f/1.8 is out of budget. It's not economical to produce fast lenses for the few APS-C users who do buy lenses. The people who are shooting fast lenses are either trying to lower DOF, or lower noise with lower ISOs. That super narrow subset of APS-C shooters aren't a viable market, and this non-viable market can easily put themselves into a viable fullframe market that will give them the thin DOF and low noise they desire at a variety of price points. From $500 used 5DmkI, $1200 used 1DsmkII for servo, $1900 6D, $2700 5DmkIII for servo, or $6000ish 1DX for all the latest bells and whistles.

Because of servo performance needs, a limited budget, I recently went with a used 1Ds2. I *really* like the quality I get, even with an ancient fullframe body. I can't see myself going back to a crop body. I don't shoot enough wildlife to justify lowering the quality of the rest of my photos. My 7D mainly gathers dust along with the T1i. If I *knew* the 6D could handle my servo needs, I'd sell all three bodies and pick up a 6D.

DxOMark show the 5DmkI outperforms your T3, and the semi-pro 7D in image quality and ISO. If you're looking for fast primes, you're looking for the advantages of fullframe. Why not get one? For the price of a lens, you can have everything you want*, just not in lens form.

(* = Assuming you weren't just hoping for something lighter!)

The trouble is that I like the T3 as a travel camera, if anything happens to it I'm only out a few hundred dollars and (as long as I don't lose my camera more than once a year) I can replace it with something better for the same cost. My motivation not to use full frame is that APS-C will always be "much" cheaper going forward, dirt cheap, so it makes sense to plan on using that format a lot. Image quality is more than good enough for personal use, the only real problem so far is the physical limitations of the format on the wide end of the focal length scale.
Maybe I just need to wait for EOS-M to displace the Rebel series, which should open up a lot of options for new lenses. The distance between the mount and the sensor is 18mm which, as I learned today, means that where an 18mm lens on an SLR is heavily retrofocused, that will not be necessary with EF-M.

[9VIII]

I don't think your ideas will get much consideration from Canon. Marketing and development may even be moving in the opposite direction. Think about this, if you want your customers to migrate to your higher end models you creat items that do not limit the migration. At present the EF-S line limits upgrades to full frame.

The APS-C line was a compromise when it was introduced. Cost of production and the lack of processing power lead to small sensors. The technology today has brought the digital camera up to what film bodies could do. As full frame sensors become more economical I think you may see the APS-C bodies slowly squeezed out. How many people would buy a crop body over FF if the cost is similar, there is over a 2x cost spread between the two now but the gap is gradually closing.

http://www.dpreview.com/articles/033...s-masaya-maeda

Now that full frame is appearing in cameras the size and price of the 6D, the future of APS-C at the semi-pro level is in doubt

This may be the quote you're thinking of?

At the same time, they also say that APS-C is moving to the new standard for high end consumer cameras, that means we're going to have a plague of cheap APS-C cameras, and it appears that EOS-M may also be part of that strategy.

[HDNitehawk]

Actually I had read the article but wasn't thinking of it when I wrote the response. Canon didn't need to say it you could see it in their actions and releases the last few years. Plus it just makes sense. If a cheap full frame sensor and processor that could give us 10 fps had existed in 1995 we would have gotten it. Instead we got this;
http://www.canon.com/camera-museum/c...=1995-2004&p=2