In my previous article I wrote about patent protection particularly of the Nikon Z system. Andy Westlake recently produced a piece on the strange incompatibility that has occurred with some third-party lenses when used on the new Sony A7 V, which raises IP issues with that system. This is ‘strange’ because Sony offers lens makers a free licence for the E-mount. This licence gives access to the ‘basic specifications’ of the lens mount system without royalty payment. I take ‘basic specifications’ to mean base-level protocols, which should allow operation with all E-mount cameras, but not later and more advanced versions of the protocol, which provide for additional functionality and faster operation.
In preparation for this article, I did a quick search for Sony autofocus patents, which revealed very little – only mechanical arrangements where the novelty was bizarrely having a mount not quite large enough for a full-frame sensor (US9131134B2). That’s not to say that there aren’t any more, but the same search with the word ‘Sony’ replaced by ‘Nikon’ or ‘Canon’ turned up hundreds. It would appear that patenting is not Sony’s major route to protecting its autofocus IP.
In general, there are two ways other than patents which are used to protect IP. One is copyright, but this is problematic when it comes to programs and protocols. There is some case law which has upheld this kind of copyright claim, but the circumstances have been quite particular. This leaves the third route to protection, which is trade secrets. Simply, detailed information is never disclosed generally. For this reason, what will follow in this article is purely informed speculation, since Sony will not tell us exactly what has happened.
For those that didn’t read the article, it appears that some third-party lenses failed to work in AF-S, AF-A and DMF when fitted on the A7 V camera. Andy suggested that the manufacturers of these lenses had reversed engineered the protocols rather than obtaining a licence. But since the licence is free, and would give them access to full implementation information, I think that it’s more likely that the A7 V has failed to operate with the ‘basic’ form of the protocol.
Some third-party manufacturers’ (the Japanese companies) lenses did operate satisfactorily, which I would take as an indication that they are using a higher level of the protocol than the free ‘basic’ one – very likely it entails a paid licence. However, these lenses are limited to operation at 15 frames per second, which suggests that there is a further level of protocol which is exclusive to Sony. I’ll discuss what the technical details might be in a later article.
How did this happen?
Another question is ‘how did this arise?’ Whilst we don’t know for sure, I can think of two possibilities which fit the observed facts. One is that the engineers writing the firmware for the A7 V simply made a mistake, and coded these functions in a way that didn’t work with the basic protocol. The second is that some part of the protocol definition was unintentionally ambiguous, and was interpreted differently by the Japanese manufacturers on one hand and the Chinese and Koreans on the other. In any case, we can deduce that it was not intentional, since the different lens manufacturers issued corrective firmware for their lenses very quickly and the same time, as if they have been issued with the knowhow by Sony.
The outstanding question is why Sony did not correct the firmware in the A7 V. Again, we do not know, but it occurs to me that to do so could have been a considerable undertaking for them, and also a possible admission of liability for the mistake. Far simpler to update (or clarify) the ‘basic protocol’ specification and let the lens vendors do the required work.
Related reading
- The real story behind the Nikon lens patent dispute with Viltrox
- Autofocus: DSLR vs Mirrorless Theory and Practice
- How does the Sony A9 III global shutter work?
