A lot of CMOS sensors have Amp glow but some of the higher end Sony CMOS sensors don’t. Many of the CMOS sensors that are picked up come from video applications where amp glow is of little consequence; take the IMX183 as an example.
The exposure time required is a consequence of pixel size, focal ratio and read noise. Low read noise, large pixel on a fast telescope (thinking GSENSE-6060) you can potentially get away with 30s exposures and being read noise limited with narrowband.
A lot of CMOS sensors have Amp glow but some of the higher end Sony CMOS sensors don’t. Many of the CMOS sensors that are picked up come from video applications where amp glow is of little consequence; take the IMX183 as an example.
The exposure time required is a consequence of pixel size, focal ratio and read noise. Low read noise, large pixel on a fast telescope (thinking GSENSE-6060) you can potentially get away with 30s exposures and being read noise limited with narrowband.
I admit -
I don't know enough about CMOS sensors.
30 seconds for a narrowband image?
I only know my KAF8300 mono CCD sensor.
With Ha it would take at least 10 minutes before I could see a good picture forming.
One frame I did was 30 minutes.
Surely the signal is going to buried in the dark frame or offset noise at 30 seconds?
In regard to these new IMX455 based cameras, if you already owned a full frame camera using the older Kodak CCD sensors, will you see any improvement?
Why bother paying $5-6K for a new camera that outputs exactly the same images that you can produce with the Kodak sensor.
In my mind the new IMX455 CMOS cameras should provide an excellent image in 2 hours compared to what you could get in 10 hours with the Kodak sensor. Or the IMX455 based camera can pull in faint distant galaxies that would be swamped by read noise on the Kodak based camera.
Otherwise I can't see the point of spending that amount of money, if you are just starting out with no frame of reference then that's different, go for it.
Here are some sample images taken with the IMX455 cameras, they don't look any better than what I have seen with Kodak sensors.
Yes Bill, I know I reached a practical limit with my ICX814 attached to a 4” on the Helix, and would like to see a similar or even better a deeper tri/quad colour narrowband image of the Helix taken with CMOS.
Is it true that you can't take long sub frames with CMOS sensors due to AMP glow?
If so then that's a problem with narrow band as a 3 minute subframe
hardly gives you any signal.
cheers
Allan
It is not really the case. 5 minute subs for instance on my ASI294 will produce a noticeable glow in several locations, but darks calibrate it out. You could do the same for ten minute subs but that just led to too many blown out areas in most images so I generally don't go that long.
The glows look hideous in isolation on a stretched sub (Light or dark frames) but if you don't stretch the data you literally end up looking at a black screen in the case of darks and the master darks generated subtract the glows effectively.
It is not really the case. 5 minute subs for instance on my ASI294 will produce a noticeable glow in several locations, but darks calibrate it out. You could do the same for ten minute subs but that just led to too many blown out areas in most images so I generally don't go that long.
The glows look hideous in isolation on a stretched sub (Light or dark frames) but if you don't stretch the data you literally end up looking at a black screen in the case of darks and the master darks generated subtract the glows effectively.
Hi Paul,
I don't know about CMOS.
Here is a 30 minute Ha single frame with my KAF8300 mono - QHY9M.
Field rotation wrecked it as I didn't set up the polar alignment accurately enough
but it was very satisfying to get a good photo with one exposure.
This is a linear stretch in NASA FITS liberator and
only one small adjustment in Photoshop curves.
You can't do that with CMOS can you?
For that target? Probably not as it would swamp it in that time, but that's not the way the CMOS cameras are normally used. I have heard of some people doing longer exposures but I don't have the links for results.
Given how bright that target is, I'd be tempted to do 30 x 1 min exposures of Ha using something like the ASI1600. I bet the integration result would be at least as good as that shot.
For that target? Probably not as it would swamp it in that time, but that's not the way the CMOS cameras are normally used. I have heard of some people doing longer exposures but I don't have the links for results.
Given how bright that target is, I'd be tempted to do 30 x 1 min exposures of Ha using something like the ASI1600. I bet the integration result would be at least as good as that shot.
Hi Chris,
Yes or maybe 6 x 5 minute sub frames stacked?
I will follow these CMOS cameras with great interest.
Its just the megapixel race to get more resolution. Also marketing. 61mp must be better than 42mp right? And in many uses it may be, just not all. I doubt these CMOS sensors will challenge the older Kodak CCDs for now.
Sony has been doing a great job of increasing the number of megapixels and improving noise levels at the same time. But they seem to have run out of tweaks at this stage or the next set of tweaks is a bigger change. They do have some tricks up their sleeves for the next round of tweaks. Quad Bayer filters for one. Possibly organic films on sensors that have much greater dynamic range.
This 61mp sensor though seems to be the same architecture as the 42mp sensor and has 1 stop worse noise performance. You can see the noise levels in the DPReview studio comparison tool. Most users of the Sony A7Riv (that uses this sensor) seem happy with the noise performance but acknowledge its worse than the 42mp A7riii but autofocus and other improvements still make it worthwhile.
Smaller pixels as the overall size of the sensor usually is the same just more in the same area.
It should be interesting to see how Canon goes with their higher megapixel sensor as in the past when they put out a 50mp sensor it was simply noisier. But their latest mirrorless APSc M6ii sensor seems to have evolved.
Its just the megapixel race to get more resolution. Also marketing. 61mp must be better than 42mp right? And in many uses it may be, just not all. I doubt these CMOS sensors will challenge the older Kodak CCDs for now.
Sony has been doing a great job of increasing the number of megapixels and improving noise levels at the same time. But they seem to have run out of tweaks at this stage or the next set of tweaks is a bigger change. They do have some tricks up their sleeves for the next round of tweaks. Quad Bayer filters for one. Possibly organic films on sensors that have much greater dynamic range.
This 61mp sensor though seems to be the same architecture as the 42mp sensor and has 1 stop worse noise performance. You can see the noise levels in the DPReview studio comparison tool. Most users of the Sony A7Riv (that uses this sensor) seem happy with the noise performance but acknowledge its worse than the 42mp A7riii but autofocus and other improvements still make it worthwhile.
Smaller pixels as the overall size of the sensor usually is the same just more in the same area.
It should be interesting to see how Canon goes with their higher megapixel sensor as in the past when they put out a 50mp sensor it was simply noisier. But their latest mirrorless APSc M6ii sensor seems to have evolved.
Greg.
Hi Greg,
It makes me wonder whether the problems of the sensors are
far behind us and it really boils down to what lens you put in front of the sensor?
All the rest is just marketing.
I think a lot of the public are dupes thinking more pixels
equals a better picture.
My wifes opinion (She was a professional/industrial photographer) when I bought her last camera was that the mega pixel wars were out of hand and it was just marketing. And that was with me buying her a Nikon D3 with 12mp and about 8.4 micron pixels.
It makes me wonder whether the problems of the sensors are
far behind us and it really boils down to what lens you put in front of the sensor?
All the rest is just marketing.
I think a lot of the public are dupes thinking more pixels
equals a better picture.
cheers
Allan
I think it depends on how many new ideas the engineers can come up with to keep improving the sensors. It must get harder to do.
My wifes opinion (She was a professional/industrial photographer) when I bought her last camera was that the mega pixel wars were out of hand and it was just marketing. And that was with me buying her a Nikon D3 with 12mp and about 8.4 micron pixels.
With mobile phones -
it's funny how they put a monster sensor with a huge number of pixels
behind a tiny lens smaller than a pea.
It makes me wonder whether the problems of the sensors are
far behind us and it really boils down to what lens you put in front of the sensor?
All the rest is just marketing.
I think a lot of the public are dupes thinking more pixels
equals a better picture.
cheers
Allan
Picked it in one.
Most commercial telescope optics deliver 10 micron (or larger) spot sizes in their central zones, and quickly bloat out to more than two or more times that figure.
Add seeing disturbances and 30 micron will be a good figure on most nights.
75Mp in a 35mm sensor might sound awesome, but the reality is: those 2 micron pixels are doing you no favors at all.
Your can over-sample blurred data as much as you like. It will still look blurred.
However an optical system that delivers say 5 micron spot sizes and superb mount (+AO) that does not smear the incoming starlight beyond what the optics can deliver is FAR more important than megapixels.