Advanced

There's very little difference between the Sigma DP2, which was released in April 2009 and the DP2s, which was announced in February 2010. The sensor and lens in the new camera are the same as the earlier model and the rather quirky user interface is unchanged. According to a company press release, there's a new AF algorithm for faster autofocusing and a Power Save mode to enable more shots per charge. The rear panel buttons are more clearly labelled, and the camera ships with the latest raw file processing software.

Otherwise, the new model is physically and technologically identical to the DP2 model, which we reviewed in May 2009. So, instead of repeating our description of the camera, we'll concentrate on the performance aspects of the new model. Our review of the DP2 can be read by clicking on this link.

To show just how similar both models are, we've reproduced pictures of the front, back and top of each model, with the DP2 on the left side and the DP2s on the right. The only change we have been able to find - aside from the model numbers on the top panels of each camera - is the re-labelling of certain rear panel buttons in red instead of pale blue.

Front views of the Sigma DP2 and DP2s cameras show them to be identical. (Source: Sigma.)


On the rear panel, four buttons on the DP2s have been re-labelled in red instead of pale blue. (Source: Sigma.)

The top panels of both cameras are identical, save for the model names. (Source: Sigma.)

The review camera was supplied with the optional VF-21 viewfinder, which clips onto the hot-shoe on the top panel of the camera body. The optional HA-21 hood adaptor, which attaches to the lens via a bayonet mounting and includes a specially-designed lens hood, was also provided. An accessory flash, the EF-140 DG, is also available as an option.

Sensor and Image Processing
We've covered the Foveon sensor used in the DP2s in our review of the DP1. Measuring 20.7 x 13.8mm, it's roughly 12 times larger than the ‘1/2.5-inch' type (5.76 x 4.29 mm) sensors found in most digicams. This chip takes advantage of the fact that red, green and blue light penetrate silicon to different depths, enabling image data to be extracted at three levels, corresponding with the blue, green and red channels that make up a digital image.

Accordingly, Foveon chips don't require Bayer filter overlays to record colour information and no colour interpolation when the image data is processed. Colours should, therefore, be more accurately and richly recorded with a Foveon chip. We found this to be the case with raw files from the DP1, although JPEGs from that camera looked rather flat.

The DP2s offers the same image size and quality settings as the DP1 and DP2. Four JPEG sizes are available with three compression ratios, along with the proprietary12-bit X3F.RAW format. RAW+JPEG capture is still not supported. This may be because the TRUEII processor (which is unchanged from the DP2) can't handle the extraction of JPEG images from the X3F.RAW files at point of capture. Typical file sizes are shown in the table below.

The DP2s places the same sensitivity limitations on JPEG files as the DP2. Users who want to shoot at ISO 1600 or ISO 3200 (the highest sensitivity supported) must shoot raw files as JPEG capture has a top sensitivity of ISO 800.

As in the DP2, movie clips can only be recorded with QVGA (320 x 240) resolution at a frame rate of 30 fps. The data recording limit for one movie shoot is 2GB, regardless of the capacity of the memory card you use. This equates to approximately one hour. Users can also add 10- or 30-second voice clips to still images via the voice recording function.

Playback and Software

Playback modes are unchanged from the DP2. A more detailed explanation can be found in our review of that camera. The supplied Sigma Photo Pro software (v. 4.0 for both Windows and Macintosh) is the only application provided. It can also be downloaded from Sigma's DP2 website (http://www.sigma-dp.com/DP2/photopro.html).

Despite being an update on the application supplied with previous Sigma cameras, the user interface and capabilities of this file browser and raw file converter are largely unchanged, except for the addition of the new noise-reduction adjustments, which appear between the colour wheel and the histogram in the adjustment panel.

Two five-point tabs enable users to separately adjust Chroma and Luminance noise when processing X3F files taken with sensitivity settings of ISO 400 and above. No adjustments are provided for lower sensitivities. The other interface changes apply only to Mac users, who gain the JPEG conversion and batch white balance processing functions that were previously Windows-only.

Version 4.0 claims a superior image processing algorithm and improved compatibility with multi core CPUs.  Both factors could explain why the new application appears to display and process files slightly faster than previous versions. We covered Sigma Photo Pro quite extensively in our review of the DP1 but we've included some screen shots here as reminders.


The browser window on Sigma Photo Pro4.0.


The editing interface.


Options for saving converted X3F.RAW files.

Performance
Sigma appears to have tweaked several performance criteria in the 13 months since we reviewed the DP2 as the review camera produced better results in most of our Imatest tests. The richness of the colour from the Foveon sensor remains the most compelling reason for anyone to buy this camera. However, you must shoot  X3F.RAW files to obtain the best results.

Although JPEG test shots were noticeably ‘punchier' than similar shots from the DP2 camera we reviewed last year (indicating Sigma has put some effort into improving JPEG performance), the dynamic range in outdoor shots from the test camera was limited. Shadows Care had to be taken to avoid even slight over-exposure as highlight clipping was an ever-present issue with both JPEGs and raw files.

Unfortunately, autofocusing was as sluggish and unpredictable in the review camera as it was in the DP2 camera we reviewed.  In bright lighting, high-contrast subjects often disrupted focus, while hunting was common in dim lighting, particularly with complex subjects that included near and far elements and subjects within about a metre of the camera.

The lack of an AF-Assist lamp remains an issue for users of this camera. Focusing manually is the only effective option at such times and the low resolution of the LCD monitor isn't at all helpful. (The clip-on viewfinder provides no assistance as it doesn't integrate with the camera.)

The AF  motor also remains relatively noisy, ‘purring' audibly as it hunts of an edge and in manual mode, this purring occurs as you rotate the focusing dial on the upper rear panel. Exposure metering was usually accurate, regardless of the metering pattern we selected.

Adding the noise reduction adjustments to the new raw file processing software gave it a slight advantage over the previous raw converter at high sensitivity settings. But, the quality of long exposures at high ISO settings above 400 was so poor this addition could only be capitalised on for short exposures.

Some improvements were seen in flash shots but, again, raw shots taken at ISO 3200 had noticeable colour artefacts that could not be suppressed with the settings in the new raw file converter. Otherwise the two applications delivered very similar results.

Imatest showed JPEG files to have levels of resolution that were well above expectations for the sensor's resolution, while X3F.RAW files produced even higher results. Resolution of JPEG files declined as ISO sensitivity was increased, although the two highest JPEG settings (ISO 400  and ISO 800) delivered very similar figures. X3F.RAW files maintained a distinct advantage and were high right up to ISO 3200. The graph below shows the results of our tests.

The lens on the review camera remained an above-average performer, despite suffering from some edge softening at wider apertures. Highest resolution was measured between f/4.5 and f/7.1, with diffraction reducing resolution from about f/9 to f/14 (the smallest available aperture. The graph below shows the results of our tests.

Lateral chromatic aberration ranged from negligible to the lower end of the ‘low' range across the aperture range of the lens. Interestingly, JPEG files produced lower CA than raw files indicating some in-processing takes place. Distortion was negligible and vignetting was barely visible. Bokeh was acceptable for the focal length of the lens and sensor area. Contre-jour shots could be flare-affected but fitting the lens hood significantly reduced the incidence of flare.

Low-light performance was only marginally better than we found with the DP2. In exposures longer than 10 seconds, colour noise became visible in raw files at ISO 400 and by ISO 800, saturation had declined and images were artefact-affected. At ISO 3200 very little colour remained and sharpness was noticeably reduced.

Flash exposures at high ISO settings showed none of the colour shifts we observed in the long exposures and contained much less image noise. However, by ISO 3200, colour artefacts could be seen and slight image softening was evident. As in the DP2, the built-in flash was too weak to illuminate an average-sized room at ISO settings lower than 400.

White balance performance was slightly better than the DP2 camera we reviewed. While the auto setting failed to correct the orange cast of incandescent lighting, it produced close-to-neutral colour rendition under fluorescent lights. Both manual pre-sets over-corrected slightly but the custom setting, although tricky to use, produced close-to-natural colours.

Video quality was much as you'd expect for a camera that can only record QVGA (320 x 240 pixel) quality with monaural sound tracks. The camera's 3:2 aspect ratio means each frame has only 320 x 212 pixels. Focus is locked at the beginning of each clip and can't be changed, even in MF mode. This prevents the AF noise from being recorded. However, the overall quality of audio recordings from the camera was uninspiring.

For our timing tests we used a 4GB SanDisk Ultra Class 4 SDHC memory card with a claimed transmission speed of 15MB/second. For both JPEG and raw files in the P shooting mode we measured a constant average capture lag of 0.7 seconds, which was about half the lag we measured with the DP2. This lag reduced to 0.1 seconds with pre-focusing.

However, processing times were longer with the new model. Each JPEG image took an average time of 3.2 seconds to process, while it took almost seven seconds to process each raw file.
In the continuous shooting mode the review camera could only record four high-resolution JPEG images in a burst of 0.8 seconds. It took only 6.7 seconds to process this burst. For raw files, the buffer limit was three files and the buffer memory filled in 0.6 seconds. It took 18.2 seconds to process this burst.

Buy this camera if:
- You're a serious photo enthusiast who is prepared to spend time learning to operate and understand this complex camera.
- You want a well-built compact camera with P, A, S and M shooting modes.
- You don't mind working with a frustratingly non-intuitive user interface and menus that are difficult to read outdoors.
- You're not interested in recording video clips.

Don't buy this camera if:
- You want a point-and-shoot digicam.
- You want fast and reliable autofocusing.
- You're not prepared to shoot and process raw files.
- You want to shoot widescreen or high-definition video.

 IMATEST GRAPHS

From unedited JPEG files.

From X3F.RAW files converted into 16-bit TIFF format with Sigma Photo Pro 4.0 software without additional adjustments.

SAMPLE IMAGES

Auto white balance with incandescent lighting.

Auto white balance with fluorescent lighting.

Recorded as an X3F.RAW file before conversion into JPEG format, during which shadowed areas were brightened. ISO 100, 1/100 second at f/14.

The same subject captured as a JPEG file showing the dynamic range limitations of the camera. ISO 100, 1/100 second at f/14.

Recorded as an X3F.RAW file before conversion into JPEG format with no additional adjustments. ISO 100, 15 seconds at f/3.2.


Recorded as an X3F.RAW file before conversion into JPEG format with no additional adjustments. ISO 800, 15 seconds at f/9.

Recorded as an X3F.RAW file before conversion into JPEG format with no additional adjustments. ISO 3200, 8 seconds at f/13.


Recorded as a JPEG file. ISO 50, 1/40 second at f/2.8.


Recorded as a JPEG file. ISO 800, 1/40 second at f/6.3.

Recorded as an X3F.RAW file before conversion into JPEG format with no additional adjustments. ISO 100, 1/40 second at f/13.


Recorded as an X3F.RAW file before conversion into JPEG format with no additional adjustments. ISO 400, one second at f/8.

Recorded as an X3F.RAW file before conversion into JPEG format with no additional adjustments. ISO 200, 1/40 second at f/7.1.

Close-up recorded as an X3F.RAW file before conversion into JPEG format with no additional adjustments. ISO 200, 1/500 second at f/8.

Backlit subject; recorded as an X3F.RAW file before conversion into JPEG format with no additional adjustments. ISO 200, 1/100 second at f/7.1.


Backlit subject; recorded as an X3F.RAW file before conversion into JPEG format with no additional adjustments. ISO 400, 1/60 second at f/7.1.


Recorded as an X3F.RAW file before conversion into JPEG format with no additional adjustments. ISO 200, 1/80 second at f/7.1.

Recorded as a JPEG file. ISO 50, 1/320 second at f/7.1.


Still frame from QVGA video clip.

Image sensor: 20.7 x 13.8 mm Foveon X3 Direct Image Sensor (CMOS) with 14.06 megapixels effective  (2652 x 1768 x 3 layers)
A/D processing: 12-bit
Lens:  24.2mm f/2.8-f/14 (35mm equivalent focal length: 41mm); 7 elements in 6 groups
Zoom ratio: n.a.
Image formats: X3F lossless compression RAW data (12-bit), JPEG (Fine, normal, basic); Movie (AVI) Voice memo to still image (10 sec), Voice recording (WAV)
Image Sizes: Stills - 2640 x 1760, 2640 x 1485, 1872 x 1248, 1312 x 880; Movies - QVGA (320 x 240) at 30 fps
Shutter speed range: 15 to 1/2000 second
Self-timer: 2 or 10 second delay
Image Stabilisation: n.a.
Exposure Compensation: +/- 3EV in 1/3EV steps
Exposure Bracketing: +/- 3EV in 1/3EV steps (three frames)
Focus system/range: Contrast Detection Type with 9 AF points (manual selection); range - 28 cm to infinity; dial type manual focus
Exposure metering/control: TTL Full Aperture Metering with Evaluative, Centre-weighted Average  and Spot Metering
Shooting modes: Program AE, Shutter Priority AE, Aperture Priority AE, Manual
ISO range: Auto - ISO 100-ISO 200; With Flash - ISO 100-ISO 400;  Manual - ISO 50, 100, 200, 400, 800; ISO 1600 and  3200 in RAW mode only
White balance: Auto, Daylight, Shade, Overcast, Incandescent, Fluorescent, Flash, Custom
Flash: Pop-up (Manual) with auto exposure control; GN 6
Flash modes/range (ISO auto): Forced flash, red-eye reduction, slow Synchro; range - 28 cm to 4.3 metres
Sequence shooting:  3 frames/second for 3 raw or 4 JPEG frames
Storage Media: SD/SDHC/MMC cards
Viewfinder: optional VF-21 finder clips onto hot-shoe
LCD monitor: 2.5-inch TFT Colour LCD with 230,000 dots
Interface terminals: USB 2.0, Video out (PAL/NTSC), Audio out (monaural)
Power supply:  BP-31 rechargeable lithium-ion battery (approx 250 shots/charge)
Dimensions (wxhxd): 113.3 x 59.5 x 56.06 mm
Weight: 260 grams (without battery and card)