A number of factors led to the use of 2MP sensors in the main imaging cameras used on NASA's Mars rover, Curiosity, says the project manager responsible for their development. The slow data rates available for broadcasting images back to Earth and the team's familiarity with that family of sensors played a part, says Malin Space Science Systems' Mike Ravine, but the biggest factor was the specifications being fixed as far back as 2004. Multi-shot panoramas will see the cameras deliver high-res images, he explains, but not the 3D movies Hollywood director James Cameron had wanted.

'There's a popular belief that projects like this are going to be very advanced but there are things that mitigate against that. These designs were proposed in 2004, and you don't get to propose one specification and then go off and develop something else. 2MP with 8GB of flash [memory] didn't sound too bad in 2004. But it doesn't compare well to what you get in an iPhone today.'

The 34mm (115mm equiv.) Mastcam from the Curiosity rover - developed by Mike Ravine and his team at Malin Space Science Systems.

The first consideration when choosing a sensor was the amount of data it would produce.

There are three ways for Curiosity, to broadcast data back to Earth - but it's only the UHF transmitter that can be used for transmitting the amounts of data required for sending back images. 'The UHF antenna transmits to two spacecraft orbiting Mars, which relay the results back to Earth. That's where most of the data is coming from. It gives us on the order of 250 megabits per day, and that's got to be shared between a bunch of instruments, so there's not much bandwidth for the cameras.' Ravine explains.

Another factor was that the same sensor had to meet the needs of four different cameras (MAHLI, the two Mastcams and MARDI, the camera tasked with capturing the rover's descent to the planets' surface). 'Everything in a project like this is sensitive to price and risk, both real and perceived. The cameras differ in terms of their optics, but by building them around a single platform, we didn't have to re-test and qualify each of them separately. This makes them more dependable and less expensive than if you have to do it four times.'

'We developed all four cameras around a common architecture so the choice of sensor was hedged across all of them. We wanted to be able to capture high frame rates, particularly with the descent camera.' he explains. MARDI, the downwards-pointing 'descent camera' had just a two-minute descent to the planet's surface, so a high frame rate was essential. The KAI-2020 chip was the smallest Kodak made capable of 720p HD video. 'We also looked at a 4MP sensor but it would have run around half as fast. And the state of CMOS sensors wasn't credible in 2004. They're an interesting option now, but they weren't then.'

The initial thumbnail image from the MARDI descent camera, showing the heat shield dropping away, just before the final descent begins. Click here to see a video of the heat shield falling away, based on these thumbnail images.
Image Credit: NASA/JPL-Caltech/MSSS

The other advantage of the Truesense Imaging chips was the team's familiarity with their behavior. 'We've built-up decades of cumulative experience of working with Kodak and now Truesense interline sensors. We know how to clock them and drive them - they're a very easy CCD to drive,' says Ravine. A similar level of confidence was needed for the cameras’ memory, he says: 'the flash we ended up using was because we had a lot of radiation test data for it.'

The full-resolution image, released August 8th, giving a clearer idea of what the 2MP cameras onboard the Curiosity rover with be capable of.
Image Credit: NASA/JPL-Caltech/MSSS

And the low (by modern standards) pixel count needn't be a problem, Ravine says, since the Mastcams will produce stitched images from multiple exposures. 'A mosaic produced from a higher pixel count camera wouldn't offer huge benefits over what we'll be getting.'

Ravine is understandably proud of the work his team has contributed to the project, on the day the rover's mast is raised for the first time: 'We're getting pictures from Mars,' he points out. But it's not hard to detect a note of disappointment in his voice when he mentions the zoom lenses that were being prepared for the project: 'they're currently sitting in a container on the other side of the building.'

Sadly, the cancellation of the zoom lens project for MSL means that it won't produce the 3D footage the team originally hoped for. 'They were going to be 6.5-100mm zoom lenses, which would allow us to set them both to the same focal length for capturing stereo images,' he says. However, problems designing the lens without using wet lubricants (which would require battery-sapping heating to ensure continued operation in Mars' extremely low temperatures), proved difficult and the development was halted.

The benefits of stitching: a two-image panorama shot with the rover's 1024 x 1024 mono navigation cameras. The Mastcams should produce higher-resolution, color images.
Image Credit: NASA/JPL-Caltech

Hollywood director James Cameron was part of the team and successfully lobbied NASA to re-start the zoom project, with talk of 3D movies showing Curiosity's progress across the Martian landscape. Sadly, although the project was re-started, even Cameron's enthusiasm wasn't enough to ensure they'd be ready in time. 'We're going to get some great mosaics with the focal lengths we've got, but we're not going to get the wide-field cinematic moments or the 3D movies from Mars that Cameron wanted.'

Another perspective - polar projection of the rover's deck and its surroundings. The image is composed of low-res thumbnails, except the two full res panels used in the image further up the page.
Image Credit: NASA/JPL-Caltech