camcorders

Nikon D90 DSLR Camera Review

If you haven\'t heard already, the Nikon D90 is the world\'s first digital SLR capable of recording video—which is why we\'re reviewing it on our camcorder site. The video feature has grabbed a lot of attention in the gadgetry world simply because it represents an innovation in the industry. It has excited videographers looking for a way to use multiple lenses, and it has piqued the interest of manual control enthusiasts who yearn for the ability to manipulate and adjust every last setting.

October 24, 2008

Our editors review and recommend products to help you buy the stuff you need. If you make a purchase by clicking one of our links, we may earn a small share of the revenue. Our picks and opinions are independent from any business incentives.

Performance

Video Performance* (7.75)*

Being the first digital SLR to record video, the Nikon D90 has garnered an impressive amount of buzz. Besides having the ability to change lenses, another big attraction of the D90 is its large CMOS sensor in comparison to consumer camcorders. The D90 boasts a huge 23.6mm x 15.8mm sensor, which converts to a 1.12-inch diagonal sensor length. In comparison, this is more than twice the size of the 1/1.8-inch sensor found in the Samsung SC-HMX20, which currently houses the largest CMOS sensor in a consumer camcorder. It's completely normal to expect larger sensors in DSLR cameras, and it's important to recognize that sensor size is really only one small aspect of a camcorder's video performance. Many other factors have a large affect on overall video quality.

Unfortunately, the Nikon D90's highest quality setting records video at 1280 x 720 rather than at 1920 x 1080, which any quality HD camcorder is capable of. The second DSLR to record video, the Canon EOS 5D Mark II, is capable of recording video at 1920 x 1080 (it also retails for $1700 more than the Nikon D90). The D90 also has the ability to record in both NTSC and PAL settings, with the option located in the Setup Menu. PAL and NTSC are two different encoding systems, with NTSC being the standard in North America (as well as Japan, parts of South America and other areas) and PAL being the standard in most of Europe, Asia, and Australia. We did some recording in both settings and found very little quality differences between the two, but it's a good option to have if you plan on using the camera overseas. Unless noted, all of our testing was done in NTSC.

The Nikon D90 records video at 24 frames per second (fps), which is different than the usual 29.97 fps rate that most camcorders utilize. This gives the footage a slower, more cinematic look (as film traditionally captures at 24 fps). This slower frame rate is intriguing for many users who desire video that captures motion in a different manner than traditional video. Many camcorders, specifically models from Canon and Panasonic, offer different frame rate modes in an attempt to simulate the 24 fps style.

With these unique specs, the D90 was often able to produce impressive results (especially in moderately low light), but it really never came close to matching the quality a true HD camcorder is capable of.

Nikon_D90_3000lux_auto_NTSC_web.jpg

Nikon D90 at 3000 lux in Auto mode

Samsung_SC-HMX20_3000_Lux_Auto_60i_web.jpg

The Samsung SC-HMX20 at 3000 lux in Auto mode

Panasonic_HDC-SD100_3000_lux_auto_web.jpg

The Panasonic HDC-SD100 at 3000 lux in Auto Mode

Beginning our testing with the D90, we shot a DSC Labs Chroma DuMonde chart at an even 3000 lux, which is very bright light. We then pulled frames from that footage and compared them to similar images captured from other camcorders that we've put under identical testing. In our bright light testing the Nikon D90 produced a decent image, although it didn't come close to matching the sharpness of consumer HD camcorders. A fact mainly due to the lack of a 1920 x 1080 full HD capability on the D90, but also probably associated with the insufficiency of the Motion JPEG compression. Compared with Canon camcorders (the HF11 and HG20), the D90 produced colors with much less saturation. We noticed the same thing when comparing with Panasonic (HDC-HS100 and HDC-SD100). The D90's closest matches, as far as color reproduction goes, are the Sony (HDR-SR12 and HDR-CX12) and Samsung (SC-HMX20), which both produced soft, less-saturated colors. The D90 does offer a picture controls feature, which allows you to adjust sharpness, saturation, hue, etc. This gives you a chance to boost colors or give your image the look you desire. The picture controls don't work wonders in video mode, but they do have an effect.

{{article.attachments['Nikon_D90_3000lux_auto_NTSC_blowup_moire_web.jpg']}}

Nikon D90 at 100% crop

{{article.attachments['Canon_HF11_3000_Lux_Auto_60i_web_blowup.jpg']}}Canon HF11 at 100% crop

{{article.attachments['Samsung_SC-HMX20_3000_Lux_Auto_60i_web_blowup.jpg']}}

Samsung SC-HMX20 at 100% crop

{{article.attachments['Sony_HDR-CX12_3000_lux_auto_web_blowup.jpg']}}

Sony HDR-CX12 at 100% crop

As hard as it might try, the Nikon D90 simply cannot record video with the same shapness as a dedicated HD camcorder. The blown up images above give you a good idea of how much sharper most HD camcorders are in comparison to the D90. The presence of moiré patterns was another prominent flaw produced by the D90 in our testing. In the first image above, notice the blue discoloration above the 2.35 marker as well as the slight yellow band towards the right side of the image. These instances of interference are called moiré patterns and they mostly occurred during our testing in areas with fine lines and grids (look at the stills from our video resolution test to see more examples). While this interference was prominent in much of our lab tests, it was rarely noticeable in footage we shot in 'the real world.'

{{article.attachments['Nikon_D90_auto_depth_of_field1_web.jpg']}}
{{article.attachments['Nikon_D90_auto_depth_of_field2_web.jpg']}}
Nikon D90 showing off its shallow depth of field Depth of field control is one of the appealing features of the D90. Most camcorders tend to have a large depth of field, which makes it easy for both foreground and background objects in the frame to be in focus at the same time. With a shallow depth of field, the focus tightens to a smaller plane—foreground, midground, background, or somewhere in between.

{{article.attachments['Nikon_D90_auto_jello1_web.jpg']}}

Panning left with the D90

(freeze frame caught in mid-pan)

{{article.attachments['Nikon_D90_auto_jello21_web.jpg']}}

Panning right with the D90

(freeze frame caught in mid-pan)

The big problem with the D90's video mode is a tendancy towards 'wobbles.' Panning and moving the D90 around while you record can produce a terrible horizontal wobble within the frame. Objects appear to tilt in the direction you are moving the camera while a pan is taking place. When you stop, the object bounces back into its straight appearance, giving off a wobble that is similar to jiggling Jell-O. This problem was very noticeable anytime we recorded quick back and forth motion and the only real way to avoid it would be to use a tripod when panning or to only record stationary shots.

Here is a sample video we took of the D90 that shows off the the terrible wobble effect:

{{product.attachments['Youtube Video']}}

The wobble effect is created from the D90's sensor processing different parts of the image at different points in time. This is called a 'rolling shutter,' which has nothing to do with the camera's shutter speed or an acutal moving shutter. What's happening is the sensor begins to expose the top of the frame and then works its way down. Because the sensor works too slowly in the D90, the bottom and top of the frame appear to move independently of each other, producing the wobble effect. Rolling shutter and wobble effects are problems with any camcorder that uses CMOS sensors, but the D90 has worse problems with these effects than most camcorders. For more information about rolling shutter and how it relates to different sensors, visit this website.

Nikon_D90_auto_trees1_web.jpg

The D90 simply doesn't have the detail of an HD camcorder.

The D90's lack of sharpness were in plain sight when we shot scenes with a lot of detail. In the shot above, of leaves on a tree branch, there's a good deal of noticeable artifacting, especially in the high contrast areas of the image.

Overall, the Nikon D90 works well in bright light, but it has a completely different quality and feel to it than a traditional HD camcorder. The 24 fps frame rate does create a unique look, and it has a more cinematic feel than the 24P mode some camcorders offer. However, the D90 has way too many problems with video performance—choppy auto exposure, no auto focus, rolling shutter (producing wobble), artifacting, moiré patterns, and lack of sharpness with images of fine detail.

It is definitely possible to record amazing video with the D90—just check out these sample videos Nikon posted on its website. It should be noted, however, that while many of these videos look very impressive, none of them involve a quickly panning camera, nearly all of them were shot using a tripod, and each of them was an expertly pre-set scene (rather than footage taken on-the-go). Also, many of these videos were taken with expensive, specialty lenses.

Video Resolution* (15.75)

*We tested the video resolution of the D90 by shooting a DSC Labs video resolution chart under a bright, even light. We then view the footage on an HD monitor to analyze it. Recording at its highest quality setting (1280 x 720 HD) the Nikon D90 produced a horizontal resolution of approximately 525 line widths per picture height (lw/ph) and a vertical resolution of approximately 600 lw/ph.

This is a fairly good score, especially when considering the camera doesn't record in full HD. The best resolution from a consumer camcorder we've seen so far this year is 700 lw/ph both horizontal and vertical, which was achieved by the Samsung SC-HMX20. The D90 came close to matching the resolution of the Panasonic HDC-SD100 (600 lw/ph in both horizontal and vertical) and the Sony HDR-CX12 (600 lw/ph horizontal, 575 lw/ph vertical).

In standard definition recording, at 640 x 424, the D90 produced an awful resolution of 300 lw/ph horizontal and 225 lw/ph vertical and the image quality was severely diminished.

Nikon_D90_reschart_moire1_web.jpg

The D90 speckled our resolutions chart with moiré patterns.

While the D90's HD resolution scores were generally decent, there were problems with moiré patterns appearing throughout the chart—producing strange yellow and blue color bands as well as interference.

Low Light Performance* (7.34)

*We test low light performance in three separate stages: comparative analysis, color accuracy/noise/saturation testing, and light sensitivity assessment. For our comparative analysis we shoot a DSC Labs Chroma DuMonde chart at both 60 and 15 lux, then compare the results with other camcorders that we've put through the same tests in our labs.

Nikon_D90_60lux_auto_ISO_up_to_1600_web.jpg

The Nikon D90 at 60 lux, auto shutter, auto ISO (up to 1600)

Samsung_SC-HMX20_60_Lux_Auto_60i_web.jpg

The Samsung SC-HMX20 at 60 lux auto mode

Panasonic_HDC-SD100_60_lux_auto_web.jpg

The Panasonic HDC-SD100 at 60 lux auto mode

The D90 produced some wonderful results at 60 lux, producing a brighter image than we're used to seeing from HD camcorders. The colors were also deeper and more saturated (especially noticeable in the reds and yellows) at 60 lux than the HD camcorders we compared it to. The D90 still couldn't compete in sharpness at this light level. Noise also started to make a slight appearance at 60 lux.

Canon_HF11_60_Lux_Auto_24P_web_b.jpg
The Canon HF11 at 60 lux in 24P mode Some camcorders, like the Canon HF11 above, offer a 24P frame rate option that attempts to recreate the effect of a 24 fps frame rate. Part of the reason the Nikon D90 has a brighter image compared to an HD camcorder is because of its slower frame rate (24 fps of the D90 vs. 29.97 fps of video). As you can see, even the 24P mode of the HF11 still came out darker at 60 lux than the Nikon D90 did under the same light. At 60 lux, we shot our DSC Labs Chroma DuMonde under a variety of different settings and none of them appeared to make any difference in picture quality. We shot with ISO set at 1600 and 3200 as well as with shutter speeds of both 1/60 and 1/30 of a second. None of these settings provided the least bit of change in brightness, sharpness, or color. The only setting that appeared to alter anything was when we switched the video mode to PAL, which produced a slightly brighter image at 60 lux. We didn't even see a difference between shooting at an ISO of 200 and an ISO of 3200. This suggests the D90's only manual settings that offer true control over video, are aperture and exposure adjustment. 
Nikon_D90_15lux_auto_ISO_to_1600_web.jpg

The Nikon D90 at 15 lux, auto shutter, auto ISO (up to 1600)

Samsung_SC-HMX20_15_Lux_Auto_60i_web.jpg

The Samsung SC-HMX20 at 15 lux auto mode

Panasonic_HDC-SD100_15_lux_auto_web.jpg

The Panasonic HDC-SD100 at 15 lux auto mode

Canon_HF11_15_Lux_Auto_24P_web.jpg

The Canon HF11 at 15 lux 24P mode

Next we shoot the same DSC Labs Chroma DuMonde chart at very low light—an even 15 lux. The D90 produced a respectable image at this light, but it didn't have the brightness advantage over HD camcorders that we saw at 60 lux. The camera still captured deep, saturated colors, although noise was very prevalent. Again, ISO and shutter speed adjustment didn't improve or alter the image one bit. Comparing the D90 with the Canon HF11's 24P mode at this light, we saw both capturing vived, dark colors, along with a similar overall image (except with the Canon being far sharper).

Our second low light test looks at color accuracy, noise, and saturation levels. After shooting an X-Rite Color Checker chart at an even 60 lux, we export frames from that footage to Imatest imaging software for analysis. According to Imatest, the Nikon D90 produced a color error of 7.22. This is an excellent score, exceeding the 9.9 posted by the Samsung SC-HMX20, which gave the most accurate color performance amongst the top tier consumer HD camcorders.

The noise percentage of the Nikon D90 at 60 lux was a low 0.7925%, another very good score compared to your average HD camcorder. Only Panasonic had a lower noise percentage, with the HDC-SD100 recording 0.738% noise at 60 lux. The D90, however, had a far higher color saturation with 103.4% at this light level—a score very close to what Canon received with the HF11.

The third, and last, stage of low light performance testing looks at sensitivity. We hooked up the D90 to a waveform monitor, which measures exposure in IREs (the standard measurement used in broadcasting). We then slowly lowered the light until the camcorder generated a peak of 50 IRE. Recording in auto mode, the D90 produced 50 IRE at 10 lux. This is really nothing special compared to some of the top-notch consumer HD camcorders. The Panasonic HDC-SD100 required 13 lux to produce 50 IRE and the Sony HDR-CX12 and HDR-SR12 both required 14 lux of light. The Samsung SC-HMX20, one of the best low-light performers, needed only 5 lux.

It comes as a surprise, considering the enormous size of the D90's CMOS sensor and it's slower frame rate, that it didn't score better in our low light sensitivity test. The Canon HF11, when recording in 24P mode produced 50 IRE at only 4 lux. We expected results like this from the D90, especially after it showed us such a bright image at 60 lux. Because the D90 showed a significant drop when we lowered the lights to 15 lux, suggests the camera is very good in moderately low light, but the quality doesn't hold up when things get really dim. The fact that changing shutter speeds and ISO didn't provide any boost in low light performance was also a big disappointment. Even so, the D90 is a strong low light performer, although not entirely living up to its expectations.

Stabilization* (8.08)*

The kit lens with the Nikon D90 is equipped with a Vibration Reduction image stabilization system. We tested its ability to reduce the shakiness of a video image by attaching it to our specialized device in our lab.

We ran stabilization tests at two different speeds. On speed one, which is comparable to the motion of an unsteady hand, the Nikkor lens reduced 80% of the shake. On speed two, which is much faster and more agitating (like that of a moving vehicle), the lens was able to reduce 83% of the shake.

The D90 and its kit lens, the AF-S DX Nikkor 18-105mm f/3.5-5.6G ED VR, performed far better in our testing than most consumer camcorders that make their way through our labs. The D90 being such a heavy piece of equipment likely assisted in maintaining its balance as well.

Wide Angle* (13.6)*

Obviously, the wide angle ability of the Nikon D90 entirely depends on what lens you are using with the camera. We did all our testing with the supplied kit lens—an AF-S DX Nikkor 18-105mm f/3.5-5.6G ED VR lens. With this lens, classified as a standard zoom lens on Nikon's website, the camera measured a wide angle of 68 degrees in video mode. This is about 20 degrees, or 40%, wider than the average camcorder that passes through our labs. This score emphasizes the biggest advantage the Nikon D90 has over every consumer camcorder on the market—a huge stock of wonderful interchangeable lenses

Our editors review and recommend products to help you buy the stuff you need. If you make a purchase by clicking one of our links, we may earn a small share of the revenue. Our picks and opinions are independent from any business incentives.

Sections

  1. Performance
  2. Format
  3. Auto / Manual Controls
  4. Still Features
  5. Handling and Use
  6. Audio / Playback / Connectivity
  7. Other Features
  8. Conclusion & Comparisons
  9. Photo Gallery

Our editors review and recommend products to help you buy the stuff you need. If you make a purchase by clicking one of our links, we may earn a small share of the revenue. Our picks and opinions are independent from any business incentives.

Advertisement - Continue Reading Below

Our editors review and recommend products to help you buy the stuff you need. If you make a purchase by clicking one of our links, we may earn a small share of the revenue. Our picks and opinions are independent from any business incentives.

What's Your Take?

All Comments