Class A: Some good points made, but I find the article is skewed towards mirrorless by omitting a number of disadvantages:
1. The camera body of a MILC body is typically smaller but adding a lens with a speed comparable to that used on a DSLR, the overall size advantage is not big. As a matter of fact, for fast / long lenses the small MILC body is an inadequate match to the lens' size and weight.
The argument that most MILC lenses are optimised for their format has limited weight, as lens size / weight is typically dominated by the lens speed, i.e., more closely related to the size of the front element rather than the size of the format projected.
2. A DSLR only turns on the sensor when needed. A MILC, on the other hand, needs to power it for composition, metering, and focusing as well which should result in a higher average sensor temperature, adding to noise in images.
P.S.: I meant to change the phrase "behaves like on full frame" to "equivalent to a ... lens on full frame" but could not edit the post anymore.
The drastically lower amount of photons captured allows the Olympus lens to have a smaller front element, etc.
Many people hold the view that "f/2.8 is f/2.8" in terms of exposure, but don't realise that "exposure" is to image quality what pixel noise is to image noise. The latter matters, the former does not.
More pixels compensate for increased pixel noise and a larger sensor compensates for less exposure.
Conversely, if you keep the pixel noise or exposure constant and increase the number of pixels and sensor size respectively, you get better image quality.
Hence the Olympus lens really *is* a lens that deserves an "f/5.6" label when using FF-currency.
In other words, it is not the case that the Olympus lens only "behaves" like an f/5.6 lens. It is always the same lens and using "exposure" to judge a lens is just not helpful. It matters how many photons go "in", not how many photons per square mm go "out", because the total area matters.
Richard: Of course the Olympus lens is a "40-150/2.8". I did not want to accuse Olympus of false advertising when I wrote that "-- in full-frame terms -- " it is only a f/5.6 lens.
Writing "in full-frame terms it is" and "behaves like on full frame" is equivalent to me.
The point is that in terms of total light collected the f/2.8 Olympus lens collects 2 stops less light than an f/2.8 FF lens. When comparing image quality, one must compare the total amount of light collected, since that determines image noise levels, as opposed to the irrelevant (since format independent) "exposure".
The Olympus lens is labelled "f/2.8" but due to the difference in sensor size that is only an "f/5.6" in full-frame terms. The Canon lens could be a lot smaller and lighter if it were an "f/5.6" lens only.
Class A: Nice photography! Shame about the uninspiring scenery, though. Also, try to move away from ACR as the standard converter. I believe ACR camera profile quality depends on the brand supported and using the embedded DNG profiles isn't ideal either, according to my experience.
Try as I might, I could never get nice colours from my K-5 II, not even by using Adobe's profile editor to tweak profiles. As soon as I switched to Capture One 9 all was good. Hope Phase One will add K-1 support soon, as their dedicated profiles are better than what you get from converting generic DNG files, AFAIC.
@Yanko Kitanov: I very much appreciate that CO9 supports DNG files without requiring a dedicated profile, but I'd say that K-1 images will most likely look even better once Phase One managed to provided a dedicated profile for the K-1.
Carey, thanks for your response.
I understand the idea behind a baseline, a level-playing field, if you will.
Unfortunately, ACR, cannot play that role as it treats RAW files from different manufacturers and camera models differently. A different black level is applied, a different tone curve is used, a different colour mapping performed, etc.
You'd have to use something like dcraw, or perhaps RAWTherapee, to get a true level playing field. With ACR you'll always also a difference in level of support Adobe provides to various manufacturers.
I understand that ACR is often referred to as an industry standard and that results are representative due to the widespread use of Lightroom and Photoshop. However, at least in my experience, ACR is not delivering the best results for some cameras, even when you try hard tweaking. It is, therefore, another valid viewpoint to look at a camera's output with a converter that gets the most out of it.
Thanks again for your response, it is much appreciated.
Nice photography! Shame about the uninspiring scenery, though. Also, try to move away from ACR as the standard converter. I believe ACR camera profile quality depends on the brand supported and using the embedded DNG profiles isn't ideal either, according to my experience.
Jacques Cornell: If they're so serious, how come they didn't put a pass-through hotshoe on top of the transmitter? Honestly, I can't fathom how a radio trigger maker can do this. Have they never heard of event photography? Y'know, with on- and off-camera flashes at the same time? What they've made is a trigger for studio use only, and I'll bet event shooters outnumber studio shooters by a large margin.Ugh. The stupidity, it burns.
Have a look at the Cactus V6. Uses a dial as well. Only manual power control, but I personally prefer that anyhow.
aris14: It would be interesting, not for photogs but for the history of management, to know some day under what circumstances and who were the guys who took the decision to stop the MZ-D launch in those days. Something tells me that it would be a nasty story...
The Contax N Digital used the same sensor and had a very bad reception. It was withdrawn after one year only and soon after the brand Contax died.
It seems that not pulling through with the MZ-D proved to be a very good decision.
Subscription model = not interested
Class A: Almost everyone unfamiliar with the Pentax lens line up severely underestimates the amount of available FF lenses.
Of course, first there are the new zooms, covering 15mm-450mm. But the list continues:
There are not only some "D-FA" lenses (like the 100mm macro) that carry a "FF compatible" designation, but also (around seven) "DA" lenses that have been sold as APS-C lenses but produce an FF image circle. The DA* 55/1.4 is just one example.
Then there are of course the three Ltd lenses (31/1.8, 43/1.9, 77/1.8) which, despite being initially developed for film, easily outresolve 36MP. These are available new, and if one extends one's search to the used market then there are a multitude of gems available, all AF capable, supporting metering and 36MP compatible.
That's not even counting third-party glass, e.g., from Sigma.
It is just not true that the lens selection is meagre. That argument only applies if one restricts oneself to silent focus lenses with a built-in motor.
I just did a count of FF lenses for the K-1. Including third-party models from Sigma, Tamron, and Samyang there are currently
12 zoom lenses, and33 primes available.
These can all be purchased brand new.
Only the Samyang lenses are MF only, the rest are all AF lenses and some of them support silent focusing.
Almost everyone unfamiliar with the Pentax lens line up severely underestimates the amount of available FF lenses.
Couldn't agree more about the difference between 'classic' and 'retro'.
Did you see my earlier post about the "four-way controller" actually being an "eight-way controller", etc.?
Class A: Dear Allison Johnson,
you wrote about the Pixel Shift Resolution system: "This does not increase linear resolution...".
However, this is incorrect, if you consider colour resolution.
All Bayer-filter sensors only have half the resolution in the red and blue colour channels. Some clever demosaicking can compensate for the lack in resolution compared to the green channel, but there are obviously limits.
The Pixel Shift system achieves optimal resolution for all colours (akin to a Foveon sensor in terms of the result).
Hence, it does increase (colour) resolution, which translates into a plain resolution increase, as long as one is shooting colourful scenes.
I understand what Allison intended to say, i.e., that there is no increase in number of pixels.
However, she needs to state it properly. The way she expressed it, she is making a statement about image details captured. It is imprecise to treat "linear resolution" synonymous to "pixel count in one dimension".
Since PixelShift increases chroma resolution, the actual linear image resolution is increased.
BTW, Pentax said they tried the pixel increasing sub-pixel supersampling as well but didn't find it effective in actually delivering more detail. Just passing on the message.
bizao: Anybody Konws what is the resolution when capture with pixel shiftI didn't find anywhere
@Zdman: The line about "linear resolution" is wrong.
What it meant to say is that there is no increase in number of pixels, but it instead makes a statement about details captured.
Since PixelShift increases chroma resolution, the actual linear resolution is increased for colourful subjects.
Allison's statement is incorrect and another DPR article on the K-1 confirms it by stating that "chroma resolution" is increased.
The term "linear resolution" does not imply that only "luminance resolution" is considered. The term "linear" simply means "along one dimension", i.e., if the sensor size increases by a factor of four by doubling both its sides then the linear resolution only increases by a factor of two (the square root of four).
Richard, thanks a lot for this excellent first impressions review!
You mention the "four-way controller" and I wonder whether you have noticed that it is, arguably, an "eight-way controller". The buttons are shaped in such a manner to support pressing two buttons at the same time for diagonal movement of the AF area selection.
Regarding the screen tilting mechanism, Pentax claims that it keeps the centre of the screen in the optical axis. Not sure whether that converts into an actual benefit in practice, but that seems to be the idea behind it.
I reckon you haven't seen the D-BG6 grip yet. I cannot understand how Pentax (again) managed to permutate the button layout compared to the camera. It was OK for the K-5 and already wrong for the K-3. This is huge blow for me as I like grips for portrait shooting.
Not sure about the significance of a "rec" button, if there is a video switch and using the shutter release button then even enables (radio triggered) remote release of video.
Dear Allison Johnson,
The approach of allowing variability in the hopes that many customers won't care certainly has merit, as long as manufacturers appropriately respond to those customers who do care.
From first hand experience I can attest that Sigma is keen to satisfy even exacting demands. Sure, I'd prefer every first copy I acquire to be perfect, but even more surely, I wouldn't want to pay a lot more money for this convenience.
On the other hand, a salesperson recently tried to explain to me that QC issues with lenses are no longer a problem in this day and age, trying to justify his policy that I wouldn't be able to be refunded for a new lens if I found it to have optical problems. Anyone still wondering why brick and mortar stores are finding it hard to survive?