Who Fancies a Drop In Digital Sensor?

[jeff_sudduth]

<blockquote>

<p><strong>I'm not saying it is economic nor marketable</strong> but my engineering is not so far out of date that I can't see <strong>it would be feasible</strong>.</p>

</blockquote>

<p>If it is not economical or marketable it is not feasible.</p>

[steve m smith]

<blockquote>

<p>If it is not economical or marketable it is not feasible.</p>

</blockquote>

<p> <br>

<br /><br />Not financially feasible but possibly technically feasible.</p>

[ljwest]

<p>A digital camera's "sensor" is so much more than simply the sliver of silicon the light hits. The processor that interprets the sensor data, applies corrections, etc. to the image, creates your JPEG (if you do that), and ultimately writes the images out to the card plays a huge part in making the magic happen.</p>

<p>Because space is at a premium, with a 35mm format DSLR, you would essentially have to rebuild the camera each time to replace all that, unlike a medium format where you could change just the back, and have the processor, etc. changed too.</p>

<p>But suppose you could just "drop in" a sensor. Say you had a nice, top-end camera with 10 megapixels, and let's say it'll do 8 FPS RAW bursts, which you use for birds in flight or sports or something where the burst performance is important to you.</p>

<p>Now, the 20 MP sensors are out, and you'd like one of those. Even supposing there was one that was a "drop in", if you only change the sensor, you'll run into one huge bottleneck. The processor(s) that are still in the camera. Now, the processor has to work on 20 MPs, and shove them all out to your card. At a minimum, you're looking at halving your burst performance, and probably it would take something more off of that, due to processing the image.</p>

<p>It's a lot like upgrading a personal computer. Sure, they can and do make boards where the processor can be upgraded, but all too often many other factors overtake the processor in importance, and, ultimately, it is easier - not to mention cheaper - to replace the whole thing. Even if you could drop a Core i7 into the same socket as a Pentium, you wouldn't want to, because the board built for the Pentium would hobble the i7 incredibly with insufficient RAM capacity, crawling bus speed and thin, slow I/O paths.</p>

<p>As for the "green" side of things, even if you could replace the sensor, now you have a shiny new sensor with an old, used shutter mechanism and mirror assembly, not to mention, the same metering and exposure system from before. If you shoot a lot, well, then, those would have to be changed at some point, too.</p>

<p>I suppose one thing we could do would be lobby the Canons and Nikons of the world to take or buy back their old gear, refurbish it, and make it available inexpensively to schools to teach the art of photography. Or, there is always the tried and true method of selling it on or giving it away when you've updated to the newer equipment.</p>

[h._p.]

<blockquote>

<p>if you only change the sensor, you'll run into one huge bottleneck</p>

</blockquote>

<p>As I suggested, above, Larry, such a product would only make sense <em>if</em> it was built with a replaceable processing block. As I also said previously, my assumption is that the camera shell would implement the lens mount, shutter, viewfinder, switches and communications bus. All the other electronics would be user replaceable.</p>

<p>Of course, it's unlikely that there's enough demand to persuade a manufacturer to produce such a device but it's always nice to consider the possibility.</p>

[q.g._de_bakker]

<i>"As I suggested, above, Larry, such a product would only make sense if it was built with a replaceable processing block."</i><br><br>Which brings us round to the very first line of the first reply in this thread.<br>Instantly replaceable (digital) sensors (that <i>"impossible dream"</i>) have been with us for some 20 years now.