Canon EF 50mm f/1.8 STM

<p>Haven't seen this discussed yet - Canon announced the <a href="http://www.usa.canon.com/cusa/consumer/products/cameras/ef_lens_lineup/ef_50mm_f_1_8_stm">next iteration</a> of their 50mm f/1.8 lens, this time with stepper focusing motor. It's great that Canon finally updated the 50mm lens, keeping it small, light and cheap (same price as the old 50mm 1.8). The optical design seems more or less unchanged - even though the press release claims improvements, the MTF curves posted by Canon seem awfully similar to the old 50 1.8. At least the aperture now has 7 blades, no more pentagonal OOF highlights.</p>

<p>They look darn near identical! (the MTF chart) </p>

<p>That said, I think I'll probably snatch one up to see how it does...</p>

<p>I have to say I am disappointed it doesn't have IS as I was expecting this in line with their other new primes (24/28/35), but perhaps IS will be reserved for a new f1.4 version. Still I don't really get this addition. It looks spiffier than the current f1.8, which may be more what it is all about at this price point.</p>

<p>I suspect that this unit's primary point is to reduce manufacturing costs overall, as it will allow obsolete processes and parts to be completely phased out. Simply by replacing this unit, I suspect they can reduce overhead and production inefficiencies. The 50/1.8 sells so many units that this optimization could result in a significant improvement to Canon's costs.</p>

<p>It appears they already quietly discontinued the 50/2.5 (though they are obviously still widely available new), and I'd suspect that any others running the old DC motors are at least up for consideration for the chopping black.</p>

 

<p>It's probably a combination of cost cutting by consolidation of manufacturing with the new STM motor and trying to appeal to the growing video crowd with the STM motor designation. </p>

<p>As for IS, you could have it but it would likely at least double the cost of the lens. One selling point for the 50/1.8 is that it's cheap. I'm actually quite surprised that the new STM isn't more expensive than the old DC motor lens, so keeping the cost down must have been a priority.</p>

<p>I suspect the optics are identical. The MTF plots are so similar that they must be. Canon do claim new coatings though.</p>

<blockquote>

<p> Canon do claim new coatings</p>

</blockquote>

<p>They always do when there is nothing much else to say. I am sure they are new, but I doubt it's really any better - I mean what really has happened in coating technology that will change the way a 50/1.8 will perform flare wise? They haven't added nanocoatings to improve ease of cleaning as are found in the new Ls. Obviously this lens has to be kept cheap. Still I was hoping for an IS version at 2-3X the price.</p>

<p>Does this have the manual override of auto focus like the 40mm pancake. I bought the 40mm pancake for that and the closer focusing. If this was out when I bought it, I probably would have bought this lens.</p>

<blockquote>

<p>Does this have the manual override of auto focus like the 40mm pancake. </p>

</blockquote>

<p>If it does, that alone makes it a much more compelling optic than the 50/1.8 II. I suspect it's AF system is going to be very nearly indistinguishable from the 40/2.8. Originally, the lack of FTM (in the 50/1.8 1/2) focus was why I bought a 50/1.4, and, despite the poor IQ WO - and repeated failures of it's AF - of the 50/1.4 never looked back.</p>

 

<blockquote>

<p>Still I was hoping for an IS version at 2-3X the price.</p>

</blockquote>

<p>I expect that the replacement for the 50/2.5 Macro will have IS - probably a proper macro w/ hybrid IS - though I doubt it'll only be 2-3x the price of the 50/1.8 ;)</p>

 

<blockquote>

<p>I mean what really has happened in coating technology that will change the way a 50/1.8 will perform flare wise?</p>

</blockquote>

<p>When the 50/1.8 II came out (in 1991) digital imaging was not even on the radar. Changing the type of coating to prevent flaring due to internal reflection off the CMOS (much more pronounced than most film reflection) makes good sense though, and is easy and cheap.</p>

 

<p>"Does this have the manual override of auto focus like the 40mm pancake. I bought the 40mm pancake for that and the closer focusing"<br>

According to B&H's website, this lens does have manual focus override and focuses closer at 14" to the older 17.8" .</p>

 

<blockquote>

<p>Changing the type of coating to prevent flaring due to internal reflection off the CMOS (much more pronounced than most film reflection) makes good sense though, and is easy and cheap.</p>

</blockquote>

<p>Yes but how can lens coating affect that? By the time it gets reflected off a sensor the light has already passed through the lens. They need to change the coating on the sensor to change this. I am just saying I doubt the flare will be any different, it's a fall back statement. The Canon FD 50/1.8s are pretty good on the mirrorless cameras and they have "very old" coatings. Yes, you may be right about a new macro 50mm. It would be nice if they made it f2 like the Zeiss Makro Planar.</p>

<p>If the lens has the MF-by-wire arrangement like the 40/2.8 then it will be not the best for MF even though it will have the manual override.</p>

<p>Lower reflectivity coatings mean that any light bouncing back off the sensor will pass back out through the lens instead of being reflected back to the sensor from the lens elements. That's the theory.</p>

<p>I suspect any difference will be negligible. I've been using a 50/1.8 MkI for the last 15 years and I've never been bothered by flare or lack of contrast. I've never had any problems keeping it clean either so I'm not sure that "nanocoating" or "fluoride" coatings on the front element are all that useful.</p>

For me, the pentagonal OOF highlights was probably the worst performance aspect of the original lens, but I do agree the main reason for the upgrade is the AF video performance of the STM motor. Also so may casual users seem to prefer to use liveview for everything, for which will work much better on the new lens.

<blockquote>

<p>Lower reflectivity coatings mean that any light bouncing back off the sensor will pass back out through the lens instead of being reflected back to the sensor from the lens elements. That's the theory.</p>

</blockquote>

<p> <br>

That's not quite correct in this case - although the effect is largely the same. In this case, 'reflected' visible light (off the sensor's filter stack, which is, itself, coated or else it would be <em>highly</em> reflective) is minimized by the effective cancellation of reflected wavelengths through superposition. Since the coating and air have different indexes of refraction, some of the light is reflected off the coating. The remainder passes through the coating, and some reflects off of the glass ( the rest passes back out the lens). The thickness of the coating through which the wave passes (both ways) imparts a 'delay' or phase shift on the reflected light. As the second portion (that which was not reflected off the coating, but was reflected off the glass) exits the coating (heading back toward the sensor), it's phase is shifted by 180 degrees, so it, and the first portion (reflected off the coating) have directly opposing sine waves. by the law of superposition, those cancel each other out (ie where the peak magnitude is on one, the opposite is on the other ((+1 )+( -1 )= 0 ). <br>

<br>

Of course this is the simplest example, and only works completely at a single wavelength (working with diminished effectiveness the further from that wavelength the reflected light actually is. Multi coatings are more effective, working across a broader range of wavelengths. <br>

<br>

But of course plain glass does not have any of this benefit. The only reason most pictures taken with film era lenses are even usable is the multicoatings on the sensor's filter stack. multi coating the back element of the lens only improves that (though obviously only marginally except in the most challenging lighting Yay for Science!<br>

</p>

<p>Do'oh! Let me try again...</p><div></div>

I suspect Canon would have to release 3 or 4 different versions of a 50mm, with and without IS etc, and then still people would be bitching. This lens look like a decent replacement of the so-called nifty-fifty and I hope that designation goes away with this one. Canon did a good job with this lens and kept the price reasonable. Personally, I have no interest in it with either my crop or full frame body but do think it is a nice addition to the Canon lineup.

<p>I know the wave optics description of the mechanism of AR coatings.</p>

<p>Using the wave optics description, AR coatings don't just reduce reflection by destructive interference, they increase transmission via constructive interference . The light which would have been reflected is transmitted, not destroyed. Another way to look at is is that the light must go somewhere, and if it doesn't get reflected, it must get transmitted. The Quantum Mechanical description involves superimposed probability waves which "interfere" with each other in regard to reflection, lowering the probability of a photon being reflected, hence increasing the probability it will be transmitted. It's actually a rather complex subject if you get down and dirty with the details. However the only thing that matters for photography is that reflection in decreased and transmission is increased and that's a fact confirmed by experiment.</p>

<p>So as I said before <em>"</em><em>Lower reflectivity coatings mean that any light bouncing back off the sensor will pass back out through the lens instead of being reflected back to the sensor from the lens elements</em>" and that's exactly what happens. <br /> <br /><br /></p>

<p> </p>

<p>Some actually like the harsh pentagonal bokeh of the older 50/1.8 lenses...</p>

<p>I too would like to see Canon release a real competitor for the Zeiss 50/2 Makro-Planar, which is a great lens in every way except for lack of autofocus. I remember seeing some news of recent Canon patents filed which showed a 50/2.8 macro lens design.</p>

<blockquote>

<p>I have no interest in it with either my crop or full frame body</p>

</blockquote>

<p>Kerry, your perception of "bitching" may be related to the fact that you're not interested in the lens. Clearly the new lens is not really made for those of us who already have a 50mm Canon lens as, unless you are mad about STM, it is the same lens as the previous version to all intents and purposes. I guess we will have to wait for a new 50/1.4 to get to see similar improvements to what they have already made in the last few years to the 24, 28 and 35 primes.</p>

<blockquote>

<p>I know the wave optics description of the mechanism of AR coatings.<br>

Using the wave optics description, AR coatings don't just reduce reflection by destructive interference, they increase transmission via constructive interference</p>

</blockquote>

<p>Sorry Bob, I didn't mean to imply that you didn't! It's just that your reply didn't really answer Robin's question, so I thought I should add a bit of wave/ray optics in there to flesh out the 'how' of the question. Also, ray/wave optics (alone) doesn't fully explain what happens.</p>

<p>However, that doesn't make the assumption made in your reply technically correct. The short answer is that in a simple quarter (L/4) wave AR coating, <em><strong>both</strong></em> the reflected light, <strong><em>and</em></strong> the transmitted light are <em>reduced</em> via destructive interference. This is clear from a wave optics visual. Certainly the AR coating material is inherently (slightly) less reflective than the glass, but only a bit, as, were it significantly less, it wouldn't be as effective (since one component would far outweigh the other, defeating the purpose) additionally doing such could significantly impact image sharpness! The effectiveness comes not by changing the transmittance, but changing the effective reflectivity by imparting a phase change on alternating reflected EM waves. (as outlined in the above graphic). In the case of this lens, a few simple (cheap) AR coatings are probably all that are done, as while a layer or two isn't usually that expensive, they add up fast! Especially when you need widely varying index of refractions! While there are (theoretically) coatings which (can) do what you've indicated, they are not deposited on a 50/1.8 nor any other camera lens because that would only make things worse from a reflected light perspective (the more light bouncing around the inside of a lens, the worse the image will look!).</p>

<p>But of course this leads to the question where does the energy go? A <em>simple</em> answer which is correct (from wave optics, EM region, <em>and</em> QM explanation) would be that the (L/4) AR coating redirects reflected light in such a way that the reflected light is absorbed by the material instead of being reflected. This is done through the phase change (represented in wave optics explanation) which causes (through superposition) a (significant) reduction in the probability of emission at each point. </p>

<p>Well that's my understanding at any rate ;) - and works with every bit of physics taught to me in my Optics, EM, <em>and</em> QM classes.</p>

<p>There are (of course) a variety of coating types and variations, which can do anything from nearly eliminating reflections, to reflecting 99.999% of incident light (these typically work through constructive interference)... But notice that nothing you can do with a coating yields more energy than was input (which, if you ignore the material absorption is exactly what happens in some coatings via constructive interference)... and they get mighty pricey mighty fast! </p>

 

<p><em>But of course this leads to the question where does the energy go? A simple answer which is correct (from wave optics, EM region, and QM explanation) would be that the (L/4) AR coating redirects reflected light in such a way that the<strong> reflected light is absorbed by the material instead of being reflected</strong>. This is done through the phase change (represented in wave optics explanation) which causes (through superposition) a (significant) reduction in the probability of emission at each point.</em><br>

<em> </em><br>

But that's <em><strong>not</strong></em> what happens. Anti-reflection coatings <strong>INCREASE</strong> transmission. The light that isn't reflected isn't absorbed/dissipated as heat or converted into some other form of EM radiation. There's no power dissipated in the interfering reflected waves. This is very easily verified by experiment. See for example http://www.astrosurf.com/luxorion/Physique/transmittance-coating.jpg</p>

<p>You could equally well call the coatings pro-transmission as call them anti-reflection. Both are true. I guess they are called anti-reflection because there's a bigger percentage change in reflection (say, 6% to 3%, a 50% reduction) than in transmission (94% to 97%, a 3% increase), so the effect on reflection is proportionally greater than the increase in transmission.</p>

<p>For a single element the effect on transmission is small, but consider a complex lens with 15 elements and 30 air/glass interfaces. Uncoated you would get maybe 50% transmission. Coated you can get 95% or higher transmsission.</p>

<p>You can't analyze the situation using simple wave optics and one reflection. That provides a reasonable looking explanation for suppression of the reflected light but it's incomplete. In fact there are an infinite number of reflection and transmissions as the light bounces back and forth within the coating layer(s). The reflected rays are 180 degrees out of phase and destructively interfere. The transmitted rays are 90 degrees out of phase and constructively interfere. You have to sum over infinity. Alternatively you can fully solve the wave equation or you can consider the quantum mechanical analysis in which you are dealing with probability waves. Lower probability of a photon being reflected means there's a higher probability of it being transmitted (not absorbed or converted into something else). </p>