I have been wondering for years why no one has made camera backs for old Nikons and other treasured cameras which have a full size CCD (though it doesn’t have to be particularly high resolution) and all the digital camera bits. A real “digital back”. The gubbins would be easily reconfigured for different hardware. The casting of the back would be different, of course, but everything else would be open source. There are so many old lenses which are pretty high quality but don’t fit/work with new bodies very well. The rest of the works in old bodies are undervalued as well.
Oh, in my prior post about the #2 selling Nikon lens I only meant to refer to your category of lenses under $200. The second-highest selling Nikkor in this category, the 50/1.8 AF-D, needs an in-camera motor. The two Nikkors that sell in larger volumes, the 35/1.8 and 50/1.8 AF-S, both have the motor in the lens. With the AF-S 50 selling in higher volumes, I have difficulty believing that people with low-end bodies that can’t drive the AF are buying this lens.
Technically, the F3AF (Nikon’s first AF body, from 1983) also lacked an AF motor in the body.
The first AF-S lenses were not zooms but the flagship tele-primes. The 300/2.8 and 600/4 both got AF-S in the summer of 1996.
There would likely be significant registration problems, not to mention dirt and sensor fragility. But there’s also the problem of powering the sensor and knowing when to begin/end capturing. I mean, while you could use the film-advance mechanism to know when to switch to a new exposure, the back has no idea when the shutter will be pressed and has to be constantly ready for the mechanical shutter to open, unless you want to keep turning it on and off. The image processing and digital specific hardware would probably result in something looking like even less elegant versions of the bulky old Kodak and Fuji DSLRs based on Nikon’s film bodies. I suspect the target audience would be tiny: those willing to pay the high prices inherent in full-frame sensors, but who have a big backlog of old lenses from obsolete mounts they’ve yet to move on from, and who don’t need that much in the way of reliability and durability.
I’m not sure anything in old bodies is undervalued. The light-tight box aspect of digital cameras probably isn’t that large a component of their price.
Point… A very important one…
We’d necessarily be working within the limits of the equipment we have access to of course. There’s no way a group of volunteer tinkerers can match Nikon or Canon’s precision.
But that’s not the aim of the exercise - at least initially. Can the concept be proved? A hackable, open hardware and usable camera system at not too high a price point?
A 300 f/2.8? Why not a 300 f/2?
Still, these were rather expensive, pro lenses. Just the thing for an entry level body to focus…
I suppose you’re right. Maybe Nikon was trying to sell the 18-55mm/d40 as inseparable pair.
1 Like
I don’t know what point you’re trying to make.
AF-S was initially an expensive feature limited to high-end lenses. AF itself was also a high-end feature when introduced. This doesn’t mean that they are out of place on entry-level products introduced later… especially when making bodies without focus motors allows you to make them cheaper. Were the 300/2.8 and 600/4 AF-S lenses made to pair with entry-level bodies from 10 years later that would not be able to focus lenses without motors? Obviously not. In fact, these AF-S lenses were actually incompatible with contemporaneous entry-level bodies, as the F60—introduced after these lenses—lacked the circuitry to drive AF-S lenses.
I also like how you describe a $5,500 300/2.8 (the 600/4 is $10,000) as only “rather” expensive, yet describe the lack of affordable options amongst Amazon’s 20 best-selling lenses as “slim pickings”—even though, at the instant I write this, none of the top 20 costs more than 10% of this “rather” expensive lens.
Maybe Nikon recognizes that although they would like to sell lenses (especially fast, expensive lenses) by the bushel to every DSLR owner, most owners of entry-level bodies are unlikely to ever buy additional lenses. I mean, if they thought the bayonets of entry-level lenses were going to get significant use, and be swapped out frequently, why would they make them in plastic?
1 Like
It’s mostly a nostalgia and “can we make it work” product. The proper point for it to be introduced was probably ten or 15 years ago.
I do it all the time. http://www.jeffrey-martin.com/#gigapixel-photography
(canon 550D, 70D, 7D, 70-200 2.8, 400mm 5.6…)
3 Likes
It’s unfortunate; but the price discrimination model used in cameras is precisely what both creates many of the problems that could so…gallingly…easily…be fixed with just a dash of software and maybe an RT5350 bodged in somewhere and provides an incentive for the manufacturers to not solve them.
The price for all the fancy moving parts, exotic glass, and slab of imaging silicon supported by a nest of lilliputian circuit boards connected by lots and lots of flex cables? An undeniably impressive value. One part technology-indistinguishable-from-magic, one part mechanical-watch-assembled-by-pixies-and-gnomes, amazingly cheap.
Now the fun part… Why doesn’t your camera support histogram display? Well, if it were 1985 it would be because of the enormous cost of the extra RAM and computational power required. Since it’s not, it’s because the mostly-identical model that costs $50 extra does support those and screw you, cheapskate.
Why does the WU-1a/b cost 50-60 dollars, support interaction with a limited number of our cameras, and only a subset of its features on many of those, when generic wifi adapters start at under $10, and you can get a wifi module and a slab of flash memory stuffed into an SD card, probably using technology stolen from time lords, for half the price of our wifi-only module? Well. Because. We. Can.
The USB port? It only acts as a mass storage device on our cheap camera, a mass storage device and a remote shutter control on our midrange model, but offers full programmatic access to basically all camera functions on our expensive one. That’s because our USB Interface descriptors are hand-carved into flash ROM by monks, and our Endpoint descriptors only grow in a single vineyard in Silicon Valley. Or, y’know, price discrimination.
4 Likes
It’s price discrimination that’s often hidden from the consumer until after the money changes hands. A flippy screen is an easier sell than the DoF preview.
Hey, it’s not like camera companies are running huge profits. And get rid of software and firmware locks and you will cut into the revenue they make from people who upgrade to gain access to those locked features: they will have to either lose revenue/profits or raise prices. Get rid of price discrimination and entry-level cameras will increase in price while pro cameras will drop in price. Price discrimination actually helps keep prices down for the cheapest customers just like it does for airline seats (for all the complaining about airlines, if there was no booking-time-based or service-class-based price discrimination the average airfare would be much higher, just as it was in the era of regulated airfare). So, cheaper cameras for working pros who can expense their equipment, but higher prices for hobbyists and neophytes who may be gaining additional access to features they don’t actually use?
I took no overall position on the price discrimination as a whole, just noted that its manifestations are overwhelmingly visible and annoying in software, as opposed to other areas, which is what makes the “What if we bring OSS into this?” idea so compelling.
In other markets price discrimination has totally different manifestations, some of which are pretty much irrelevant to software (hardcover books costing more mostly because they release first, not because of production cost, staggering theater and DVD release dates, etc, etc.) and some that are almost the opposite of the camera market: Computers, say, have nearly zero software differentation; but CPUs with capabilities disabled, motherboards with only half the DIMM slots soldered on, and assorted other hardware-based price discrimination techniques are all over the place.
I wasn’t really looking for an argument on price discrimination in general, just noting that the way it works with cameras creates a lot of situations where the hardware is pretty amazing but the software is clearly messing with you, which is where being able to modify the software in your interests is most attractive. If the situation had shaken down differently, and camera makers competed brutally on firmware quality, it’d likely be the case that only Richard Stallman would be interested in OSS camera firmware; but the low end would have some other deliberately galling defects and people would be trying to kickstart or 3d print their way out of those.
Just hack the hell out of the stuff and take the additional features by force if they don’t want to give them to us peacefully. I hear @fuzzyfungus way too well.
The Canon CHDK works on many older models and allows precisely these add-ons. We need more of the same on more cameras and more brands. Or, you know, modchips - possibly daughterboards with different microprocessor but identical pinout in form of e.g. a strip of flexible polyimide foil with soldering spots matching the BGA pinouts of the most common camera processors. Desolder the old one, attach the new one, and voila, much higher model with opensource software. The imaging chips are usually built around relatively standard interfaces, and a FPGA with support for LVDS lanes could interface with them fairly easily; same goes for the displays and the buttons are trivial. Reverse-engineer what goes to which pin of the CPU, and let the FPGA with an add-on processor or SoC (or even internal “soft” one programmed in the FPGA, depending on the cost of either solution) do the rest.
Given that there are way fewer interfacing standards to CMOS sensors, displays, and memories, and only a few major camera controllers, the modkit could be relatively universal, differing only in configuration of the pins and the polyimide foil adapter…
Not anything like what I’m talking about, but a great example of marketing drone’s idea of disruptive technology. It’s just like film! And diggitle! Gimme money! Interesting link, though.
That’s basically the entire digital media price discrimination system. DRM makes it just difficult enough that some will opt for iTunes, Netflix, Hulu Plus, Amazon Prime, or whatever. And the next level is physical media.
Well, they do compete on firmware quality, just not only on firmware quality. I mean, Nikon and Sony share the same sensors. Firmware and software makes up much of the differentiation between their bodies. Sure, lens mounts are a huge factor, as are ancillary hardware considerations as well as ergonomics and the like, but firmware and software is likely the major differentiating factor in terms of IQ.
I’m not so optimistic that fully open-source firmware would result in the innovation you hope for, given the huge investments necessary to develop a lot of the core functionality. I’m not sure that the necessarily higher costs of entry into an unlocked system would make these open-source implementations cost effective. In fact, without ad-supported revenue streams available for developers, I’m not sure that there would be much incentive to develop on this kind of platform.
Out of curiosity, how many people are actively making software for this, and how useful are their creations? Do they do anything more than simply unlock features already developed by Canon, or can they do something really interesting like add lens corrections to an older camera?
This has to be (and is, and will be) fought.
There is a different sort of market where the prices are not just above roof but straight in the stratosphere - lab instruments. Many kinds are just an image sensor with some electronics around, essentially a modded camera. I expect the core of the R&D to come from here, from the sub-segment where there is hightech knowledge and lack of money. Possibly some smaller universities.
Don’t know how many but the stuff is quite useful for automation of things. The scripts are easy to be custom-written for any application. The wiki has some low level but steady chatter.
http://chdk.wikia.com/wiki/Special:WikiActivity
Here are the supported commands:
CHDK Scripting Cross Reference Page | CHDK Wiki | Fandom
The scripts commonly used are for bracketing (focus depth, exposition), timelapses, motion detection (waiting for an event), automation of all kinds of operations.
The firmware makes the camera functions available to the scripts. The scripts then can do anything the functions themselves can support. There is even support for USB communication via PTP, so the script can talk with an attached computer.
No lens correction - not directly. But you can unlock the RAW format, and then do the corrections in postprocessing.
This topic was automatically closed after 5 days. New replies are no longer allowed.
