[Paul MaudDib]

Are you fluent in python & perl, do you know Linux, are you NAC'd, can you work from home and what's your hourly rate?

Yes, yes, yes, not willing to do clearance work right now (did a service academy, no thanks), prefer a real job that I can come in to every day, and as of today I'm no longer working hourly. Respectively.

As a guy in the hiring seat in the real world service business, I used to see a lot of applicants touting their compsci ticket.

Many turned out to be paper tigers who learned a lot of ancient history, some current theory, and a lot of misguided info about futuristic implementation, and were fluent in a language or two (a current relevant language maybe if you were lucky), depending on the date they got their ticket punched.
Eric

So, are you willing to bet your "real-world service business" on quantum computing? Yes or no will do.

I mean, it's a pretty straightforward test to see who's an idiot in the field: if you think QUANTUM is the answer to all your problems, you're an idiot who's going to run your business into the ground.

Your "internet real business" means as much as my "internet real qualifications". Diddly squat, if what you're saying doesn't make sense.

As to how any of this relates to camera CPU's leaves a lot to be desired...heck my cell phone is a quad core nowadays...but the bottlenecks are always the lowest common denominator... bus and cache memory and well as storage memory speed that has to be able to keep up & power and heat which in total has to be within reason for LSI in something as small as a camera running on a little battery that's expected to be good for at least 300 actuations with a flash thrown in ocassionally on a single 1Hr charge.
Eric

So give me a rough estimate in terms of computations per watt on how much an ASIC consumes relative to a general-purpose phone CPU? Should be interesting, because it's orders of magnitude difference.

If you want to get a lot done on a little power - you program an FPGA or get an ASIC burned, same as it's been for the last 40 years.

My guess is that large files will be easier to handle at about the same rate as the electronic technology progresses to handle the throughput, meaning it will be a net break even with modest gains over time in FPS as new sensors come out. The PC tech will follow the same curve, and the latest output images will be handled within reasonable limits on current generation PC's or the manufacturer wouldn't bother making a camera for the average consumer that couldn't handle it.

Agreed - the person who said that new CPUs and memory are advancing over sensors by a factor of 2 is absolutely correct. The PCs aren't the limiting factor here, it's the diffraction limit. Quantum doesn't do crap for cameras. If your computer is derpy in TYOOL 2014 - it's your fault.

---

Last edited by Paul MaudDib; 10-31-2014 at 08:27 AM.

[AquaDome]

If Sony is making a 50 megapixel sensor, then somebody wants a 50 megapixel sensor badly enough that it has become worthwhile to produce it. But who ordered it?
The engineers work for the bean-counters, and bean-counters like to count things. Higher ISO, more megapixels, bigger this, smaller that. It makes for a lot of visitations to the old yardstick. If we left it to the engineers (I'm an electronics engineer), I'm sure we would have a better sensor before a bigger sensor, plus better auto-focus.

A 50 megapixel sensor is not "next generation". It is a revision of the current generation, injecting steroids into the part that didn't look big enough last time. Increasing the number of pixels means a larger data-dump, requiring a faster bus and more processor power. It could be done right now if you don't mind holding a hot brick that plugs into the wall and has a cooling fan in it. If you want it in a hand-held, portable, battery-operated device, it will require "next generation" lower-power parallel processors and beyond-the-horizon batteries.

[ElJamoquio]

If you want it in a hand-held, portable, battery-operated device, it will require "next generation" lower-power parallel processors and beyond-the-horizon batteries.

That seems to be a bit of an exaggeration, as there are already 36MP cameras smaller than a Pentax MX.

[patarok]

Hey, so did I. And my step brother did the first traffic control system in Toronto using 6502s and machine code. He's moved on to Google among other places, I've moved on to retirement.

I suggest, you guys join the pentax hack developers....

[Poit]

If Sony is making a 50 megapixel sensor, then somebody wants a 50 megapixel sensor badly enough that it has become worthwhile to produce it. But who ordered it?

Nikon.

The first thing that came into my mind when this thread began was "Nikon ordered this sensor in response to the 645z". It seemed like a pretty obvious conclusion to me.

A Nikon FF with a 50MP sensor would be a far cheaper (for Nikon) development option than producing their own medium format system (with new lenses etc etc). AND it would likely retail for half the price of a 645z.

[RonHendriks1966]

Nikon.

The first thing that came into my mind when this thread began was "Nikon ordered this sensor in response to the 645z". It seemed like a pretty obvious conclusion to me.

A Nikon FF with a 50MP sensor would be a far cheaper (for Nikon) development option than producing their own medium format system (with new lenses etc etc). AND it would likely retail for half the price of a 645z.

If it's for Nikon then they call it D4x and Retail it at D4s price. A great studiocamera like D3x once was......you back in time where 24 megapixel was enormous!!

[Poit]

If it's for Nikon then they call it D4x and Retail it at D4s price. A great studiocamera like D3x once was......you back in time where 24 megapixel was enormous!!

That would make sense. I actually had in my mind something along the lines of a D800, but a D4 spec'd camera makes more sense now you mention it. And still cheaper than 645z...

[Big Dave]

My comment was pertaining to a desire for sensor makers to improve the IQ of sensors, not simply cramming more pixels into a given size.

You have to wait for the next refinement or generation to get improved IQ with given pixel size. Increasing pixel size and format will give you improved IQ also.

[luftfluss]

You have to wait for the next refinement or generation to get improved IQ with given pixel size. Increasing pixel size and format will give you improved IQ also.

I don't think you are comprehending the context in which my comments are made. For example, the 16MP APS-C sensor that was been in use for the past few years has been superseded by a 24MP sensor that offers more pixels, but not necessarily better IQ in other measureables - and in some aspects, the 24MP sensor is a slight step back. I, along with some other people, are in the camp who would prefer the MP not increase and things like sensitivity and dynamic range improve. This is completely separate from sensor size and format.

[Big Dave]

I don't think you are comprehending the context in which my comments are made. For example, the 16MP APS-C sensor that was been in use for the past few years has been superseded by a 24MP sensor that offers more pixels, but not necessarily better IQ in other measureables - and in some aspects, the 24MP sensor is a slight step back. I, along with some other people, are in the camp who would prefer the MP not increase and things like sensitivity and dynamic range improve. This is completely separate from sensor size and format.

How can it be separate? You yourself said that the smaller pixel of the 24mp asp-c were inferrior. You can keep those nice big 5 micron pixels and get more resolution by increasing the sensor size. My 16mp K5 II has the same pixel size as my 36mp A7R.

[luftfluss]

How can it be separate? You yourself said that the smaller pixel of the 24mp asp-c were inferrior. You can keep those nice big 5 micron pixels and get more resolution by increasing the sensor size. My 16mp K5 II has the same pixel size as my 36mp A7R.

That is *not* what I said. I said, "... in some aspects, the 24MP sensor is a slight step back". That statement implies that there are some aspects where the 24MP sensor is equal to - or superior to - the 16MP sensor.

And I'm talking about the technology of the sensor, not the size nor the pixel pitch.

[Big Dave]

That is *not* what I said. I said, "... in some aspects, the 24MP sensor is a slight step back". That statement implies that there are some aspects where the 24MP sensor is equal to - or superior to - the 16MP sensor.

And I'm talking about the technology of the sensor, not the size nor the pixel pitch.

I guess you are correct, the K3 24mp is newer then the K5 16mp, so it should have advances in tech. The BSI tehnology is a way to reduce noise in high density sensors like the newer asp-c units. Samsungs new 28mp camera and the Sony RX1 benefit from this. I was just talking about a general characteristic if all other things were equal.

[LensBeginner]

I suggest, you guys join the pentax hack developers....

Is there such a community? I'd love to have a hacked firmware for the K-01, with some of the features that were excluded available...
Many of them are just software limitations...

[Mikesul]

There is some new activity which you can follow here:
https://www.pentaxforums.com/forums/6-pentax-dslr-discussion/250555-resurrect...e-hacking.html

However, I do not recall seeing anything about the K-01. Check it out.

[LensBeginner]

There is some new activity which you can follow here:
https://www.pentaxforums.com/forums/6-pentax-dslr-discussion/250555-resurrect...e-hacking.html

However, I do not recall seeing anything about the K-01. Check it out.

Hmmm... looks like it's mainly about the K30...
The K-01 would probably be the last camera to be touched by such a project... it's definitely niche...
LoL... "Shodan"... memories!