Musings on Photography

So, what would a decent metering system look like?

Posted in equipment by Paul Butzi on May 2, 2007

Plenty of interesting comments on my post about the bad performance of metering systems in digital cameras that are used in camera raw mode.  This makes me think it would be worthwhile laying out what I think a decent metering system would be like, and why I don’t think it would be outrageously expensive for camera manufacturers to incorporate such a system into digital SLRs.

StephanB commented that:

I still use my Pentax Digital Spotmeter. I find the brightest part, expose it a Zone VII and that’s it. If the contrast range is larger than 8 stops I decide if it is OK to lose the shadows or not. If not I consider making two exposures. That method never fails.

This is the core to any decent metering system for a digital camera recording images in raw mode – find the brightest spot(s), and calculate the exposure so that those spots fall at the high end of the image without clipping.  This, as the saying goes, is not exactly rocket science. 

So what we need is for the camera to have the ability to meter the scene, breaking the scene down into samples with a radius of 1 degree or smaller.  Just for yucks, let’s posit that Stephan can control exactly where the 1 degree measurement of his Pentax meter to within .25 degrees, and that we’re using a lens with a normal focal length.  A 50mm lens on an EOS-5d will cover an angle of view of about 40 degrees horizontally and about 26 degrees vertically.  If we were to break this scene down into samples (aka pixels) that covered .25 degree, we’d end up with an array of samples 160 samples wide by 104 samples high.  This is four times the spatial resolution of Stephan’s Pentax meter, so I’m pretty confident that this metering matrix would suffice.

How fancy would such a device have to be?  It need be no more fancy than a Bayer style sensor.  How big would it have to be?  160 pixels by 104 pixels, or a whopping big 16.6 kilopixels, or .016 megapixels.  Would such a sensor be expensive?  Well, taking a quick look at, I see that the cheapest webcam has a sensor that’s 320 x 288, or double the resolution of what we just decided we needed.  Cost for the webcam?  Ten bucks.  That’s ten bucks including the packaging, the webcam housing, the cable, the software, the whole schmear.  Figure cost of goods sold is, what, 10% of the retail price, and the cost of the sensor is something like half of the COGS, and we end up with the sensor costing 50 cents. 

Sure, this webcam is probably noisy, and we’d want less noise.  But the point here is not that a webcam sensor would work well, it’s that a sensor on the order of what we’d need to build a decent meter would not be an esoteric, cutting edge, expensive part.  Any camera manufacturer would, of course, incorporate this metering system into every camera they build, so they’d amortize out the part development cost pretty quickly. Heck, the same part that has the sensor can do the computational work of finding the brightest and darkest spots, and calculating the correct exposure for the various camera modes.  Light meter on a chip, as it were.

I don’t think the reason why camera manufacturers are putting such cruddy metering systems in high end digital SLR’s is that the technology is infeasible.  I think it’s that they’re just too blinkered by the history of SLR’s to realize that there’s a much better way, and it would cheap, and it would be dramatically better than what they’re doing now.  Nikon are almost there with their ‘matrix metering’, but they just can’t seem to close the conceptual gap and conclude that calculating exposure for an image that’s captured in raw mode is fundamentally different from calculating exposure for JPEG recording.

The first manufacturer to wake up, smell the coffee, and offer a metering system and metering mode that is tailored to raw capture should win market share.  Now, I just hope that Nikon, Canon, Sony, et al read my blog.

Get with it, guys.  I’m tired of waiting already.

2 Responses

Subscribe to comments with RSS.

  1. StephaneB said, on May 8, 2007 at 4:28 pm

    Hmmm, I follow your reasoning but I am not sure I agree with its line.

    Paul, like me you come from a view camera. For what I can see on your website, you are more than apt at measuring light. I can’t find any exposure problem in your 5D shots either. So I am sorry I fail to get what the problem is.

    Regarding your proposed ultra matrix meter, Nikon has had something like 1000 pixels in its metering system for years and it does not seem to make much of a difference.

  2. Wish List « Musings on Photography said, on November 30, 2007 at 9:08 am

    […] either. In fact, I think they could do it and give it out free. I’ve written before (here and here) about the stupidity of metering systems in digital cameras. A lot of the technical hurdles to […]

Comments are closed.

%d bloggers like this: