I was just wondering what kind of quantitative improvement should we expect from the D7100 compared to the D7000 & D600, regarding
1) the noise 2) the dynamic range at high ISO
Law of physics cannot be broken of course, but is there really any space left to significantly improve the performance of the D7000 DX sensor, which is already quite good ? Should we rather initiate the move to FX 24 MP like D600 ?
Sat 02-Mar-13 04:21 PM | edited Sat 02-Mar-13 04:52 PM by Chris Platt
Compared to the D7000. Thom suggests the sensor is the same one in the D5200. If that is the case you shouldn't see much (any) difference in noise and maybe 2/3 EV improvement in dynamic range at ISO 3200 judging by DXOMark measurements.
>Thom suggests the sensor is the same >one in the D5200.
And even if it turns out to be a newer sensor, I wouldn't expect more than a small improvement. Keep in mind that the sensor used in the D5200 is the highest-rated DX sensor by DXOMark. That sensor isn't very old, and I wouldn't expect a big jump in performance over that.
The raw measurements will not be much different but the visual impression will probably be noticeable in the fact the noise will have a finer grain that seems to not be as noticed as course grained noise. The advantage of having enough pixels to down sample, like we do with the D800, removes a lot of noise by summing surrounding pixels so signal adds and non-correlated noise will partially cancel. The sharper more detailed files of the D7100 should allow more noise reduction in post before detail is noticed to be lost. I expect the camera to be capable of producing images second to none in the DX class, no matter what Sony and Canon try to counter with. Stan St Petersburg Russia
Most probably you're right for noise cancellation , but this is with the hypothesis that each pixel generates the same amount of noise than the D7000. If it is not the case, and the "individual pixel noise" is greater on the D7100, then what is diminished by summing up random noise is compensated by the higher noise intensity. Otherwise, why not having zillions of pixels ?
For the Dynamic range at high ISO, what do you think will be the effect ?
Random noise is well distributed across the spectrum and really does not add the way correlated data does so there is a real increase in signal to noise ratio. This principle is used in all sorts of information theory applications in communications, weak signal work and optical systems. Integrating a number of detections of a signal buried in noise, the signal raises to a greater degree than random noise in fact. Binning pixel data of large arrays of tiny pixels results in significant improvements in signal to noise ratio. Nokia has a tiny sensor in the 808 model phone that has a 41mpx camera that allows combining pixels to improve DR and noise performance. The pixels are really tiny 1.4 micron yet by binning they are able to get excellent DR at moderately high ISO I notice the improvement in every D800 image I downsample. You will notice it also. Stan St Petersburg Russia
Sun 03-Mar-13 11:47 PM | edited Sun 03-Mar-13 11:48 PM by jbloom
>I say this in part because the D5200 tested significantly >better with DxOMark than the D3200, also with a 24mp sensor.
I'm not sure I would say "significantly," but I don't want to get bogged down in semantics. Being a sports guy, I tend to look at the low-light ISO numbers of DXOMark as a guide. For the D3200, D5200 and D7000, they are 1181, 1284 and 1167, respectively. The difference between the D7000 and D3200 is about 1/7 stop. Those are close enough to be considered essentially indistinguishable to me. The other metrics are similarly close. There have been small incremental improvements in sensors over the past few years, and I see no reason to think that will change. (Plus, of course, there is the possibility that Thom Hogan is right in thinking the sensor may be the same one as is used in the D5200, in which case the performance should be essentially the same, too.)
Mon 04-Mar-13 12:34 AM | edited Mon 04-Mar-13 12:37 AM by billD80
The difference between the D7000 and >D3200 is about 1/7 stop. Those are close enough to be >considered essentially indistinguishable to me. The other >metrics are similarly close. There have been small incremental >improvements in sensors over the past few years, and I see no >reason to think that will change.
Here's where I disagree... The D3200 improvement seems to be minimal over the D7000, except when you factor in the additional 8mp's that the D3200 has (and the much lower price)...
The D5200 goes even farther.
Yes, on just the ISO (leaving mp's out of it) there hasn't been a "leap", but when these improvements are still forthcoming with a 24mp sensor, I think we're talking a significant advance.
My hunch, and it is only that, is, within reason, the D7100 will be even better than the D5200, on ALL counts, and with everything factored in, a good deal better than the D7000. It's a hope...
It is remarkable if Nikon just manages to achieve the D7000 ISO performance at 24 MPs. Personally I think the well is likely tapped for ISO performance on DX by and large. The reality is that the sensor size is likely at the point of bottlenecking substantial upgrades.
I would note some caution in using DXOmark's measures respecting ISO performance. DXO bases their score on "the highest ISO setting for a camera that allows it to achieve an SNR of 30dB while keeping a good dynamic range of 9 EVs and a color depth of 18bits".
Some have suggested this methodology is flawed as the numbers are fairly arbitrary. Note what is does not take into account, how higher ISO photos look. What happens when you try and recover detail from the shadows, etc.
I personally don't find that either the D3200 or D5200 have in actual fact, better ISO performance than the D7000.
That being said they did manage to squeeze around 2/3rds of a stop between the d90/300s and the d7000 so we'll see.
>It is remarkable if Nikon just manages to achieve the D7000 >ISO performance at 24 MPs.
They've already done that with the D5200.
Making a sensor higher-MP doesn't affect noise performance that much. What you lose in noise performance of the individual pixels you make up in the fact that you are averaging the noise out over more pixels.
What matters is the amount of noise in the output image, not the amount in the individual pixels. That's determined by how much noise is generated in the sensor and how much light it gathers. Overall noise generation is determined by the sensor technology; how much light is gathered is determined by the size of the sensor. Megapixels aren't really a significant factor.
>Personally I think the well is >likely tapped for ISO performance on DX by and large. The >reality is that the sensor size is likely at the point of >bottlenecking substantial upgrades.
I think so, too. Barring any breakthrough technologies, that is. (Maybe this one?)
>Some have suggested this methodology is flawed as the numbers >are fairly arbitrary. Note what is does not take into >account, how higher ISO photos look. What happens when you >try and recover detail from the shadows, etc.
DXOMark's approach is arbitrary, but I find that it pretty well matches my arbitrary standards. There is no real line-in-the-sand measurement for how much noise is acceptable, but having a consistent basis of comparison is useful. If you find that the DXOMark numbers are consistently low or consistently high for you, adjust accordingly. I would be surprised if you find them very inconsistent, however. At the level of fine detail, there may be differences between sensors and processing systems caused by such issues as color noise vs luminance noise that impact shadow detail, but for the most part those are second-order effects to me.
>I personally don't find that either the D3200 or D5200 have in >actual fact, better ISO performance than the D7000.
Which matches up nicely with the DXOmark numbers -- that 1/7-stop difference I mentioned as being essentially indistinguishable. As you go to higher ISOs where the SNR drops, the performance of the systems may diverge a bit, but probably not that much.
>Which matches up nicely with the DXOmark numbers -- that >1/7-stop difference I mentioned as being essentially >indistinguishable. As you go to higher ISOs where the SNR >drops, the performance of the systems may diverge a bit, but >probably not that much. >
I guess I was hoping for some alchemy on Nikon's part, not science... darn.
I find my D7000 to be more than a full stop better than my D90. To me the IQ and high ISO performance is a combination of low noise plus DR and detail in big crops. When you factor all of that in, my D7000 produces images that are equal to the D90 at more than 1.5 stops higher ISO.
If we get another 2/3 of a stop out of the D7100 over the D7000, I'll be elated. If we only get 1/3 stop of noise performance, I'll be equally happy, because the higher MP and better AF system will make the camera at least a stop more usable.
That said, there WILL be a learning curve with this new camera. The D7000 requires far better focus technique than the D90 to achieve sharp images. I expect something similar with the 24MP D7100.
When I was looking to buy the D7000 and reading various magazine reviews etc I noted the comment that to get the full benefit of the sensor would require the use of Nikon's professional lenses. If this is true does the increase to 24mp make this even more relevant? Also if this is the case then why put large sensors on on cameras lower down the range which are beginner/improver models?
As usual apologies in advance if I am talking rubbish.
>When I was looking to buy the D7000 and reading various >magazine reviews etc I noted the comment that to get the full >benefit of the sensor would require the use of Nikon's >professional lenses. If this is true does the increase to >24mp make this even more relevant? Also if this is the case >then why put large sensors on on cameras lower down the range >which are beginner/improver models? > >As usual apologies in advance if I am talking rubbish. > >Alan
MP's are an easy marketing tool, so pocket point & shoot cameras with tiny sensors were packed with 16mp years ago.
As to the lens aspect, yes, a more densely packed, properly executed sensor will reveal poor technique and lenses when images are viewed at 100%. However, Nikon is not the only manufacturer of tack sharp lenses. The resolution on my Tamron 17-50; Sigma 8-16; Sigma 150 macro, is phenomenal... The new Sigma 50-150/2.8 is said to be one of the sharpest zooms some reviewers have ever tested.
>As usual apologies in advance if I am talking rubbish.
Not rubbish at all, but perhaps some might view my reply as such!
>benefit of the sensor would require the use of Nikon's >professional lenses. If this is true does the increase to >24mp make this even more relevant?
Short answer, sort of. Long answer, the dirty little secret they don't tell you.
The D7100 has a pixel density of around 3.9 microns. System resolution is usually expressed in lp/mm (line pairs per millimeter). If I calculated correctly the pixel pitch would work out to about 128 LP/mm. For 35mm format the resolution of the average consumer zoom is about 60 to 80 lp/mm with excellent lenses at about 100 to 120 lp/mm. So yes to take full advantage of the 24MP, you need the best lenses.
But there's another wrinkle, diffraction. To get the most out of the 24MP, you'll need to shoot at fairly wide apertures. If we assume twice the density for resolving detail, then the D7100 would become diffraction limited approximatly above f/6.3 and the D7000 at above f/8.
>then why put large sensors on on cameras lower down the range >which are beginner/improver models?
There is a number of reasons why. One is to improve noise performance. I believe sensor technology used in the D7100 is fairly mature ( there is a new technology in the wings though that might increase sensor performance). So while the higher pixel density of the D7100 probably makes individual pixel performance a bit worse than the D7000, the pixels are smaller and less offensive in the print. Also if you down sample the 24 MP to 16 MP you average out the noise and the D7100 will provide better image quality then the D7000.
At least I think I got that all worked out correctly!
I understand diffraction limiting in general terms, but I wondered how much of an impact it would have in real prints. Say we took the same shot, with the same lens, on a D90, a D7000, and a D7100, which have different pixel densities and therefore different diffraction limits. If we shot all 3 at f11, which is probably just below where I expect the D90 to start suffering from diffraction, which camera would produce the best 12x18 print, after down-sampling the d7000 and d7100 shots. In other words, does diffraction more than make up for the extra pixels at small f-stops, or do the higher pixel densities still provide a net benefit?
You pose an interesting question for which I would say it depends on how you define quality.
First, I'd like to start by stating that I know just enough to make myself dangerous and get myself into trouble!
But I'll try my best!!
In regards to diffraction, there's two separate issues I believe. The affect on perceptive sharpness and ability to capture spatial frequency. We can correct for the lack of perceived sharpness caused by diffraction; but we cannot recover lost detail due to it.
In regards to perceptive sharpness it correlates to the amount of enlargement of the captured Airy Disk. So in that regard all three cameras are bound by the same amount of enlargement. For DX cameras and an 8x12 enlargement viewed at about 18 inches that is said to be around f/11.
As to spatial frequency, if the diffraction spot is the size of two pixels, it's still considered resolvable. It doesn't become limiting until the spot reaches three times the pixel pitch. As such by f/11 there probably isn't much difference in fine detail between the D7100 and D7000. But most likely a bit better than a D90. Of course all this assumes there is small enough detail in the scene to matter.
But in your poser, you suggested down sampling. So in essence you are throwing away any fine detail advantage in the higher pixel density camera, so in essence the result should be about the same. But in that case the significant difference would be in the SNR which was covered in an earlier post by Stan.
I understand that the D7100 would be diffraction limited at f/11, but the D90 wouldn't be. That's why I posed the question as a shot at f/11. In this case, would the D7100 shot show more detail than the D90 shot, or would diffraction cause enough blur to make them about equal, or make the D90 shot superior?
The higher resolution sensor will always record more (or finer) detail than the lower resolution sensor, even after the effects of diffraction become visible. I don't believe that there is any theoretical model that can tell us how much more.