Motorola Droid RAZR Review - A Better Clad Bionic

Camera - Still

I guess it shouldn’t come as any surprise at this point that the RAZR, you guessed it, uses the same OmniVision OV8820 8 MP CMOS sensor as the Bionic. That means 1.4µm square backside illuminated pixels. However, as we’ve discussed in the past there really are four parts to the whole smartphone imaging chain: the sensor (CMOS), optical system, Image Signal Processing (ISP), and finally software on the OS talking to the ISP. What’s different on the RAZR versus the Bionic is optical system and software.

The Bionic included F/2.8 optics with a 4.6mm focal length. The RAZR keeps that 4.6mm focal length and includes F/2.4 optics according to EXIF. It’s entirely likely that the focal length field hasn’t changed, but given that from what I can find commercially available OV8820 F/2.4 packages have a focal length around 4.95mm. However, it’s reasonably close. Both of these designs are likely 4P (4 plastic elements) as well.

We’ve done the usual thing and taken photos with the phone under test in our smartphone test locations (which are a bit difficult to control and might have slightly different than usual lighting due to seasons changing) and in our controlled indoor tests.

As a reminder, in the smartphone bench samples only locations 3-7 remain available for testing.

The RAZR does decently well in our controlled testing. Distortion is minimal, the lightbox with light on sample is sharp and looks accurate, and colors look similar to the Bionic. Despite being a half stop wider on the ISO12233 chart we see no more spatial frequencies in the horizontal or vertical (meaning performance is clearly not diffraction limited, which isn’t a surprise). In the lights off test, the RAZR still doesn’t illuminate the scene in the dark, and instead defaults to focusing to infinity which produces a blurry image in our box.

I guess while we’re on the subject of focus, this is a continual problem in the bench test. Focus is soft or missed focus entirely in locations 3 and 4, and in our sample bench video as we’ll show later. I’m not sure what the problem is here, but I’m confident I allowed the AF routine to run properly before capturing - this is just the position the software decided was best focus.

I mentioned software because the RAZR’s camera software subjectively seems less stable than the Bionic’s. I experienced a crash or two in the course of taking bench samples and normal test photos, and like other Motorola camera apps had UI elements disappear sometimes. Obviously not having a physical shutter button makes having a working UI even more important, and the RAZR just needs a bugfix update to address the camera app stability.

Camera - Video

I also shot video at the test location, and here if you look at 1:1 zoom you can see that the RAZR does appear to miss focus despite running its continuous auto focus routine a few times. I shot this video a number of times after a reboot expecting different results but never got a completely sharp video.

The positive part of video recording on the RAZR is that it still uses the same bitrate and H.264 features as the Bionic - 15 Mbps high profile for 1080p30, and 10 Mbps high profile for 720p30. Audio is two-channel stereo AAC at 128 Kbps. You can again pull all of these out of build.prop just as shown below.

ro.media.camcorder.1080p=mp4,h264,30,15000000,aac,128000,44100,2 ro.media.camcorder.720p=mp4,h264,30,10000000,aac,128000,44100,2 ro.media.camcorder.d1NTSC=mp4,h264,30,6000000,aac,128000,44100,2

So that’s a good thing, and again thanks in part to OMAP4’s excellent video encoder, though we see most of the high end smartphones shooting 1080p based on OMAP4 and Exynos using high profile.

Rear Facing 1080p30 Video Sample

Front Facing 720p30 Video Sample

In addition we’ve uploaded the raw rear facing camera video sample without YouTube’s transcoding which you can grab from us in a big zip here. Again the encode quality of the videos is above average, but sharpness would be much better were it not for these focus issues.