If canon is going to keep pushing EFS, they need higher-end EFS lenses

[mikekelley]

10-22 is sharp as a tack but slow as a dog.

10-2x with constant 2.8 would be just fantastic.

Some nice EF-S primes would be great too. I'm sure that they could be made significantly cheaper (less glass) and super-sharp.

How about some EFS super-teles? Again, they could be made cheaper, lighter and just as sharp as their EF equivelents, if they kept within the EFS build standard.

Less variable aperture crapola. More serious lenses for EFS mounts, please.

I know that they are capable, I just don't get why they continue to pump out consumer grade minutiae.

[RDKirk]

Remember that Canon considers EFS to be a consumer grade product line.

[jobe1492]

Just want some weather sealing.
Canon needs a 35mm 1.8 EFS. A 20mm EFS would be nice too.
EFS super teles - LOL no way. They lens would snap in half.

[mikekelley]

Remember that Canon considers EFS to be a consumer grade product line.


Then by this logic, only people with 1d bodies (pro bodies) need non-consumer grade glass.

As far as I am aware, the only thing considered "pro" is the 1d line.

[mikekelley]

Just want some weather sealing.
Canon needs a 35mm 1.8 EFS. A 20mm EFS would be nice too.
EFS super teles - LOL no way. They lens would snap in half.
Remember that, when tailored for the 1.6 sensor size, the image circle is smaller, and thus, less glass is used, and thus, it weighs a bit less. So I don't think that is so plausible.

[JeffreyG]

I don't see much point in EF-S primes. For the discriminating 1.6X shooter there already exists the 14L, 24L, 35L, 50L etc. etc. And to be honest, owning a handful of those pricy L primes for exclusive use on 1.6X bodies makes little sense as their speed advantage could be more than compensated for by getting the slower standard primes and a used 5D.

Where the EF-S line is deficient is in giving the 1.6X shooter a high quality, fast wide angle zoom (such as the EF-S 17-55) that is also built well. IMO the 17-55 is a nice lens (I used to have one) but it costs too much for the way it is built. It also needs better resistance to veiling flare.

The new EF-S 15-85 looks interesting, but the variable max aperture and the build again leave it wanting against something like a 5D with a 24-105L.

[Mike Deep]

Not agreeing or disagreeing with the OP, I just want to get this in:
EFS super teles - LOL no way. They lens would snap in half.
What kind of nonsense is this? An EF-S lens could be produced to the same build quality as an L. Just because Canon won't, it doesn't mean they can't. A smaller image circle doesn't somehow force the manufacturer to adhere to plastic and suck.

[DaveSt]

Personally, I feel just the opposite. I think Canon needs to release more lower cost EF-S lenses. As was mentioned earlier, there are plenty of high grade Canon EF mount lenses that work just fine on 1.6 crop bodies. If the EF-S line is meant to be a consumer grade line then Canon should actually make some lenses that the typical consumer is comfortable paying for. Look at some of the lenses in the lineup. The 10-22 costs $700+ and the 17-55 costs over $1000. If those are consumer level prices I'll eat my shorts. I really hope the new Tamron 17-50VC turns out to be a great lens. I probably won't buy it but I would love to see sales of the 17-55 drop in a big way to send Canon a message. Chances are neither will happen, but it would be nice. If Canon decides to release any prime EF-S lenses, I would hope they would shoot for something more like the Sigma 30 f/1.4 rather than a 35L.

**Edit: I should note that I feel the EF-S 55-250 is more the way I think the EF-S line should go. There is a very useful lens priced at a level many consumers and amateurs can stomach.

[jbhswim]

I would love to see more ef-s lenses, but thats only because i have a 1.6x crop camera.

the next guy who comes along will want more ef lens for his 5d mark2..

canon will find the right amount..

[KenjiS]

What kind of nonsense is this? An EF-S lens could be produced to the same build quality as an L. Just because Canon won't, it doesn't mean they can't. A smaller image circle doesn't somehow force the manufacturer to adhere to plastic and suck.

See Olympus and their 300mm f/2.8

Which is the equivalent of a 600mm f/2.8 :)

Also their 150mm f/2...Thats a spanking great lens right there...

Oh! What about the 90-250...

Look, Criticize 4/3rds, but they do have some amazingly cool lenses [And optically, Those lenses are all the equal of Canon/Nikon..they're impressive]

I really hope the new Tamron 17-50VC turns out to be a great lens.

MSRP for the Tamron 17-50 VC with its crappy AF and build is $750...And it will likely sell for $750 [Tamron sticks to their MSRP, i see little price deviation on their part at different stores]

Note you can get a used 17-55 in great shape for that..Not worth it IMHO to get the Tamron..

ok...Now im going to get flogged for this, but whats the big deal with variable aperture...Seriously? Just because its variable aperture doesnt make it a poor lens [See, my 100-400 for instance]

[Andrushka]

Just want some weather sealing.
Canon needs a 35mm 1.8 EFS. A 20mm EFS would be nice too.
EFS super teles - LOL no way. They lens would snap in half.

Need a 35mm? try the 28 f1.8 or the 35mm f2
Need a 20mm? try the 20 f2.8

Not saying these are the best lenses (well the 28 1.8 has a decent rep) - but all of these are non-L's and have a "consumer" price tag. You say you wan't these in EF-S but that doesnt really make sense as they are kind of standard lengths. Not like a 17mm on EF-S which is equivalent view to 28mm on full frame for instance.

I have a Sigma 30 1.4 and it is great and is a similar field of view to 50mm on full frame - which is what i wanted. Theres pretty much something out there to cover almost everything even on a 1.6 crop body

[wickerprints]

No EF-S super-teles along the same lines as the 300mm and longer L lenses, because even with a smaller image circle, you still need a lot of glass. Remember that f-number = focal length / entrance pupil diameter. This formula is true regardless of how large an image circle is cast by the lens. Since front element diameter > entrance pupil diameter, this sets a lower bound on how small you can make the front of the lens.

So then it comes down to the question of why anyone would purchase an EF-S 400/2.8 IS. Granted, its field of view would be equivalent to a 640/2.8 lens in 35mm format. So it would make more sense to make, say, an EF-S 400/5.6 or 200/2.8 but we already have EF versions of these lenses.

See how that works out? If you want to reap the benefit of a more narrow equivalent field of view due to the smaller image circle, then you pay with a larger and heavier lens, or you give up some max aperture. For instance, 200/2 on APS-C would give equivalent field of view to a 320/2 lens and you pay for it in weight, but there isn't such a lens for 35mm format. Or you could get the 200/2.8 for APS-C giving you the 35mm equivalent field of view of a 320/2.8 and now you've benefited with a smaller, lighter lens compared to the 300/2.8 on the 35mm format.

Therefore, the reason why we don't see EF-S super-teles is that the existing EF lens selection is more than adequate. They are expensive, yes, but as I have pointed out, this range of focal length is constrained by a simple relationship between focal length and f-number, and less by the size of the image circle.

[mikekelley]

Personally, I feel just the opposite. I think Canon needs to release more lower cost EF-S lenses. As was mentioned earlier, there are plenty of high grade Canon EF mount lenses that work just fine on 1.6 crop bodies. If the EF-S line is meant to be a consumer grade line then Canon should actually make some lenses that the typical consumer is comfortable paying for. Look at some of the lenses in the lineup. The 10-22 costs $700+ and the 17-55 costs over $1000. If those are consumer level prices I'll eat my shorts. I really hope the new Tamron 17-50VC turns out to be a great lens. I probably won't buy it but I would love to see sales of the 17-55 drop in a big way to send Canon a message. Chances are neither will happen, but it would be nice. If Canon decides to release any prime EF-S lenses, I would hope they would shoot for something more like the Sigma 30 f/1.4 rather than a 35L.

**Edit: I should note that I feel the EF-S 55-250 is more the way I think the EF-S line should go. There is a very useful lens priced at a level many consumers and amateurs can stomach.


I mean, I don't think they need to make any more variable aperture lenses. The 18-55 IS and 55-250 are plenty for the average consumer who is going to leave it on a basic sports zone/green box and wants some zoom.

See Olympus and their 300mm f/2.8

Which is the equivalent of a 600mm f/2.8 :)

Also their 150mm f/2...Thats a spanking great lens right there...

Oh! What about the 90-250...

Look, Criticize 4/3rds, but they do have some amazingly cool lenses [And optically, Those lenses are all the equal of Canon/Nikon..they're impressive]



MSRP for the Tamron 17-50 VC with its crappy AF and build is $750...And it will likely sell for $750 [Tamron sticks to their MSRP, i see little price deviation on their part at different stores]

Note you can get a used 17-55 in great shape for that..Not worth it IMHO to get the Tamron..

ok...Now im going to get flogged for this, but whats the big deal with variable aperture...Seriously? Just because its variable aperture doesnt make it a poor lens [See, my 100-400 for instance]


If it was like 2.8-4, there'd be a case for variable aperture lenses in the 15-100mm range. But theyre all 3.5-5.6, which, in that range, is pretty much useless to me (and a lot of people who like a lot of control over their photos)

Variable aperture to 5.6 at 400mm is a lot different than variable aperture at 55mm and 5.6. One is terrible, the other, not so much.

I do think that olympus's funky lenses are freakin' awesome. I'm glad that there is a company making some waves, I just wish there was a bigger following and more support.

[KenjiS]

If it was like 2.8-4, there'd be a case for variable aperture lenses in the 15-100mm range. But theyre all 3.5-5.6, which, in that range, is pretty much useless to me (and a lot of people who like a lot of control over their photos)

Variable aperture to 5.6 at 400mm is a lot different than variable aperture at 55mm and 5.6. One is terrible, the other, not so much.

Good point, I didnt even think about that...I'm still sorta partial to the 15-85 though...I actually dont use my 28-75 wide open very frequently because of DOF it seems...For times i want control i wanted a 50mm f/1.4 anyways :)

I guess its different strokes for different folks thats all

I do think that olympus's funky lenses are freakin' awesome. I'm glad that there is a company making some waves, I just wish there was a bigger following and more support.

Ever held the E-3? That thing is built to survive a war...I played with an E-3 and the 12-60 SWM one day, Honestly, I was floored by it more than the D700 I played with not some 5 minutes before...And the deal they were running was DAMN good [It was sometihng like $2300 for the E-3, 12-60 SWM and one of the high end Olympus speedlites...Shocking good deal really because the E-3 is Olympus' highest end professional camera..]

[ed rader]

Then by this logic, only people with 1d bodies (pro bodies) need non-consumer grade glass.

As far as I am aware, the only thing considered "pro" is the 1d line.


5d and 5d II are considered pro too.

ed rader

[KenjiS]

5d and 5d II are considered pro too.

ed rader

7D is as well isnt it?

[ed rader]

7D is as well isnt it?


i'm not sure but i don't think so.

ed rader

[msowsun]

Canon is calling their 7D a "mid range" camera.

http://www.usa.canon.com/templatedata/pressrelease/20090901_eos7d.html

CANON U.S.A. REDEFINES MID-RANGE DSLR CATEGORY WITH THE NEW EOS 7D DIGITAL SLR CAMERA

The 5D doesn't have weather sealing. Is it really considered professional?

[KenjiS]

Canon is calling their 7D a "mid range" camera.

http://www.usa.canon.com/templatedata/pressrelease/20090901_eos7d.html



The 5D doesn't have weather sealing. Is it really considered professional?

In my book if the 5D Mark II is considered pro so is the 7D...

[wickerprints]

Why does it matter whether a particular body is considered "pro?" In particular, what material advantages does such a designation confer? ???

[KenjiS]

Why does it matter whether a particular body is considered "pro?" In particular, what material advantages does such a designation confer? ???

Mainly build quality...Thats all

And by that definition, the 7D has at least the same level of sealing as the 5D II, a shutter unit rated to the same cycles and very robust-looking construction...

[Hogloff]

10-22 is sharp as a tack but slow as a dog.

10-2x with constant 2.8 would be just fantastic.

Some nice EF-S primes would be great too. I'm sure that they could be made significantly cheaper (less glass) and super-sharp.

How about some EFS super-teles? Again, they could be made cheaper, lighter and just as sharp as their EF equivelents, if they kept within the EFS build standard.

Less variable aperture crapola. More serious lenses for EFS mounts, please.

I know that they are capable, I just don't get why they continue to pump out consumer grade minutiae.

Canon has stated on more than 1 occasion that the EFS sytems are consumer oriented...thus the consumer grade lenses. You want professional specs, by their L glass.

[_aravena]

What!?! WHY!?! How? Have you seen the prices of the EF-S and you want them more expensive. Yes they have cheap ones but only starters lenses then it jumps!

[msowsun]

Mainly build quality...Thats all

And by that definition, the 7D has at least the same level of sealing as the 5D II, a shutter unit rated to the same cycles and very robust-looking construction...

I thought I read somewhere that the 7D has the same weather sealing capability as the 1D series, but the 5D Mk II only has the same as the 40D.

Am I wrong?

[mikekelley]

If it delivers results I will pay for it - same goes for many people out there.

As far as I was aware, the 5dII was a full-frame prosumer camera.

The 1dsIII is the full frame 'pro' camera.

[mikekelley]

Re: weather sealing of cameras...

It is obvious by reading this thread and others that NOBODY (and myself included) on this forum knows exactly how weather sealed any body except a 1d is. One thing is for sure and that is that a 1d is more weatherproof than every other body.

Someone has to sacrifice a 40d, 50d and 5d2 and then we'll find out for sure. Who likes to burn money?

[ed rader]

I thought I read somewhere that the 7D has the same weather sealing capability as the 1D series, but the 5D Mk II only has the same as the 40D.

Am I wrong?


i heard the 7d is similar to the 5d II.

ed rader

[wickerprints]

Mainly build quality...Thats all

And by that definition, the 7D has at least the same level of sealing as the 5D II, a shutter unit rated to the same cycles and very robust-looking construction...

But here's the question: Is "pro/consumer" something that Canon specifically designates for each of its products and then builds that product to a higher standard because of that designation? If so, then there is a way to objectively call something "pro" and a reason to do so, because build quality, as you put it, would then be driven by what Canon calls it.

But if photographers or someone other than Canon calls a particular product "pro," then how does that relate to build quality as you claim? If they say something is "pro" because it has higher build quality, then it doesn't really make any sense to use that nomenclature because it adds nothing of value. One would simply say "higher build quality."

My point is that trying to split the product line into "pro" versus "consumer" is not a useful distinction because it is not clearly made by Canon, and if only one functional advantage is evident (e.g. build quality) then what is the purpose of giving what basically amounts to another name?

[RPCrowe]

I love my 17-55mm f/2.8 IS lens but, the major gripe about this lens is that people don't want to pay that high of a price for a non-L lens.

I love to use my 17-55mm along with my 70-200mm f/4L IS lens on a pair of 1.4x cameras.

However, an even better combination would be a 50-150mm f/2.8 IS EFS lens which has image quality equal to the 70-200mm f/4L IS. This lens could be made considerably lighter in weight than the present 70-200mm f/2.8L IS lens and the 50-150mm would be even nicer in combination with the 17-55mm f/2.8 IS lens.

By the way, there is a pretty darn nice 11-16mm f/2.8 Tokina for 1.6x wide angle needs. Additionally, the 12-24mm f/4 Tokina isn't quite shabby either.

[msowsun]

Canon says the 7D has the same weather sealing as the EOS-1N.

http://www.canon-europe.com/For_Home/Product_Finder/Cameras/Digital_SLR/EOS_7D/index.asp

Magnesium alloy body with environmental protection
The magnesium alloy body offers weather sealing equivalent to the acclaimed EOS-1N

The 5D Mk II has some "sealing materials" , but so does the 40D.

http://www.dpreview.com/reviews/CanonEOS5DMarkII/page5.asp

Construction and environmental seals
As you can see from the first image below the body is made up of three pieces of magnesium alloy, the only plastic elements being the sides and the base. With the advent of the Mark II Canon are finally talking about the dust / water resistance of the body, the second image below shows these seals, Canon's description: "The battery compartment, memory card door, LCD and the camera buttons are all fitted with sealing materials (indicated in red). In addition the adoption of high precision split-level alignment of the magnesium-alloy external covers, high precision dial construction and external rubber grip covers (indicated in green), has improved the camera's dust and water resistance."

[DaveSt]

What!?! WHY!?! How? Have you seen the prices of the EF-S and you want them more expensive. Yes they have cheap ones but only starters lenses then it jumps!

That was kind of my thinking as well. Then again, perhaps Canon prefers to keep the quality(?)/price high realizing that there are plenty of options already available for lower cost crop lenses (Sigma, Tamron, Tokina). Those other companies are doing a pretty nice job of producing good lenses at what I would consider consumer level prices.

[Wilt]

Why does it matter whether a particular body is considered "pro?" In particular, what material advantages does such a designation confer? ???
perceived status among other Canon photographers :confused:

[secabeen]

etter combination would be a 50-150mm f/2.8 IS EFS lens which has image quality equal to the 70-200mm f/4L IS. This lens could be made considerably lighter in weight than the present 70-200mm f/2.8L IS lens and the 50-150mm would be even nicer in combination with the 17-55mm f/2.8 IS lens.


I agree with this. Sigma has a decent 50-150 f/2.8 HSM, but it doesn't have IS. I'd happily pay L-level prices for a 50-150 f/2.8 IS USM.

Once the 7d comes out, it will be interesting to see what Canon does. They clearly are committed to the crop body, so I could certainly see a hole in their lineup in the mid-zoom range for crop sensors. (Yes, you can use a 70-200, but that's not ideal for crop just like the 17-55 range isn't ideal for a FF)

[ed rader]

perceived status among other Canon photographers :confused:


not exactly.

http://www.usa.canon.com/consumer/controller?act=MultiMiscPageAct&key=CPS_Member&keycode=CPS&fcategoryid=111

ed rader

[happy hopping]

In the "Canon DSLR, THe Ultimate Photographer's guide by Chris Grey", he said EF-S lens is just a cheap ver. of EF lens.

Read the other thread on 17 - 85 len w/ Error 99, myself included, we have 3 members in 1 days posting the same problem. Turns out it's the Diaphragm that fails.

[wickerprints]

not exactly.

http://www.usa.canon.com/consumer/controller?act=MultiMiscPageAct&key=CPS_Member&keycode=CPS&fcategoryid=111

ed rader

See, I understand the CPS thing, but I don't get the impression that's what other people generally mean when they refer to "pro" bodies or lenses. After all, the CPS requirements are much more lenient. Almost all EF lenses currently in production are considered "pro" lenses, for example.

I suppose you could say only the "platinum" level of CPS would be counted as "pro" but this seems arbitrary and not at all implied by Canon or CPS.

That's why I asked the question the way I did. Do you call something "pro" because it satisfies a certain set of criteria? Or does it satisfy certain criteria because it is called "pro?" These are not equivalent. And who has decided on these criteria anyway?

[Wilt]

not exactly.

http://www.usa.canon.com/consumer/controller?act=MultiMiscPageAct&key=CPS_Member&keycode=CPS&fcategoryid=111

ed rader

Ed, I don't think most photographers even have a clue what Canon thinks is satisfactory qualification for CPS...what matters to most is more simple criteria..."oh, he uses a 1Ds camera" and maybe now "oh, he uses a 7D"...'perceived status'

[mikekelley]

Ed, I don't think most photographers even have a clue what Canon thinks is satisfactory qualification for CPS...what matters to most is more simple criteria..."oh, he uses a 1Ds camera" and maybe now "oh, he uses a 7D"...'perceived status'

Yep.

That.

[XterraJohn]

As long as CPS is being brought into the discussion, please note that the 60 macro, 10-22, and 17-55 EF-S lenses are all listed as "pro" lenses.

[toxic]

I think they need to sort out the midrange lenses first...an EF-S 20 or 30 would be nice, though I doubt it'll arrive. I really hope they come out with an f/3.5-4.5 or f/2.8-4 standard zoom slotted below the 17-55, but I dunno if that's happening since the 15-85 was just released.

Other than that, a 50-150/2.8 would be great. And a 17-55 II to address flaring and dust issues, and smooth the bokeh a little bit....

[Chez Wimpy]

Some nice EF-S primes would be great too. I'm sure that they could be made significantly cheaper (less glass) and super-sharp.

Yes, but then why would you upgrade to FF? ;) Canon doesn't want to make it *that* easy on the consumer.

See Olympus and their 300mm f/2.8

Which is the equivalent of a 600mm f/2.8 :)

Well, to be accurate, the equivalent of a 600mm f5.6 lens on FF... lest Olympus claims their C-2100UZ has a "380mm" (70mm) f3.5 zoom (!!!).


It is obvious by reading this thread and others that NOBODY (and myself included) on this forum knows exactly how weather sealed any body except a 1d is.

And not even that... Canon doesn't publish those specs.

[Joaaso]

I agree with this. Sigma has a decent 50-150 f/2.8 HSM, but it doesn't have IS. I'd happily pay L-level prices for a 50-150 f/2.8 IS USM.

Once the 7d comes out, it will be interesting to see what Canon does. They clearly are committed to the crop body, so I could certainly see a hole in their lineup in the mid-zoom range for crop sensors. (Yes, you can use a 70-200, but that's not ideal for crop just like the 17-55 range isn't ideal for a FF)
+1
I'm dreaming of a EF-S 50-200 f2.8 IS USM, but in the lack of such a lens, I've been debating with myself wether I should get a 200 f2.8L or 135 f2L instead (or both), but a 50-150 with IS could be a decent compromise...

[LightRules]

How about some EFS super-teles? Again, they could be made cheaper, lighter and just as sharp as their EF equivelents, if they kept within the EFS build standard

Don't know if anyone's clarified, but see specifically posts 33 - 41:

http://photography-on-the.net/forum/showthread.php?t=239903&highlight=ef-s+telephoto

[mikekelley]

Interesting enough.

[Daniel Browning]

10-22 is sharp as a tack but slow as a dog.

10-2x with constant 2.8 would be just fantastic.


Agreed. Unfortunately, one of the problems with such a lens is the ratio between back focus distance and image circle. Compared to FF, the APS-C has a much larger ratio (despite the few mm that Canon shaved off for EF-S, which means "short back focus"). That makes it impossible to scale the design of the 16-35 f/2.8 L down to a 10-22 f/2.8. The 10-22 needs more (relative) retrofocus, which increases cost (or aberration), and that can offset some of the gain from only needing to cover a smaller image circle.


Some nice EF-S primes would be great too. I'm sure that they could be made significantly cheaper (less glass) and super-sharp.


They can be cheaper, but perhaps not as much as you think (for the wides), again because the backfocus ratio is different than full frame.


How about some EFS super-teles? Again, they could be made cheaper, lighter and just as sharp as their EF equivelents, if they kept within the EFS build standard.


That's a common misconception. As others have pointed out, there is little cost savings for lenses longer than 85mm, and none for lenses longer than 150mm.


Less variable aperture crapola. More serious lenses for EFS mounts, please.


Seconded.


I know that they are capable, I just don't get why they continue to pump out consumer grade minutiae.

I think part of it is purposeful market segmentation, and part is just catering to the existing market divisions. Now that full frame bodies are getting cheaper and cheaper ($2000 A850!), they may go the other way and encourage any prime/quality shooters to go full frame. But hopefully not. The 7D may give them more reason to build quality EF-S primes like an EF-S 15mm f/2.

I don't see much point in EF-S primes. For the discriminating 1.6X shooter there already exists the 14L, 24L, 35L, 50L etc.


Those lenses are not the ideal replacement of EF-S 9mm, 15mm, 22mm, and 32mm.


etc. And to be honest, owning a handful of those pricy L primes for exclusive use on 1.6X bodies makes little sense as their speed advantage could be more than compensated for by getting the slower standard primes and a used 5D.


Agreed. But if Canon made EF-S primes in those focal lengths, their use would make sense. A large part of their cost goes to the larger image circle.


Where the EF-S line is deficient is in giving the 1.6X shooter a high quality, fast wide angle zoom (such as the EF-S 17-55) that is also built well. IMO the 17-55 is a nice lens (I used to have one) but it costs too much for the way it is built. It also needs better resistance to veiling flare.


Agreed.


The new EF-S 15-85 looks interesting, but the variable max aperture and the build again leave it wanting against something like a 5D with a 24-105L.

Agreed.

If Canon decides to release any prime EF-S lenses, I would hope they would shoot for something more like the Sigma 30 f/1.4 rather than a 35L.


I think it's more likely that they will just copy Nikon and do an EF-S 35mm f/1.8.


**Edit: I should note that I feel the EF-S 55-250 is more the way I think the EF-S line should go. There is a very useful lens priced at a level many consumers and amateurs can stomach.

So far it appears that Canon agrees with you.


ok...Now im going to get flogged for this, but whats the big deal with variable aperture...Seriously? Just because its variable aperture doesnt make it a poor lens [See, my 100-400 for instance]

Agreed. I did't mind variable aperture either (back when I had those kinds of lenses). It's just a trade-off.

See Olympus and their 300mm f/2.8

Which is the equivalent of a 600mm f/2.8


300mm on Four Thirds has an angle of view equivalent to 600mm on FF35.
f/2.8 on Four Thirds has light gathering power, depth of field, and diffraction equivalent to f/5.6 on FF35.
It's equivalent to 600mm f/5.6, not 600mm f/2.8.

(Most people get hung up on f-number, not realizing that FF35 at ISO 1600 has the same noise level as Four Thirds at ISO 400. FF35 can have less light per area and more light to come out to the same noise level in the end.)


Also their 150mm f/2...Thats a spanking great lens right there...


The equivalent FF35 lens (a 300mm f/4) is half the price and includes I.S.


Oh! What about the 90-250...


Pretty nice except for the $5,000 price. The equivalent FF35 lens would be 180-500mm f/5.6.


MSRP for the Tamron 17-50 VC with its crappy AF and build is $750...


I love my 17-50. The only bad thing about the AF is how loud it is. It's lightning fast and accurate. I wish they would add VC for less than $300, but it will probably still sell like hot cakes.

Need a 35mm? try the 28 f1.8 or the 35mm f2
Need a 20mm? try the 20 f2.8


The Nikon 35mm f/1.8 and Sigma 30mm f/1.4 blow Canon's old EF lenses out of the water. And for good reason: EF lenses have to cover an image circle that has 2.5X more area! A dedicated Canon EF-S 35mm f/1.8 for $200 or less is just what the doctor ordered.

I'm hoping that Canon will build a wider variety of EF-S lenses, but I'm not holding my breath.

[msowsun]

Originally Posted by KenjiS http://photography-on-the.net/forum/images/buttons/viewpost.gif (http://photography-on-the.net/forum/showthread.php?p=8595824#post8595824)
See Olympus and their 300mm f/2.8

Which is the equivalent of a 600mm f/2.8


300mm on Four Thirds has an angle of view equivalent to 600mm on FF35.
f/2.8 on Four Thirds has light gathering power, depth of field, and diffraction equivalent to f/5.6 on FF35.
It's equivalent to 600mm f/5.6, not 600mm f/2.8.

(Most people get hung up on f-number, not realizing that FF35 at ISO 1600 has the same noise level as Four Thirds at ISO 400. FF35 can have less light per area and more light to come out to the same noise level in the end.)


Originally Posted by KenjiS http://photography-on-the.net/forum/images/buttons/viewpost.gif (http://photography-on-the.net/forum/showthread.php?p=8595824#post8595824)
Also their 150mm f/2...Thats a spanking great lens right there...


The equivalent FF35 lens (a 300mm f/4) is half the price and includes I.S.



I always assumed that an Olympus 4/3 lens was just as fast as the f/stop says on the lens.

Daniel, are you saying that when I mount a lens on my 40D, it is no longer the f/stop that it says on the lens?

So my 70-200 f/2.8 is really equivalent to a 112-320 f/4.48?

I know teleconverters cause a loss of light. But can you explain how this loss of light happens due to crop factor?

[DaveSt]

The Nikon 35mm f/1.8 and Sigma 30mm f/1.4 blow Canon's old EF lenses out of the water. And for good reason: EF lenses have to cover an image circle that has 2.5X more area! A dedicated Canon EF-S 35mm f/1.8 for $200 or less is just what the doctor ordered.

I'm hoping that Canon will build a wider variety of EF-S lenses, but I'm not holding my breath.

I couldn't agree more. I would order that lens in a heartbeat.

[Drozz119]

7d is a Sports/Wildlife camera.. Right??

You would think, they would put out a lens that would fit it perfectly..

The EF-S 120-300mm USM IS 2.8 (192-480mm)

Imagine, as a sports shooter, to have both the 300 2.8 and the 400 2.8 in one lens!

It would probably be the size of the sigma 100-300 f4, and would fly off the shelves @ $2000..

Are they scared it would take away from the sales of the 300l and 400l?

[Daniel Browning]

I always assumed that an Olympus 4/3 lens was just as fast as the f/stop says on the lens.

It is. And the focal length is exactly what it says on the lens too: 300mm.


Daniel, are you saying that when I mount a lens on my 40D, it is no longer the f/stop that it says on the lens?


No, that's not what I'm saying at all.

The focal length never changes, no matter what body you put a lens on. However, the FF35 *equivalent* focal length does. For example, 300mm on APS-C has an angle of view equivalent to 480mm on FF35.

Similarly, the f-number never changes, no matter what body you put a lens on. However, the FF35 *equivalent* f-number does. For example, 300mm f/2.8 on APS-C is equivalent to 480mm f/4.5 on FF35 for light grathering capability, depth of field, and diffraction.


So my 70-200 f/2.8 is really equivalent to a 112-320 f/4.48?


Yes. Try it yourself and see how similar the shots are:


100mm f/2.8 ISO 200 1/500 on APS-C
160mm f/4.5 ISO 640 1/500 on FF35.


The angle of view, depth of field, and diffraction will be the exact same. The brightness and noise will be the exact same as well.

Demonstration using the 5D and 20D (http://forums.dpreview.com/forums/read.asp?forum=1032&message=16107908)


I know teleconverters cause a loss of light. But can you explain how this loss of light happens due to crop factor?

The total amount of light *per area* depends solely on f-number. f/2.8 on a 6x4.5mm digicam has the exact same exposure (total light per area) as f/2.8 on 60x45mm medium format. But as everyone knows, digicams capture less light because they have 100 times less total area. (The size of the lens, too, hints at how much less light they gather.)

The total light gathering capability depends on exposure *and* area (for digital formats). Since FF35 has four more area than Four Thirds, it gathers four times more light (two stops) when given the same exposure (light per area). Therefore, to get the same amount of light on full frame, you have to reduce exposure by two stops. That means less light per area, but the same total amount of light.

HTH HAND.

[LightRules]

The EF-S 120-300mm USM IS 2.8 (192-480mm)

Imagine, as a sports shooter, to have both the 300 2.8 and the 400 2.8 in one lens!

It would probably be the size of the sigma 100-300 f4, and would fly off the shelves @ $2000..

Are they scared it would take away from the sales of the 300l and 400l?

<sigh> See post #43 above.

[msowsun]

I metered a scene with my 70-200 2.8 on my EOS 40D (1.6 crop) and lSO with my EOS 630 (old full frame film camera).

Both cameras gave the exact same shutter speed and aperture at the same ISO. So where is there a loss of light?

[mikekelley]

I don't think that there is a loss of light at all.

[msowsun]

So my 70-200mm 2.8 is equivalent to a 112-320mm f/2.8

And an Olympus 300mm f/2.8 is equivalent to a 600mm f/2.8


YES?

[mikekelley]

Yes.

I think he might be confusing it with the change in depth of field.

But there is no light loss.

[toxic]

I metered a scene with my 70-200 2.8 on my EOS 40D (1.6 crop) and lSO with my EOS 630 (old full frame film camera).

Both cameras gave the exact same shutter speed and aperture at the same ISO. So where is there a loss of light?

The exposure is the same. The amount of light (photons hitting the sensor) is not.

[msowsun]

Yes.

I think he might be confusing it with the change in depth of field.

But there is no light loss.

He said:

300mm on Four Thirds has an angle of view equivalent to 600mm on FF35. f/2.8 on Four Thirds has light gathering power, depth of field, and diffraction equivalent to f/5.6 on FF35. It's equivalent to 600mm f/5.6, not 600mm f/2.8.

He did say "light gathering". That is the part I had trouble with.

But it did make me think for a minute. :confused:

[tkbslc]

It's just semantics.

Imagine light is rainfall. You have a 5x5' tub and a 5x10' tub. In 10 minutes twice as much rain hits the 5x10' tub, but the average rainfall is the same for both tubs. And if your goal is to fill up each tub with 6 inches of water, both will fill up in the same amount of time, so neither system has an advantage.

[mikekelley]

This has nothing to do with rumors, so why was it moved?

[wickerprints]

I metered a scene with my 70-200 2.8 on my EOS 40D (1.6 crop) and lSO with my EOS 630 (old full frame film camera).

Both cameras gave the exact same shutter speed and aperture at the same ISO. So where is there a loss of light?

I don't think that there is a loss of light at all.

So my 70-200mm 2.8 is equivalent to a 112-320mm f/2.8

And an Olympus 300mm f/2.8 is equivalent to a 600mm f/2.8


YES?

Yes.

I think he might be confusing it with the change in depth of field.

But there is no light loss.

Daniel is correct and has already explained the situation precisely.

The amount of light per unit area is the same for the same exposure. But one sensor is physically smaller than the other.

Total light gathered = (light per unit area) x (area of sensor).

If one is happy to print/display the resultant images at the same relative size of the sensors that produced them (i.e., same enlargement ratio), then there is no difference in equivalent exposure, DOF, etc. They would be identical in every way except that (1) the smaller one would appear center cropped, and (2) any differences in pixel density may show as a relative difference in sharpness depending on shooting conditions.

But in order to realize the longer focal length equivalent of APS-C compared to 35mm, your enlargement ratio must be 1.6x greater. And this is where you "lose light," as Daniel pointed out.

[tkbslc]

Daniel is correct and has already explained the situation precisely.

The amount of light per unit area is the same for the same exposure. But one sensor is physically smaller than the other.

Total light gathered = (light per unit area) x (area of sensor).

If one is happy to print/display the resultant images at the same relative size of the sensors that produced them (i.e., same enlargement ratio), then there is no difference in equivalent exposure, DOF, etc. They would be identical in every way except that (1) the smaller one would appear center cropped, and (2) any differences in pixel density may show as a relative difference in sharpness depending on shooting conditions.

But in order to realize the longer focal length equivalent of APS-C compared to 35mm, your enlargement ratio must be 1.6x greater. And this is where you "lose light," as Daniel pointed out.

You are not losing light, you are magnifying each pixel more. Each pixel does not get darker the more you magnify it.

[wickerprints]

You are not losing light, you are magnifying each pixel more. Each pixel does not get darker the more you magnify it.

True, but then what is the equivalent sensitivity at the moment of capture if this is done?

[msowsun]

True, but then what is the equivalent sensitivity at the moment of capture if this is done?

Who cares?

The main point I want to make here is that crop factor does not effect the f/stop.

If you need a fast low light 2.8 lens for fast shutter speeds to stop action, your lens will always be f/2.8 for exposure purposes and you will always be able to use a fast shutter speed.

SO.... An Olympus 4/3 300mm f/2.8 WILLbe equivalent to a 600mm f/2.8 (35mm FF)

If some one is shooting sports with an Olympus 300mm 2.8, and there is a guy beside him with the 5D and 600mm 2.8, they both will be using the same shutter speed, and both will have the same relative image.

[mikekelley]

Who cares?

The main point I want to make here is that crop factor does not effect the f/stop.

If you need a fast low light 2.8 lens for fast shutter speeds to stop action, your lens will always be f/2.8 for exposure purposes and you will always be able to use a fast shutter speed.

SO.... An Olympus 4/3 300mm f/2.8 WILLbe equivalent to a 600mm f/2.8 (35mm FF)

If some one is shooting sports with an Olympus 300mm 2.8, and there is a guy beside him with the 5D and 600mm 2.8, they both will be using the same shutter speed, and both will have the same relative image.

Yes! This. I do not care about "total" light hitting a sensor. OBVIOUSLY it is going to be more if the sensor is bigger, but it doesn't change the exposure.

[Wilt]

SO.... An Olympus 4/3 300mm f/2.8 WILLbe equivalent to a 600mm f/2.8 (35mm FF)

I just wanted to alter the wording so as to most precisely state what is so prone to misinterpretation by beginners:

Using a 300mm f/2.8 lens on an Olympus 4/3 body WILL be equivalent in the Field of View which is captured by using a 600mm f/2.8 on 35mm FF body which WILL be equivalent in the Field of View which is captured by using a 375mm f/2.8 lens on APS-C body

[wickerprints]

Who cares?

The main point I want to make here is that crop factor does not effect the f/stop.

If you need a fast low light 2.8 lens for fast shutter speeds to stop action, your lens will always be f/2.8 for exposure purposes and you will always be able to use a fast shutter speed.

SO.... An Olympus 4/3 300mm f/2.8 WILLbe equivalent to a 600mm f/2.8 (35mm FF)

If some one is shooting sports with an Olympus 300mm 2.8, and there is a guy beside him with the 5D and 600mm 2.8, they both will be using the same shutter speed, and both will have the same relative image.

It may not make much difference from a practical standpoint to you, but for the purposes of understanding, what Daniel said is correct regardless of whether or not you want to believe it.

Please read his post and actually click on the link he provided, before you care to argue this any further.

[pwm2]

The Olympus lenses are freaking awesome. And freaking expensive. Which kind of proves the fact that just having a smaller image circle does not make them cheap.

Canon should release some really good wide lenses and very fast lenses in EF-S, since the 1.6x crop requires wide lenses that does not exist in the EF lineup. I also feel that Canon should have made the 7D fully weathersealed and created an L-grade weathersealed 17-55 lens. And they should consider an L-class 10-22/2.8. And they could produce a number of cheaper really wide primes in EF-S. For the really wide lenses, the smaller image circle should make a significant price difference, even if that doesn't applies to tele lenses.

For the tele side, there really is no real advantage to release EF-S lenses. Look at the price for the Olympus 300/2.8. Well, take a closer look at the price of the Canon 300/2.8. Do buy that one, and you will get the same depth-of-field. There isn't a reach difference between the 1.6x cameras and the 2x cameras. Just manually crop the photos you get from the camera, and you will have exactly the same reach and depth of field as the four-thirds have. The biggest advantage with four thirds is that the camera body can be made smaller, and the wide lenses too. For tele, the disadvantage is that your heavy lens will be just as heavy, but with a smaller body sticking out on one side.

[tkbslc]

It may not make much difference from a practical standpoint to you, but for the purposes of understanding, what Daniel said is correct regardless of whether or not you want to believe it.

Please read his post and actually click on the link he provided, before you care to argue this any further.

It could easily be inferred that the differences are due to magnification factor and pixel density, rather that "light gathering ability".


And to get back on topic:

Olympus uses a 2x crop factor, so that would require 1/2 the diameter of a FF lens in theory:

Olympus 4/3 only 70-300 f4-5.6 is 80x127mm
Sigma 70-300 DG Full frame 70-300 f4-5.6 is 76x119mm.

So olympus can't even make their 2x crop 70-300 any smaller than Sigma's Full frame. Why should Canon bother?

[pwm2]

I thought I read somewhere that the 7D has the same weather sealing capability as the 1D series, but the 5D Mk II only has the same as the 40D.

Am I wrong?
To my knowledge you are. The 5Dmk2 and the 7D has the same weather sealing as an ancient-time 1-series body.

When I bought the 40D, Canon also pushed a lot for the weather sealing, making people excited.

I'm a bit surprised that Canon spends time pushing weather sealing like that. If Canon figured that weather sealing was important to the customer group interested in the 40D or the 5Dmk2 or the 7D, they should instead consider actually improving the weather sealing. You get happier customers if you deliver the goods, instead of just letting the mouth go off.

A large number of APS-C owners already owns a lot of sealed L lenses, and have spent a lot of money on rain gear for their cameras. That should make it obvious to Canon that people who buy the xxD or xD are prepared to spend a little extra on getting the rain protection integrated in the camera in the first place. A $50 increase in price would be enough for significant improvements of the weather sealing. Buying Kata or AquaTech or maybe surviving with a plastic bag that you suddenly notices a hole in is way more expensive for us buyers.

[JeffreyG]

So my 70-200mm 2.8 is equivalent to a 112-320mm f/2.8

And an Olympus 300mm f/2.8 is equivalent to a 600mm f/2.8


YES?

No,

Two main issues that I think are clouding this.

You are focused on attaining the same shutter speed at the same ISO. But note that the larger sensor will deliver the same noise at much higher ISO.

Also, the larger sensor will give much less DOF at the same framing.

So, a 300/2.8 lens on 4:3 sensor at say, ISO 400 will give the same framing, DOF and noise as a 600/5.6 lens at ISO 1600 on a FF sensor.

This is why 300/2.8 on 4:3 format really is equivalent in all senses to using 600/5.6 on FF.

[pwm2]

Everyone here has heard that a FF35 has more narrow field of view than a APS-C?

No one questions it, because you have heard it many times?

But with the same lens mounetd on a 5Dmk2 or a 50D, the depth of field is the same when the image is projected on the sensor. It is just that the depth-of-field relates to the size of the sensor - or the size of the crop, if you crop your photo.

If I take the photo with a 1Dsmk3 or a 5D/5Dmk2 with a 300/2.8 and print the full image, it will have more narrow DOF than if a 50D or 7D had taken the shot with the same lens. The 50D or 7D would have captured just the center part of the scene, so the print would only show the center part. The 50D print would look like it had less DOF.

If I now take the full image from the 1Ds3 or 5D and crop it with a factor 1.6 and print, I will have produced the same image as the 50D could capture. And with the same DOF.

That is why a 300/2.8 on a four thirds does not directly scale to a 600/2.8 on a FF35 body.

That is also why a APS-C or APS-H don't have more reach than a FF35 body. When you crop the same part of the image, all cameras will have the same reach. The question is just how many $$$ it will cost to get enough pixels on that cropped section.

[wickerprints]

Olympus uses a 2x crop factor, so that would require 1/2 the diameter of a FF lens in theory:

No. I see this "theory" repeated numerous times but it is incorrect. Crop factor does not proportionally relate to front element diameter. There is a weak relationship between the sensor size and the total amount of glass, but it is dominated by the size of the front element as the focal length gets sufficiently long.

f-number = focal length / entrance pupil diameter.

front element diameter > entrance pupil diameter.

Therefore,

front element diameter > focal length / f-number.

There is no way around this. When focal length is sufficiently large or f-number sufficiently small, the front element diameter must be very large, independent of the sensor size.

Olympus 4/3 only 70-300 f4-5.6 is 80x127mm
Sigma 70-300 DG Full frame 70-300 f4-5.6 is 76x119mm.

So olympus can't even make their 2x crop 70-300 any smaller than Sigma's Full frame. Why should Canon bother?

Indeed, that's why Canon doesn't bother. The size of these lenses is limited by a multitude of factors, one of which I have described, and other limitations being the requirement to include the AF motor and the zoom barrel itself, among others.

[wickerprints]

That is also why a APS-C or APS-H don't have more reach than a FF35 body. When you crop the same part of the image, all cameras will have the same reach. The question is just how many $$$ it will cost to get enough pixels on that cropped section.

Correct.

And this gives me an idea...(half joking) Maybe Canon should create an APS-C/35mm hybrid sensor! The total sensor area is 24x36mm, but make it high pixel density in the center APS-C rectangle, and low pixel density in the periphery. If you want the full frame picture, you set the camera to process the high density center to give the low-density equivalent pixels. Otherwise, just throw away the periphery for APS-C shots. That way, APS-C people can brag about how they get more reach, and the FF people can say they get the advantages of the full image circle. :p

[DDCSD]

See Olympus and their 300mm f/2.8

Which is the equivalent of a 600mm f/2.8 :)

Also their 150mm f/2...Thats a spanking great lens right there...

Oh! What about the 90-250...


So, you want a $6,000 300 f/2.8, a $2,200 150mm f/2 and a $5100 90-250 f/2.8 that will only work on 1.6 crops cameras instead of the $4,100 300 f/2.8, $1350 135 f/2 and $1,200 70-200 f/2.8 lenses that Canon already makes that fit on all EOS cameras?

[Daniel Browning]

I metered a scene with my 70-200 2.8 on my EOS 40D (1.6 crop) and lSO with my EOS 630 (old full frame film camera).

Both cameras gave the exact same shutter speed and aperture at the same ISO. So where is there a loss of light?

As I said before, the "loss of light" is from the loss of area. In the scenario you described above, there is more light falling on the film in the EOS 630. However, some of my comments (e.g. brightness and noise) only applied to digital cameras, because film has a nonlinear relationship between film area and total light (and total noise).

I don't think that there is a loss of light at all.

You are correct that the light intensity remains the same, but there is a loss in the total amount of light (and therefore an increase in noise).

So my 70-200mm 2.8 is equivalent to a 112-320mm f/2.8

And an Olympus 300mm f/2.8 is equivalent to a 600mm f/2.8

YES?

No.

It's just semantics.

Imagine light is rainfall. You have a 5x5' tub and a 5x10' tub. In 10 minutes twice as much rain hits the 5x10' tub, but the average rainfall is the same for both tubs. And if your goal is to fill up each tub with 6 inches of water, both will fill up in the same amount of time, so neither system has an advantage.

That's a great analogy, thanks. However, I think the larger system has an advantage. The amount of light (water) causes a difference in the amount of noise. A drizzle with a large tub (f/14 FF35 ISO 3200) will give you as much water as a downpour in a tiny bucket (f/2.8 digicam ISO 100). That is one advantage of FF35 over digicams.


The main point I want to make here is that crop factor does not effect the f/stop.


Nobody said it did. Just like nobody said that crop factor affects the focal length. (It doesn't.) The crop factor does affect the FF35-equivalent focal length, though. And it does affect the FF35-equivalent f/stop.


If you need a fast low light 2.8 lens for fast shutter speeds to stop action, your lens will always be f/2.8 for exposure purposes and you will always be able to use a fast shutter speed.


On film that was correct. Film had a nonlinear response so that a large amount of area did not perfectly compensate for lower intensity per area (especially so for film of the same ISO rating). On digital, however, the relationship is completely linear: a two-stop decrease in light-per-area (f/5.6 -> f/2.8 ) is perfectly compensated by a two-stop increase in area (Four Thirds -> FF35).

In other words, if 150mm f/2 ISO 400 1/500 is what you used on Four Thirds, then you can get the exact same photo with:


190mm f/2.5 ISO 500 1/500 on Canon APS-C
300mm f/4 ISO 1600 1/500 on FF35.
480mm f/6.4 ISO 5000 1/500 on MFDB (56x37mm)




SO.... An Olympus 4/3 300mm f/2.8 WILLbe equivalent to a 600mm f/2.8 (35mm FF)


Not in any of the ways that matter to photographers, including noise, DOF, and diffraction. They are similar is that they happen to have the same intensity per area, but that is not as important.


If some one is shooting sports with an Olympus 300mm 2.8, and there is a guy beside him with the 5D and 600mm 2.8, they both will be using the same shutter speed, and both will have the same relative image.

No. The FF35 camera has far less noise at f/2.8. You have to stop it down to f/5.6 and increase ISO two stops to get the noise level up to the amount that Four Thirds has.

Yes! This. I do not care about "total" light hitting a sensor.

Yes you do.


OBVIOUSLY it is going to be more if the sensor is bigger, but it doesn't change the exposure.

It does not change the exposure (light per area), but it changes the total amount of light (exposure times area).

The total amount of light is important because it affects noise. And you *do* care about the noise. High ISO is the one that most people think of, but there are many other effects as well:


dynamic range
tonal gradations
color depth
underexposure latitude
aberration correction
post processing


If you didn't care about any of those things, you would be using a mobile phone-sized sensor (6x4mm). An f/2.8 lens costs less than a dollar for that size of sensor.

But most photographers do care about that stuff (including you, I'd bet), and that's why most of them use sensors that have 20-35X more area than a mobile phone sensor.

[XterraJohn]

As I said before, the "loss of light" is from the loss of area. In the scenario you described above, there is more light falling on the film in the EOS 630. However, some of my comments (e.g. brightness and noise) only applied to digital cameras, because film has a nonlinear relationship between film area and total light (and total noise).



You are correct that the light intensity remains the same, but there is a loss in the total amount of light (and therefore an increase in noise).



No.



That's a great analogy, thanks. However, I think the larger system has an advantage. The amount of light (water) causes a difference in the amount of noise. A drizzle with a large tub (f/14 FF35 ISO 3200) will give you as much water as a downpour in a tiny bucket (f/2.8 digicam ISO 100). That is one advantage of FF35 over digicams.



Nobody said it did. Just like nobody said that crop factor affects the focal length. (It doesn't.) The crop factor does affect the FF35-equivalent focal length, though. And it does affect the FF35-equivalent f/stop.



On film that was correct. Film had a nonlinear response so that a large amount of area did not perfectly compensate for lower intensity per area (especially so for film of the same ISO rating). On digital, however, the relationship is completely linear: a two-stop decrease in light-per-area (f/5.6 -> f/2.8 ) is perfectly compensated by a two-stop increase in area (Four Thirds -> FF35).

In other words, if 150mm f/2 ISO 400 1/500 is what you used on Four Thirds, then you can get the exact same photo with:

190mm f/2.5 ISO 500 1/500 on Canon APS-C
300mm f/4 ISO 1600 1/500 on FF35.
480mm f/6.4 ISO 5000 1/500 on MFDB (56x37mm)



Not in any of the ways that matter to photographers, including noise, DOF, and diffraction. They are similar is that they happen to have the same intensity per area, but that is not as important.



No. The FF35 camera has far less noise at f/2.8. You have to stop it down to f/5.6 and increase ISO two stops to get the noise level up to the amount that Four Thirds has.



Yes you do.



It does not change the exposure (light per area), but it changes the total amount of light (exposure times area).

The total amount of light is important because it affects noise. And you *do* care about the noise. High ISO is the one that most people think of, but there are many other effects as well:

dynamic range
tonal gradations
color depth
underexposure latitude
aberration correction
post processing
If you didn't care about any of those things, you would be using a mobile phone-sized sensor (6x4mm). An f/2.8 lens costs less than a dollar for that size of sensor.

But most photographers do care about that stuff (including you, I'd bet), and that's why most of them use sensors that have 20-35X more area than a mobile phone sensor.

I'm afraid that I've lost track of what point you were trying to make. :confused: Do you think that Canon should start adding weather-sealing to their EF-S lineup?

[tkbslc]

No. I see this "theory" repeated numerous times but it is incorrect.

Agreed, that's why I called it a theory and debunked it with actual lens measurements. Olympus' 70-300 for a 1/2 size sensor is actually slightly larger than Sigmas.

[KenjiS]

It's just semantics.

Imagine light is rainfall. You have a 5x5' tub and a 5x10' tub. In 10 minutes twice as much rain hits the 5x10' tub, but the average rainfall is the same for both tubs. And if your goal is to fill up each tub with 6 inches of water, both will fill up in the same amount of time, so neither system has an advantage.

Exactly, its just depth of field, Not metering

And this relates to the same composition, For the same composition and perspective you have less depth of field with a larger sensor...

IE, if you take a scene with a 50mm f/1.4 on a 1.6x and a 85 f/2 on a full frame, The image [provided you composed the same] will be pretty much identical....

However, 4/3 cameras are also a lot worse in High ISO, about 2 stops worse generally....and yes, this is because their sensor is tiny, I understand what he's saying about light gathering effectiveness, But this is just in relation to how the sensor and the photosites read light

Basically, a bigger photosite is more sensitive to the incoming light because it can collect more light and focus it onto a point, Thus a less dense sensor is more sensitive to begin with and requires less amplification to be more sensitive...Less amplification contributes to less noise, This is what i theorized they needed to do to improve sensor technology [make things a lot more sensitive to begin with...And thats what Canon did with the 7D]....

And all this technical crap makes my head hurt x.x a lot ....

[Shadowblade]

Given their announcement of the partly-weather-sealed 7D, and a possible fully-weather-sealed pro body in the future, I wouldn't mind seeing some EF-S L lenses (probably zooms) to go with these pro/prosumer bodies. After all, people will buy these bodies for speed/action photography, and partly for lighter overall system weight (for travel photography, adventure sports photography, etc.) - it makes little sense to carry an EF lens that weighs more than twice as much as necessary.

At present, it's not the optical quality of EF-S lenses that sets them apart from EF L lenses - rather, it's the build quality and sealing. EF-S 10-22mm certainly equals the 17-40mm f/4L in optical quality, while the 17-55mm f/2.8L IS arguably equals the 24-70mm in image quality when used on the same camera (it wouldn't be fair to compare a full-frame sensor directly with a crop sensor for image quality).

I could certainly see the following lenses being popular - particularly long zooms, for the wildlife/birding crowd (who often like the crop factor and preferentially shoot with crop cameras):

EF-S 10-20mm f/2.8L IS USM - wide zoom
EF-S 17-55mm f/2.8L IS USM - normal zoom
EF-S 50-150mm f/2.8L IS USM - medium telephoto zoom
EF-S 150-500mm f/5.6L IS USM - long telephoto
(maybe) EF-S 400-800mm f/5.6L IS USM - wildlife

[tkbslc]

Given their announcement of the partly-weather-sealed 7D, and a possible fully-weather-sealed pro body in the future, I wouldn't mind seeing some EF-S L lenses (probably zooms) to go with these pro/prosumer bodies. After all, people will buy these bodies for speed/action photography, and partly for lighter overall system weight (for travel photography, adventure sports photography, etc.) - it makes little sense to carry an EF lens that weighs more than twice as much as necessary.



Not sure you read the rest of the thread, but any EF-S lens longer than about 50mm is not going to be any smaller than a FF lens. And if they add L build and sealing in to it, it certainly won't be lighter.

Out of your list, Can maybe see the 10-20 f2.8 and 50-150 being appealing. They already have the 17-55 IS, although a better built mk2 wouldn't hurt. The long lenses would be such specialy lenses that there would be no point making them EF-S and limiting their market- especially because they would still be the same size (read my earlier post about the 4/3 70-300 vs a FF one)

[Shadowblade]

Not sure you read the rest of the thread, but any EF-S lens longer than about 50mm is not going to be any smaller than a FF lens. And if they add L build and sealing in to it, it certainly won't be lighter.

Out of your list, Can maybe see the 10-20 f2.8 and 50-150 being appealing. They already have the 17-55 IS, although a better built mk2 wouldn't hurt. The long lenses would be such specialy lenses that there would be no point making them EF-S and limiting their market- especially because they would still be the same size (read my earlier post about the 4/3 70-300 vs a FF one)

I didn't actually read the rest of the thread...

Why would a long EF-S lens be just as large as an EF-lens. Sure, the front element will need to be just as large (around 90mm for f/5.6 @ 500mm) but every other internal element could be smaller, essentially producing a cone-shaped lens.

Long lenses are specialty lenses anyway, EF or EF-S...

[nphsbuckeye]

In my book if the 5D Mark II is considered pro so is the 7D...
Agreed. If one is pro, then so is the other.

[tkbslc]

I didn't actually read the rest of the thread...

Why would a long EF-S lens be just as large as an EF-lens. Sure, the front element will need to be just as large (around 90mm for f/5.6 @ 500mm) but every other internal element could be smaller, essentially producing a cone-shaped lens.

Long lenses are specialty lenses anyway, EF or EF-S...

Well if the mount, front element and lens length are all about the same, then what is the difference? If Canon can make an EF-S lens that is 8" long and 4 inches in diameter and 5lbs (Assuming the 150-500), or an EF lens that 8" long 4 inches in diameter and 5.5lbs, why would they make the EF-S? It save a negligible amount of weight and limits the user base. That is if it is even possible to make it ANY lighter.

The example I gave was the Olympus 70-300 4/3 lens. It is actually a few percent larger in all regards (Even weight) vs the Sigma 70-300 FF lens.

[Daniel Browning]

Exactly, its just depth of field, Not metering


It's not just depth of field. It's also total light, noise, diffraction, and more.


Basically, a bigger photosite is more sensitive to the incoming light because it can collect more light and focus it onto a point, Thus a less dense sensor is more sensitive to begin with and requires less amplification to be more sensitive...Less amplification contributes to less noise,


No. Even if the larger sensor had smaller pixels, or the same size pixels, it would still be true, because it has nothing to do with the pixel size.

If you want to prove that large sensors have the same light (and same noise) with slower f-stops, here are some simple instructions to prove it. Plus they prove that it's not related to pixel size, either:

Use any camera and zoom lens (e.g. 50D+70-200)
Take one picture at 70mm, f/4, ISO 100, 1/500
Take another picture 140mm, f/8, ISO 400, 1/500
Apply a 2X crop factor to the 70mm shot (matching FOV)
Apply the same raw conversion and post processing.
Downsample both to factor out the difference in resolution.
Compare noise
You will see that brightness and noise are the same. Cropping the 70mm shot by 2X gives it the same AOV as 140mm, and simulates a sensor that has a 2X crop factor (e.g. Four Thirds vs FF35). This simple experiment is a good simulation of sensor size difference that can be tested by anyone with a camera. It shows how intensity can be traded for area. It also shows that it is not related to pixel size. More on that is described here:

Small pixel sensors do not have worse performance

[tkbslc]

If you want to prove that large sensors have the same light (and same noise) with slower f-stops, here are some simple instructions to prove it. Plus they prove that it's not related to pixel size, either:
Use any camera and zoom lens (e.g. 50D+70-200)
Take one picture at 70mm, f/4, ISO 100, 1/500
Take another picture 140mm, f/8, ISO 400, 1/500
Apply a 2X crop factor to the 70mm shot (matching FOV)
Apply the same raw conversion and post processing.
Downsample both to factor out the difference in resolution.
Compare noiseYou will see that brightness and noise are the same. Cropping the 70mm shot by 2X gives it the same AOV as 140mm, and simulates a sensor that has a 2X crop factor (e.g. Four Thirds vs FF35). This simple experiment is a good simulation of sensor size difference that can be tested by anyone with a camera. It shows how intensity can be traded for area. It also shows that it is not related to pixel size. More on that is described here:

Small pixel sensors do not have worse performance

How do you explain the completely different noise handling performance between sensors of the same size? eg. 10D vs 40D or D100 vs D90? It can't all come back to surface area and light collection.

[nphsbuckeye]

How do you explain the completely different noise handling performance between sensors of the same size? eg. 10D vs 40D or D100 vs D90? It can't all come back to surface area and light collection.
That's a brave question.

[Shadowblade]

Well if the mount, front element and lens length are all about the same, then what is the difference? If Canon can make an EF-S lens that is 8" long and 4 inches in diameter and 5lbs (Assuming the 150-500), or an EF lens that 8" long 4 inches in diameter and 5.5lbs, why would they make the EF-S? It save a negligible amount of weight and limits the user base. That is if it is even possible to make it ANY lighter.

The example I gave was the Olympus 70-300 4/3 lens. It is actually a few percent larger in all regards (Even weight) vs the Sigma 70-300 FF lens.

Only the front element - the light collecting part - and the aperture diaphragm need to be full-sized. I'm sure a lens can be made such that the aperture ring is directly behind the front element - after all, in the human eye, the pupil (created by the iris, analogous to the aperture diaphragm) lies directly behind the cornea, in front of the lens (the main element for adjusting focus). What happens after that all depends on what you intend to do with the light that has been collected - funnel it into a small image circle using smaller elements, or funnel it into a larger circle using larger elements.

The full-frame lens would need to be more cylindrical, with larger elements to produce an image circle with 1.6x the diameter. Naturally, if you design the crop lens the same way as the full-frame lens (with the aperture diaphragm somewhere in the middle of everything) you'll end up with large elements throughout and minimal saving in weight and cost, since it's not really practical to have small elements between the front element and the large aperture diaphragm. But if you can put the aperture diaphragm directly behind the front element, I don't see why the rest of the elements (i.e. most of the lens) couldn't be of much smaller dimensions, with the lens taking the overall shape of a truncated cone - large aperture diaphragm and front element, with everything else behind much smaller. Cones have a smaller volume than cylinders of the same radius. Odd looking, but it'd probably work... plus, you could house the whole thing within a largely-empty (and therefore lightweight) cylinder, giving a more practical shape.

[Daniel Browning]

How do you explain the completely different noise handling performance between sensors of the same size? eg. 10D vs 40D or D100 vs D90?


Different sensors have different performance. The newer one is the same or better, 95% of the time.

It can't all come back to surface area and light collection.

Something I said must have given you the wrong impression, but I assure that technology does have an effect. Try the experiment I described above. It allows you to simulate two different sensor sizes with the exact same technology.

[wickerprints]

Only the front element - the light collecting part - and the aperture diaphragm need to be full-sized. I'm sure a lens can be made such that the aperture ring is directly behind the front element - after all, in the human eye, the pupil (created by the iris, analogous to the aperture diaphragm) lies directly behind the cornea, in front of the lens (the main element for adjusting focus). What happens after that all depends on what you intend to do with the light that has been collected - funnel it into a small image circle using smaller elements, or funnel it into a larger circle using larger elements.

The human eye is biological, a camera lens is mechanical. Not the same principle. A human iris opens and closes through muscular action. The cornea and the aqueous and vitreous humors have refractive power in addition to the lens. The lens is also capable of flexing in order to focus. By contrast, a camera lens diaphragm has rigid leaves that cannot collapse in the way a muscle can shorten its length. Therefore, unless one invents an entirely novel mechanism for the diaphragm that is fast (much faster than the ability of a human iris to dilate or contract, btw) and enables it to completely open, it is not possible to place the diaphragm directly behind the front element without having to make that element even larger. There simply would be no space for the leaves to retract to. Furthermore, the glass in a camera lens is rigid. The only way to adjust focus is to move the various elements along the optical axis. This increases the length of the system and further restricts how narrow the barrel may be made, since the internal motion of these elements is achieved through a system of precision-milled nested cylinders.

Look at the internal design of the super-tele primes and tell Canon how they could make it half the size without changing the front element diameter.

[Shadowblade]

The human eye is biological, a camera lens is mechanical. Not the same principle. A human iris opens and closes through muscular action. The cornea and the aqueous and vitreous humors have refractive power in addition to the lens.

The cornea and aqueous/vitreous humours refract light, but have no ability to adjust their refraction. This is why some elderly people, who have lost their ability to focus effectively with their lenses (i.e. unable to adjust focus distance) become long-sighted, with focus fixed near infinity, and require reading glasses for close subjects. In other words, these elements of the eye are essentially a fixed focal length, fixed focus distance lens - replaceable by a single piece of glass. You could potentially place this piece of glass either in front of or behind the aperture diaphragm/pupil. In fact, you could replace it with either the front element itself, or, in lenses with a protective front element, with a single glass element directly behind the front element.

A (fixed focal length) camera lens requires many elements to enable it to alter its focal distance. If it were to only have a single plane of focus, one could just as easily use a magnifying glass (indeed, that's exactly what a magnifying glass is - a lens of fixed focal length and fixed focus distance).

The lens is also capable of flexing in order to focus.

The lens can't focus in any other way. It alters the covergence distance of incoming light by altering its geometry - the contracting ciliary muscles pull on the suspensory ligament of the lens, releasing their pressure on the anterior surface of the lens and allowing it to become more convex, decreasing the focus distance (which is why a ciliary muscle which cannot contract effectively results in an ineffective lens that can't focus at close distances).

So does a camera lens. Since glass/fluorite elements do not flex, a camera lens alters its geometry by having a number of mobile glass elements which, by moving back and forth, essentially simulates a single element of variable geometry (e.g. the lens of the eye) and allows inflexible glass elements to focus at multiple distances (unlike a magnifying glass). By adding more elements, one can also alter the effective focal length of the lens, as in zoom lenses. By adding an element which is completely asymmetrical, or symmetrical only along one axis, one can even simulate astigmatism; in the eye, this is usually (and undesirably) caused by asymmetry of the cornea (a 'fixed element', if you will). Allow for rotation of this element and you can also rotate the axis of astigmatism.

By contrast, a camera lens diaphragm has rigid leaves that cannot collapse in the way a muscle can shorten its length. Therefore, unless one invents an entirely novel mechanism for the diaphragm that is fast (much faster than the ability of a human iris to dilate or contract, btw) and enables it to completely open, it is not possible to place the diaphragm directly behind the front element without having to make that element even larger. There simply would be no space for the leaves to retract to.

My EF 70-200mm f/2.8L IS USM has a front element diameter of 77mm and a body diameter of around 80mm. In order to achieve an aperture of f/2.8 @ 200mm, a lens will need a maximum aperture diameter of 71.5mm. One clearly does not need a huge amount of space to retract diaphragm blades...

The blades in a lens' aperture diaphragm do not retract radially - rather, each blade essentially forms the edge of a circle, the centre of which is incongruent with the centre of the aperture and with the centres of the circles which form the other blades. When retracting, each blade rotates along its own circle, moving obliquely away from the centre of the aperture and widening it without greatly increasing the distance of the most distal point of the blade from the lens axis. In other words, the most proximal part of the blade moves a long distance from the centre, but the most distal part doesn't move much further (in fact, it may even come closer to the lens axis at the widest apertures).

In any case, a wide front element and aperture (no wider than for a full-frame lens, though) would be acceptable, since we expected it to be wide anyway (it forms the fat end of the cone) and the focusing elements of the lens would be behind it, and much smaller.

Furthermore, the glass in a camera lens is rigid. The only way to adjust focus is to move the various elements along the optical axis. This increases the length of the system and further restricts how narrow the barrel may be made, since the internal motion of these elements is achieved through a system of precision-milled nested cylinders.

Clearly they don't require a very wide barrel, as there are many lenses out there with diameters of 52mm or less, as opposed to >77mm for many EF lenses...

[pwm2]

Don't expect the light rays to do two 90 degree bends inside the lens, just to quickly move into the center of your cone.

[Shadowblade]

Don't expect the light rays to do two 90 degree bends inside the lens, just to quickly move into the center of your cone.

Possible with a convex front element. Would have to be ULD, though.

Isn't that basically what fisheyes do?

Also possible to separate the front element and the aperture with a few centimetres of air, which weighs approximately nothing (apart from a bit of empty outer shell) and reduces the angle required.

[wickerprints]

The cornea and aqueous/vitreous ...

You haven't said anything that I don't already know, and you haven't countered what I have pointed out. The fact remains that your proposed design--to put the diaphragm directly behind the front element--is incorrect and reflects an ignorance of camera lens design principles.

Oh, and by the way, when you can show me an unaided normal human eye that has the resolving capability of a 500mm lens, then you might have the faintest semblance of a reasonable comparison.

[Shadowblade]

You haven't said anything that I don't already know, and you haven't countered what I have pointed out. The fact remains that your proposed design--to put the diaphragm directly behind the front element--is incorrect and reflects an ignorance of camera lens design principles.

How have I not countered what I pointed out? Either you haven't read the reply, or failed to understand the optics behind it.

These were your points:

1. Cornea/aqueous humour/vitreous humour have refractive properties.

Fixed focal length with fixed plane of focus. A window also has refractive properties. Easily replaced with 1 piece of glass. You can put that piece of glass anywhere, depending on its shape.

2. The lens of the eye is capable of flexing to focus.

A camera lens does the same thing by moving multiple rigid elements, essentially creating a multipartite lens that 'flexes' by moving each element individually. There's no optical law saying these elements have to be big.

3. The leaves of a camera's aperture diaphragm are rigid.

And they can also rotate out of the way. By design, a wide-open diaphragm doesn't take up any more space than a fully closed one. This is part of the wide end of the cone.

4. You need a long barrel so that the elements can move.

This is the same whether you're using large elements or small elements. Smaller, more strongly curved elements generally don't need to move as far. Even if they did, you're still saving on diameter, even if not in length.

5. There is a certain minimum width, since internal motion is achieved through a series of nested cylinders.

These nested cylinders clearly do not need to be very wide, as there are many perfectly good lenses with small diameters. In other words, it is not the limiting factor for barrel width in longer lenses - the limiting factor is minimum aperture diameter to achieve a certain f-stop.

[Shadowblade]

Oh, and by the way, when you can show me an unaided normal human eye that has the resolving capability of a 500mm lens, then you might have the faintest semblance of a reasonable comparison.

1. Not a fair comparison. The human eye produces an image on the retina, which is smaller than a crop sensor. Obviously a larger sensor will require a larger image circle, which requires a larger lens. I am not advocating a 500mm lens the size of a human eyeball.

2. Not a fair comparison. A larger sensor will usually have greater resolving power than a smaller one. The retina is smaller than a crop sensor. The fovea (the part that's actually sharp) is many times smaller. This is a sensor limitation, not a lens limitation.

3. Not a fair comparison. The neurons supplying the retina enter the eye at the optic disc, or 'blind spot'. They then travel in front of the retina to their position, entering the retina from the front. Naturally this somewhat diffracts the light and reduces resolution. Just like placing a very strong anti-aliasing filter in front of a camera sensor. Some animals do not have this problem.

4. Not a fair comparison. The human eye is not a precision-made machine. There are remnants of the developmental hyaloid artery within the aqueous humour. Corneas are rarely perfect. Every copy is different. Just like Sigma lenses. If one came up with an ideal design and replicated it every time, you could have an extremely high-resolution system. It's all about shape and precision, not size. You can do that with camera lenses. With human eyes, you're stuck with what you have.

5. Fine. Give me a hawk eye.

[Wilt]

Q: "Can lenses be made smaller?"
A: "Yes, simply look at Olympus OM lenses compared to Canon EF lenses for a real Mutt and Jeff comparison!"

Admittedly EF lenses have AF mechanisms to incorporate and that adds bulk. But fast lenses do not have to be as bulky as they are! Compare Tamron f/2.8 AF lenses with Canon f/2.8 AF lenses to convince yourself.

[KenjiS]

Q: "Can lenses be made smaller?"
A: "Yes, simply look at Olympus OM lenses compared to Canon EF lenses for a real Mutt and Jeff comparison!"

Admittedly EF lenses have AF mechanisms to incorporate and that adds bulk. But fast lenses do not have to be as bulky as they are! Compare Tamron f/2.8 AF lenses with Canon f/2.8 AF lenses to convince yourself.

Or the Micro-4/3rds lenses...

[krepta]

I wish Canon would make a 24-70 or 24-105 equivalent (in FF 35mm) zoom lens for 1.6x crop that has a small, constant aperture throughout the focal range. It would be something like "EF-S 15-45mm f/2.8 USM" or "EF-S 15-65mm f/4.0 [IS] USM" (IS being a a nice addition, but not absolutely necessary).

I know about the new, upcoming EF-S 15-85mm f/3.5-5.6 IS USM; it has an ideal focal length range, but the variable aperture makes it less tentalizing [I]for me (I have no doubt it will be welcome by many --> see KenjiS :cool:). Even if it was just a constant f/4.0 (I'd prefer 2.8, of course), I would be willing to lose one stop on the wide end to gain three stops on the tele end.

As for the build, somewhere along the 17-55's build is fine with me, although I would certainly appreciate L quality build. Still, the optics are way more important than the build, at least in my case (I take good care of my stuff and don't try too many "stunts" like shooting directly under heavy rain without a cover).

But let me make it clear that by no means do I want or expect L quality on a EF-S lens for cheap. I happily spent ~$1,000 for my 17-55, and I'm willing to dish out in the range of $1,200 - $1,500 for one of the two wished-for lenses I detailed above if they have L, or close to L, quality optics and build, and perhaps IS as well.

Obviously, in the long run I intend to go FF (5D MkIII, maybe?) and both the 24-70L and 25-105L are lenses I want in my kit (I will also consider any updates to those if it happens in the timeframe when I am ready to buy). But at the moment I feel a little limited with the choices of quality lenses offered by Canon for 1.6x crop. Maybe I should look into Sigma or Tamron. I have also considered the 17-40 f/4L, which comes close to the 15-45 range I mentioned aboved.

In the meantime, my two mainstay lenses are the EF-S 17-55mm f/2.8 IS USM and EF 70-200mm f/2.8L IS USM, so I have a coverage gap between 55mm and 70mm. I have the EF 28-105mm f/3.5-4.5 II USM, but I stopped using it after getting the 17-55; the quality just isn't there for me with that lens. I use a very simple solution to deal with the coverage gap, and it's inexpensive: walk up closer to the subject/scene when using the 17-55, or step further back from the subject/scene when using the 70-200. :lol:

[KenjiS]

As for the build, somewhere along the 17-55's build is fine with me, although I would certainly appreciate L quality build. Still, the optics are way more important than the build, at least in my case (I take good care of my stuff and don't try too many "stunts" like shooting directly under heavy rain without a cover).



Actually i did that with my 17-55 once...it didnt hurt it..

oddly enough, for a lens renowned for sucking dust its pretty waterproof o_O

[AngryCorgi]

5d and 5d II are considered pro too.

ed rader

They are marketing the 7D as such too.

[krepta]

Actually i did that with my 17-55 once...it didnt hurt it..

oddly enough, for a lens renowned for sucking dust its pretty waterproof o_OI'll have to take your word for it, as I wouldn't dare test it out! :lol:

Yes, I have heard so many stories about the 17-55 being a dust vacuum, but I have never had that problem. I was probably lucky to get one that did not have the dust problem.

[KenjiS]

I'll have to take your word for it, as I wouldn't dare test it out! :lol:

Yes, I have heard so many stories about the 17-55 being a dust vacuum, but I have never had that problem. I was probably lucky to get one that did not have the dust problem.

Mine didnt either :)