p.2 #1 · EF–R Pinout - will 3rd party lenses be locked out from RF mount?
AJSJones wrote:
The bayonets for the two mounts are incompatible, so any extra pins on new EF lenses wouldn't be able to reach the new pins on the camera, right? New EF lenses might possibly have things (processors) in that the RF adapter could communicate with that older EF lenses don't, so "older EF pins" might have updated functions - such as new IS algorithms where the canera provides information and refines the IS group movement. If that's the case, I think R lenses will be developed on Canon's normal (slow) timeline while EF lenses that benefit from "tweaks" (rather than full conversion to RF only models) will complete whatever EF roadmap has been in place for years. That would maintain support for EF but might encourage folks to buy RF bodies to get the most benefit. AKA - they don't think their EF body base will switch over quickly - the white paper seems clear that EF will continue for some time to support them.
You mention "new EF 400 and 600 mm lenses? They are marketed as compatible with RF" - does this mean something dfferent from "All EF lenses work on the RF adapter"? I've not seen anything special about them do you have any more detais?...Show more →
You are right, there are only 8 pins on the lens side of all the adapters. I got it from melcat's post something I interpreted as also the lens side of the adapters having 4 extra pins. That's why I thought there might be extra pins also on the new EF lenses, but obviously not.
p.2 #2 · EF–R Pinout - will 3rd party lenses be locked out from RF mount?
Mike_5D wrote:
The problem with encryption is that it gives the manufacturer a legal weapon against 3rd parties. At least in the US, even attempting to break encryption, no matter how weak, is illegal. Of course, this work could be done outside the US, but the resulting product would probably be illegal to sell here.
True, but I mostly opined on status quo - i.e. unencrypted protocol. New pins might transfer new kind of information.
p.2 #3 · EF–R Pinout - will 3rd party lenses be locked out from RF mount?
alundeb wrote:
You are right, there are only 8 pins on the lens side of all the adapters. I got it from melcat's post something I interpreted as also the lens side of the adapters having 4 extra pins.
Yes, I just checked both in B&H's product shots and the white paper, and I was wrong about that. There are 8 pins, only, on the EF side of the adapter. I will insert a note in my original post.
However, I also checked my video source for the number of pins on the body side of the plain adapter, and it's definitely 12. It's at 6:19 in this video:
&frags=pl%2Cwn
and unpausing from there you can clearly see it's a plain adapter, without the control wheel.
Sadly, not only does that limit R support in the two new big whites, it also negates one of my two benign reasons for those pins.
p.2 #6 · EF–R Pinout - will 3rd party lenses be locked out from RF mount?
melcat wrote:
I am finding it very hard to think of a good reason why those 3 pins for the old EF data are still there if it is not for Canon to keep some parts of the protocol private to itself. Why have 3 more pins, 3 extra points of failure?
The RF-series lenses (and one of the EF to R adapters) has an extra control ring. Looks like you can define it to control just about anything, so that you can have aperture, shutter speed and ISO (or some other combo) each on its own dial/ring controller (one on the lens, two on the camera). That'll take extra pins.
p.2 #8 · EF–R Pinout - will 3rd party lenses be locked out from RF mount?
Based on the Canon white paper about the EOS R and RF mount, it sounds like they want to create lenses with more controls on them. The old pin style limited their ability to add electronic controls and functions. Now with more pins, we see that there is a control ring on RF lenses. I wouldn't be surprised if we see other functions added (mappable buttons for example) on longer lenses. Since a hand is typically tied up holding a long lens, it makes sense to add some finger controls to the lens to help with camera control.
I could easily see a control ring near where one would hold a lens, along with a couple of buttons. Map a button to an exposure setting, or AF mode for example, turn the control ring to change that value...
p.2 #10 · EF–R Pinout - will 3rd party lenses be locked out from RF mount?
jcolwell wrote:
The RF-series lenses (and one of the EF to R adapters) has an extra control ring. Looks like you can define it to control just about anything, so that you can have aperture, shutter speed and ISO (or some other combo) each on its own dial/ring controller (one on the lens, two on the camera). That'll take extra pins.
Canon don't need to add more pins to add more control doodads to the lens, and although they chose to use the 4 new pins for the control ring it's unlikely they needed them just for that. Indeed, some existing EF lenses have AF stop buttons without requiring dedicated pins for that.
That's because the processors in the lens and camera body format the information into messages. The messages will be distinguished in some way from each other, typically by the first byte or two being a number. For example, when a lens sees power come on, it might send message type 1, containing the lens's name, speed and focal length range, which the body can use for the EXIF. When you take a picture, the body will command the lens to stop down, and the lens will do that and reply with a message (say #3) saying it did, and the focus distance and zoom setting when it did. When you press or release the AF stop button on a lens that has one, those will also be messages, say 4 and 5. And so on: you could define messages 6 and 7 for control wheel clockwise and anticlockwise. There will be other messages to say the lens has established IS lock, completed a request to focus to a given distance etc. Similarly, there is a different set of messages sent to the lens by the body.
The message numbers, format of the records, and the way the body and lenses respond to messages from each other together make up the protocol. This is what third party lens makers like Sigma have to reverse engineer.
Maybe the old synchronous data link is very slow and sending lots of messages when you turn the ring is enough to delay more important messages about AF and IS going from the lens to the body, but it doesn't seem likely. Just how fast can you turn a control ring?
The new asynchronous link will be faster, perhaps much faster, and probably also packetised so that a long message like DLO data doesn't block short messages. That was why I thought they'd just send everything now using the new link.
But I'm not a hardware engineer, just a software guy who's written drivers to operate hardware like this, and I tend to think of things in ways that make them operate as simply as possible, because that's what's right for software. I think in this case I forgot to consider that they probably have lots of existing lens modules and designs they want to drop into the new RF lenses, like focus motors and aperture mechanisms, and maybe it's just easier for them to drop those into a design that uses the old EF data pins.
p.2 #11 · EF–R Pinout - will 3rd party lenses be locked out from RF mount?
melcat wrote:
Canon don't need to add more pins to add more control doodads to the lens, and although they chose to use the 4 new pins for the control ring it's unlikely they needed them just for that. Indeed, some existing EF lenses have AF stop buttons without requiring dedicated pins for that.
That's because the processors in the lens and camera body format the information into messages. The messages will be distinguished in some way from each other, typically by the first byte or two being a number. For example, when a lens sees power come on, it might send message type 1, containing the lens's name, speed and focal length range, which the body can use for the EXIF. When you take a picture, the body will command the lens to stop down, and the lens will do that and reply with a message (say #3) saying it did, and the focus distance and zoom setting when it did. When you press or release the AF stop button on a lens that has one, those will also be messages, say 4 and 5. And so on: you could define messages 6 and 7 for control wheel clockwise and anticlockwise. There will be other messages to say the lens has established IS lock, completed a request to focus to a given distance etc. Similarly, there is a different set of messages sent to the lens by the body.
The message numbers, format of the records, and the way the body and lenses respond to messages from each other together make up the protocol. This is what third party lens makers like Sigma have to reverse engineer.
Maybe the old synchronous data link is very slow and sending lots of messages when you turn the ring is enough to delay more important messages about AF and IS going from the lens to the body, but it doesn't seem likely. Just how fast can you turn a control ring?
The new asynchronous link will be faster, perhaps much faster, and probably also packetised so that a long message like DLO data doesn't block short messages. That was why I thought they'd just send everything now using the new link.
But I'm not a hardware engineer, just a software guy who's written drivers to operate hardware like this, and I tend to think of things in ways that make them operate as simply as possible, because that's what's right for software. I think in this case I forgot to consider that they probably have lots of existing lens modules and designs they want to drop into the new RF lenses, like focus motors and aperture mechanisms, and maybe it's just easier for them to drop those into a design that uses the old EF data pins. ...Show more →
I know how electronic communications work. I was merely offering a plausible reason to explain why there are more contacts.
p.2 #12 · EF–R Pinout - will 3rd party lenses be locked out from RF mount?
Mike_5D wrote:
The problem with encryption is that it gives the manufacturer a legal weapon against 3rd parties. At least in the US, even attempting to break encryption, no matter how weak, is illegal. Of course, this work could be done outside the US, but the resulting product would probably be illegal to sell here.
Nope!
Don't worry about this. This may be sort of in the DMCA but don't fret. It doesn't apply to stuff like this.
It ONLY applies to stuff where the content is what has value and is being protected. Movies, etc.
It does not apply to things like this. So don't fret! It's all good!
It's been tested in court. Lexmark tried this with printer ink. They then tried to use the DMCA to sue people who made 3rd party ink.
They lost. They were basically laughed at by the court.
Reverse engineering a lens protocol would be just fine.
There's also an older court case about garage door openers btw. I don't remember the name.
I do remember that the generic company that was makign competing garage door openers won easily. So..again.. It's all good. The DMCA is bad, but 'tis not THAT insane. :P
