I have (for the past 3 years plus) the D700 and am not planning on a trade for the D600/D610 at all. The D700 has less MP's, but the sensor is actually bigger than the D600. The 700 body is a semi-pro body and has many of the professional features, where as the 600/610 body is the "pro-summer" body and you have to go into menus to get things that you can get at a push of the button on the 700.
The weird thing is the D700 still contends somewhere between the 600 and the 800. Neither of these later bodies are a "direct replacement" to the D700.
I'm not sure what you're shooting (or actually, what you're wanting to shoot in the future), but the D700 is no slouch. My bet for what it's worth is, in the hands of someone knowing what they're doing (professional of very capable enthusiast, you'd see very little if any difference in images (even MP vs. MP) unless you're doing wall murals.
I own a D700 and am very happy with it. There's nothing wrong with a camera which produces images of 'lesser megapixels' (lower resolution). If you're not planning on making large prints (bigger than A3 size), there's really no need to have a camera which produces 24MP or more. The 12MP from a D700 is, in my opinion, more than enough. You'll also get more photos per memory card
As a side note, I've produced A3 prints taken with a 5MP Nikon camera. Results are very satisfactory.
Like the others, I think the D700 is still very capable amongst the latest Nikon offerings. I'd only consider disposing of it in favour of the more expensive top of the range professional level cameras
I bought the D700 when it first came out and it is still serving me well on all my professional assignments. I'm a fill the frame with your subject and composition type person so the D700's 12mps haven't been a limiting factor. It's enough resolution for publication and for making an excellent 16x20 print.
Love the D700 and I resisted the D600 for a long time, holding out for the D800. Last week, the price of the D600 finally got me and I bought one. Guess what..?
I could easily see the difference in IQ on the very first shoot. I fully intend to keep the D700 and use it as well, but if someone offered to trade it for the D600 even up, it would be gone...
I'm not saying that you can't get wonderful images with the D700. It's a camera I could be happy with the rest of my life.
But as much as you love your D700, just don't kid yourself (like I did...) that it's somehow better than the D600. It's not... JMO.
PS: The shutter sound in the D600 is almost silent compared to the D700. I've heard some folks have issues with the menus/controls, but I found it pretty similar and Nikon lets you reprogram most of the exterior buttons to almost anything.
I think there is a lot of other factors like the proven reliability of the D700. For me, I have no problems with the D700 and producing wonderful prints with my Espon 3880 printer, that's good enough for me.
There has been too much emphasis on IQ/MP but the real bottom line is that will it really make you a better photographer?
But really not really sold on digital age for the way it looks. I am in the process in going back into the film age, mostly for the b&w side. All I have to do get my darkroom back into action and get a good scanner. I bought a cheap F5 for 250, that would go with my FM2 and my Makina 67....and D700 for digital....that's all I really need for cameras.
I know film is extra work, but I think it's worth it for the end result. I don't have worry about the latest MP camera......but I must have to admit that the Nikon DF has got my attention somewhat!
Sat 16-Nov-13 09:21 PM | edited Sat 16-Nov-13 09:43 PM by Leonard62
I've made signal to noise measurements in RF systems, audio systems and video systems in my 40 years as an electrical engineer. They can be measured with a vacuum tube voltmeter, a spectrum analyzer or even an oscilloscope. What you do is feed the desired signal into the measuring device to set a reference point. You then remove the signal source and measure the remaining signal. In most cases it's noise. You can then compute the signal to noise, in db, by subtracting the remaining signal when the source is removed from the reference signal. The remaining signal or the remaining noise when the signal is removed is the noise floor.
edit: in a poor system when the noise floor is high the noise floor will overlap the desired signal and the system will have poor s/n. As new sensors are being developed, the noise floor at the different ISO settings is being improved and overall noise performance is improved.
A link below and Len's explanation. The main advantage of lowering the noise floor for a sensor is that it can retrieve more information out of the shadows. That allows for increased DR below middle gray and means you can think differently about exposure of a high contrast scene, especially when you are shooting RAW. A for instance - on my D600, I have found I can underexpose sunset scenes by at least 2 stops to retain all highlight data, then in Lightroom, recover the shadows for a noisless image. With my old D200, a two stop underexposure resulted in significant shadow noise and even banding.
>It is interesting that by looking at images without the in >camera noise reduction (i.e.. RAW), the difference in noise >appears a lot closer than you might think.
Actually it is closer. If you look at the Dxomark snr 18% measurement graph, the difference is a 1/2 stop between ISO 600 to 800 and a 3/4 stop between ISO 6400 and 12800. The noise floor drop at ISO 12800 is 2 db for the D600. One stop equals 3 db.
Just picked up an D700 3 weeks ago, I have to say...high iso performance is not that much better than my D7000. The only advantage of a D600 is probably the high iso performance, other than that, you will miss everything that D700 had to offer that can't be found on D600, i think the focusing module is a big turn down on d600/610.
Kevin, it is my understanding that the low-light performance of D700 vs D7000 is a full stop better. I skipped over D7000, so I base my opinion on the data I found online (http://snapsort.com/compare/Nikon_D700-vs-Nikon_D7000). I also checked the comparison on the above mentioned link: http://www.dpreview.com/reviews/nikon-d600/20 where I added D700 and D7000 to the comparison. I like to shift the comparison region towards a darker area (this is where you would typically use high ISO). Notice that when you are doing comparison in JPEG D700 and D7000 seam rather close indeed. But when you switch to RAW you can clearly see the difference. So it seams that D7000 is using more aggressive noise-reduction, which suggests that with proper post-processing of RAW file from D700 you can get a much better result than from D7000. My experience with D700 certainly confirms it. Vitaliy
>Two stops?! Where are you getting this data? > DxoMark. At each camera's base ISO a full two stop difference based on SNR, (that noise floor thing again). By ISO 6400, a stop. Then the D600 goes on its ISO 25600 limit.
"Extraordinary claims require extraordinary proof " - Carl Sagan
>>Two stops?! Where are you getting this data? >> >DxoMark. At each camera's base ISO a full two stop difference >based on SNR, (that noise floor thing again). By ISO 6400, a >stop. Then the D600 goes on its ISO 25600 limit. > >
David I think you're mixing up db with stops. At most there is 1 stop between the D600 and D700. 3 db is 1 stop. At ISO 200 the D600 is at 42db, the D700 is at 40 db. That's .67 of a stop.
Look at the chart again and tell me what you think. The D600's base ISO is 100, not 200. The measurement by DXO is 45.4 DB, or 14.24 stops. The D700 at ISO 200 is 41.2 db, or 12.15 stops.
After that, the D600 stays anywhere from .67 to nearly a stop ahead until it pushes into territory the D700 cannot go by having more DR at 25600 than the D700 at 12800. It is a noticeable improvement and then there is the noticeable resolution increase.
I am not denigrating the D700, it is a great camera, but the D600 has a next generation sensor, its AF system has been optimized for FX with f/8 on the central 11 af points, it has a 100% viewfinder, a shutter rating that is the same as the D700, an equivalent, (not identical), build quality, an extremely flexible dual SD card system, and a superior auto ISO functionality that matches shutter speed to focal length, (and that is customizable as well). It is more like the D800 than the D700, (I shot my brother's D800 extensively for a week back in June, while comparing the cameras closely).
I realize the difference in flash synch, the frame rate and some other features I never use like PC sync. That said, the D600 is no slouch.
Just a perspective,
"Extraordinary claims require extraordinary proof " - Carl Sagan
Thu 21-Nov-13 03:14 AM | edited Thu 21-Nov-13 03:15 AM by Leonard62
I agree with everything you just posted. However when I compare noise I reference the same ISO. I didn't overlook your comparison at base ISO but I felt it was misleading. In this thread the question was about high ISO performance. In that case you would compare the snr at like high ISO.
As you know from other threads that I will be getting the Df. It was my plan to sell the D700 to get the Df which has much of it's performance the same as the D600/610. But now I'm thinking I'll keep the D700. It's such a good camera as a backup to the Df.
>Another elementary question, please: what is the math behind >converting dB to stops? And conversely?
There is not a simple calculation for db to stops. Here is a paper from Stanford which explains the relationship. Page 13 is significant and is the bottom line where 3 db is equal to 1 stop comes from.
>Has anyone published a conversion table: dB to f-stops?
George, If you want to know the difference in f stops when comparing two snr levels as computed by DxOmark, you already have it. 1 db = 1/3 f stop, 3 db = 1 f stop.
However in the broader sense a db is not an absolute value but the difference between two signals measured under the same conditions. An absolute value would be qualified by adding a suffix to db such as dbu, dbm, and dbmv. In my field 1 mv = 0 dbmv and was measured across a 75Ω cable.
The Stanford lecture indicates that to optimize the S/N ratio, minimize the shutter speed (limited by motion blur) and then maximize the ISO (limited by saturation. The latter would imply a "full" histogram, which conforms to the usual recommendation. But, further, isn't there a statement, or at least a clear implication (I don't recall which), that the analog amplification degrades the S/N ration less than the subsequent digital amplification?
Do you agree?
And, for the D700, at what point then does the analog amplification cease and the digital amplification begin? -- and, is that point only a function of ISO, or are there other factors?
Finally: as to relative v. absolute: when DxO publishes a DR figure, is that relative to their measurement of the "noise floor"?
I hope someone can comment on the analog vs digital amplification switchover point for the D700.
Anyway, scoping out what speeds are available to freeze motion with the D700, I am sort of amused to note that NOWHERE does the Nikon D700 Manual specify what shutter speed settings are available! (At least, a search on the PDF did not find any. )
Sure, I can go pick up the camera But still, a clumsy oversight, I think.
As specified on page 429 of the manual, you have the option of 1/3-stop, 1/2-stop or 1-stop steps for the shutter speed. Is that what you meant?
If you choose 1-stop steps, then the shutter speed sequence goes 1/8000 - 1/4000 - 1/2000 - 1/1000 - 1/500 - 1/250 - 1/125 - 1/60 - 1/30 - 1/8 - 1/4 - 1/2 - 1 - 2 - 4 - 8 - 15 - 30. Going for 1/2-stop steps adds other values half-way in between those. These are the standard series, and aren't specific to a D700.
I sent a note to Prof. Levoy at Stanford, who gives the course the slides for which, re S/N, Len posted a link on this forum.
Prof. Levoy suggested the above-linked article re data as to when the camera's digital multiplication takesa over from the analog amplification. He said he had no data for Nikon, but had info rer Canon, so I asked for the article on Canon, in the hope that someone has similar data for the Nikon. The above-linked article is not mathematical (unlike the course slides), but is to the same practical effect.
Ok I'm confused. What is the relationship between dynamic range, which I suspect is what you are referring to, and noise? DR of D600 is 2EV more than D700, I understand that, and it can be useful. But I don't quite understand how that relates to noise, if it does at all.
> But I don't quite understand how that >relates to noise, if it does at all.
I believe it does.
Noise, for the most part, becomes very visible in the shadows, less so in midtones and even less so in highlights on any image, but especially if you underexpose for various reasons. The more you can drop into the shadows to extract clean detail, the more DR the camera has. Most of the improvements in DR for digital cameras have come from dropping the noise floor, which directly affects noise and DR.
So the D600 and the D700 can do an interesting thing, say, compared to my venerable D200. You can expose for the highlights in a high contrast scene and recover the shadows in post processing. This is assuming you are shooting RAW.
I stopped shooting HDR for landscape once I got my D600, because it has enough DR to allow me to get what I want in a single shot. I have not used the bracket feature once on this camera, whereas on my older cameras like my D200 and D70s, on landscape scenes, I used it all the time.
On the high ISO side, that same noise floor reduction means that the ambient noise of the system has been reduced and increases the DR as well. Its the same effect, ie, the better the SNR, the higher the ISO can be for useable shots. The obvious king right now is the D4.
"Extraordinary claims require extraordinary proof " - Carl Sagan
That's only (as I think David and Len have agreed above) when you measure them at their base ISO, which is 100 for the D600 and 200 for the D700. At equal ISO values and higher up the scale, the two are much closer.
Since the original topic is about high ISO image quality, I first though that when the +2EV improvement in noise was mentioned above, that you can shoot, say, at ISO 6400 with D600 and get a quality similar to ISO 1600 from D700. That is about the difference between my D200 and D700. In fact I found the difference between D200 and D700 to be closer to +3EV in noise.
If Nikon ever produces a non-flagship camera that improves high ISO noise performance of D700 by +2EV, I will be the first one to press the 'buy' button
>Ok I'm confused. What is the relationship between dynamic >range, which I suspect is what you are referring to, and >noise? DR of D600 is 2EV more than D700, I understand that, >and it can be useful. But I don't quite understand how that >relates to noise, if it does at all.
David's explanation is very good. While dynamic range and signal to noise are two different parameters, they are closely related. Actually dynamic range is the result of signal to noise performance. As snr improves so does DR.