sonyalpharumorsLog in

It is a common misconception that reverse engineering is illegal. In fact, it is perfectly legal to reverse engineer stuff to see how it works and to fix compatibility problems. What would be illegal is to steal code. It would also be illegal to utilize foreign patented technology without licensing, but that’s something completely different. In some jurisdictions, it is (since a couple of years) also illegal to break encryption for commercial purposes (but not for scientific or journalistic purposes), but yet again, this is not the same. The lens protocol is not encrypted, anyway.

So, there’s nothing wrong with Sigma reverse engineering the lens protocols per se. In fact, this never has been any secret, as Sigma openly admits the fact that they do it in interviews. They are proud of it, I have been told by Sigma representatives.

The problem is, that they have no reason to be proud of it, not because of the reverse engineering, but because of the lousy job they do. Instead of trying to really understand how it works, they just seem to fiddle around a bit until it works – for a given lens on a current body. That’s why there’s a high risk, that they will continue to run into problems if only minor things get changed or enhanced by Sony.
Nevertheless, Sigma has proven to be totally immune to any insights on how to get it gone right in the past 25 years or so. The learning curve is plain out flat. It’s been about a dozen times now that a large bunch of their lenses stopped working (in various ways) with newer camera models released. The lenses listed at Sigmas web site reflect only a fraction of it because they don’t list lenses affected, but lenses for which they (still) have a fix. If the fix becomes unavailable after a number of years, the lens gets removed from the list, so you will have to check old mirror announcements in newsgroups and forums, not Sigma’s web site to get a better picture of the situation. I guess, all in all, more than 95% of all Sigma A-mount lenses in the wild are affected by some kind of compatibility problems, while they offer fixes for some 1 or 2 percent of them. Interestingly, some people continue to praise Sigma’s “free service”.
Unfortunately, you can be almost sure that a Sigma lens will stop working after a number of years. And if the lens is then no longer a current model and out of warranty, Sigma will not provide any chip upgrades (as they must for legal reasons for as long as the lens is under warranty) – so, the “free service” argument is a very weak one IMHO (and it is not free anyway, unless Sigma would pay compensation for the shipping costs and time as well). Also, according to them, Sigma does no longer test their fixes on older camera models, in particular not on film bodies, which are an integral part of the A-mount system and still make up the largest part of it – and will continue to do so for years.

For example, when the new HSM lenses were introduced by Sigma, I was trying to use them on my Dynax 9 SSM in 2008 and found that the AF was not working okay. It would first focus to infinity before running back into the point of sharp focus all the time – even if the subject has been in perfect focus already. I reported this problem to Sigma multiple times in mail and in person but they insisted my camera must be in error because the AF would work fine on the DSLR-A200 (a camera they used for testing). In fact, the tested lens sample did work fine on the DSLR-A200, but not on the Dynax 9 SSM according to my own testing. I told them, my camera would be fine (all original Minolta/Sony SSM lenses work without problems) and that whilst there were no apparent problems with current bodies, the fact that it did not work on the Dynax 9 SSM clearly showed that their implementation of the lens protocol or data must be incomplete and faulty, and that this may not only affect the Dynax 9 SSM but also future camera bodies. So, even if they’d not care about it for the Dynax 9 SSM specifically, they still should in order to ensure compatibility with future bodies. Nevertheless, they continued to deny the fact and refused to list this compatibility problem (which was meanwhile also approved by others) on their web site.
In 2010 I retested some of their lenses (including newer samples of those lenses used in the original test) and found them to be working perfectly fine on the Dynax 9 SSM now. Even their OS image stabilizer worked fine on film bodies (except for on the first generation of camera bodies, which does not provide the necessary power supply). Clearly, they had silently changed the implementation, so that there are now samples of the same lens model in the wild not working on older cameras and samples working fine.
Reporting the good news to Sigma, they now told me, no, this cannot be true, according to their data, their lenses were incompatible with SSM-capable film bodies, but that they would not provide a fix even for their current lenses, because they do not officially support film bodies any more. Also, they could not provide serial numbers in help distingiush bad from good samples, because, according to them, they never implemented a fix. #-|

Then we saw the compatibility issues with the SLT-A33/SLT-A55 in 2010 – for which Sigma does not offer fixes for the majority of older lenses. Ever guessed why the did not publish a list of lenses affected? Because they would have to list several hundred models then, and in some cases they would even have to distinguish between different versions of the same model, some working and some not.
Nevertheless, the nice side effect is, lenses marked “a33/a55 compatible” are likely to be compatible with SSM-capable film bodies as well.

Now, we see problems with the SLT-A65/SLT-A77. Again, the list they give is incomplete and only shows lenses, which are still current. Owners of older lenses are out of luck.

The bottom line is, you buy Sigma lenses for a specific camera body. Don’t expect it to work on older bodies and don’t expect it to work on future bodies – and don’t expect the problems to be fixed by Sigma, neither for free or for cash. Finally, test the lens sample and camera combination you want to use yourself, because the information given out by Sigma has proven to be incomplete, misleading and often plainout wrong, unfortunately. If this is your idea of a “system”, you will be fine investing money in Sigma lenses, however, my idea of a system is quite different. I expect a 20 to 25 year old lens to work on a current body just as fine as a new one, and I also expect a new lens to work on a 25 year old body (perhaps not support all functions, but at least work).