A digital camera from scratch?

C

cometarossa

New member
Hello everyone,
I'm just wondering if is it possible to build a digital camera assembling together cmos, lens and a computer as storage and cam software?
To be more precise what are the problem we will ride over?
thanks in advance :)
 
It's possible, the question is WHY though? What is it you want to achieve that's not already available? Surely it CAN'T be done cheaper than what's commercially available now?
 
vidled said:
It's possible, the question is WHY though? What is it you want to achieve that's not already available? Surely it CAN'T be done cheaper than what's commercially available now?
Click to expand...
Ok then.. i want a videocamera full HD with a digital data flow uncopressed with photographics optics lens... i guess why spend 17000$ for the red camera if i could build it myself?
 
I see.

R&D into something like that goes into the millions. There's NO WAY you could make one for less than 17k$ I reckon.
But hey, I might be wrong, and this is the DIY section after all; there'd be many who'd be keenly interested in the progress of such an undertaking, just not sure how realistic that'd be. Best of luck though.
 
vidled said:
I see.

R&D into something like that goes into the millions. There's NO WAY you could make one for less than 17k$ I reckon.
But hey, I might be wrong, and this is the DIY section after all; there'd be many who'd be keenly interested in the progress of such an undertaking, just not sure how realistic that'd be. Best of luck though.
Click to expand...

Thanks friend ;)
I hope someone could explaine why is it impossible or too expensive :)
 
I am really interested in what you come up with. I hail from te school of common scense. I believe if you have the time and tools, give it a try. You might just succeed.
 
I hope someone could explaine why is it impossible or too expensive
Click to expand...

Many many factors. If you look at DIY stuff it is almost always low-tech ... roller skate wheels on plywood for a dolly. Maybe the highest tech you might get would be a zoom controller or a LED light... neither of these come close to the complexities of a digital camera

look at this for how a digital camera works:
http://electronics.howstuffworks.com/digital-camera.htm

Which doesn't even begin to talk about the tolerances involved (how precisely the distance from lens to imager has to me machined.

Impossible... nothing's impossible

extreamly impractical.... You bet!
So if you:
1. Are a master machinest
2. Are an optics wiz
3. Can write image processing and data moving software with eaze
4. have a fabrication plant to creat the custom integrated circuits (or at least can find one that will do a small run)

Then you might have a snowball's chance, but cost?

A. You can't get the parts for anywhere near the price as a company that buys by the thousands
B. They have optimized for mass production to keep costs down

and the result.

likely far worse image quality than the cheapest cell phones.


Not what you wanted to hear, but somethings are great canidates for DIY and some are just so far beyond the realm of practial.
 
cometarossa said:
Thanks friend ;)
I hope someone could explaine why is it impossible or too expensive :)
Click to expand...

Impossible? No, it isn't. It's VERY possible. Some persons are doing it right now, as you know.
Too expensive? Yes, probably. Why? Because of R&D. That simple: you buy a CMOS, and then it's not the right one. You buy another and fry it because of a faulty power supply, you buy another, and it get's lost in an uninsured package....see where I'm going here? Nothing is simple, and such an undertaking especially so.

But I repeat myself when I say "best of luck"!
 
(1) Its not impossible, in fact there is an entire section devoted to it over at Dvinfo.net, there was much interest before the consumer HD "affordale" cameras starting popping up, currently while some are still pursing it, it has a lot less movement now than before.

Actually the Silicon 2k camera was the result of efforts of "DIY" HD, and others systems as well.

(2) It requires a lot of (a team of electronic engineers + programmers) to get good results, however the cost of R&D the real gamble, you may come out with something fruitfull or just fail.

By all means if you have the drive and ambition and money then go for it.

I too was amongst the little guys pre "DIY" HD days, but soon gave up, and I'm a mixed bag of Mechanical Engineer, electronics hobbyist and indipendent programmer all rolled into one, and even for me its a scary uphill to climb.

On the other hand if your happy to delve into learning masses of optical mathematics as well as imaging science (eg learning about DSP and debayer algorithms), coupled with Microprocessor + FPGA + PIC + PSOC + whatever else you decide you need. Lets not forget you may actually need to build a complete hardware that goes into capturing and processing that picture. Most people have gone the route of actually cramming a custom built PC (because it saves from having to build one from scratch) but this brings further complications of its own, size vs performane vs power considerations.

Ok, ok I'm not saying you can't do all of that, heck you may know more than me, but at least see it this way:

out of the numerous "DIY" HD guys. how many have actully succeded?

do not take my comments in a negative way :)
 
Last edited:
Would you gain anything by starting with a cheap camera and then trying to modify it? Could you, for example, just wire in new chips and hack the software to work with them?

Perhaps another approach would be to take something like the HV20 and hack it so that it has real external controls (Iris, focus, etc) and passes the uncompressed signal out the HDMI cable to a laptop for capture. The ultimate would be if you could make it a 3 chip, since it's cmos is true HD res.

This approach might also be a good way to learn how the camera works inside in case you later want to try to build one from scratch.
 
You must log in or register to reply here.
Back
Top