Fredy Ross
Veteran Member
Hope somebody has the answer to this as my searching only talked about the adapter with the M1 from a few years ago. Thanks in advance for the answer.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Does the ef/efs adapter slow down the Af of the M3?
- Thread starter Fredy Ross
- Start date
-
- Tags
- canon eos m3
verum lens
Leading Member
- Messages
- 709
- Reaction score
- 303
Unfortunately, just because Canon patents something doesn't mean it will ever hit retail.
Fredy Ross
Veteran Member
thanks and I think they will produce it if it is a handicap.
It isn't the adapter, but the lenses that make the difference. Some EF-S and EF lenses can be almost as fast as native EF-M lenses. Others are drastically slower.
EF-M, EF-S, and EF lenses all communicate with the camera using the same protocol, but with different pin assignments. The EF to EF-M adapter has no internal "brains" it is basically just a tube with wires connecting the front to the back.
EF-M, EF-S, and EF lenses all communicate with the camera using the same protocol, but with different pin assignments. The EF to EF-M adapter has no internal "brains" it is basically just a tube with wires connecting the front to the back.
Coyote_Cody
Senior Member
Help me understand why a set of 10 wires(whatever the #), 2 inches at most long, from cam via contacts thru adapter to EF/EF-S lens via contacts, wire only now, no electronics, slows down AF ? How ?
How can you even know, no way for direct comparison is there - M w & w/o adapter - same lens - HOW? Can't use two diff AF cams in a comparison you know - that's always apples vs orangutangs.
The above mentioned new adapter has lens involved, does it not, so how does this make AF faster ? Metabones and such are NOT used for this purpose, so Canon needs lens to make AF faster - seems very ODD to me.
So the slower AF is purely due to the longer distance from rear lens to M sensor, or an extra thousandsth of an ohm in the wire? Is this a 'brightness' issue or some bizarre 'angle of light' hitting the sensor OR ?????
M AF is a closed loop system and just this little length of wire makes this big of diff, or this slightly diff in light ?
Don't get it !!
Someone please explain, use tech terms if possible, no Harry Potter plz !!
How can you even know, no way for direct comparison is there - M w & w/o adapter - same lens - HOW? Can't use two diff AF cams in a comparison you know - that's always apples vs orangutangs.
The above mentioned new adapter has lens involved, does it not, so how does this make AF faster ? Metabones and such are NOT used for this purpose, so Canon needs lens to make AF faster - seems very ODD to me.
So the slower AF is purely due to the longer distance from rear lens to M sensor, or an extra thousandsth of an ohm in the wire? Is this a 'brightness' issue or some bizarre 'angle of light' hitting the sensor OR ?????
M AF is a closed loop system and just this little length of wire makes this big of diff, or this slightly diff in light ?
Don't get it !!
Someone please explain, use tech terms if possible, no Harry Potter plz !!
The adapter is only some wires and a hole, all it does is put the lens a bit farther from the sensor (increase the flange distance to what it would be on a DSLR).
The autofocus speed will depend on the lens used.
In my experience, the M3 is faster to focus with the adapter and EF/EF-S lenses, when compared with the M1.
To be more technical and precise, the M1 moved the lens almost random in one direction or another, it got the the end of the focus throw, and returned (only if it didn't guess the right direction), and then it hunted a bit near the correct focus. The M3 goes in the right direction, and when it gets there it might hunt a little or not at all (In my opinion the camera clearly uses phase detect at first and then contrast AF to fine tune).Here is a video I made regarding autofocus vs M1:
The autofocus speed will depend on the lens used.
In my experience, the M3 is faster to focus with the adapter and EF/EF-S lenses, when compared with the M1.
To be more technical and precise, the M1 moved the lens almost random in one direction or another, it got the the end of the focus throw, and returned (only if it didn't guess the right direction), and then it hunted a bit near the correct focus. The M3 goes in the right direction, and when it gets there it might hunt a little or not at all (In my opinion the camera clearly uses phase detect at first and then contrast AF to fine tune).Here is a video I made regarding autofocus vs M1:
Coyote_Cody
Senior Member
That is what I said above.
That is obvious, HOW does the adapter compromise this or lessen this ?
I/we expected/demanded this, and how could Canon NOT improve the AF of M3 w/o angering EVERYONE !?
Your above M1 AF description really has nothing to do with an/the adapter, that is how ALL AF systems will work in some conditions, ANY/ALL AF systems must first 'determine' where to move to improve focus by testing the AF, once 'direction' is determined, then it proceeds, M1 just searched SLOWLY TOO much (a hunting dog!!), & does it with native lens OR via adapter in my experience.
So how does an adapter decrease AF speed ???
PPL seem to assume this is a given, and it is NOT ! Unless someone shows me otherwise.
THx!!
With the use of the adapter, there is now an extra set of contact points and extra distance the AF signal must pass, between both the adapter and lens, then back to the camera's processor.
The extra distance and contact points introduce resistance through which the signal must pass.
This slows a signal down by mere nano/pico seconds.
So, yes, it does slow it down.
I highly doubt any person's eye/brain function can detect it, though.
The extra distance and contact points introduce resistance through which the signal must pass.
This slows a signal down by mere nano/pico seconds.
So, yes, it does slow it down.
I highly doubt any person's eye/brain function can detect it, though.
Last edited:
verum lens
Leading Member
- Messages
- 709
- Reaction score
- 303
I can't tell if this post is being made in jest.
Also the reason focusing is slower on the M is because it uses an entirely different mechanism to find focus.
All the lens knows is "focus closer" or "focus further", the speed that happens in the same on the EOS M or the 5D Mark III. The amount of time it spends in the the state of "focus closer" or "focus further" is what varies between the cameras.
Last edited:
Why would you feel my reply is in jest?
It's a very specific answer to the very specific question posted in the topic.
No other parameters were given.
Coyote_Cody
Senior Member
IIRC, 11" of high quality wire is ~1 nano-second (propagation speed), early supercomputers (cray) used 'wire length' to time its signals (that's why I remember this fact) !!
I simply do NOT think a few mili-ohms OR 0.15 ns will make MUCH/ANY diff to the AF of ANY camera, assuming more than a few tenths of a volt are used to carry the 'driving voltage' to the AF motor in lens (certainly much more than this is used) !!
A very small voltage (one of the most significant things about Li-on batteries is the relatively constant voltage until a large drop-off at near total discharge) from battery MIGHT cause AF slowing, but then all functions of cam will likely slow down or quit.
No, I think ALL this jibber-jabber (like that phrase?) about the adapter effecting the AF is rather UNscientific/factual until I see some engineering/science/empirical evidence to prove it. (climate change debate anyone?? religion debate ?)
Exactly.
Entropy512
Veteran Member
I know this is resurrecting a somewhat old thread, however - I'm a Sony body owner who has done some reverse engineering of the E-mount protocol, and for whatever reason this thread came up when searching for additional details on the EF-mount protocols beyond my personal favorite reference at https://pickandplace.wordpress.com/2011/10/05/canon-ef-s-protocol-and-electronic-follow-focus/ - specifically regarding command 0x50 (motor speed control) mentioned in the comments of that blog post. (This is important because I believe 0x50 is critical to live view/CDAF compatibility, and appears to be missing from nearly all Canon EF to Sony E-mount adapters per http://www.dpreview.com/forums/post/56692298 )
The key being - Does an EF/EF-S lens on an M3 perform significantly different from the same lens on an SLR body with Hybrid CMOS AF III in liveview mode?
As to Canon's patent regarding an AF adapter with improved "speed" - it looks like the patent is in a foreign language (Japanese?) and perhaps someone/something misinterpreted the definition of "speed" here. Why? The patent describes optics in the adapter - there is no reason for this for EF to EF-M unless the adapter is a focal reducer.
What's a common name for focal reducers? Speed Booster
Is there anyone here over in Canon-land that has done any EF mount reverse engineering/control with a microcontroller? Can anyone confirm for me that there is indeed a focus motor speed control command (0x50?) - does it work on the EF85/1.8 USM? It would save me a lot of time wiring up a Canon extension tube with pigtails to drive my 85/1.8 (although I will have to do this eventually...)
Hi Freddy Ross,
I have used the Eos M and the M3 with my EF 100-400mm L IS II and compared the AF speed of this lens to the Canon 70D. My long lens has faster AF with the 70D and has fast tracking AF as well.The AF on the M3 with this lens and adapter is dependent on having good light. This combination( M3 + long lens) does NOT AF in servo mode( tracking AF).
I find that my ef-m lenses will perform in servo AF mode with the M3 but not as well as the 70D with EF or EF-S lenses. I am only offering my practical user experiences from the past year. For me, Single shot AF with the M3 + EF 100-400mm L IS II is not too bad and usable.
I do not offer any technical explanations/speculations as I believe that Canon knows their AF systems better than I do.
I would really like a Canon mirrorless body that has AF and tracking AF that is as good as Canon's 70D. Maybe the Eos M4 will be as fast a Canon's mid range DSLRs.( I am hopeful)
I can say that the the image quality of the M3 is very comparable to the 70D (IMO
) I have produced many beautiful images with this setup ( M3 + EF 100-400mm L IS II).
Cheers
Murray
Black Necked Stilt silhouette
Lilly Pilly blossoms turning into buds
Yellow Striped Flutterer dragonfly.
Forest Kingfisher
The Eastern Dwarf Tree Frog is 12-15mm at full adult size.
Adult Eastern Dwarf Tree Frog.
I have used the Eos M and the M3 with my EF 100-400mm L IS II and compared the AF speed of this lens to the Canon 70D. My long lens has faster AF with the 70D and has fast tracking AF as well.The AF on the M3 with this lens and adapter is dependent on having good light. This combination( M3 + long lens) does NOT AF in servo mode( tracking AF).
I find that my ef-m lenses will perform in servo AF mode with the M3 but not as well as the 70D with EF or EF-S lenses. I am only offering my practical user experiences from the past year. For me, Single shot AF with the M3 + EF 100-400mm L IS II is not too bad and usable.
I do not offer any technical explanations/speculations as I believe that Canon knows their AF systems better than I do.
I would really like a Canon mirrorless body that has AF and tracking AF that is as good as Canon's 70D. Maybe the Eos M4 will be as fast a Canon's mid range DSLRs.( I am hopeful)
I can say that the the image quality of the M3 is very comparable to the 70D (IMO
) I have produced many beautiful images with this setup ( M3 + EF 100-400mm L IS II).Cheers
Murray
Black Necked Stilt silhouette
Lilly Pilly blossoms turning into buds
Yellow Striped Flutterer dragonfly.
Forest Kingfisher
The Eastern Dwarf Tree Frog is 12-15mm at full adult size.
Adult Eastern Dwarf Tree Frog.
Entropy512
Veteran Member
I still can't understand why Canon doesn't use their flagship on-sensor PDAF technology (dual-pixel AF) in their mirrorless cameras. They have EVERYTHING in place technology-wise to make a massive push into mirrorless, but instead Canon users get nonsensical bodies like the 80D and mirrorless bodies that lack DPAF.
The Dual-Pixel sensors make the most sense for mirrorless and video/cine cameras - but Canon stated the 80D was "primarily for stills" when explaining the lack of 4k - why stick a sensor that's going to shine the most when recording video in a mirrored SLR that is "primarily for stills"?
I'd still be interested to see a comparison of the M3 vs. one of the mirrored bodies that has HCAF III (such as the T6i/T6s) in live view with the same lens in terms of AF performance.
This purely speculation, but I wonder if the shorter flange distance of the M system (and resulting sharper incident light angles) wreaks havoc with the DPAF.
Entropy512
Veteran Member
Interesting theory, which would indicate that DPAF's approach is somehow more sensitive to that than HCAF III and Sony's OSPDAF approaches - although the impression I got from at least one presentation on HCAF III (unfortunately, it was photos of presentation slides from a conference somewhere so it was very low-res, and I also don't have a link) kind of implied that HCAF III sites were similar in design to DPAF sites - except that Canon only made some pixel sites AF sites.
There was also a link I saw a while ago (again, can't find it now, sorry...) indicating that Fuji did it by masking off half of each of two pixel sites and having them work together - similar to DPAF but obviously NOT usable on every site. Speculation is that Sony's approach is probably similar.
A comparison of an EF lens on an HCAF III mirrored body such as one of the T6 versions and the M3 with the EF-EFM adapter would probably answer a lot of questions - I suspect most of the limitations of the M3 with EF/EF-S lenses are just fundamental HCAF III limitations and would go away if a DPAF sensor were used.
Keyboard shortcuts
- f
- Forum