JackM: 5D3's AF is behind the times? LOL. I guess the 1DX's is too then.
Face recognition is only relevant for the very few potential customers that primarily shoot portraiture.It appears that the personal preferences of those who should remain objective are coming through in a pretty big way.We should hold those who publish reviews to a higher standard than that. As the commoners we're allowed to comment with bias and emotion, but who wants to read reviews knowing very clearly that there is a slant to the preferences and analysis of those who are supposed to be neutral.That's really what we've lost thanks to the internet: a sense of separation between the opinionated and nameless rabble and those who are producing content that should present fact, and present it in a balanced way. Fact presented in such a way as to support a single perspective is still fact, but because something is fact does not make it objective. Selective truth telling is a clever way to push a subjective viewpoint (read: bias).
shutterbud: The last gasp from a once-great manufacturer. This model admits that Canon is so far behind the times that they have to give you everything they have to make the purchase of their latest APS-C camera even worthwile. Compare a D700 to an A7R. Then compare a 7D to this.It really is quite sad.
Bud,I'm pretty sure he meant on the cameras he owned the primary concern wasn't having the best focus system money could buy. Today that's basically the 7DII and 1D X. After his exaggerated "unlimited dynamic range" I'd say he was talking about the hypothetical "perfect camera" and not an actual one. And your name is Bud. Can you read your name? Would you prefer Shutter or must we write it out completely and as is, shutterbud. Are you really an improper noun...at least he capitalized "Bud." He didn't call you pal or chief or buddy...he called you your name.
mpgxsvcd: It is remarkable that despite all of the valid issues that Dpreview and its readers have pointed out the owners of Canon cameras continue to produce impressive results.
Does this mean that Canon camera owners have to work harder to get great images? Does it mean they have to avoid difficult situations more than other camera owners would have to? Or does it simply mean that they don’t find themselves in situations that often where the disadvantages of their camera are relevant?
Wait, pictures are composed of something other than dynamic range? Judging by most comments and posts that isn't the case. If it isn't a 14EV DR shot then it's CRAP! (Imagine that in the same tone as "Welcome tee all things Scottish. If it's not Scottish, IT'S CRAP!"). But seriously, thanks for mentioning that there are other things to consider when taking a picture.
Yep, and with little need to upgrade companies like Amazon create partnerships with DPR an DxO in order to show differences and then exaggerate the heck out of them. Not that reviewers are doing that kind of thing out of malice or greed because when was the last time you met a scientist, engineer, or photographer who wasn't the type that would argue the tiniest detail until the end of time?Televisions went to LED-backlit, 1080p, 240Hz, 60" screens and now they are pushing 4k televisions because older televisions are in a crowded marketplace where they are a commodity and have narrow profit margins. Yeah, buy a 4k television and then sit two feet away from it to really tell a difference. Yeah, there is a difference, but my parents, friends, and significant other already say "I can't tell a difference between a DVD and Blu-ray." Holy crap. I can, and as much as I enjoy technology it's time to move on to enrich our souls and clean up the planet now.
What it means is there really is a threshold of usefulness here as there is in any other part of life. At some point the advancement curve starts to level out dramatically and while other sensors may be "years ahead" it just so happens that those years took place on a part of the curve that was barely more than level. Also, boiling a camera down to one or two small (obsessive-compulsive, ADHD) complaints doesn't change the fact that overall the camera has technology and performance that others don't have.I remember looking around for my first DSLR. Resolution was so low, noise was such a problem, and AF systems were so rudimentary that the balance of all those features really did make every little difference pretty significant. Now, there are difference, but since we're past the point where those difference are required for a tool to be truly useful places like DPR have to split hairs just to remain relevant. Same with everything else....phones, tvs, tablets. Reaching.
mpgxsvcd: Noise performance of the Canon 7D mkII is not its biggest issue. Actually the noise performance in low light is quite good. Even the DXO tests show that its low light performance is less than ½ a stop behind its competitors. Take a look at the low light scene near the bottles at ISO 1600-6400 and you will see the 7D mkII noise is handled quite well.
Dynamic range is the issue with the 7D mkII. It is almost 2 stops behind its competitors in dynamic range. The studio scene can’t demonstrate that difference because the scene simply does not have enough dynamic range to exceed what the 7D mkII can display. If the dynamic range of the scene does not exceed the dynamic range capacity of the camera’s sensor you can’t draw any conclusions about dynamic range for that image.
Barn, it was a rhetorical question that I immediately answered. We all know how to preserve our highlights. Quit being so bitter that Canon is still in business. This is the dumbest discussion I've ever been part of and I'm sorry I ever said anything. It's like being in a Three Stooges bit.
Okay....here's the problem. "More data" on the high end (to the right) is just more photons. "More levels" just means that's where more and more photons can be gathered to improve SNR and overpower the noise that exists at the lower end. As far as I can tell all your arguing is terminology, but in fact we're talking about the same thing.
And that's ISO100...only....after that all of this is hot air.
Preserving highlights? Who does that? Everybody who needs to, that's who. And I never said that the net affect of capture and display wasn't linear. The histogram is a JPEG histogram. Always. Always. Never have we seen a RAW histogram on a camera. So JPEG encoding matters whether you shoot RAW of JPEG.
I get to be confrontational. I don't work here.
Highlights don't need protecting all the time. You guys act like every shot has a tonal range spanning the entire histogram. Sometimes you do, and it's just as valid to butcher your shadows as it is to butcher your highlights depending on where you want your detail. Of course HDR is an option if you need both. HDR is the most abused post processing effect and usually looks unreal in a bad way.If you are underexposing to preserve highlights you're assuming a wide tonal range and the "underexposure" is there because most of the range isn't highlights. So yes, then you have to lift shadows. OF COURSE there are applications where this happens far more often and every bit of sensor performance counts, but having to lift the lowest shadows by three or four stops I'd say is rarely necessary. And when you underexpose to preserve highlights (highlights still as far to the right as possible) how much recovery does the camera have? Canon usually has more than others.
Discuss physics and shot noise and read noise all you want, Rishi, but what it amounts to is there is noise in the shadows. I had no idea (dripping with sarcasm). And it's that noise which limits the ability of detail to be lifted out of the shadows without also lifting the noise (resulting in no real additional detail.....yielding less dynamic range....yielding the results you claim to have that we have yet to see).So, use your levels in RAW the way you should since JPEG shooting is what it is and if you clip your highlights there they are gone. ETTR is useful in RAW because the histogram is lying to you. Now preach more physics and the practice of exposing is still the same. Don't underexpose your images because you can always drag your shadows back down but you can't get more light after the fact. Unless of course you don't actually want that shadow detail, which is also perfectly valid from a creative standpoint. But for post processing the more light the better.
"You ETTR because the human eye is more sensitive to shadows, so the linear response of the sensor has to be pushed to the right which stretches the left side of the histogram and compresses/truncates the right."Editing RAW recovers what the JPEG shows as gone to the tune of 1 to 2 stops of highlights. ETTR uses the high end that JPEG says isn't there."Rishi, you do push to the right to use more bits because otherwise the shadows would be way underexposed and the linear capture would be skewed by the stupid JPEG histogram leaving the brightest stops unused." How is this untrue?"You ARE using more bits because the RAW shadow need less stretching and highlights makes use of ACTUAL stops on the high end."..that the JPEG histogram compresses and/or clips."There are more levels open at the bright end and if you don't use as much of the brightest stops as possible you will be stretching very thin shadow levels." How is this different than avoiding noise at the low end by underexposure?
And Rishi..."@Joseph Black: You are wrong. That is a fallacy, from old texts, propagated to this day by, apparently, people like yourself. " Who exactly are people "like myself"? Your tone is confrontational and condescending. You preach SNR and then ignore the signal. Just because SNR is the same doesn't mean underexposing yields a signal that has the same information in the same places, which make pushing and pulling a whole different ball game. Yes, where the signal is on the histogram matters in post.
RaghavBaijal: Just to give you guys a point of reference to understand the value of 2 Stops of DR:
The difference of 2 Stops of DR is greater than changing from Crop Sensor Cameras to Full Frame Sensors!!!
In other words, Landscape photographers would benefit more by moving to Sony/Nikon Crop sensors than by moving to Canon Full Frame Sensors!!! If you jump from Canon Crop Sensor camera to Nikon/Sony Full Frame, you will gain almost 3 Stops of DR!!! That's almost like HDR Pictures!
Its just sad... Canon has a lot of catching up to do in Sensor tech...
This DR discussion reminds me of someone moving from a point and shoot for the first time talking about DSLR's as though the lens had nothing to do with detail resolution. They would talk about some camera having a 10MP sensor and how much more resolution there was compared to an 6MP sensor, and then go on to talk about their brand new $200 70-300 f/5.6 lens. The DR difference according to DxO only (so far as I can tell) exists at ISO 100 exclusively. Context people, context. And what Karl said, too.
Raghav, you obviously have no idea what HDR means.
LOL rw. Case...rested.
Rishi, as brightness doubles at the low end you go from 2 to 4 to 8 to 16 and those are full stops of brightness increase. On a given 14 bit channel you can do the math as to how many levels exist in the highest two stops. Reducing shot noise is always desirable, and that includes not underexposing the shadows. You still shouldn't underexpose the highlights, but the shot noise is OF COURSE more noticeable when the shadows stops have so little information and shot noise can easily overwhelm them. And of course tone mapping is required because of the way the linear capture would be displayed on our devices....because it isn't reality and that's what images are there to simulate, requiring all manner of digital trickery. Post eighteen comments now to teach me again why I'm wrong.
You ETTR because the human eye is more sensitive to shadows, so the linear response of the sensor has to be pushed to the right which stretches the left side of the histogram and compresses/truncates the right. Rishi, you do push to the right to use more bits because otherwise the shadows would be way underexposed and the linear capture would be skewed by the stupid JPEG histogram leaving the brightest stops unused. You ARE using more bits because the RAW shadow need less stretching and highlights makes use of ACTUAL stops on the high end.There are more levels open at the bright end and if you don't use as much of the brightest stops as possible you will be stretching very thin shadow levels.
JordanRusev: Nikon D7100 gives more detail and is cheaper. I say it as a canon fan, it hurts me but is truth.
It's actually not true. They appear to be almost identical, and in the neighborhood of detail one should keep in mind the role of the lens. As for noise, they are very similar. And in terms of judging a camera based on two variables: get crucial, man. The difference in cameras these days has more to do with performance and the system. You're fortunate to be no more demanding than having two variables with which to make a camera buying decision.
EduPortas: To everyone saying that DR is only average on the 7D Mark II is ask you this: what good is picture taken with 13 or even 14 stops of DR if it is not in focus? This is a camera for people who need a large AF area on their sensor, mainly journalists who actually use their lateral AF points when composing with a subject on either the extreme left of right of the frame. As a pro, I can appreciate Canon making AF much easier for us. I will gladly sacrifice a bit of DR for this. (Not a Canon fanboy here. I've owned Nikon, Canon, Sony and Fuji gear, BTW).
And full frame is a niche market for journalists. Actually, I doubt journalists prefer full frame. They may prefer speed, in which case a full frame 1D X is the way to go, but that's now got crop competition from the 7DII.The only camera markets left are niche markets. Quit using that term like it means anything. Right now the new niche markets are those that used to be small and have gotten smaller, like medium format digital or large format film or film of any kind. Even the primary market for cameras now, smartphones, is a niche market for the phone itself. Most pictures taken now are only taken because people are forced to have a camera along with them built in to their phone. Most people don't give a flying leap about ever taking a picture for the sake of taking a picture, unless they are on vacation. Most people only take pictures now for instant Facebook/Instagram/Twitter validation in order to replace the human contact we used to have way more of.
Can't wait to see actual test results. Until then tl;dr.
If it's tough to show then the relevance and extremity must not be as relevant or as extreme as you imply.
Anyway, just wishing there was something other than a chart to go by.
Elaka......holy crap on a stick. Get real. That little example wasn't much to go on. That guy hates everything Canon makes. This hole idiotic back and forth isn't about "loyalists" or science or sensors. It's all about overblowing small difference and pretending they are mind blowing and world changing.
Talk about the tree until you're blue in the face. Others will spot the forest.