Canon 2x extender opinions needed
- Thread starter adam*
- Start date
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Not much, except in price. Check out the lens/extender reviews and test chart crops on www.thedigitalpicture.com
- Messages
- 5,872
- Name
- Richard
- Edit My Images
- No
Canon Bob
Loves the Enemy
- Messages
- 9,754
- Name
- Bob
- Edit My Images
- Yes
I can't say that I use the 2xIII with my 300/2.8 but I did try a few comparisons when I first got the MkIII. Very little difference to see except some improvement in the corners of FF images.....but, in reality, they're not usually on the focal plane anyway.
Bob
Bob
odd jim
Flimsiest Lambresta
- Messages
- 9,208
- Name
- Jim
- Edit My Images
- Yes
Canon Bob said:
Is the 75% figure based on a mathematic average of using an x2 TC or specific to the Canon TC in question?
Just curious as the AF speed of my Sigma 70-200 f/2.8 with Kenko MC-7 x2 TC, although reduced, doesn't seem quite as much as 75%?
- Messages
- 3,383
- Edit My Images
- No
I do not know but suspect it is could be down to the loss of light Jim and a 2x tc will two lose stops of light irrelevant of make.
Although 75% sounds drastic remember with the 300/2.8 you are starting off with a blindly fast AF - at times it will just slow it down from near instant without the tc to just very fast with it!!. To put it another way, if the bare lens would normally focus on something in 0.2 of a sec then a 75% reduction in speed still means that it will focus in 0.375 of a second (if my maths is correct!).
By far the biggest aid to counteraction the slow-up in AF speed with a 2x tc on board is by using a focus limiter in my experience, something I always did when using a 2x tc.
Last edited:
Canon Bob
Loves the Enemy
- Messages
- 9,754
- Name
- Bob
- Edit My Images
- Yes
That's an interesting point Jim.....
It's a firmware based restriction when using converters...the AF drive speed is reduced to .75x (2x) or .5x (1.4x). Is it in the lens firmware or the body...I don't know. It maybe that the Sigma doesn't follow the rule (lens based) or the Kenko doesn't report (lens or body based). Of course, it may be that the realtime difference is barely noticeable as borne out by the figures that Roy quotes.
Bob
- Messages
- 3,383
- Edit My Images
- No
Now you mention the firmware restriction Bob I seem to remember reading a Canon white paper which said that the slowdown was to maintain focus accuracy - so it could be that they could make it faster but maybe it would not be as accurate!
Canon Bob
Loves the Enemy
- Messages
- 9,754
- Name
- Bob
- Edit My Images
- Yes
That's precisely the reason Roy but Canon's explanation doesn't specify whether it's lens based or body based imposed limit.
Bob
- Messages
- 3,383
- Edit My Images
- No
Slightly off topic but I remember a few years ago comparing 1.4x tc's when attached to a 400/5.6 on a 40D body. I taped the pins on the Canon 1.4x and a Kenko Pro 1.4 x and then compared each of them against a cheap non reporting Kenko (1.5x tc) and the non reporter was quite a bit quicker to focus (trouble was the IQ was not as good as the two reporters). BTW The two taped tc's were just about the same. In non of the cases did the Camera know a tc was present so why did the 'cheapy' perform so much quicker ?
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
It's confusing isn't it! Re your last comment, it sounds like the camera did somehow know that a TC was present, even though you'd taped some pins?
Check this press release when Kenko released the new DGX telecons with improved reporting, and see if you can make head or tail of it! See the note at the bottom too http://www.thkphoto.com/news/news-pr012010-9.html
It seems like the reporting protocol is both complicated and differs between cameras/lenses. A bit like Canon's E-TTL flash system which is notorious for playing differently between models when third party guns and triggers are attached, when their reverse engineering gets caught out and the guns need new firmware. They keep changing it.
Another strange thing I have found. 100-400L with Kenko 1.4x Pro 300 DG telecon, would not focus on at all a 40D, but would on a 350D, albeit rather slowly. Others have found the same, and also with the 7D which you might expect to do better.
There's more. The Kenko 1.4x works fine on my 5D2 and 70-200L 4 IS, and appears to report everything correctly
It also fits the 24-105L and that too will AF, but does not report the changed f/number. When I try it with a 50 1.8, it neither reports the correct f/number and will not focus at all - not on anything. It gets to the approximate focus position, then shunts back and forth either side of the sharp position endlessly - it just goes on for ever but never locks on under any circumstances. All I can say on that is the 70-200L is designed to take extenders, and has three extra contacts, which the extender also has pins for (the ones you tape) but camera does not. But exactly what the heck is going on, I have no idea
Canon Bob
Loves the Enemy
- Messages
- 9,754
- Name
- Bob
- Edit My Images
- Yes
Richard, Roy,
The basis for aperture and focal length reporting is simply the three additional pins....one common and two returns, each dedicated for 1.4x or 2x extenders. Any lens without the three additional pins simply won't see an extender. It is the lens that modifies its message to the body based on seeing these contacts closed.
Part two of the equation seems to revolve around the newer versions of Canon and Kenko units where it would seem that the extender also delivers something on the data stream to the body (without the lens being party to it) unlike the simple pass through contacts of earlier versions.
I suspect we'll see that the 500/4 II and 600/4 II's paired with MkIII extenders will not work with taped pins as the bodies will prevent AF attempts even though the lens will still report 600/4 instead of 1200/8.
Bob
The basis for aperture and focal length reporting is simply the three additional pins....one common and two returns, each dedicated for 1.4x or 2x extenders. Any lens without the three additional pins simply won't see an extender. It is the lens that modifies its message to the body based on seeing these contacts closed.
Part two of the equation seems to revolve around the newer versions of Canon and Kenko units where it would seem that the extender also delivers something on the data stream to the body (without the lens being party to it) unlike the simple pass through contacts of earlier versions.
I suspect we'll see that the 500/4 II and 600/4 II's paired with MkIII extenders will not work with taped pins as the bodies will prevent AF attempts even though the lens will still report 600/4 instead of 1200/8.
Bob
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Yes, thanks Bob. It's the "Part two of the equation" bit that's confusing.
I'd also like to know why I, and others, could not get a 40D to AF with taped pins on the Kenko and 100-400L, yet it does with the (older, cheaper) 350D. Just different camera I guess, but then Roy says he managed to get AF in this way on a 40D with a 400L 5.6!
And why won't the 50 1.8 work at all? Though having said that, it struggles to focus consistently even without an extender LOL
Edit: Bob, off topic, but do you have a source which explains which pins do what on Canon lenses? Double edit: googling, just found this http://en.wikipedia.org/wiki/Canon_EF_lens_mount
Last edited:
Canon Bob
Loves the Enemy
- Messages
- 9,754
- Name
- Bob
- Edit My Images
- Yes
Richard,
I'm going to hazzard a guess that the difference in 40D 400/5.6 operation would be down to the setting of Cfn III-1 for lens drive.
If the value is 0 then AF will only be attempted if focus is already close to the correct distance. If the focus is not close enough to make a move calculation then no move will be made until either a manual focus input is made or the lens is directed at something closer to the current focus distance. If the value is 1 then the lens will perform an end to end search and inevitably get to the target.
This was selectable on the 40D but simply not available on the 350D.
Bob
I'm going to hazzard a guess that the difference in 40D 400/5.6 operation would be down to the setting of Cfn III-1 for lens drive.
If the value is 0 then AF will only be attempted if focus is already close to the correct distance. If the focus is not close enough to make a move calculation then no move will be made until either a manual focus input is made or the lens is directed at something closer to the current focus distance. If the value is 1 then the lens will perform an end to end search and inevitably get to the target.
This was selectable on the 40D but simply not available on the 350D.
Bob
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Possibly Bob, cheers. It was a while ago now and I haven't got the kit to replicate any more.
Re the pins protocol, after a quick google I can't find anything much beyond what's on that Wiki link with the pins ID. It seems like others have tried, and also not got far.
- Messages
- 3,383
- Edit My Images
- No
Taped converters with either the 100-400 or 400/5.6 seem to work reasonably well up to and inclusing the 30D but from the 40D onwards they are almost non starters (on the centre focus point anyway - they still worked reasonably well on one of the outer focus points of the 40D) I have always thought it was something to do with the centre focus point being enhanced for better performance with a f2.8 lens on the 40D and onwards.
Last edited:
- Messages
- 3,383
- Edit My Images
- No
Are you sure you have not got the wrong way around Bob!, if you set it to 0 (the default) then focus search in ON, with it set to 1 the focus search will be off and will not focus unless there is something near as you say.
I have played around with this many times when I had the 400/5.6 but it did not make a lot of difference as far as AF with a taped tc in my experience. It did stop 'hunting' when set to 1 but did not obtain focus any better with it on or off. From lots of stuff I have read for the 40D and taped tc's the 400/5.6 seemed to do better than the zoom for some reason.
Last edited:
Canon Bob
Loves the Enemy
- Messages
- 9,754
- Name
- Bob
- Edit My Images
- Yes
Comms are 19200 baud with 8 data bits and 1 stop bit but no parity.
The protocol is a series of Hex bytes with interogation and response
The body initiates this with id<13h>
The lens responds ok<13h>
followed by <focal length>,<aperture><13h>
After this initial wakeup, the body then goes on to command the lens to absolute AF positions and apertures as shooting takes place.
Each lens has a unique (almost unique) code that the body can insert into the Exif data.
Any deeper and I'll start talking gibberish.....hmm, "Talk Gibberish", a new section maybe?
That be the case Roy....a small plonker moment was just had
Bob
- Messages
- 23,200
- Name
- Richard
- Edit My Images
- No
Comms are 19200 baud with 8 data bits and 1 stop bit but no parity.
The protocol is a series of Hex bytes with interogation and response
The body initiates this with id<13h>
The lens responds ok<13h>
followed by <focal length>,<aperture><13h>
After this initial wakeup, the body then goes on to command the lens to absolute AF positions and apertures as shooting takes place.
Each lens has a unique (almost unique) code that the body can insert into the Exif data.
Any deeper and I'll start talking gibberish.....hmm, "Talk Gibberish", a new section maybe?
That be the case Roy....a small plonker moment was just had
Bob
