>>> 1) shorter registration distance means no retrofocus, which means better >>> image quality at wider angle of view. >> >> That does not play out in reality, or is at least inaccurately stated. Wides >> are a challenge on digital sensors, non-retrofocus wides in particular. The >> best performing wides on any digital sensor are retrofocus or at least >> semi-telecentric designs, to date. It's a matter of getting an orthogonal >> light path to the sensor all the way to the edges rather than retrofocus >> design really. A short mount registration means you have more room to add >> corrective optics behind the primary element groups and less worries about >> mechanical interference. > > Interesting. I had heard that the reason that all of my wide angles needed > the retrofocus elements was because of the registration distance, that you > can't put the lens any closer than the mirrorbox.
This is true for any SLR lens: the moving mirror poses a limit on how close you can site the lens to the imaging plane. Retrofocus lenses permit short focal lengths for SLR mount registrations because the primary nodal point is projected behind the physical lens optics, allowing them to clear the mirror. For a "mirrorless" camera, the moving mirror is not the issue, but sensor sensitivity to the incident angle of light hitting the sensor at the edges and corners becomes the larger factor in obtaining a good mating of lens and sensor stack. The ray trace through a retrofocus design lens helps alleviate this problem, even without additional corrective elements in the lens design. > I had also heard that the latest generation sensors weren't nearly so > sensitive to problems with light not coming in on a perpendicular path. The more recent sensors have reduced sensitivity to off-orthogonal ray traces, yes, due to using shallower photosite wells and more sophisticated optics. But reduced by how much is the question. The ideal is still orthogonal, perhaps some sensors can tolerate 5° off axis to the photosite now where in the past problems would be evidenced at 2° or greater. >>> 3) No mirror bouncing around means no mirror bouncing around. >> >> Shutter vibration is still an issue, note all the hysteria about "shutter >> shock" lately. A fully electronic shutter will conquer that issues, but non >> work 100% to date. > > Is shutter vibration alone less of an issue than shutter vibration plus > mirror bounce? If not, should I stop bothering with the mirror lockup on my > DSLR? Not sure I understand your question. Mirrorless camera shutters have more actions per exposure cycle (assuming no electronic first curtain). The first curtain must close, the sensor is re-set, then the standard sequence of curtain movements proceed, then the data is read from the sensor, then the sensor is reset, after which the first curtain re-opens. The critical first curtain close/standard cycle opening juncture in a mirrorless camera can generate as much vibration as an SLR in some circumstances—it all depends on the specifics of a particular design's timings, masses, total mass of the camera, etc. SLR mirror/shutter mechanisms are very mature designs and generally operate very smoothly indeed. I wouldn't change what you do with a DSLR for vibration control on the basis of what a mirrorless camera does. >>> 4) Live view works much better for low light manual focus etc. What you see >>> is what you get. With a mirrorless you don't get the huge shutter lag >>> while everything bounces around between live view and actually taking the >>> photo. >> >> There is certainly less mechanical banging about going on. But it's amusing >> to hear you talk about 'much better for low light manual focus' when so many >> of the SLR diehards claim exactly the opposite. :-) > > I'm just speaking from personal experience. In bright light, I'm very happy > with optical and a split prism. It depends on the quality of the optics and focusing screen, but I'm quite happy with good SLR reflex viewfinder systems too. In dim light, I find EVF type viewfinders are easier to work with, like you do. > >>> There are certain ergonomic advantages to the pure optical path when >>> looking through the viewfinder. >> >> Ergonomic advantages? Can't imagine what you mean by that. And what is a >> "pure" optical path? > > I mean just optics, not going through a sensor and electronic viewfinder. > The biggest one is no lag between real life and what you see. Also, some > EVFs don't have the same level of image resolution in bright light that a > good optical viewfinder has. The 'speed of light' responsiveness of a reflex optical viewfinder is its biggest asset. And of course when it comes to resolution and dynamic range I'm referring to the latest, current crop of EVFs as seen in the Olympus E-M1, Sony A7/r/s, and Fuji XT-1. IMO, the Olympus is the best of them ... They all use the same or very similar panels, but I think the Olympus optics and their brightness adaptation is the best of breed at present. The Sony A7 become difficult to view with in bright light sometimes, as does the Fuji XT-1, but the Olympus E-M1 doesn't. It feels for all the world just as responsive and snappy to view through in bright light as an optical finder, and degrades the least in low light too. > The only engineering issue I can see with EVF and image quality is that if > the sensor is always active, it'll get warmer and therefore noisier. Ultimately, it's possible but a minor issue in practical implementation for still cameras used in most circumstances. Modern sensors in continuous capture mode run pretty low power and thus pretty cool. And auto-switching of the EVF also reduces power consumption. This is part of the debate between in-body IS vs in-lens IS as well. Stands to reason that if you want to keep thermals constant, anchoring the sensor to a lot of mass to absorb and radiate heat makes sense, but that's exactly what doesn't work if you want good IBIS. Olympus in the E-M1 is focusing on still photography primarily with excellent multi-axis IBIS, so it has some heat issues with long video captures. The Sony and Panasonic mirrorless bodies focus a lot more energy on motion capture, note that neither provide IBIS as that would compromise their ability to provide a massive heat sink for thermal stability. Both the A7/r/s and Panasonic GH4 are darn sophisticated movie machines for consumer cameras. They're not optimized up to Black Magic or Red standards, of course, but they do a pretty darn terrific job. Godfrey -- PDML Pentax-Discuss Mail List [email protected] http://pdml.net/mailman/listinfo/pdml_pdml.net to UNSUBSCRIBE from the PDML, please visit the link directly above and follow the directions.
