Canon Stupidly perhaps, I'm going to try the EOS-M

wt21

Hall of Famer
Just for something new, I'm going to try the EOS-M. The kits are available again in the US Amazon page. I've seen some very nice shots coming from the lenses on the M.

For $299 it's about the cheapest fast 35mm equivalent out there. It's less than the X100, less than the Olympus 17mm 1.8 lens (by itself!) Less even than a EPM2 + 20mm 1.7.
 
You could do worse. Might not be any "future" in it depending on what Canon does. But at that price it's darn near "disposable" by todays standards -- and you do get a heck of a build quality at that price.

I know it would frustrate the heck out of me though, so I'm not getting one... unless they drop to $199 then I might be tempted, lol!
 
I had pre-ordered one before it was released. The original firmware was awful. The camera could hardly focus. I know they have improved the firmware somewhat but overall it was the worst camera I had ever touched my entire life and returned it within a few days. It felt like holding a bar of soap. A better buy would be a GX1 that Adorama has had for $199 recently.
 
If nothing else, you get a very nice 22mm f2. Canon can't afford to not participate in the mirrorless CSC space, so the lens will likely find use in a future body.
 
The EOS M can be worth it depending on your intent. I don't think it competes well against other CSCs. It still isn't that cheap since you can get M43 and NEX cameras right around $300 as well. But if you think of the EOS M with the 22mm as a fixed lens APS P&S camera, then that's another thing entirely. At $300 it competes very well against the X100, Coolpix A and the Ricoh GR.

The focusing is now Fuji X speed. That is when you can get it to focus at all. Here's a funny Kai video demonstrating the new firmware.

[video=youtube;wwZDStFUixQ]
 
Hahah... poor Bill, yet another system to tear you in another direction :)

Hope you like it... the 22mm looks like a pretty good lens and I've seen some nice crisp shots from the EOS-M. They may have underwhelmed in the AF department but as I've said with the Fuji line, this just shows how hard of a problem to solve this really is and how far m4/3 has come to be as fast as it is.
 
They may have underwhelmed in the AF department but as I've said with the Fuji line, this just shows how hard of a problem to solve this really is and how far m4/3 has come to be as fast as it is.

Is it hard though? Afterall every P&S camera does it. Seems that the problem is the exception and not the rule.
 
Is it hard though? Afterall every P&S camera does it. Seems that the problem is the exception and not the rule.

Yes and no - smaller sensor seems to make a difference as well, possibly not least of which because with increased DoF focus doesn't need to be as exact. Then add in complexities of interchangeable lenses and other design elements.

Earlier m4/3 cameras & lenses, Sony NEX, Fuji, the RX1 and various other larger sensor compact cameras all struggle with AF compared to modern m4/3 that's had several generations of iteration. Heck, focusing my Nikon DSLR in live view mode is painful. Given all that I have to believe this is a harder problem than people give it credit for.
 
Hahah... poor Bill, yet another system to tear you in another direction :)

Hope you like it... the 22mm looks like a pretty good lens and I've seen some nice crisp shots from the EOS-M. They may have underwhelmed in the AF department but as I've said with the Fuji line, this just shows how hard of a problem to solve this really is and how far m4/3 has come to be as fast as it is.

As lenshoarder notes, think of it as the cheapest fast 35mm going. X100 is $600 used (motto "now with more SAB!"). EPM2+17/1.8 is about $700.

I'm getting the M from Amazon, so it it's truly awful, it'll go back, but I am pretty impressed with what I've seen from this lens, and also with the Canon 70D PDAF-on-a-chip. If that's where EOS-M is headed, then that's where I'll want to be.

I am betting Canon might make a very nice M body with their next release, with just a couple of lenses. Some of the walk around variety. Unlike Oly or Panasonic, they don't have to make a "complete" kit, and that's OK for my uses. I just need 3 lenses: an ultra-wide zoom (which they have) a normal (I'd prefer 50mm eq, though) and a 100mm eq. macro (which they don't have yet).
 
Yes and no - smaller sensor seems to make a difference as well, possibly not least of which because with increased DoF focus doesn't need to be as exact. Then add in complexities of interchangeable lenses and other design elements.

Earlier m4/3 cameras & lenses, Sony NEX, Fuji, the RX1 and various other larger sensor compact cameras all struggle with AF compared to modern m4/3 that's had several generations of iteration. Heck, focusing my Nikon DSLR in live view mode is painful. Given all that I have to believe this is a harder problem than people give it credit for.

Bigger glass takes more power to move in/out, and you have to do that quickly with CDAF. If PDAF-on-a-chip works, though, then some of this could be negated (at least in good light).
 
Yes and no - smaller sensor seems to make a difference as well, possibly not least of which because with increased DoF focus doesn't need to be as exact. Then add in complexities of interchangeable lenses and other design elements.

Earlier m4/3 cameras & lenses, Sony NEX, Fuji, the RX1 and various other larger sensor compact cameras all struggle with AF compared to modern m4/3 that's had several generations of iteration. Heck, focusing my Nikon DSLR in live view mode is painful. Given all that I have to believe this is a harder problem than people give it credit for.

I've had a NEX 3 since they came out. It's never had a problem with AF. While the current NEX models are a little faster, I wouldn't call the difference a struggle. Also, I don't think that sensor size has much to do with it since larger sensor cameras tend to focus faster than their smaller sensor cousins. A relatively large sensor RX100 focuses faster than say a much smaller sensor Canon 100hs. In fact, while the EOS M was/is slow to AF for a CSC. It's right in line with the speed of many P&S cameras.

DOF doesn't really come into play since CAF is relative, not absolute. So whether there is a lot of DOF or none at all, it's all the same. CAF is really simple. The focus is deemed to be where there is the most contrast. So it moves the lens through a range and iterates down to where the contrast is the highest. Offhand, I can think of two things that limit it. Processing power of the camera and the speed at which it can move the lens. It's really a simple operation so processing power isn't much of a concern. In the case of the EOS M, it shares the same chipset as the DSLRs so that can't be the limiter. That leaves the control of the lens as the problem. What is surprising is that the higher mass zoom lens focuses faster than the lighter 22mm. Probably because the pancake has a less powerful power/weight motor.
 
I think that's a big part of it -- how big is the glass relative to how big is the engine. Kind of like horsepower and cars. You can have a big engine, but how big's the rig it's pulling? I am sure they work together, along with algorithms and processor power.
 
Bigger glass takes more power to move in/out, and you have to do that quickly with CDAF. If PDAF-on-a-chip works, though, then some of this could be negated (at least in good light).

If you look at Nikon Series 1 - PDAF on a chip does work, and rather brilliantly. No, not another Canon-Nikon thread. :) I suspect it has a lot to do with processing power - Canon may have dipped their toes into the market to see what the possibilities are - and in so doing shaved some corners on their normally very leading engineering to minimize investment layout.
 
Sure, lens design is a big factor; my point was just that if it were that easy of a nut to crack, everyone would be doing it right out of the gate on their first try. Nobody *wants* to produce and market the camera with the slowest AF, right? It's in their best interest to produce sharp lenses, good performing sensors, and fast focus.

The firmware improvements to AF speed mean that in addition to lens design, it's also the algorithms in use that are a big factor. Yes, it's "just" contrast detection, but how you implement that, prioritize, and classify areas is all important. Coming from a software background, it's easy to understand how it takes time to build up an effective algorithm with incremental improvements, bugfixes, and performance enhancements over time.
 
Back
Top