What does the future hold for CCD's

[DavidU (Dave)]

I was thinking how the CCD technology has improved in recent years.
What does the future hold? In 5 years will we have 50 mega pixel Kodak CCD's 4" across? Will they be run at -230C fed by it's own liquid nitrogen store?
The amount of storage and processing power would have to be huge.What about telescope design? How does an amature fully illuminate a 4" CCD?
Whats your thoughts?

[multiweb (Marc)]

I think the larger the CCDs, the wider the field and the more demanding on the optics, so scopes that can take advantage of those large chips are going to be very expensive or well corrected.

[DavidU (Dave)]

Yes ! So true, however do premium imaging systems become un affordable or will the Chinese nail the triplet Apochromats & RC's as well as DK's?

Originally Posted by multiweb

I think the larger the CCDs, the wider the field and the more demanding on the optics, so scopes that can take advantage of those large chips are going to be very expensive or well corrected.

[mithrandir (Andrew)]

It wouldn't take much growth for amateur scopes to be unable to illuminate a bigger chip.

A 24 megapixel - 6064*4032 - chip with 7.8 micron pixels (four times the pixels of a QHY8 and the same pixel dimension) would have about a 56mm diagonal.

In my opinion deeper wells, fewer defects and lower noise would be more useful.

[DavidU (Dave)]

Originally Posted by mithrandir

In my opinion deeper wells, fewer defects and lower noise would be more useful.

Thats where I'm coming from. The populace are fueling the Uber pixel fire.

[Peter Ward]

To improve over today's CCD's, future chips will need to have

1)Higher QE
2)Lower noise
3)Higher resolution
4)Higher dynamic range

Unfortunately in general terms, point 3 usually means smaller pixels, and point 4 larger pixels.

Similarly High QE devices are great at picking up all signals, including noise with higher efficiency. These are very tricky problems that at best are seeing incremental improvements.

Two interesting points of investigation are *orthogonal transfer CCD's* and *clipper chips* (suggest you Google them). These promise adaptive optics on a chip and sub electron read noise.

Chip size is very much a double edged sword
as there are very few telescopes available with well corrected & flat photographic fields in the 50mm round region, let alone the 65mm square required (read 3" fittings) by soon-to be released large format self-guided systems.

Sure there are a few 4" focusers out there, but I suspect it will be quite a while...indeed...maybe never, due lack of demand...before we commonly see 90mm back ends on affordable telescopes.

While CMOS continues to improve leaps and bounds, I find it interesting to see that the APS format DSLR has not been swamped by 35mm chips...and suspect the cost of well corrected DSLR optics is the sticking point.

Going back through the literature, just a handful of amateur astrophotographers used anything larger than 35mm film. 2 1/4" was a rarity. Sure there were some notable professionals using up to 14" plates
but they also had instruments optimized for these plates, and hey..the AAT Focal length is pretty long.

Just my 2 cents worth....

[AlexN]

I've got to say I'd like to see innovations over the general stuff that we all know and love...

with the range of 8, 11 and 16mp CCD's on the market, I dont see going to higher megapixels to be of that greater importance.. more so, smaller pixels with higher QE.

Personally I'd like to see an APS sized chip, somewhere around the 11mp mark, that would give something like 5.2um pixels... of course, this is great for scopes like mine, where the image circle demands a smaller APS sized sensor, however its focal length requires smaller pixels to achieve good resolution... I suppose, for me, anything with the KAF-8300 sensor would be fine...

I dont see anyone needing a 4" CCD... Even something twice the size of the APS sensors would be difficult to fully illuminate for majority of amateur scopes... things like the AP140 that have a 100mm corrected imaging circle might do the trick.. but how many of us own one of those..

I look forward to seeing what comes of the SBIG STX line of cameras that are coming soon. I've used self guiding for a while with previous cameras I've owned and I must say, a lot of people don't like it (most, dont like the idea of it, as they've not actually used it) I personally think its about as awesome as it gets.

As it says in my signature, Simplicity is the ultimate sophistication. The new SBIG cameras, if they are as easy to use as the ST9 and ST10 that I've owned, should prove to be the ultimate in sophisticated imaging systems...

[AlexN]

Oooh.. I just noticed that there will be an STX camera with the KAI 10100 chip in it.. perfect for my needs... 4.75um pixels, 10.1mp.. shame its colour.. Thems the brakes...

[g__day (Matthew)]

How about a colour technology behind each pixel using a microscopic prism array so rather than a Bayer matrix - each pixel detects LRGB across four distinct counters. I believe such ICs have already developed and successfully tested?

That and a materials technology that doesn't need very lower temperatures to have low read noise would be great.

An algorithm that intelligently switched each pixels' pixel well from an integer to a float (once the well as an integer would otherwise be swamped) I am sure is on the cards.

[gregbradley]

Originally Posted by AlexN

I've got to say I'd like to see innovations over the general stuff that we all know and love...

with the range of 8, 11 and 16mp CCD's on the market, I dont see going to higher megapixels to be of that greater importance.. more so, smaller pixels with higher QE.

Personally I'd like to see an APS sized chip, somewhere around the 11mp mark, that would give something like 5.2um pixels... of course, this is great for scopes like mine, where the image circle demands a smaller APS sized sensor, however its focal length requires smaller pixels to achieve good resolution... I suppose, for me, anything with the KAF-8300 sensor would be fine...

I dont see anyone needing a 4" CCD... Even something twice the size of the APS sensors would be difficult to fully illuminate for majority of amateur scopes... things like the AP140 that have a 100mm corrected imaging circle might do the trick.. but how many of us own one of those..

I look forward to seeing what comes of the SBIG STX line of cameras that are coming soon. I've used self guiding for a while with previous cameras I've owned and I must say, a lot of people don't like it (most, dont like the idea of it, as they've not actually used it) I personally think its about as awesome as it gets.

As it says in my signature, Simplicity is the ultimate sophistication. The new SBIG cameras, if they are as easy to use as the ST9 and ST10 that I've owned, should prove to be the ultimate in sophisticated imaging systems...

I tend to agree Alex. I have seen images from a Proline 39mp one shot colour and they weren't impressive. Also huge file sizes so you'd need an i7 chipped computer with huge and fast memory to handle it.

But it seems that is the way CCD technology is going. More mp.

There are a lot of point and shoot cameras around that are 12.1mp. What sensor are they using? Perhaps its a Sony sensor and so most likely very low noise. Perhaps that is a CMOS chip.

I'd like to see Sony enter the larger chip market more as they usually put out killer products.

Greg.

[AlexN]

I'd love to see sony release something around the size of the KAI-16000M (35mm format, 16mp) giving 7.4um pixels, nice big field to play with..

Would be great if they could do it with similar noise characteristics as the ICX-453 found in the QHY8.

[Peter Ward]

Originally Posted by AlexN

I'd love to see sony release something around the size of the KAI-16000M (35mm format, 16mp) giving 7.4um pixels, nice big field to play with..

Err....yeah...it's called a STX16000

[AlexN]

wink wink, nudge nudge I wonder if anyone can do me a good deal on an STX camera when they come out... hrmmm...

[mithrandir (Andrew)]

Originally Posted by gregbradley

There are a lot of point and shoot cameras around that are 12.1mp. What sensor are they using? Perhaps its a Sony sensor and so most likely very low noise. Perhaps that is a CMOS chip.

I'd like to see Sony enter the larger chip market more as they usually put out killer products.

My Sony a200 has a 10Mpixel CCD - 3872x2592 with 6.1um pixels. It's almost the same size as the chip in the QHY8, so might be a possibility.

I think I read the chips in a700 and a900 are CMOS.

[Hans Tucker (Hans)]

Originally Posted by Peter Ward

Err....yeah...it's called a STX16000

So SBIG are using Sony Sensors in their STX16000 rather than go with Kodak Sensors I didn't think that Sony made sensors in that Mega pixel range.

[AlexN]

Andrew - Do you know the model number of the CCD in the A200? Could be very interesting.

Hans - No, the STX16000 has the Kodak KAI-16000M sensor...

[rally]

Many of the issues we face have already been dealt with in existing technology, its just amatter of time befoire they can deliver these types of chips in commercial quantities at a price point that mere mortals can afford.

Back illuminated chilled sensors can provide high Qe with low noise - they just cost more at present.

Although I suspect that many of the near future advances are going to come along by way of using existing technology - smart ideas using existing technology to push the frontiers.

I hope Black Silicon can bring the promises they make to fruition - Qe- 10,000%, spectral sensitivities from 400 to 1500nm (even 2500nm is being stated)
http://www.sionyxinc.com
They make some extraordinary claims that appear to be well backed up by Harvard Research and they are being well funded.

This recent multi CCD imaging system patent by Olympus may be of interest ?
http://www.google.com/patents?id=dwS...age&q=&f=false

It uses a gapless multiprism design to separate the various light channels into three channels - one for each of three CCDs
It appears that the amount of light being reflected and transmitted at each layer is close to 100% hence almost no losses.

This would mean no antialiasing filters, no bayer filters or debayering algorithms and a vastly improved imaging system to current RGB techniques for Astro - just swing that out of the way for narrow band !

As for having telescopes which have flat fields across these larger image planes - many already exist now, although most are higher end scopes, they are not out of the reach of amateurs.
Its really a matter of supply and demand - as soon as the demand arrives the supply willl catch up.

It will all happen at its own pace and we will all yearn for something even better when it arrives, when in fact most of us are not yet capable of maximising our use with the current technology devices !

[Peter Ward]

Originally Posted by rally

This recent multi CCD imaging system patent by Olympus may be of interest

..................

As for having telescopes which have flat fields across these larger image planes - many already exist now, .............

3CCD Video cameras have been out for a while...I am surprised by the Olympus patent app for what has been for over a decade a (video) industry standard....but who knows?

I have two telescopes with fairly wide *and* flat fields..Taka FSQ and AP155 (4" field flattner).....my RCOS is OK....AP130 less so...I'd love to know what else is out there. (Don't say CDK's IMHO they don't cut it)

[mithrandir (Andrew)]

Originally Posted by AlexN

Andrew - Do you know the model number of the CCD in the A200? Could be very interesting.

I tried to find that before I posted my previous reply. No luck.

[DavidU (Dave)]

Black silicon looks like the "next big thing".