melcat
Offline
• • • •
Upload & Sell: Off
|
 p.1 #2 · EF–R Pinout - will 3rd party lenses be locked out from RF mount? | |
Now for what we know about the contacts on the RF lenses and adapters.
Canon added 4 more pins, for a total of 12.
We can assume from Canon's claims about higher speed that some or all of these pins are an asynchronous link, meaning that there is no clock signal. The electronics of the 1980s meant that synchronous links were faster, whereas the reverse is true now.
It may be that they use 2 of the 4 pins for an asynchronous link that can transmit in one direction at once (like USB 2), and the other 2 for 3.3V power and ground, but seems more likely to me that it's full-duplex (like USB 3), using all 4 pins for data and allowing transmission in both directions at once.
Obviously the RF lenses still use the 6V motor power on the first 3 pins of the EF mount, and given my speculation that the RF data link is full-duplex probably 2 more of those pins for 5.5V and its logic ground. That makes up in all pins 1, 2, 3, 4 and 8 of the EF mount. Pins 5, 6 and 7 are the data and clock of the old slow synchronous EF data link.
So the first odd thing is that pins 5, 6 and 7 are actually still there on the lens. Could there be a mechanical reason for that? Or do these new lenses still speak the old EF protocol for all their basic operation, with just the extras going over the new pins? If the latter, third party lens makers don't need to ever figure out the RF protocol. But they could not implement a control ring, send DLO data to the body, or use Canon's new closed-loop IS unless they did. Without locking out third parties completely and the attendant bad publicity, Canon would nonetheless have an advantage over them.
The next surprising thing is that all the adapters, including the plain one, include the 4 extra RF pins on the body end. This is disturbing because it raises the possibility that the body will refuse to work if there isn't something talking over those pins, either an adapter or lens. It would mean that third parties could not make short-register lenses that fitted directly onto the body without reverse-engineering the protocol. And that could be a problem if there is encryption involved, as printer manufacturers have done with cartridges.
I should add that none of the well-known reviews, brochures, white paper or online shops I originally looked at showed the lens side of the plain adapter. I eventually was able to see it by freeze-framing a promotional video from a non-US retailer.
The final surprising thing is that the 4 extra pins are present on the lens side of the EF–R lens adapter, even though no EF lens has them. EDIT: this is incorrect. I can think of two explanations:
- Canon just make 12-pin contact blocks to save costs and inventory, and the extra pins aren't used. This is encouraging because it would be a benign explanation of the extra 4 pins on the body side of the plain adapter.
- Canon want to keep open the option of adding RF features like the DLO data transfer to EF mount lenses in the future, perhaps ones meant for video with the drop-in filter adapter. In that case, they wouldn't want adapters out there that don't work with such lenses. But this seems very unlike Canon to make a product which would mount but not work on an EF camera.
(end of incorrect info and speculation)
In summary, it might be that the new mount locks out third party lenses completely, or hobbles their functionality, or does neither. Until we know, I wouldn't be buying. Ideally, Canon would follow Sony and give an undertaking to supply the lens interface to third parties.
Edited on Sep 11, 2018 at 05:18 AM · View previous versions
|
