April 16, 2014, 02:10:06 AM
This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.
Messages - ecka
« on: March 19, 2014, 06:43:55 AM »
A hypothetical Samyang 50mm F1.2 with autofocus that works well, good sharpness and contrast at F1.2, and costing less than $1000 is a dream. In fact, I'd be happy with a 50mm F1.4 that is so good with Zeiss Otus, but costing only $500.
I'll get it even if it only has an AF confirmation chip if it'll be as good as the Zeiss otus but $500 or less.
I think that AF confirmation chip would be too little (it'a DIY thing) and autofocus is a bit too much to expect (for the price).
How about electronic aperture?
eckaThe lenses you mentioned are not really "tele" lenses
"because there are no small CSC tele lenses without huge compromises in IQ"
I would dispute that claim Olympus & Fuji both make very good and high IQ lenses for CSC cameras. Lenses like the Olympus 12mm f2 ED, 60mm f2.8 ED, 75mm f1.8 ED, Leica 45mm f2.8 in micro four thirds or the Fuji XF 27mm f2.8 all perform excellently.
Well, not the 12mm, but taking that crop factor into account, current primes get you to the equivalent of 150mm (I would add the 45mm Olympus to the list), with very high image quality. As for Micro 4/3 zooms, they tend to compare quite favorably with their dslr equivalents and are, of course, much smaller and lighter.
Well, for me, the cheaper m4/3 output is not good enough to choose it over a good P&S camera or two, or three (G1 X, RX100, RX10) and the expensive one just makes no sense.
150mm on FF is a short telephoto, nothing powerful really , just like anything else in 100mm to 200mm range. So, for m4/3 it must be 100mm+.
dead end is a generous description. this system was still born - Canon shoved it out the door ... Now you're in
a "pasting feathers on a turkey" mode and you might be better off kissing it off and starting from scratch.
The EOS M was the second best-selling MILC in Japan last year. One country's meat is another one's spoiled turkey...
I just wonder how many Japanese buyers buy it with only one lens and will never even put another lens on it. Of course, as I recall, the average number of lenses owned by Canon DSLR buyers is less than two.
There are many people who would buy it without any EF-M lenses. They only care about adapters for lenses they already have, like Canon EF. That makes a lot of sense for telephoto, because there are no small CSC tele lenses without huge compromises in IQ. EF-S 55-250mm IS STM can be considered "near native" lens for EOS-M (via the adapter) - fast AF, nice optics, stabilized, compact (for 88-400mm equivalent ) and affordable. Many are using old manual lenses and they love it.
So the only important bit is to educate people on the apparently simple concept of magnification.
At the very essence of all this is:
1. How big is the object as you now see it in relation to how big it is in real life.
2. How big was the aperture opening.
The first allows for everything involved in the reproduction; focal length, distance to object, coc, sensor size, crop, print or screen size, and viewing distance. Each of those affects the magnification. Plugging these values into a dof calculator just allows it to calculate the magnification taking the print or screen size and viewing distance as standards for a set CoC, some calculators actually allow you to change the CoC and magically your DOF changes, even though the image is already taken.
The second dictates the amount of blur in relation to the magnification.
Hyperfocal is a side issue with no merit, stuff falling within the hyperfocal is still not as sharp as the plane of focus. magnify it the same and it is just as blurred (try it with the images on the LL link, I did for a thread a long time ago). Indeed lenses marked hyperfocal scales are historically at least one stop wider than accepted norm CoC figures because they used a different value to calculate them. But the important bit is that hyperfocal is just another manifestation of magnification, it isn't sharp, it is just small enough to give the illusion of sharp.
It is all about the illusion of sharp. The actual plane of focus is always at it's thinest, which is the diffraction limit of a lens. However, when the CoC becomes as small as a single pixel of your camera sensor, it is perfectly sharp from that point and smaller. That's how you gain the depth (of field ). You will cross the line when it is impossible to magnify it enough, because the resolution is too low.
It is just magnification and aperture, once you accept that, however convoluted the route to getting there, then it is easier on the mind.
Start to think magnification and aperture and all the inconsistencies and complications fall away.
Well, there are other things involved in "magnification" which are not taken into account by most people who are trying to learn how things actually work, so it may be an oversimplification. This rule may not work that well when shooting something farther away, because UWA lenses would go hyperfocal, while the tele lens would still produce some blur in the background (due to stronger magnification) and that's the rare situation when the CoC thing becomes important before you actually take a picture. You have to take the convoluted route first, and then, if you survive , you can calculate using the magnification and aperture, because you'll know the exceptions.
A lot of people seem to have difficulty grasping this concept. Including some who think they know better in this thread. Its shown quite well in this article.
Note that no where in the luminous landscape article did the guy refer to subject isolation.
DOF and subject isolation (or background blur - or do you mean something else?) are not necessarily the same.
Taking pictures with a short focal length and short subject distance vs with a long focal length and longer subject distance (for the same framing, using the same camera and aperture) results in the same dof, but the latter will give you a much blurrier background.
How about - "For the same magnification of the area in focus (different lenses, different distances), longer focal length at the same aperture will give you same DoF and more background magnification".
"Wide open" vs "wide open" is pointless.
I see your point, but I beg to differ concerning the conclusion: If you want the thinnest dof (or fastest speed) with any lens/camera combination you'll use "wide open" a lot in reality, so it's perfectly valid to look at this performance no matter the synthetic equivalence.
Look? - yes. Compare? - no. If you want the thinnest dof, then you should get a FF camera in the first place. The fact that 10-22 can't do f/1.8 (which is the FF equivalent for f/2.8 ) doesn't justify your logic. There is no f/1.8 UWA for APSC (for now, maybe the mighty Sigma will make one later ) and that's another reason to go FF.
Sony needs to release uwa asap for a7 system. ...otherwise, it could be "dead end" as eos-m.
FE 16-35/4.0 zeiss/sony is next up for 2014 according to the roadmap.
Current 24-70 f4 size is kinda big for this system. They need to release compact prime at f4.
Yeah, for an UWA 14mm (or even 20mm) f/4 would be nice, but for 85mm ... wee..ell I think f/2 is small enough for me ... I could live with that .
16-35mm f/2.8 L II on a 5D3 Vs 10-22mm on a 70D, which theoretically would give the better image?
The often criticized "problem" of current Canon ff lenses is the lack of corner performance esp. wide open (if you care about that). Ignoring that, the ff sensor can always generate higher res yadayadayada (see posts above), this is visible in the iso crop.
Another potential fact to keep in mind is that you can buy nearly 3x 70d+10-22 for 1x 5d3+16-35 :->
For fair comparison it should be 16-35L@16/5.6 vs 10-22@10/3.5, or at least choose the best performing aperture number from each lens. "Wide open" vs "wide open" is pointless.
I think that "excellent detail and sharpness" is a matter of opinion. I thought my 7D+Sigma 50/1.4 was excellent, until I tried 5D2+50/1.8'II (I hated the build, pentagon aperture and AF noise, but it was SHARP!). FF wins thanks to either bigger pixels or more of them.
I'm going to be blunt because I've shot both combinations (owned both lenses at one time, now just the Sigma): at low ISO and wide apertures, if your final 7D+Sigma file doesn't look better then your final 5D2+NiftyFifty file something is wrong.
By SHARP! I mean f/2.8+ sharp . Sigma 50/1.4 produces much nicer bokeh (I think it is one of the best 50s), but I don't think it is sharper wide open on crop, than 50/1.8'II on FF. However, sometimes wide open images from 50/1.8'II are a bit dreamy with glowing high contrast edges.
Ecka you are right. In the real world with good light I would not hesitate to use a crop sensor camera with a good lens. The problem is noise at high ISO when light levels drop. This is where a full frame body comes into its own. The maximum ISO I use on my 7D is 800 whereas on my 5Dmk111 I will go to 4000 ISO. I do not rely on test charts to tell me what the best images are from a camera, I rely on experience and the final image displayed on my computer.
Actually, I was talking about low ISO , but yes, FF wins the high ISO battle by 2 stops.