Sigma Warns of Lens Aberration Correction Bug with Select Sigma Lenses and Canon EOS 6D Mark II, EOS

privatebydesign

EOS-1D X Mark III
CR Pro
Jan 29, 2011
9,682
4,185
Re: Sigma Warns of Lens Aberration Correction Bug with Select Sigma Lenses and Canon EOS 6D Mark II,

Romain said:
Firmware update needed... 5dIV has the same kind of issue at launching before the first firmware update, if i remember well... No panic mates!..

Yes, and the point is it is Sigma firmware update needed not a Canon one. Ergo it is a Sigma issue not a Canon one.
 

Mt Spokane Photography

I post too Much on Here!!
CR Pro
Mar 25, 2011
16,601
1,564
Re: Sigma Warns of Lens Aberration Correction Bug with Select Sigma Lenses and Canon EOS 6D Mark II,

Romain said:
Firmware update needed... 5dIV has the same kind of issue at launching before the first firmware update, if i remember well... No panic mates!..

Unless a user wants to use his camera, of course. Many only own a lens or two, and if they would not work, I'd panic. Of course, there is a work around, but its been almost two weeks before Sigma admitted the issue and the work around.

Sigma erroneously misstates the release date for the 6D MK II, and the issue applies to many cameras like the 77D which was released long long ago.

There is no mention of a firmware update, I don't think they know how.
 

Romain

Click by chance, everyday...
Dec 20, 2016
27
28
Re: Sigma Warns of Lens Aberration Correction Bug with Select Sigma Lenses and Canon EOS 6D Mark II,

privatebydesign said:
Romain said:
Firmware update needed... 5dIV has the same kind of issue at launching before the first firmware update, if i remember well... No panic mates!..

Yes, and the point is it is Sigma firmware update needed not a Canon one. Ergo it is a Sigma issue not a Canon one.
 

rrcphoto

EOS R6
Jun 20, 2013
2,505
147
Re: Sigma Warns of Lens Aberration Correction Bug with Select Sigma Lenses and Canon EOS 6D Mark II,

factor said:
This is not Sigmas fault - seems like Canon crippled their cameras first, and third party lenses support now...
So mutch attention to protect self pastly ... there is no other arguments more for Canon ... ? :(

that's rediculous. the problem is how Sigma mimics their lenses to the camera body. they give an ID number of an existing canon lens that may not even be relevant. For instance their 24-70 may be mimiced to a canon camera as being a 14mm 2.8.

at times they give an ID of lenses that are no longer supported by canon. this happened with tamron years back. caused all sorts of problems because the camera body didn't have an idea on what AF points the lens supports.

In this case, the body takes a look at the lensID number, and *assumes* it's a canon lens of a particular type - it then applies the lens aberration / vignetting correction for that lens. and lo and behold, it looks like sheit because it was actually a sigma lens instead.


care to tell me how that's canon's fault?
 

neuroanatomist

I post too Much on Here!!
CR Pro
Jul 21, 2010
24,650
2,183
Re: Sigma Warns of Lens Aberration Correction Bug with Select Sigma Lenses and Canon EOS 6D Mark II,

rrcphoto said:
care to tell me how that's canon's fault?

It just is. Trust him.
 

hne

Gear limits your creativity
Jan 8, 2016
314
34
Re: Sigma Warns of Lens Aberration Correction Bug with Select Sigma Lenses and Canon EOS 6D Mark II,

Valvebounce said:
Hi weixing.
My understanding is that Sigma (and other 3rd party manufacturers?) use a code for a lens (not necessarily in the same focal length range?) that has similar drive requirements for the AF and IS systems so that the camera drives these systems at the desired rates?
Using an unallocated number might result in the body identifying the lens as "no lens attached" or even worse, driving the focus system in to the end of the travel at full speed due to not having any parameters set?
I'm not certain of any of this hence marking them as questions so that hopefully one or more of the knowledgeable members will correct or verify my understanding of the situation.

Cheers, Graham.

I think you've got that wrong. If the camera had to know something about the lens so as not to drive the AF at full speed into the end (and if that was actually potentially damaging for the lens), we would be warned about having to upgrade firmware in cameras before mounting new lenses on old cameras. How many of the lenses in my signature do you think would be harmed if I mounted them to my old announced-1990 EOS 1000 (original Rebel in the US)? My guess is zero. Of all the lenses I own, only the 50/1.8 II is old enough to have been announced the same year as or earlier than the EOS 1000.

Lens aberration correction is using data stored in the body, most probably looked up by lens code. The lens codes used by Sigma/Tamron/Cosina/... are only some times values for similar Canon lenses. For example, the 35Art uses a 17-35/2.8 code and the 18-35Art uses the 85/1.2L code: https://sno.phy.queensu.ca/~phil/exiftool/TagNames/Canon.html#LensType
Modern cameras are delivered with profiles for a bizarre number of lenses. My 5DmkIV was delivered with all profiles known to the canon lens registration utility installed.

Somewhere I once found a page with detailed protocol information written by someone who had painstakingly reverse-engineered quite a bit of the EF lens protocol as a hobby project. The lens has to be surprisingly smart for a late 1980s protocol. The camera body tells the lens to refocus a number of steps in a direction and can then query the new distance (for some newer lenses, list current as of ETTL-II introduction in 2008 here: http://www.bobatkins.com/photography/eosfaq/ettl2.html )
 

LonelyBoy

EOS RP
Feb 18, 2015
745
0
Re: Sigma Warns of Lens Aberration Correction Bug with Select Sigma Lenses and Canon EOS 6D Mark II,

Mt Spokane Photography said:
Romain said:
Firmware update needed... 5dIV has the same kind of issue at launching before the first firmware update, if i remember well... No panic mates!..

Unless a user wants to use his camera, of course. Many only own a lens or two, and if they would not work, I'd panic. Of course, there is a work around, but its been almost two weeks before Sigma admitted the issue and the work around.

Sigma erroneously misstates the release date for the 6D MK II, and the issue applies to many cameras like the 77D which was released long long ago.

There is no mention of a firmware update, I don't think they know how.

If you only have a lens or two, and sell your old camera before getting a new one and making sure it works for you (ideally inside the return window), my sympathy is rather low.
 

Valvebounce

EOS R5
CR Pro
Apr 3, 2013
4,536
431
54
Isle of Wight
Re: Sigma Warns of Lens Aberration Correction Bug with Select Sigma Lenses and Canon EOS 6D Mark II,

Hi hne.
Good point re the older bodies not knowing about new lenses, I guess my supposition about unused codes causing harm was wrong.
Maybe canon know how to get their new lenses to tell older bodies what they are and what is required of the body to a level commensurate with the intellectual ability of each body, they are after all mini computers that communicate with each other!
Maybe Σ hasn't reverse engineered that part correctly?
On the other hand Σ don't spoof Canon lens ID's to get the lens corrections ;D, we know they don't work ::), so they must spoof the ID's for some other reason leaving AF and IS parameters as the likely reasons, using unassigned ID's must have some issue or they would likely use them. Maybe the hazard with unused codes is that Canon may use that code next for a lens with completely different requirements from the Σ lens that is using it? Imagine what problems that could cause!

Cheers, Graham.

hne said:
Valvebounce said:
Hi weixing.
My understanding is that Sigma (and other 3rd party manufacturers?) use a code for a lens (not necessarily in the same focal length range?) that has similar drive requirements for the AF and IS systems so that the camera drives these systems at the desired rates?
Using an unallocated number might result in the body identifying the lens as "no lens attached" or even worse, driving the focus system in to the end of the travel at full speed due to not having any parameters set?
I'm not certain of any of this hence marking them as questions so that hopefully one or more of the knowledgeable members will correct or verify my understanding of the situation.

Cheers, Graham.

I think you've got that wrong. If the camera had to know something about the lens so as not to drive the AF at full speed into the end (and if that was actually potentially damaging for the lens), we would be warned about having to upgrade firmware in cameras before mounting new lenses on old cameras. How many of the lenses in my signature do you think would be harmed if I mounted them to my old announced-1990 EOS 1000 (original Rebel in the US)? My guess is zero. Of all the lenses I own, only the 50/1.8 II is old enough to have been announced the same year as or earlier than the EOS 1000.

Lens aberration correction is using data stored in the body, most probably looked up by lens code. The lens codes used by Sigma/Tamron/Cosina/... are only some times values for similar Canon lenses. For example, the 35Art uses a 17-35/2.8 code and the 18-35Art uses the 85/1.2L code: https://sno.phy.queensu.ca/~phil/exiftool/TagNames/Canon.html#LensType
Modern cameras are delivered with profiles for a bizarre number of lenses. My 5DmkIV was delivered with all profiles known to the canon lens registration utility installed.

Somewhere I once found a page with detailed protocol information written by someone who had painstakingly reverse-engineered quite a bit of the EF lens protocol as a hobby project. The lens has to be surprisingly smart for a late 1980s protocol. The camera body tells the lens to refocus a number of steps in a direction and can then query the new distance (for some newer lenses, list current as of ETTL-II introduction in 2008 here: http://www.bobatkins.com/photography/eosfaq/ettl2.html )
 
<-- start Taboola -->