20D and 3rd party lens

[CanonUser]

Hi All,
Certain features found in Canon's newer model such as ETTL-II (and may be the 9 points focus) require that the lens feeds distance data back to the camera. Would Tamron or Sigma lens be able to meet these requirements?

Thanks in advance for your thoughts.

Alan

[Chris1le]

It's funny you brought this up. I was going to start a thread titled
"Canon announces release of 20D, Sigma rechippers surrender." :twisted:

[Jay Todd]

Right. Why is it exactly that the older Sigma APO lenses are not useable on digital bodies?

[cmM]

Yes, any lens with an EF / EFs mount will work. Only a few new lenses have the ability to send distance information for ETTL II purposes. That's nice, but not necessary. Not all canon lenses have that, only some new ones. I think even ETTL II will work without the input from the lens, that's just a plus.

[RichardtheSane]

Right. Why is it exactly that the older Sigma APO lenses are not useable on digital bodies?

Sigma reverse engineered the EF mount toproduce their lenses.

Canon changed the way the EF mount communicates with lenses a while back. In doing so they made sure all the older Canon lens line up will still work.

So the older sigma lenses cannot communicate with the newer EF mount.

[dtrayers]

Yes, any lens with an EF / EFs mount will work. Only a few new lenses have the ability to send distance information for ETTL II purposes. That's nice, but not necessary. Not all canon lenses have that, only some new ones. I think even ETTL II will work without the input from the lens, that's just a plus.

Actually, quite a few of Canon's lenses have the distance data:

http://photonotes.org/articles/eos-flash/#distancedata

I was surprised that the 85mm f/1.8 supports distance data but the much higher end 85mm f/1.2 doesn't. Same-same for the 50mm f/1.4.

[cmM]

wow, there are a lot of them.
I though there are fewer lenses capable of that.

The first lens with a distance decoder came in 1990.... that's even more surprising. WTF, I was 7 :!: :P

[Tom W]

Yes, any lens with an EF / EFs mount will work. Only a few new lenses have the ability to send distance information for ETTL II purposes. That's nice, but not necessary. Not all canon lenses have that, only some new ones. I think even ETTL II will work without the input from the lens, that's just a plus.

Actually, quite a few of Canon's lenses have the distance data:

http://photonotes.org/articles/eos-flash/#distancedata

I was surprised that the 85mm f/1.8 supports distance data but the much higher end 85mm f/1.2 doesn't. Same-same for the 50mm f/1.4.

The 85/1.2L and the 50/1.4 are older designs. The big concern with third party lenses is whether Canon has made any subtle changes to the mount that would render them useless on the 20D. Somebody is going to have to test that aspect - perhaps someone ought to e-mail one of the testers that has a copy of the 20D.

[RDKirk]

>>Canon changed the way the EF mount communicates with lenses a while back. In doing so they made sure all the older Canon lens line up will still work.
<<

I saw something from Chuck Westfall a while back stating that Canon has never changed the EF communication protocol. If they had, either older lenses would not communicate with newer bodies or newer bodies would not communicate with older lenses.

Sigma failed to perfectly reverse engineer the comms protocol in the first place. Other companies such as Tamron did a better job of it.

[DocFrankenstein]

I have a grand invested in sigma. I hope it'll work. :(

[Cadwell]

I saw something from Chuck Westfall a while back stating that Canon has never changed the EF communication protocol.

And of course Chuck has no vested interests here.

[RichardtheSane]

>>Canon changed the way the EF mount communicates with lenses a while back. In doing so they made sure all the older Canon lens line up will still work.
<<

I saw something from Chuck Westfall a while back stating that Canon has never changed the EF communication protocol. If they had, either older lenses would not communicate with newer bodies or newer bodies would not communicate with older lenses.

Sigma failed to perfectly reverse engineer the comms protocol in the first place. Other companies such as Tamron did a better job of it.

If Canon changed nothing, then the surely sigma lenses would still work?

The change may only have been adding new features... but that is still a change. I fail to see how the sigma lenses suddenly stopped working on the newer bodies if nothing was changed.

[RDKirk]

I saw something from Chuck Westfall a while back stating that Canon has never changed the EF communication protocol.

And of course Chuck has no vested interests here.

Why should Canon lie about whether or not they changed the comms protocol? They could certainly point to reasons why technological upgrades might require it.

But the fact is that the very earlies EF lenses operate as well as (or even better) than they ever did on the very newest EOS bodies. If Sigma had perfectly reverse engineered the comms protocol capability of the EF 50mm f1.8 Mark I, all their lenses would work perfectly on every EOS camera, just as that Canon lens does--and as most other 3rd party lenses do.

[RDKirk]

>>If Canon changed nothing, then the surely sigma lenses would still work?

The change may only have been adding new features... but that is still a change. I fail to see how the sigma lenses suddenly stopped working on the newer bodies if nothing was changed.<<

The example of what Canon did exists right in their older FD lens mount. The FD lens mount had a pin labled "Reserved" in the documentation. Nobody knew what that pin was for. None of the earlier FD cameras used it, not the F-1, not the F-1n, not the AE-1, not the A-1, not the T-50. So none of the 3rd party manufacturers had that pin on their lenses.

Canon didn't use the pin until the T-90, where it provided information to their early matrixed metering system (it wasn't evaluative, but it was matrixed so that partial-area metering could be done from a number of different viewfinder areas).

If Canon had not immediately afterward abandoned the FD mount, but had continued to make cameras that used that pin, the 3rd parties would have had to scramble to "re-pin" their lenses. I think in the same way, the EF mount had some reserved capabilities that were not used until much later. Sigma (unlike others--remember that Tamron and Tokina have had very few problems) did an imperfect job of making sure their lenses had the complete comms protocol of EF lenses.

Remember that when a 3rd party reverse engineers the mount, they can NOT copy it slavishly. That would violate Canon patents. They can "work alike" but they can not "be alike" (this is why AMD can legally make processors that have the same pin-outs as an Intel processor, as long as the internal structure is different).

[Volatile]

It's funny you brought this up. I was going to start a thread titled
"Canon announces release of 20D, Sigma rechippers surrender." :twisted:

sounds like you read fark

[Chris1le]

sounds like you read fark

You are correct! :wink:

[chris maddock]

>>Canon changed the way the EF mount communicates with lenses a while back. In doing so they made sure all the older Canon lens line up will still work.
<<

I saw something from Chuck Westfall a while back stating that Canon has never changed the EF communication protocol. If they had, either older lenses would not communicate with newer bodies or newer bodies would not communicate with older lenses.

Sigma failed to perfectly reverse engineer the comms protocol in the first place. Other companies such as Tamron did a better job of it.

If Canon changed nothing, then the surely sigma lenses would still work?

The change may only have been adding new features... but that is still a change. I fail to see how the sigma lenses suddenly stopped working on the newer bodies if nothing was changed.

I have heard that they reduced the voltage supplied to the aperture mechanism, to improve battery life. The older Canon lenses are OK with that (else they wouldn't work either) but the Sigmas are not.

Whether it's true or not it is a plausible explanation, after all other third-party lenses don't have so many (if any) issues with this so the pinout of the mount isn't changed nor the commands that are transmitted through it.
Many of the Sigma problems seem to be related to the aperture so, if it isn't getting enough "oomph", it won't close down properly and won't report back that it has succeeded.

KRs
Chris

[RDKirk]

<<I have heard that they reduced the voltage supplied to the aperture mechanism, to improve battery life. The older Canon lenses are OK with that (else they wouldn't work either) but the Sigmas are not. >>

That sounds like a possibility.

It's a shame that when problems have simple causes and cures, manufacturers don't just explain what the past problems actually were, and how they've made sure those problems don't happen again.

[matty]

I just bought a 20d thinking I would be able to you use all of my quantarray lenses that i have (all manufactured by sigma). Well guess what? Nope, they all give me err99 which the manual states you will get if you use a non Canon lens. The aperture voltage sounds like a possible source. I wish I knew this prior. Is there any way around this problem?

[Marshall]

I just bought a 20d thinking I would be able to you use all of my quantarray lenses that i have (all manufactured by sigma). Well guess what? Nope, they all give me err99 which the manual states you will get if you use a non Canon lens. The aperture voltage sounds like a possible source. I wish I knew this prior. Is there any way around this problem?

Speak to sigma about the possibility & cost of having the lenses re-chipped.