More on usable dynamic range

After my previous post, I got a couple of questions on usable (or useful) dynamic range. To start with, the DxOlabs‘ version of dynamic range, the total range, is utterly useless. It measures the range up to where the signal/noise ratio is 1, i.e. the noise is as strong as the signal or image. That’s NOT a pretty image, it’s horrible. You’ve lost it long before that. That’s why I find the imaging-resource.com testing much more interesting. They test the dynamic range with different signal/noise ratios, i.e. how much less must the noise be than the signal/image for you to have an acceptable image. It seems that around 5-6 times stronger image signal than noise yields a nice image. Lower than that, you would not like to go.

In the imaging-resource.com tests (up until lately), they provide measurements for four signal/noise ratios: 1, 2, 4, and 10. Since we would rather have the dynamic range evaluated at about a rate of 5-6, I use the average of the measurements for the ratios 4 and 10. Since ((1/4 + 1/10) / 2) is around 1/5.7, that will give us the most reasonable figure on usable dynamic range. I have compiled that figure from many of the tests on imaging-resource.com and rounded the results to the nearest 0.2 f-stops (reasonable measurement error margin), so the table is courtesy of them. The measurements are all at native (base) ISO, which ranges from 80 to 200 on the cameras tested. The table shows the figures from RAW image files converted using Adobe Camera Raw (the version available at the test of each camera).

Nikon D5200 11.4
Nikon Coolpix A 11.4
Nikon D600 11.2
Ricoh GR 11.0
Nikon D4 10.8
Nikon D7100 10.8
Nikon D7000 10.8
Nikon D5100 10.8
Pentax K-5 10.8
Nikon D800 10.6
Sony A99 10.4
Nikon D3X 10.4
Nikon D5000 10.2
Nikon D3S 10.2
Nikon D700 10.0
Sony A33 10.0
Sony A900 10.0
Pentax K-x 10.0
Nikon D90 10.0
Fujifilm S3 Pro 9.8
Sony A55 9.8
Sony NEX-5 9.6
Nikon D300S 9.6
Nikon D40x 9.6
Sony A230 9.6
Canon 1D Mark IV 9.6
Sony NEX-3 9.4
Sony A330 9.4
Canon 5D Mark II 9.4
Canon EOS-1Ds Mark III 9,4
Nikon D3 9.4
Canon EOS-1D Mark III 9,4
Sony A380 9.4
Nikon D3000 9.4
Pentax K20D 9.2
Nikon D300 9.2
Sony A200 9.2
Nikon D40 9.0
Nikon D60 9.0
Sony A100 9.0
Pentax K100D 8.8
Pentax K200D 8.8
Canon 7D 8.8
Canon EOS 50D 8.8
Canon EOS-1Ds Mark II 8.8
Nikon D3100 8.8
Sony A350 8.8
Pentax K10D 8.8
Olympus E-P2 8.6
Olympus E-P1 8.6
Canon T2i 8.6
Canon Rebel XS 8.6
Canon EOS 5D Mk III 8.6
Samsung NX10 8.6
Panasonic DMC-GH1 8.6
Canon EOS 5D 8.6
Canon Digital Rebel XTi 8.6
Canon EOS 40D 8.4
Olympus E-3 8.4
Nikon D80 8.4
Canon Rebel T1i 8.4
Pentax K-7 8.4
Canon Rebel XSi 8.4
Nikon D200 8.2
Canon 1D Mark IV 8.2
Olympus E-500 8.2
Nikon D2Xs 8.2
Olympus E-PL1 8.0
Panasonic DMC-G2 8.0
Olympus E-510 8.0
Olympus E-420 8.0
Panasonic DMC-G1 8.0
Olympus E-520 8.0
Panasonic DMC-GF1 7.8
Panasonic DMC-L10 7.6
Olympus E-410 7.6

The Nikon D70, the subject of this blog, is not in the table. The closest we come is the D40 which uses the same sensor (the D40, D50 and D70/D70s share the same ICX453 sensor while the older D100 has the ICX413 sensor, both from Sony). At 9.0 f-stops, the usable dynamic range of the D40/D50/D70 is ok but not very good. To gain some considerable range, one need to look at full stop gains. At around 10 f-stops, we find the previous generation of Nikon full-frames as well as the D90/D5000/D300s using the Sony IMX038 sensor (the D300 uses the older IMX021 sensor and is rather close to the D70). To gain one more stop, up to around 11, you need to consider the latest sensors such as D7000/D5100 (Sony IMX071 sensor), D4, or D7100/D5200 (Toshiba 5051 sensor). The D5300 is unfortunately not tested in this way (but has the same Toshiba 5051 sensor as D7100/D5200) and the D800 is lagging behind (that design had other priorities).

Note that the measurements in the table are real-life figures. That includes not only the sensor but read-out electronics and in-camera image processing. Yes, that’s right – processing. There is a lot of in-camera processing even for “raw” data, also at base ISO. Even more at higher ISO, otherwise your images, especially CMOS, would look horrible. Further, the figures are after conversion in Adobe Camera Raw since you as the end-user must make a raw conversion in order to access the images.

As I wrote in a previous post, the most notable fact is that full-frame cameras are not ahead of half-frame (DX or APS-C) ones from the same era. Full-frame has a lot of advantages, such as high-ISO noise being lower, but dynamic range at base ISO is not one of them.

These are RAW file data, figures from JPEG image files show much less dynamic range. Typically around 1.2-1.5 stops difference, which is one good reason why you should shoot in RAW, at least in scenes with a high natural contrast. The absolutely best of the latest DSLRs have the same usable dynamic range in JPEG (around 9 f-stops) as the D70 has in RAW. Another way of saying this is that if you are a D70 JPEG shooter, you would gain the same in real-life dynamic range from either upgrading to one of the latest-and-greatest Nikon DSLRs or simply switching to using RAW on your D70.

Usable dynamic range

Don’t confuse total dynamic range with usable dynamic range. The total range is what is measured by test rigs as the range between lightest highlight and darkest points lost in noise. The useful range is rather how low you can go until the noise becomes objectionable, i.e. anything further is not useful. I think that imaging-resource is among the best testers of useful dynamic range, so let me borrow a paragraph from them:

A key parameter in a digital camera is its Dynamic Range, the range of brightness that can be faithfully recorded. At the upper end of the tonal scale, dynamic range is dictated by the point at which the RGB data “saturates” at values of 255, 255, 255. At the lower end of the tonal scale, dynamic range is determined by the point at which there ceases to be any useful difference between adjacent tonal steps. Note the use of the qualifier “useful” in there: While it’s tempting to evaluate dynamic range as the maximum number of tonal steps that can be discerned at all, that measure of dynamic range has very little relevance to real-world photography. What we care about as photographers is how much detail we can pull out of the shadows before image noise becomes too objectionable. […] What makes most sense then, is to specify useful dynamic range in terms of the point at which image noise reaches some agreed-upon threshold. (from http://www.imaging-resource.com/PRODS/K5/K5IMATEST.HTM)

In the same article, they go on to explain how to define usable range and how to test it. Thereafter comes the really interesting part. I have chosen an older test, that of the Pentax K-5, since it contains the Nikon D70 – the topic of this blog – in its reference tables.

The useful dynamic range table in that article is a REALLY interesting read. It explains why countless folklore stories of superior dynamic range of full-frame cameras never seem to materialize in real-world images and real-world pulling of shadows. The Nikon D3S and D3x have worse useful (real-world) dynamic range than the D7000. At least 1/2 stop and even up to a full stop worse. The D3 and D700 are even further behind. The latter ones are equal to the good old D90. Make sure you swing by the URL above and have a look at that revealing table, look in columns 0.25 and 0.1, no others (the table is sorted by the 0.1 f-stop column).

In the same table, you can see the D70 fare pretty badly. One unsatisfying thing about the table is that only the JPEG range is shown for the D70 while for many others, both JPEG and RAW are shown. RAW beats JPEG by a wide margin, and the most indicative figure for the D70’s RAW performance ought to be the D40 figure of 8.30 (beating the D3100 and being close to the D300). The table underlines the well-known truth that you should shoot RAW at least in scenes with higher dynamic range and that you can pull a lot more out of the sensor than JPEG engines can, especially a ten year old one. Let a modern RAW converter such as Capture NX-2 or Lightroom 5 do the work and like magic you have bypassed ten year old software in the D70 camera body. Another unsatisfying fact is that the table lists the camera bodies at base ISO, which for the D70 is ISO 200 making it look much worse than some of its rivals. It’s a full stop behind the full-frames D3S, D3x and D700, and two stops behind the D7000 which is the usability king in this respect (confirming real-life experiences). Nowadays, the site does not publish a table, only individual scores. While the D800 did only fairly well (above D3x but below D7000), the D600 scored the best so far by 1/3 stop. The D4 equalled the D7000 as did the D7100. This all underscores the truth that you should choose half- or full-frame firstly based on your desires regarding depth-of-field, and secondly on your interest in high-ISO shooting. Dynamic range is not a factor in this.

Speaking of the D90, even better figures are achieved by its slightly younger and much lighter sibling, the D5000. An overlooked gem that you can pick up for $150, often with little mileage. It is the modern counterpart of the Nikon FM, FM2 and FM3, and of course the FE and FE2. Small, light, excellent. Pair it with the smallest 18-55 kit lens and a couple of small primes and you have a fantastic real DSLR travel and take-anywhere kit. The prime selection is actually quite simple: Nikkor DX 35 f/1.8 G and Nikkor 85 f/1.8 G. Both truly outstanding in image quality and fairly small. If you want something wider, that’s a little bit problematic. I suggest you stay with the Nikkor 18-55 f/3.5-5.6 G VR. It’s good, especially at wider focal lenghts, but not outstanding. Or go with the Voigtlander 20 f/3.5 SL-II. Better image quality than the Nikkors, both autofocus and manual focus versions. Especially good on DX cameras. It has manual focus but the D5000 recognizes the chip in the lens and thus it’s able to meter automatically in the normal way. Very small and light, plus quite easy to focus manually since it is an ultrawide lens. Huge depth of field on a DX camera.

Another notable entry in the table is the D3100 which I own and write about in the sensor blog entry. Its old Toshiba CMOS fares worse than the CCDs of the D40 and D60. This is also consistent with how it works in practise, pulling shadows in some photo editor. A final note is that the table confirms the long held view that both the D300 and the D300s are inferior to the D90 and D5000 when it comes to useful dynamic range. The virtues of the D300 bodies rather lie in durability and handling, among other things.

The Canon bodies are behind the equivalent Sony equipped bodies (most Nikons, Sony, Pentax). The Toshiba sensor of the D3100 is behind similar Sony sensors. The old Fuji S3 Pro is impressively placed before all Canon bodies and not that much behind the best Nikon ones, even if the gap up to the D7000 is considerable.

It’s a surprisingly little discussed fact that DxOmark‘s dynamic range measurements are made using total dynamic range instead of useful range. I would love to see the imaging-resouce table plotted in a 2D graph for every ISO value. I am sure such a diagram would correspond more closely to how different cameras are percieved and felt during post processing. But as of lately, imaging-resource have given up their measurements because of their test instruments being surpassed by recent sensor developments. Unfortunately, they now use DxOlab’s figures and charts that only correspond to the least informative column of imaging-resouce’s tables, the leftmost one showing unusable (total) dynamic range.

Perhaps this rather long entry sheds some light on why my sensor blog entry is true even though a lot of people wish it wasn’t. Or perhaps it does not…in which case I have to try and rephrase my thoughts once more…

Nikon D70 summary

Once again welcome to this Nikon D70 blog. The last post will appear first so I will make a short summary here. This set of posts shows that the Nikon D70 is one of the best DSLR cameras ever produced when judged in its entirety. The posts were written over some time and are now republished in this format. They go through ergonomics, functionality and sensor/image quality. Several prejudices are defeated. The main message is that image creation is much more fun and important than pixel peeping or tech spec peeping will ever be, and image creation is indeed supported in an excellent way by the D70. If you are interested in photography, read on.

Exposure and tone curves

There are two different ways of creating image files with the D70 (or any other DSLR). Either you shoot in JPEG or you shoot in RAW/NEF. (Yes, you can set the camera up to produce both kinds of files at the same time.) A good camera should support both ways equally since they are for different purposes and occasions. If you shoot in JPEG, the camera does the image processing, converting pixel read-outs from the sensor to an image file. This involves operations like interpolation, demosaicing, white balancing, image correction, and more.

For one thing, you want the image to be exposed correctly. Nikon cameras, especially the early ones, have been accused of underexposing. This is of course not an error but a design choice. Among the serious amateur cameras, it was especially evident on the Nikon D100 which tends to underexpose down to −1/2 stop. The D70 rather tends to underexpose about −1/3 stop on average. This is inconvenient for JPEG shooters since the images often wind up darker than expected and you have to constantly apply exposure compensation to deal with it. This also leads to not paying attention to the compensation symbol +/− which would otherwise warn you about non-standard settings. For JPEG, this is a poor design choice (not a manufacturing error or malfunction). On the other hand, if you shoot RAW/NEF, the meter’s behavior is made for you. Underexposing preserves highlights which can be recovered in post-processing. You can bring the skies, for example, up in post-processing and recover details that would have been lost in a standard exposure. In 2000, Nikon were newbies in digital and worried too much about lost highlights. They designed the D100 with too big a margin. They shrunk the margin with the D70 but it is still there. I would say they got it right for RAW on the D70 but not for JPEG. The real mistake made was not making the exposure behavior configurable with each image file format. There are times when you want to shoot one format or the other. It’s not that one format is “better” than the other. They are different and made for different purposes. I would like more standard exposure when shooting JPEG but keeping the underexposure behavior for RAW/NEF.

So what did people do? Some clever people tried to use a function in the D70 called a custom tone curve to obtain an overexposuring offset to the inherent underexposure. In essence, a custom tone curve is a new contrast curve that replaces the in-camera contrast curves. At first, the idea might seem good. Back in 2004-2005, there was a number of curves being created by those people and discussed on several web forums. After some initial fuzz, they fell into oblivion because they didn’t work as intended. A longer discussion on tone curves (primarily for the D100) is found here.

There are two uses for a tone curve, the first being to offset exposure tendencies and possibly try to change the dynamic range as you go along. This is unfortunately bound to fail; the D70 does a better job than you can in automatically selecting contrast according to the image and the exposure. Thus, the only way to handle the JPEG underexposure tendencies of the D70 (and others) is the good old exposure compensation button and dial. Even toy cameras like the D3100 have that accessible through a button on the top close to the shutter release button.

The other use is much more interesting when you want to shoot JPEG and have reasonably ready images out of the camera that requires a minimum of post-processing. The D70 (and some other Nikons) have been accused of having a “dull” rendering. I would rather call it accurate, but it is not always that you want a very accurate rendering. Sometimes, you rather want a more colorful, vivid “slide film like” rendering, although not as extreme as for example Fuji Velvia which I personally find too bold and punchy. There are a lot of options possible on the D70 when it comes to image rendering. These options are called Image Optimization by Nikon and are found in the menu system. While the settings available are not bad, they are a bit conservative. The Vivid option, for example, is not very vivid. This led people to develop tone curves for this second purpose, and they can again be found on the web. Here is an example of a repository for curves for both purposes. Many of the curves available at that time mixed the two purposes up – adjusting exposure and changing the tone. Others went, in my opinion, either too far (for example trying to mimic Velvia) or missed the target. Another misconception at the time was that it should be possible to find a curve that fitted all shooting situations. There will never be such a curve.

All this led me to develop a curve with the single purpose of changing the tone balance, not compensating the exposure system of the D70. The neutral rendering of the D70, being an asset in some images, needs to be supplemented in some situations with a curve that feels more like slide film without going to Velvia extremes. The curve has been developed over many years and is, to my knowledge, the only curve still being maintained today. The curve has been derived from thousands of images of all kinds. In my view, it works well with everything from nature to street, from buildings to ants, from cars to ships, but not for people, products, or paintings that require either duller or as neutral rendering as possible. For the latter ones, it is easy to select other image options on the D70 and still keep the custom tone curve for later. The curve is called D70 Chrome and is available for download.

If you have a D70, try it and enjoy. The download to the camera requires Nikon software (Camera Control Pro) and the USB cable that came with the camera. You need not buy Camera Control Pro, the trial version is enough. You are given a 30-day trial period of which you need a couple of minutes. Once downloaded to the camera, the tone curve stays there and you need only to activate it when you want to use it just like any other, built-in tone curve.This is the way to download it from a PC:

1. Download Camera Control Pro 2 from a Nikon website to your PC. Any version from 2.5 up to 2.11 will do. From 2.12, the D70 (along with the D2-series, D200, D100, D80, D50, and D40) are no longer supported. Note that you cannot have multiple versions installed, such as 2.11 for your D70 and 2.12 for your D600 (which is only supported from 2.12 onwards). A catch-22 if you have an older camera (such as the D70 or D200) and a newer (such as the D600 or upcoming models).

2. Install Camera Control Pro and accept the trial period. If you have Windows 7, make sure the application runs in Vista SP2 mode, or you might get problems with connecting to the camera. Drivers are not fully compatible with Windows 7.

3. Restart the PC to have all drivers installed properly.

4. Download the D70 Chrome custom tone curve and save it in a directory on your PC.

5. Before connecting the D70 to the PC, make sure you use the PTP protocol. You select PTP mode under USB in the tools menu on the D70. This is a two-way protocol that allows the PC to actively talk to the camera, not only receive files as you can do also in Mass Storage mode.

6. Follow these instructions. Never mind the Japanese, your dialogue windows will look the same except for the text.

7. Enable the custom tone curve in Shooting menu > Optimize Image > Custom > Tone comp. > Custom. After this selection, you MUST select Done or your setting will not have any effect! Enjoy your shooting.

Sensors in deep

Let’s assume for a moment that you mostly agree with me on the ergonomics of the D70. Quite a few people actually do. Further assume that you mostly agree with me on the functionality of the D70. Again, quite a few people actually do. Sure, we all miss some function, but that is true of all cameras. The D70 fills the spec of the serious amateur surprisingly well. Nikon learned a lot in designing the D1 series and the D100 which they benefit from all the way until today. The learning curve was much steeper back then since digital was relatively new. Not that much happens functional-wise today. Finally, assume that you find the body housing, the grip, the mirror, the prism, etc to be reasonably ok. Sure, the finders of FX cameras are nicer but their weight offset is not.

Now we come to the last factor, the sensor. Here, most of you agree on the sensor of the D70 being outdated. “Everyone knows” that sensors go out of fashion every two to three years. Last generation’s sensor is today’s backyard technology. Some even claim that the best pocket cameras of today can match “old” DSLR sensors like the D70’s. In the film days, models lasted ten years or more. The Nikon F was introduced in 1959, its successor the F2 in 1971, and then again its successor the F3 in 1980. You could use any kind of film (slide or negative, color or b/w, any brand) in any camera body, making updating your equipment a matter of ergonomics and functionality. In the digital era, the sensor is what the film was, but it is fixed. Thus, to the delight of the manufacturers, it seems that you must replace your camera ever so often since sensor progress is so swift. Or is it?

Let’s turn to one of the premier and most well-known sensor labs, DxO, and their sensor (and lens) testing facility DxOMark.  There, you can compare camera sensors in an interactive fashion. Comparisons can be made in two ways that relates to two different aspects of sensor quality, both being important. First, you can compare pixels to see how far the development of individual pixels has come from 2004 until today. Second, you can compare sensor areas, where sensors with higher pixel density will have higher readings since more pixels can contribute to the results. In both comparisons, you compare signal-to-noise ratio, dynamic range, tonal range, and color sensitivity. DxOMark has far more elaborate explanations of these concepts, but essentially, signal-to-noise ratio is how much noise there is in the image, dynamic range is how much lightness and darkness can be accommodated in an image at the same time, tonal range is how many different shades of gray the image can contain, and color sensitivity is the number of reliably distinguishable colors in an image.

Beginning with comparing pixels, the D70 is compared to the newest and most high-tech Nikon at the time of writing, the D800, and the current Leica flagship M9. Looking first at the signal-to-noise ratio, the three sensors are almost indistinguishable. The D800 has a wider range of ISOs, but for any given ISO the results are the same. The D70 performs as well as the two rivals. Noticeable is that the “purist” M9 does a lot of noise smoothing behind the curtain, smoothing that is impossible to turn off. Only having used the M8, I was not aware of that.

Next, looking at dynamic range, the D800 stands out with 2 stops better performance, while the D70 is almost exactly as good as the M9. Which is not that bad since the M9 is considered one of the top cameras which satisfies very high demands.

Continuing with tonal range, it is indistinguishable between the three sensors. Pixels from the three cameras can contain equally many different shades of gray.

Finally, color sensitivity is almost equal, with D70 and D800 being a bit better than the M9. The number of reliably distinguishable colors in pixels from the three sensors is close.

To sum up the pixel comparisons, all cameras perform almost equally except that the D800 has a better dynamic range. The D70 and the M9 are practically equal on all grounds.

Moving on to comparing sensor areas, we compare the D70 to the Leica M8 instead of the M9 since it has about the same sensor size as the D70, and to one of the best pocket cameras, Canon S100, since it is sometimes claimed that today’s pocket cameras have caught up with eight year old DSLRs like the D70. Looking first at signal-to-noise ratio, the M8 is slightly better than the D70 but the S100 is far behind. The difference between the M8 and the D70 is about 1/2 stop while there is a more than 2 stops distance to the S100.

Continuing with dynamic range, the three sensors are fairly equal. The M8 is slightly better at lower ISOs, and the S100 compares surprisingly well to its rivals. Almost equally many levels of light and darkness can be accommodated in images from the three sensors.

Next, looking at tonal range, it is again fairly equal between the D70 and the M8. Images from the two cameras can contain about the same different shades of gray. The S100, though, is almost two stops behind.

Finally, color sensitivity is almost equal between the D70 and the M8. Images from the S100 will suffer a lot from a lesser number of reliably distinguishable colors than the former two.

To sum up the area comparisons, the D70 and the M8 perform almost equally while the S100 pocket camera is far behind. And to sum it all up, the D70 pixels do as well as a new top-of-the line DSLR except in dynamic range, where it lags behind the D800, although it performs equally to the Leica M9. The D70 sensor area does almost as well as the Leica M8 and easily outperforms one of the most highly acclaimed pocket cameras. Through field use of the D70, M8, and S100, I can safely confirm these scientific lab measurements in real shootings. I definitely miss some dynamic range in the D70 compared to, say, a D3x, but other than that, the D70 sensor is doing very well.

The most common reaction I get to all this is disbelief. Can the D70 really measure up to these modern rivals the way it is displayed here? Surely, all the hype about newer and newer sensors cannot be untrue? Surely, the manufactures are not taking us for a ride? Well, the scientific results from DxO tell something else.

As I say in other places too, don’t take my word for it. Listen to others that spend much more time and effort in finding things out. DxO is an organization I find credible and objective. But they are cautious, too. They have a popularized simple overall score that is derived only from some extreme values of their measurements. This overall score greatly amplifies small differences to make it more “interesting” and in line with expectations. The way the score is calculated is not published, so we cannot know exactly how it works in detail, but it’s obvious that is works more like the common perception of sensor development. Thus, the Nikon D800 and D800E receive the highest scores ever, 95 and 96 respectively. The Leica M9 receives 69 and the Nikon D3100 67. The D70 obtains a score of 50 which is exactly what the Canon S100 also does. But it is evident from the graphs above and from shooting them that the S100 is clearly inferior to the D70 in terms of image quality. Thus, the DxOmark measurements are very interesting but their popularized overall scores are not. Rather, they are misleading.

But again, although DxOMark is a fascinating site, I wish for you to spend more time creating great images than comparing camera sensors. My main point here is merely that you need not worry about your equipment aging, just go out and have some fun. Nobody will ever be able to tell whether you used a D70 or something newer (unless they peek into your EXIF data).

All graphics come from the DxOMark website, thus their copyright.

Megapixels and diffraction

The D70 is a 6 megapixel (MP) camera as was the norm some years ago. Since then, the manufacturers have nourished the megapixel myth – that pixel count correlates more or less exactly to image quality. On the web, many have written about the myth. Ken Rockwell, who I find a bit flamboyant at times, has written a page on the myth which also contains some fun pointers to a megapixel street test and a completely bogus site that fuels the megapixel myth by an untrue interactive zoom function. The differences in image quality from differences in pixel density are of course nowhere like what they try to imply.

Every pocket camera nowadays has 10 MP or more, phone cameras have up to 5 MP. But their image quality is not up to the standard of any modern DSLR like the D70 or newer. It has to do with a number of factors, photo site size (pixel size) being one of the most important.

The concept of picture element is also being misused in the manufacturers’ world. Pixels are not interesting, picture elements are. It takes four pixels in a Bayer arrangement (two green, one red, one blue) to make up a true color picture element. Thus, a 12 MP camera using the Bayer pattern (which most sensors do) does have 6 million green pixels, 3 million red and 3 million blue. This makes up 3 million true color picture elements, but interpolation and demosaicing brings this up approximately to around 6 million. If you shoot JPEG, this processing is done in-camera and you are totally dependent on the manufacturer’s algorithmic skills. Different manufacturers are differently skilled, fortunately Nikon are top-of-the-form. If you shoot RAW (NEF), you do the processing outside of the camera which widens your choice of algorithms

There is a trade-off between pixel count, resolution, and image quality. At first, it seems that more pixels should be attained at any cost. And it is true that more pixels could bring more resolution. But it also brings a number of drawbacks. First, more pixels will increase the file size of your images. This is the easiest drawback to live with; it will eat up more disk space and require more primary memory and processing power for your image handling. Second, more pixels on the same sensor area will decrease the size of each pixel, making it more susceptible to noise. With fewer photons captured, natural variance plays a larger role, which gives you noise. Since of course this noise can be interpolated away, you will never get worse image quality from more pixels on the same area, but not as much better as you thought (and the manufacturer seemed to promise). Third, diffraction limits your usable apertures considerably.

Of these, diffraction is probably the least understood one. There is a nice website that explains diffraction much better than I have the time to. On the new Nikon D800, diffraction limits the image quality of any lens at f/8 and smaller apertures. Which is the majority of available apertures for most lenses! And since most lenses perform at their best stopped down a couple of steps, many lenses are unable to perform their best at any aperture on a D800. Nikon is not very keen on telling you this. For a D70, diffraction starts hurting at f/16 while for cameras with the same sensor size, more pixels lead to more diffraction. The D90 is affected at f/11 and the D7000, having a very similar pixel density to D800, is affected already at f/8. Thus, while the D90 and the D7000 in theory have clearly better image resolution than the D70, this is diminished in practice by diffraction. Diffraction does not ruin the image but eats into the advantage of high pixel counts. And it does not show in the reports on new sensors.

Nikon i-TTL flash system

In 2003, Nikon introduced the CLS flash control system – and there was light. CLS stands for the Creative Lightning System, and it is indeed a very good system. It brought Nikon in front of everyone and everything, until 2009 when a modification of i-TTL (a part of CLS) made the concept take a turn for the worse.

Rewinding a bit to film SLRs, Nikon had an – at the time – good TTL flash control system, that measured the light from the flash as it bounced off the film during exposure. TTL means through-the-lens, as opposed to measuring the general light level attained by firing a flash (which is what simpler flashes do). Nikon’s system was good, but for example Minolta had an as good system back then.

Then, starting in 1999 with the launch of Nikon’s first digital SLR (appropriately named D1), Nikon users were surprised at the modifications made to the flash system. The digital sensor and the anti-aliasing filter are more shiny and do not reflect light in the same way as 35 mm film. Thus, Nikon had to modify the TTL flash system as it was no longer meaningful to measure the exposure by the light reflected. Instead, Nikon had to rely on a set of pre-flashes to determine how much flash is needed. The new system was called D-TTL, as in D for digital. The pre-flashes are bounced off the focal plane shutter instead of off the film. To correct for the light in the scene, the shutter blades on D-TTL cameras are painted gray as a gray-scale card. This was an obviously kludgy solution, and Nikon knew early on that they had to design a new system for digital, not merely patching the old one. Another sign of this is the set of flashes that accompanied the D-TTL system. There were only three flashes ever produced that were compatible with D-TTL and they soon became obsolete as the new i-TTL system was quickly launched. To add insult to injury, Nikon decided that new cameras (except for the D2 series) would not be backward compatible with flashes only a couple of years old.

The new system called i-TTL (i for intelligent, implying that the old D-TTL system was indeed unintelligently conceived) was intended for the Nikon D100 in 2002 but was not ready in time for its launch. Thus, the D100 was left with the old dysfunctional D-TTL system. The first camera to incorporate the i-TTL system was instead the D2H in 2003 and the SB-800 Speedlight i-TTL flash was introduced with it. The D70 of 2004 was the first of the non-professional Nikon cameras to have the i-TTL functionality.

The i-TTL functionality was state-of-the-art and was essentially unmodified until 2009 when some engineers back at Nikon decided to “improve” the functionality without any notice to the user community. The “new i-TTL” system tends to choose extremely high ISO under the auto-ISO option when it is not called for to make a correct exposure. The reasoning behind the change is that as cameras became better at handling high ISOs, this should be used to make images have more exposed backgrounds. While this is not wrong as an observation, the mistake made was to force this into the functionality as a non-option. In the “original i-TTL” system, you could crank up baseline ISO for the auto-ISO option manually whenever you felt this was called for depending on the image effect you were after. You are in the driver’s seat, especially if you have a camera where the ISO is assigned to a button like the D70. In the “new i-TTL” system, this is done for you in a way you cannot anticipate or control. You have become a passenger in a flash unit driven by algorithms you cannot understand. A black box of lightning. A much longer argument against “new i-TTL” is made in this blog post. For example, owners of the Nikon D300s find themselves with a completely different i-TTL than those of the otherwise very similar D300. Those poor ones that have one of each are, to say the least, confused. What Nikon did was adapt the i-TTL behavior to novice users, trying to save some of their need to think. But they did this on all models, from toys like the D3100 to cameras targeted at advances users like the D300s. Big mistake. What they should have done? 1. either applied it only to toy cameras, or 2. better yet, made it configurable. “Original i-TTL” vs. “New unpredictable i-TTL”.

This is not a lecture on flash lightning, there are a number of good tutorials out there. Essentially, you must learn to master the balance between two kinds of light sources, flash and ambient. These can be of different luminance and color, even mixed. They have differing direction, duration, etc. But in all this, you want your camera to help you. The original i-TTL system does an excellent job of doing that provided you know how it works. Just like ordinary (non-flash) exposure, there are two modes of operation with flash. In non-flash exposure, you either use the PSAM modes where you, even though you could have some automation, it can be overridden since you are in the driver’s seat. Or you use the preconceived scene and green auto modes (scene modes are called vari-programs in the D70). The scene/green modes contain ready-made settings for exposure. These modes help inexperienced users to set the camera up and you are the passenger.

The same goes for flash exposure. Apart from the manual mode, there are two TTL modes. This is not well understood by most i-TTL users. Either you use Standard-TTL mode where you are in control and can apply corrections to the flash values. If you want -1EV for the flash compensation, it has the intended effect since you are the driver. Or you use TTL-BL (where BL stands for Balanced Light) where the camera logic does what it thinks is best for you regardless of what you asked for. Sometimes your corrections are applied, sometimes not. You are a passenger. This is unknown to many and explains some reports on erratic flash exposure by some Nikon DSLR users. It’s not an error but a design choice. One that has been poorly communicated, though. You need to know how you switch between the Standard-TTL and TTL-BL modes. The selection is implicitly done by your choice of metering and automation modes. If you select M = manual exposure or if you select Spot as your metering mode, Standard-TTL is selected for you. Otherwise, TTL-BL is selected for you. You cannot select the mode explicitly. This is of course a mistake by Nikon. It should have been selectable, just as exposure mode is.

To sum up: if you want control, use Standard-TTL mode. If you want to be a passenger (and sometimes you want), use TTL-BL mode. Make sure you know how to switch modes. If you want to be in control of ISO baseline values when using flash automation, stick with Nikon cameras designed prior to 2009 that have the original i-TTL system. If you don’t want to be in control, just being a passenger and accept what’s being handed to you, then you could be as happy with the new i-TTL system. This is much more important than you might realize at first. You should have the option to use flash as fill-flash in daylight scenes all the time, and that option is severely limited in the “new i-TTL” since you must go all manual to have control over the ISO value. In the “original i-TTL” you simple stay in automatic Standard-TTL and select baseline ISO for auto-ISO manually. I very much prefer the “original i-TTL” of the D70. Apart from TTL mode selection, the D70 is flash ergonomics at its best.