Kiril Karaatanasov: Congrats to Olympus. I hope they stay on the profitable side. Indeed their newest cameras and the high end lenses are quite convincing.
To keep everything in perspective, here is what the difference between the GX8 and EM1 files look like, with A7rII for scale.
It's only a 12.5% difference in linear resolution, which is not significant in real world use. 25% difference in area again sounds noticeable difference on paper but it is not.
Plus, your math is not correct. The 20mp sensor provides 25% more resolution, but you can only crop it 20% before you get down to 16mp. 20*0.75 = 15, 20*0.8 = 16
Your macro/DOF statement is also incorrect. The more you crop, the shallower your DOF becomes, because DOF is inherently linked to magnification/output size, and cropping changes that.
The GX8's 20MP sensor has no significant advantage over current gen Olympus M43rds cameras other than a slight difference in resolution. 25% difference in resolution looks good on paper but in practice it's not really an issue.
The difference between printing at 300 dpi: GX8: 17.28" wide, EM1 15.36" wide, so you're gonna get about a 16x12 out of either.
Color Depth: GX8: 23.5, EM1: 23DR: GX8: 12.6, EM1: 12.7ISO: GX8: 806, EM1: 757
Just a fraction of a stop difference, and the EM1 has slightly better DR at that, but it's not like 0.1 stops of DR is something anyone can actually see.
Earth Art: I know that type of weather all too well. No camera is going to have much pop in overcast lighting. With that said, the image quality looks like nothing too special compared to other offerings. If I paid that much money for it, then of course I would be capable of also convincing myself the images have a special "magic" that only a Leica can produce.
Ming did a review of the camera and I honestly think it has some really cool features. The viewfinder alone seems like a quantum leap over any previous EVF to date.
Thanks for posting these DPR.
SL better at high ISO than the A7RII? Some people will believe what they want to I guess. Look at the studio comparison, at best the SL is on par with the A7RII, and at very high iso (12800 and higher) its noticeably worse in the mid-tones, probably about a stop or so worse.
That's not to say the SL's raws look bad, they're about what I would expect for a current gen sensor, but the sensor performance doesn't justify the cost. Not to say that is particularly noteworthy, no Leica is a good value when judged by sensor performance.
Jefftan: Is it possible to use F0.95 for landscape by focusing on infinity?or the razor thin DOF made it impossible?Anyone knows? Thanks
It's possible but sharpness is usually pretty poor, especially into the corners at 0.95 with most lenses this fast. If you don't care about that, and you're shooting a distant landscape focused to infinity, DOF won't be a problem, unless you have nearby foreground elements that you want in focus as well, then stopping down will be your only option.
D200_4me: Sort of a hard sell if you look at it closely. You could get a m4/3 interchangeable lens camera with a kit lens in that price range or less ($650 for GX7 with kit lens), depending on what you want. Then you'd have a much more versatile camera (because of the lens options) and of course better image quality.
Well, you can also attach a wonderful portrait lens like the 45/1.8, or an ultra wide zoom like the 7-14/4, 9-18mm, etc, and any number of other specialty lenses. The 12-50mm rather than the 12-42mm wouldn't add much more weight in any case.
So yes, adding more lenses costs more money and means more weight, but it gives you a lot of options that a fixed lens camera does not. I fail to see the logic where interchangeability is a bad thing, it's simply different. May be better or worse depending on needs. Different doesn't mean incomparable either, if there weren't differences, there would be no point to compare in the first place.
A Nikon 1 system camera with 70-300 would merit comparison as well if you're looking for extreme reach tele in a small package.
mr_landscape: CaNikon say bye-bye to most of your trusty fanatics;)
Please inform us clearly stupid and ignorant masses, which shooting situations do you frequently find yourself in that 42MP is unacceptable but 50MP is perfect?
Difference between 42mp and 50mp is about 8% linear resolution, in other words, basically imperceivable to the human eye. You'll see as much (if not more) difference in resolution simply by observing sample variances between two copies of the same lens.
If 42MP isn't suitable for your work, 50MP isn't either, and you should probably pick up an 80MP medium format back, or shoot large format film.
SteB: The 90mm macro, looks rather long for a 90mm macro lens. It looks more like a DSLR design with a built in extension to fit a short back flange distance. The Tamron 90mm macro, both the old and new additions, prove that a high performing 90mm macro lens for FF doesn't have to be big. In fact I think it's not unlikely this Sony lens may be based on the Tamron 90mm macro. I'm not saying it is, because we need to see the specifications. However, it would hardly be the first time a camera manufacture has done this.
The 90mm lens looks like a DSLR lens with a tube built in, because that is essentially what it has to be. When you get into the telephoto range, there is no getting past physics. Telephoto lenses of the same focal length an image circle (in this case 90/2.8 FF lenses) will actually tend to be longer, not shorter, when we're talking about mirrorless lenses. This is because the FF DSLR has a longer flange distance, thus the camera itself provides some of the extension required for a lens to physically have a 90mm focal length. You can see this with the FE 70-210/4, is actually a little longer than the EF 70-210/4. Not because Sony is stupid and doesn't realize people want small lenses, but because they can't break the laws of physics to accomplish it.
HowaboutRAW: 8 or 10 raws for download, please.
These jpegs look decidedly unimpressive on a 15" MacBook retina. Hoping the raws would be better.
Contrast ratio is a useless marketing term and most manufacturers make up/have their own standards for the measurements. Its not a reliable spec because you can't compare it directly, it would be sort of like if some camera manufacturers measured ISO in a different way.
Contrast ratio shouldn't have any bearing on monitor selection when color accuracy is the goal, many manufacturers pump up the contrast so the screen looks more "vivid" to the average consumer, but the color accuracy on these displays tends to be worse. Lower end panels often have higher contrast ratio ratings, because its an easy spec (even if it means very little) to sell to a consumer.
What your monitor's contrast ratio spec is, is really neither here nor there, and has no bearing on color depth.
32 bit is typically what your color setting will say(as apposed to 16/24 bit), which is 8+8+8+ 8 bit alpha. I highly doubt it says 32 bit *per channel*, if it does, take a screenshot.
Virtually every monitor made in the last 10-15 years can run in 32bit "truecolor" mode (even if some can't display all the colors, ie cheap 6bit per channel TN/e-IPS monitors)
"As of 2012 many modern desktop systems (Mac OS X, GNOME, KDE, Windows XP/Vista/7, etc...) offer an option for 24-bit truecolor with 8 bits for an alpha channel, which is referred to as "32-bit color" or the RGBA color space. Switching to an 8/16/24-bit color option in those systems generally disables transparency/translucency effects, and the only reduction in color depth is seen when going to 8/16-bit color."
Ah, here is a PDF from HP that should clear it all up. They specifically reference 30 bit (in reality 10+10+10) as apposed to 24 bit (in reality 8+8+8) when talking about "30 bit" dreamcolor monitors. http://www.hp.com/united-states/campaigns/workstations/pdfs/lp2480zx-30-bit-panel.pdf
It also references 18bit (in reality 6+6+6) bit monitors, which most cheap TN based monitors are, and even some e-IPS monitors like the dell U2311H etc.
No place in this PDF does HP claim 90/96 bit depth.
I will be happy to provide sources of information about the 6/8/10 bit depth per channel of various different panel types.
Again, a 10bit per channel (30bit total) monitor means you have a high quality, professional monitor, so I am sure it looks nice. But you do not have anything exceptional or rare, as Dell(U2413) and Eizo also offer hardware with the same specs. Possibly even the same panel, as only a few companies actually manufacture monitor panels (Samsung and LG are a couple examples).
Please provide reference for the specific software (and where) it claims this is 96 bit. As again, no other resource on the internet makes the same claim (if you can find that, I would be interested as well).
Even more THIRD PARTY sources listing the LP2480zx as a 10-bit per channel (30 bit total) monitor.
Make sure to read past the "from HP" marketing info. 12-bit LUT, 10-bit interface.
I've yet to find a single source that lists this monitor as "90bit". Please provide me with such a source if you intend to keep this up.
http://www.bhphotovideo.com/c/product/1010658-REG/hp_gv546a4_aba_hp_lp2480zx_24.html 12 bit lookup table, 10bit/per channel color, again similar to every other high end monitor on the market.
Again, computer monitors are typically specified by color per channel, not total color for all 3 channels, which is the only difference between my 10bit and your 30 bit number.
If you want to call it "30bit" that's fine I guess, but you're already adding up all 3 10 bit channels. You don't multiply it again by 3, because that makes no sense.
32 bit float gives you a data range of -3.4E+38 to +3.4E+38, which is a much, much, MUCH higher value range than you can represent on any current hardware. But somehow you think you've got a 90/96 bit monitor? How many values can your monitor display? What GPU can actually push anything close to that amount of values to the display?
Do you realize that you're basically the only person on the internet making this claim? Go ahead, google 90/96bit monitor.
Also that HP monitor is a 10Bit per-channel monitor, equivalent to most high end panels, they're just doing the same 3x math you are for marketing purposes and calling it 30 bit. Nothing new there. When you talk about bit-depth for monitors, the standard term is per channel. You can't just 3x everything and pretend like its something new. Look up the specs on B&H, clearly 10/12 bit like all other pro type monitors.
And which monitor do you have again?
Not sure what fantasy world you live in, but you do not have a 96bit monitor, because such a thing does not exist. Most high end monitors use 8-10bit panels, and even the most expensive professional monitors are 12bit at most.
revio: DPR says: "The D3300 continues to be one of the smallest and lightest digital SLRs on the market"
How can a NEW camera "continue to be one of the smallest..."???
Since it´s not been existing until now, it can´t of course...
The D3300 seems, anyway, to be a nice and capable camera, and notably smaller than its predecessor was.
A. The Nikon D3XXX line has always been on the small side, this is the latest in that series.B. ballwin12 - What part about "one of the smallest" confused you? It is *one* of the smallest DSLR, I haven't seen anyone claim it is smaller than the SL1.
Lucas_: Although a bit heavy for an f4, the 70-200 would be a great option for Sony folks, if priced right. Currently there are only huge/expensive f2.8 choices on that range from Sony, Sigma and Tamron.
The Tamron 70-200/2.8 is 1112g/201mm vs this Tokina lens at 1020g/168mm, so barely any size/weight gain.