>What does the 57% crop mean ? that it was taken in DX mode ?
He's trying to apply the same level of magnification to both the D700 & D800 images, which requires some downsampling of the D800 images.
Some may find that makes for an unsatisfying test, but in the context of comparing output of both cameras at an identical print size, it's appropriate. For those hoping the higher D800 pixel density allows them to print larger, it doesn't provide much info. But, it looks like you can be confident the D800 isn't a step backwards from D700 in terms of low-light shooting, which is something some people feared.
Great comparison. If I may, I'd like to ask a question which will take this off topic. If you'd prefer I start another thread please let me know.
I'm very close to placing an order for the D800. Of course I recently purchased a D7000 which I love (upgrade from my D200). My dilemma is I'm also considering purchasing a Nikor PCE-E 24mm tilt shift lens. The problem is I can't afford to buy both, so which one do you think would provide me with the greatest benefit?
I shoot landscapes and typically enlarge to 20 x 24. All opinions are welcome.
Buy the PC-E 24mm. The D7000 has all the resolution you need for superb 20"x24" and larger prints. Unless you're planning to have people inspect your prints from 12" away (in which case they'll hardly even know what they're looking at), the high resolution of the D7000 is more than sufficient I think.
The D800's full resolution of 36mp is a photo retoucher's dream. Those dogs need lots and lots of pixels to push around in order to make models, products and old actors look, respectively, slimmer, better and younger while keeping everything looking real. That's what 36 mp are for IMO.
The larger the print, the great the distance at which it must be viewed for full impact. At 12.3mp several years ago, I think, we already passed the point at which more megapixels mattered for big prints.
Thank you for posting! I had suspected something like this, but had not seen examples. So yes, it looks like the D800 will be a great camera for a variety of use cases: Get very high resolution when shooting low ISO scenes like landscapes, and getting at least D700-equivalent images when shooting in bad light / high ISO situations.
Many thanks for the responses. I thought the lens would be the smarter move - just needed some expertise to help me make the right decision. I'm sure the cost of the bodies will continue to provide more pixels per dollar in years to come anyway. Thanks! Doug
I was at the UK "Focus on Imaging" show yesterday and Nikon repeated what i had read published by them. That being that the D800 will equal the D700 in terms of noise reduction. It would appear from the images on the link posted that this is certainly the case up to ISO3200, but after that it appears to my eye that the D800 is cleaner. This is indeed very promising except I would like to see the same comparison with both camera images taken in DX mode.
<I would like to see the same comparison with both camera images taken in DX mode>
The center of the frame will show the same relative result. You are just taking a crop from both cameras.
Both cameras in DX mode show more noise if you are comparing to FX images. The impact of going to a smaller size image with the DX crop means you then have to upsize the D700 DX crop which magnifies the noise. On the D800 DX crop, you lose some of the beenfit of downsizing since you are starting with a 15+ mp image. The best indicator of the D800 with a DX crop is a D7000 - same pixel pitch, DX , and same processing.
Thanks Eric, Steve etc, for the explanation, except that if I take an image with my D700 in DX crop (which I've only done for experimental purposes) at ISO 1600 and cropped to 50%, the noise appears to be far less than an image from my D7000 cropped to 50%. Perhaps it's just me or my eye sight, but could pixel density in the sensors vary noise wise, dependant on some other factors such as sensor manufacturers, processors etc?
I suppose I'm trying to see if I could justify the D800 in IQ to use as both a landscape, portrait (FX) and wildlife, birding in DX. I suppose the answer will undoubtable be to wait for a D400, as I am supposing that it's a big ask for the D800 FX camera to do all of those things equally well
>Thanks Eric, Steve etc, for the explanation, except that if I >take an image with my D700 in DX crop (which I've only done >for experimental purposes) at ISO 1600 and cropped to 50%, the >noise appears to be far less than an image from my D7000 >cropped to 50%. Perhaps it's just me or my eye sight, but >could pixel density in the sensors vary noise wise, dependant >on some other factors such as sensor manufacturers, processors >etc?
This is to be expected Richard (although I don't know about far less). I think Bill's chart shows a .75 stop advantage to the D700 over the D7000 at ISO 1600.
as a collective of people who deal with imagery, it's often more telling to see a picture, than read a chart.
I agree the the final arbiter is the human eye.
However, it's my experience that the human eye is not necessarily a good judge of absolute values such as exposure or Signal to Noise Ratio (SNR). In this case I offer the chart as "confirmation" that what the eye judges seems reasonable. Had it been otherwise I would be questioning why; as I trust the measured data more than the human eye.
>Perrone, > >as a collective of people who deal with >imagery, it's often more telling to see a picture, than read a >chart. > >I agree the the final arbiter is the human eye. > >However, it's my experience that the human eye is not >necessarily a good judge of absolute values such as exposure >or Signal to Noise Ratio (SNR). >In this case I offer the chart as "confirmation" >that what the eye judges seems reasonable. Had it been >otherwise I would be questioning why; as I trust the measured >data more than the human eye. > >Regards,
>I think that most people will be using their eyes to make >judgements? Bill did you not use your eyes to make the charts?
I have just been given a warning for the above remark because it was deemed personal. I would like to state that it wasn't and merely an observation to the remark that Bill thought his information is better than the member's eyesight. I am not the only member that has questioned the validity of Bill's statements. Angry.
Wed 07-Mar-12 09:58 AM | edited Wed 07-Mar-12 10:00 AM by richardd300
In truth if we could trust our eyes and the job the brain does of interpreting colours, we wouldn't need monitor calibrators or need paint and colour charts. It's not about clear, sharp vision it's about processing colour and noise visually. As an example I have seen many posts where people have stated that the image is soft, there is chromatic abberation or excessive noise and such. In many cases I have failed to see the problem, but that doesn't mean it doesn't exist
I don't think however that charts can take the place of the power of our eyes. Without eyes we wouldn't even see the charts. I approach photography from a practical point of view. Theory is needed but it shouldn't override practical aspects. Monitor calibrators calibrate monitors and not our eyes so I don't see - no pun intended the connection.
There's an element of truth in that, however the camera and lenses are not connected to our brains and not always what we see, or think we saw, are always a true representation. Perhaps let's approach it another way, we tend to zoom to a distance with a lens or in software, at which the eyes can only see a fraction without that intervention and then with little detail. Without the assistance of science, I am sure the results would be extremely disappointing.
I do appreciate where you are coming from, but there has to be more to getting images correct and so we can enjoy them. Science and mathmatics have their part to play. After all the charts are the only visual result we have without having to understand the science behind them.
I do appreciate where you are coming from, but there has to be more to getting images correct and so we can enjoy them.
Of course there is more but usually one aspect of getting the image "correct" is to avoid apparent noise. My measurements of apparent noise are objective and not confused by different scenes, lighting, different apertures, different sharpness, etc. This is why I favor them to compare cameras. I don't expect everyone to adopt my view; but I also don't expect it to be attacked as invalid without a good argument.
Science and mathematics have their part to play. After all the charts are the only visual result we have without having to understand the science behind them.
The science and mathematics are well known and documented, even at my site. If someone doesn't care to understand the material (which is fine) then I guess they have to take the findings at face value. Or, they can disbelieve the information but (IMO) are in no position to criticize it publicly from a position of ignorance.
It will be great when the cameras are actually out and there will be a wealth of hands-on experience. For now we "make do" with available information.
FWIW, I didn't take it as too personal, just a bit snippy.
Bill did you not use your eyes to make the charts?
So here is the direct honest answer, not intended to appear "defensive".
No I did not use my eyes to make the charts! (This implies I judged something with my eyes.) A sightless person could do it. The charts are generated automatically using data that was gathered statistically by a computer program that is fed NEF files.
If I hook a "stereo" up to some equipment to measure it's power that doesn't mean I shouldn't also listen to it but it also doesn't devalue the objective measurement of the equipment. (At least in my opinion.)
>There's an element of truth in that, however the camera and lenses >are not connected to our brains and not always what we see,
The eyes are connected to our brains and if we were sightless then all of this would be for naught or at least very very difficult. Information captured by our brains via our eyes is absolutely essential to all decision making. We can't work cameras or computer programs without them or create meaningful Nef's. We wouldn't even be able to create these things without sighted persons. Even if the computers create "perfect" graphs we need eyes to see them. What came first? The chicken or the egg?
>Information captured by our brains via our eyes is >absolutely essential to all decision making.
Not all, no. Our other senses can equally give our brain information that allows it to make decisions. I can decide whether to go to the beach tomorrow after listening to the weather forecast on the radio; I can decide whether a pint of milk os "off" by sniffing it...
Now, can we please return to the original topic. Both visual assessment and measurement have their place in determining two cameras' relative performance, so let's just leave it at that.
Does anyone have an idea how the D800 compares to a D3x...???? Or better yet add a D3x down sampled to 12mp and compare all three.... I think a D3x would be a better camera to compare D800 images to. I fell it's been proven that a D800 down sampled to 12mp will look better then a D700... but what about down sampled to 24mp (D3x)????
Does anyone have an idea how the D800 compares to a D3x
When we say "downsample", it's a way of saying that we want to compare at the same output size and viewing distance. PDR is normalizing in that way. So this chart shows the D3X, D700, and D800 PDR curves. Note that we don't know for sure what the curve will be for the D800 lower than ISO 400. Also note that it is the horizontal differences between curves that is generally more important.
I for one, really appreciate the effort put into this data collection and presentation. Data is one thing that often gets downplayed because it is not well understood by many people who are happy with impressions based on perceptions alone. Our brains, which has more to do with impression than eyes do, are not good indicators of what a camera is actually doing and what it can be expected to do in various conditions. The same with sound, our brain completely synthesizes the perceived experience based on very low bandwidth, low quality data with extremely high noise to signal ratios(highly reverberation fields are mostly noise-artifacts of the original wave...like in any real acoustic space), and often fool us as to what is really there. Even the best stereo systems can't fool a 5 year old wearing a blindfold, as to which is a real sonic experience and which is a representation of one in recording. A more experienced listener is often easier to fool because of cultural factors and familiarity with what to expect. Same with photos, a casual observer of an image has an entirely different impression of an image, as a rule, than an experienced photographer because of what additional expectations and viewing habits skew impressions. That is where data and proper testing really shines, to remove the interpretation of image taking to level the playing field for less biased comparisons or indicators of predictive performance. These charts are more useful than all the web based example of test shots. The same scene might be interpreted very differently by viewers with different experiences. I was viewing a portrait on another site and liked it, it gave me a really nice impression of the person, environment and would have been pleased to have taken it. But all the comments posted were all over the map, several saying it was terrible because of a color cast, another said it was not sharp and all I could assume was that my monitor was not high enough res or color tuning but I suspect more that is was because the overall impression of the likeness and likability of the image trumped technical flaws which I never could see. If there were technical flaws, I would not have cared but it would have been nice to know what they really were based on camera characteristics so steps prevent it the next time. I suspect there was no real flaw with the image, that 99% of non-photographer, non-pixel peepers would have liked it. Data helps more than subjective impressions of camera performance overall by isolating camera performance from the environmental issues like light, materials, lens etc. But visual impressions of a particular image, with a unique time element, environment etc and all that impacts subjective impression of a specific image are also important. Both forums of information makes up a more useful and predictive collection of information.
Overall, the D800 is measuring very very good. I will be very interested in the data points below 400ISO to see it if mimics trend line of the D7000 at 100, if so, the D800 will be the undisputed studio and landscape DSLR king for dr alone, even before adding the higher res into the mix. Stan St Petersburg Russia