Patent: New Canon Mount Coming?

jolyonralph

Game Boy Camera
CR Pro
Aug 25, 2015
1,423
944
London, UK
www.everyothershot.com
Addressing a few points here.

1) The mount in the image has seven data pins. The EF mount has eight pins (the Extenders have a few more though and lenses that work with the extender have the pads to connect to these). The EF-M mount has nine pins.

So, the mount in the image is a mock-up of a generic mount. Nothing more.

2) These extra pins for the Extenders are probably what made Canon choose to avoid the simple solution of just adding a couple of new pins for the new communications channel.

3) For those who see no reason to use APS-C lenses on a FF body, that's your choice. There are many of us who would do this in certain cases. A 50mpx sensor for example gives an APS-C crop of a perfectly useable 20mpx, so if you're simply not able to buy all the FF lenses you want you can continue to use your regular EF-M lenses until you replace them.
 
Upvote 0
How about this idea:

It is not an M mount at all. It is a new mount they will use on new bodies, you will be able to use all of your existing lens.

Now the part where this makes Canon a ton of money, you buy a new lens it will only be compatible with the new bodies. If you want the latest new nifty lens you will have to shell out for the latest body.

It will make people mad and they will still buy the new lenses. Canon will make a ton of extra money off of it. Life goes on.
 
Upvote 0
jolyonralph said:
Addressing a few points here.

1) The mount in the image has seven data pins. The EF mount has eight pins (the Extenders have a few more though and lenses that work with the extender have the pads to connect to these). The EF-M mount has nine pins.

So, the mount in the image is a mock-up of a generic mount. Nothing more.

not really the communication between the mount and lens could need less pins than before. and they aren't all data pins. the EF mount has 3 ground pins. Two power and one Digital.

it entirely depends on how they interconnect, or even have a EF-X to EF adapter to support teleconvertor power,etc.

and even besides that, it's clear that canon is at least looking at EF 2.0 in this patent.
 
Upvote 0

ahsanford

Particular Member
Aug 16, 2012
8,620
1,651
fullstop said:
"auto crop" was introduced as a "stop gap fix" by Nikon when after years of preaching "DX is more than enough" they changed course in mid-stream and introduced FX bodies. So they a lot of customers sitting on DX lenses and wanted to tell them they can continue to use them on new FX cameras as well. Yes, technically they can. But no, in my opinion it still does not make sense in real life. It is a waste of precious sensor real estate and basically a "marketing trick".

I agree in broad strokes with using all of the sensor you paid for, but what you call a marketing trick others might call a far less steep cliff to climb to get into FF. Some folks are financially constrained and need to wade through the upgrade process over time.

Consider: Canon only sells two truly inexpensive EF lenses -- the 40 pancake and 50 f/1.8. If you had a 55-250, you're looking at a $1900 100-400L to replace that zoom range. If you use a standard zoom, Canon doesn't sell one in EF for less than $599. These things make the jump to FF a painful proposition.

...or if you are with Nikon you just keep snapping away with your DX glass until you can afford FX glass. By offering this, the full-frame cliff is a little less daunting. I don't see that as a negative -- not at all.

- A
 
Upvote 0

ahsanford

Particular Member
Aug 16, 2012
8,620
1,651
canonnews said:
fullstop said:
"auto crop" was introduced as a "stop gap fix" by Nikon

that's why both Sony and Nikon do it, and canon is the only one that does not.

just because you don't like it doesn't make it a valid use case.

+1. All day.

And the genesis behind an idea should diminish the value of the idea.

- A
 
Upvote 0
Apr 23, 2018
1,088
153
all the Nikon shooters i know where quite excited at first and used their DX lenses a few times on FX bodies. But not for very long. All of them sold their DX lenses rather quickly and got "real glass" for FF sensors.

Again, I'd rather rent an FF lens if I would not want to buy it, rather than using crop lenses on FF sensors.

There may well be a few people who see it differently, but from what i see in the real world, it is a small nich segment - definitely not a mainstream thing, that Canon should cater to in designing their new mirrorless FF camera system. But no problem here, if they - provided it does not get in my way and i don't have to subsidize it with purchase price. :)
 
Upvote 0
Mar 25, 2011
16,848
1,835
The translation of this paragraph is pretty rough, but it tries to explain the need for a faster protocol. The camera must wait for the lens to respond (gets a busy signal), and that limits how fast a lens can operate. It did not refer to the need to pass more data, just a limit to exchanging information. I like the "evil" bit!

I think it may have the potential to become a new communication protocol for all lenses. It might be interesting to see if it appears in the new 70-200 lenses soon to be announced. Unless they announced it, someone would have to look for the bump or new hole that would match a future camera. It seems unlikely, but products have been introduced with hidden capabilities before.


"Calling Busy the state where this waiting time has arisen, a lens does not receive the communication from a camera between Bus(ies). Therefore, even if it raises the frequency of a clock, in order that the microcomputer on the camera side may communicate Busy release of a lens with waiting, there is a limit in improvement in performance of operation. Since the communication interrupt from a camera occurs frequently in the microcomputer on the lens side on the other hand, and communication interrupt processing for a Busy signal output and release must be performed preferentially each time, Since it becomes the evil of performance improvement of operation also here, the proposal which is not based on communication but judges a lens type exception is made."
 
Upvote 0
CafferyPhoto said:
jolyonralph said:
this patent is just about communication protocols
What protocols do the other companies (Sony, Nikon) use?


From an electrical point of view I am assuming a 2-bit, 3-bit or 4-bit parallel protocol much like the old-time Centronics printer ports but with less pins! In this case I suspect depending upon the number of pins on various mounts that Canon uses 2 data pin + 2 ground pins, or 3 data pins and 3 ground, or 4-pins for data and 4 pins for ground.

It could also be a serial protocol such as RS-232 or RS-422 like they used on old-style video cameras and editing decks. In terms of speed the Baud rate (in bits per second) Canon should be able to get between 56k/baud up to 384k/baud and depending upon the error correction protocol used which needs extra bits, it could be between 6000 bytes per second up to 48,000 bytes per second which is PLENTY to get and set various lens parameters at up to 120 frames per second! This means it would be CHEAP to produce (less than $3 per lens) because if they tried doing a micro-USB style serial communications system at MANY megabytes per second, it would cost between $10 to $150 extra per lens depending upon the lens parameters needed to be get/set and powered!
 
Upvote 0
Mar 25, 2011
16,848
1,835
HarryFilm said:
CafferyPhoto said:
jolyonralph said:
this patent is just about communication protocols
What protocols do the other companies (Sony, Nikon) use?


From an electrical point of view I am assuming a 2-bit, 3-bit or 4-bit parallel protocol much like the old-time Centronics printer ports but with less pins! In this case I suspect depending upon the number of pins on various mounts that Canon uses 2 data pin + 2 ground pins, or 3 data pins and 3 ground, or 4-pins for data and 4 pins for ground.

It could also be a serial protocol such as RS-232 or RS-422 like they used on old-style video cameras and editing decks. In terms of speed the Baud rate (in bits per second) Canon should be able to get between 56k/baud up to 384k/baud and depending upon the error correction protocol used which needs extra bits, it could be between 6000 bytes per second up to 48,000 bytes per second which is PLENTY to get and set various lens parameters at up to 120 frames per second! This means it would be CHEAP to produce (less than $3 per lens) because if they tried doing a micro-USB style serial communications system at MANY megabytes per second, it would cost between $10 to $150 extra per lens depending upon the lens parameters needed to be get/set and powered!
There is quite a bit of information about the existing lens protocol in the patent, it is described as a serial protocol. Since the communication protocol is from the 1980's, I suspect that its deadly slow. As to cost, its likely pennies per lens, it appears in even low end lenses like the 50mm 1.8 that costs about $10 to build.
 
Upvote 0
HarryFilm said:
CafferyPhoto said:
jolyonralph said:
this patent is just about communication protocols
What protocols do the other companies (Sony, Nikon) use?


From an electrical point of view I am assuming a 2-bit, 3-bit or 4-bit parallel protocol much like the old-time Centronics printer ports but with less pins! In this case I suspect depending upon the number of pins on various mounts that Canon uses 2 data pin + 2 ground pins, or 3 data pins and 3 ground, or 4-pins for data and 4 pins for ground.

It could also be a serial protocol such as RS-232 or RS-422 like they used on old-style video cameras and editing decks. In terms of speed the Baud rate (in bits per second) Canon should be able to get between 56k/baud up to 384k/baud and depending upon the error correction protocol used which needs extra bits, it could be between 6000 bytes per second up to 48,000 bytes per second which is PLENTY to get and set various lens parameters at up to 120 frames per second! This means it would be CHEAP to produce (less than $3 per lens) because if they tried doing a micro-USB style serial communications system at MANY megabytes per second, it would cost between $10 to $150 extra per lens depending upon the lens parameters needed to be get/set and powered!

I think I have read that camera-lens-data exchange runs via I2C (maybe not Canon) and - I checked the following link:
https://photo.stackexchange.com/questions/61959/does-anyone-have-any-information-on-canons-vl-mount-regarding-pinout-and-protoc
Where the following contact scheme is provided:
Contact shape / Function
Dot / Battery voltage - supply for all actors
Bar shaped / Analog ground (minus) for actors like motors
------------------------------------------------------------------
Dot / Digital supply voltage
Dot / Digital data signal out from Camera to Lens
Dot / Digital data signal in from Lens to Camera
Dot / Digital clock signal to "align" the bits for both information channels
Dot / Digital ground (minus)

Which seems a well done arrangement for a mixed digital / analog system.

So it's definitely not RS-232/-485 (no common clock) nor I2C (only one data wire). If it is SPI (a synchronous serial interface using a common clock) they have no chip select (selection of a device) lines, maybe they use the cpu in the lens to interpret commands and send the info to the correct subdevices in the lens for e.g. focusing and electr(on)ic aperture.

Maybe Canon prepares to use I2C which needs one contact less. I2C is a wide spread communication protocol for sensors, displays and some actors. I2C uses address based bus devices (very flexible), allows more than one controler on the bus and can reach data transmission speeds up to
>>>>>
– Standard-mode (Sm), with a bit rate up to 100 kbit/s [~10 kByte / s]
– Fast-mode (Fm), with a bit rate up to 400 kbit/s [~40 kByte / s]
– Fast-mode Plus (Fm+), with a bit rate up to 1 Mbit/s [~100 kByte / s]
– High-speed mode (Hs-mode), with a bit rate up to 3.4 Mbit/s. [~340 kByte / s]
<<<<<< from https://www.i2c-bus.org/specification/

A 100 Hz readout of an acceleration sensor wired up by I2C with an Arduino Nano and transfer to a Serial line via USB is no problem for me as a medium experienced former programmer - what if well experienced programmers use I2C and optimize it in a well designed environment?

And what for? Who knows but a lens using at least 3 linear drives (with 3 independent motors) for focusing could adjust lens group tilt by itself (via DPAF) or do some tilt to extend the content which is in focus. This would at least triple the amount of data for the focusing operation.
 
Upvote 0
Apr 25, 2011
2,510
1,885
mb66energy said:
If it is SPI (a synchronous serial interface using a common clock)
It's reportedly SPI running at ~80kHz clock sync. No need for slave select because back in the 80s no one expected to have more than 1 MCU in the lens.

Actually, I'm surprised that negotiation for higher speeds done at lower speed of the same serial protocol is patentable at all.
 
Upvote 0

WoodyWindy

On the road again!
Jul 20, 2010
105
32
mb66energy said:
While the main content of the patent is about communication protocols the most interesting detail for me is
the ASPECT RATIO shown for the (possible) sensor (area). Maybe some preparation for larger than 36x24 mm sized sensors for future medium format sensors or multi-format large area to exploit the 43mm image circle of FF lenses in e.g. square format (31x31mm, very hypothetical I think).

I was wondering if I was the only one who had noticed that. But, I agree with others pointing out that the primary purpose of this patent is the communication protocols (which, really, is questionably patentable material given serial communications' long history of protocol negotiation - but that's a whole 'nother story.) The imagery is probably just placeholder, and meant to convey the general concept of application
 
Upvote 0
Oct 26, 2013
1,140
426
WoodyWindy said:
mb66energy said:
While the main content of the patent is about communication protocols the most interesting detail for me is
the ASPECT RATIO shown for the (possible) sensor (area). Maybe some preparation for larger than 36x24 mm sized sensors for future medium format sensors or multi-format large area to exploit the 43mm image circle of FF lenses in e.g. square format (31x31mm, very hypothetical I think).

I was wondering if I was the only one who had noticed that. But, I agree with others pointing out that the primary purpose of this patent is the communication protocols (which, really, is questionably patentable material given serial communications' long history of protocol negotiation - but that's a whole 'nother story.) The imagery is probably just placeholder, and meant to convey the general concept of application

I agree that the image is very generic and with no dimensional information or scale bar it is meaningless.
 
Upvote 0
Mar 25, 2011
16,848
1,835
Kit. said:
mb66energy said:
If it is SPI (a synchronous serial interface using a common clock)
It's reportedly SPI running at ~80kHz clock sync. No need for slave select because back in the 80s no one expected to have more than 1 MCU in the lens.

Actually, I'm surprised that negotiation for higher speeds done at lower speed of the same serial protocol is patentable at all.
Its all about the manner in which the speed change is done without adding a additional contact as was required in the previous patent. Since the existing lenses must work with it, the method of detecting and switching is what sets it apart.
 
Upvote 0