Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

The Future of Imaging - Is everything heading OSC


Recommended Posts

7 minutes ago, Adam J said:

They already do :), but only bother when they think they can sell sufficient numbers to make a batch as they cant change the whole production line over for only 1000 chips. 

As a interesting fact there was not IMX183 mono until ASI / QHY persuaded sony to do a run.

Adam

Learn something new everyday here.

Shame it's not as simple as just telling the system to skip this step for x number of times as a Friday night special.

Link to comment
Share on other sites

2 minutes ago, MarkAR said:

Shame it's not as simple as just telling the system to skip this step for x number of times as a Friday night special.

If only it were that simple - But the reality is that manufacturers wont make changes unless there is enough money it in for them to make it worth their while.

Imagine you are a car manufacturer and you only make red cars by the thousands - then one day someone comes along and asks you to make 5 blue cars at the same price. To make the blue cars, you need to buy blue paint and then remove all the red paint from all the machines & clean down and flush out all the pipes to get rid of the red paint, and swap all the machines over to blue paint and then do it all over again once you've finished making the 5 blue cars - so you've went to all that effort and instead of making money you lost money during all the downtime making the changes between red paint & blue paint and then back again.

  • Like 1
Link to comment
Share on other sites

So here are another 4 cooled mono sensors in astronomy cameras that have been added to the list i posted previously .

IMX 273

IMX 252

IMX 250

IMX 249

IMX 428

A couple of those are of a size suited to DSO imaging the IMX 428 being part of the ATIK ACIS range and sold by FLO. Not seen any images from one yet mind you.

Adam

Edited by Adam J
Link to comment
Share on other sites

If Atik / QSI comes back with a price where it's not economically viable to repair my beloved QSI then i think i will just get an ASI1600mm-Pro as a lot of the current sony mono chips are a step backwards for me  - They are much better with respect to noise but i loose too much real estate.

I also stumbled across a very interesting real-world comparison between the QSI 683(Kaf8300) and the QSI690 (IMX814) from Sara Wager - showing her direct comparisons of her own data she found that the KAf8300 actually produced cleaner images with less noise - despite the specs of the camera specs saying otherwise. This is very interesting coming from one of the words most respected imagers.

https://www.swagastro.com/real-world-comparison---kodak--sony-chip.html

Rich.

Edited by Northernlight
Link to comment
Share on other sites

2 hours ago, Northernlight said:

If Atik / QSI comes back with a price where it's not economically viable to repair my beloved QSI then i think i will just get an ASI1600mm-Pro as a lot of the current sony mono chips are a step backwards for me  - They are much better with respect to noise but i loose too much real estate.

I also stumbled across a very interesting real-world comparison between the QSI 683(Kaf8300) and the QSI690 (IMX814) from Sara Wager - showing her direct comparisons of her own data she found that the KAf8300 actually produced cleaner images with less noise - despite the specs of the camera specs saying otherwise. This is very interesting coming from one of the words most respected imagers.

https://www.swagastro.com/real-world-comparison---kodak--sony-chip.html

Rich.

Yes sensor size matters. Just be sure that you can live with the microlensing on the ASI1600mm pro. 

Link to comment
Share on other sites

1 hour ago, Northernlight said:

Well it's seems good enough for imagers producing some top class images and seeing there is currently no alternative available in a similar size format then i dont have much choice at the moment.

 

I do agree I own one myself and its great, just makes for things like Gama Cas and Horses head having ugly stars.

Narrower filters with good out of band blocking help reduce the effect as the attenuate the star brightness. I have found that my 5nm Ha Astrodon is much better than the 7nm I had before. More so than you would expect from just the difference is bandwidth so I am assuming the 7nm was leaking some out of band too.

Also the effect seems dependent on F-ratio. As you get faster it seems to reduce.

I guess the next size down would be the IMX253 in or IMX304 in the Moravian, ATIK and StarlightXpress offerings, its actually quite a bit bigger then the 460ex sized CCD sensor, but still quite a bit smaller than the panasonic chip.

Adam

Edited by Adam J
Link to comment
Share on other sites

7 hours ago, Adam J said:

Yes sensor size matters. Just be sure that you can live with the microlensing on the ASI1600mm pro. 

Well i lived with it on the QSI 683, as does every other Kaf8300  owner as i'm pretty sure the chip has Microlenses

Link to comment
Share on other sites

Are you saying  a new QSI683 or Moravian is better than a QHY268 ? I was saving for a KAF8300 but read some fud on another US site which put me off and was looking at the QHY268 instead.

My rationale  being that it would equate to a modern 6.5Mp - 8Mp Mono camera taking into account the bayer  filter but the QHY268 would have better sensor characteristics and circuitry  and longer  support

Money wise the Moravian  is similar price to the QHY (then I was thinking if I went with the Moravian why not spend a bit more for the QSI)

Link to comment
Share on other sites

1 hour ago, Northernlight said:

Well i lived with it on the QSI 683, as does every other Kaf8300  owner as i'm pretty sure the chip has Microlenses

The microlensing effect on a KAF8300 is nothing like the artefacts produced by the 1600. 

  • Like 1
Link to comment
Share on other sites

1 hour ago, billhinge said:

Are you saying  a new QSI683 or Moravian is better than a QHY268 ? I was saving for a KAF8300 but read some fud on another US site which put me off and was looking at the QHY268 instead.

My rationale  being that it would equate to a modern 6.5Mp - 8Mp Mono camera taking into account the bayer  filter but the QHY268 would have better sensor characteristics and circuitry  and longer  support

Money wise the Moravian  is similar price to the QHY (then I was thinking if I went with the Moravian why not spend a bit more for the QSI)

I would be shocked if a KAF8300 based camera did not outperform the QHY268 in narrowband so long as you can do the 30min subs that it requires.

Edited by Adam J
Link to comment
Share on other sites

2 hours ago, Adam J said:

I would be shocked if a KAF8300 based camera did not outperform the QHY268 in narrowband so long as you can do the 30min subs that it requires.

 

2 hours ago, tooth_dr said:

The microlensing effect on a KAF8300 is nothing like the artefacts produced by the 1600. 

Thanks, I  had read of microlensing as an artefact of the KAF8300, I  was using the  panny micro 4/3 back  in 2012 which is the same chip as the 1600. Not that I noticed noticeable microlensing but I've seen some examples posted on the  interweb

I believe the KAF8300 was an original 4/3 chip (predecessor of micro 4/3) , I hadn't heard of a direct comparison saying the 1600 (micro 4/3) was worse in this respect.  Amp  glow and banding was my issue with the panny chip 

 if you took arbitrary more shorter exposures with the QHY268 would they become comparable? The QSI and Morovian seem to have better cooling, maybe that helps?

Just a  pity there isn't a side by side comparison. I ended up  comparing downloading images from Astrobin in comparing and tweaking similarly  in Pixinsight, QHY268 didn't look bad to me which is what got me  thinking (actually I wanted to go with CCD but the CMOS examples look really good)

Link to comment
Share on other sites

4 hours ago, Adam J said:

I would be shocked if a KAF8300 based camera did not outperform the QHY268 in narrowband so long as you can do the 30min subs that it requires.

Adam, the QSI 683 doesn't need 30 mins subs to get good results - I live under bortle 6 skies and have managed some decent images.

I took all of these with 600s subs on my qsi 683 -  and they were all about 10-12  frames per channel. - and i'm no expert at processing by any means but managed to get do ok using 600s subs, well except for the last image - got a bit carried away with the processing and it's a bit overcooked and and a bit in your face - never the less - QSI683 produces some nice images with relatively small amounts of data.

rich_elephantstrunk_RGB_dbe_cc_HT_nr_sho__lrgb_ct_final

 

V2_Final

 

RGB_Final_Crop_V2

 

 

 

Link to comment
Share on other sites

1 hour ago, billhinge said:

 

Thanks, I  had read of microlensing as an artefact of the KAF8300, I  was using the  panny micro 4/3 back  in 2012 which is the same chip as the 1600. Not that I noticed noticeable microlensing but I've seen some examples posted on the  interweb

I believe the KAF8300 was an original 4/3 chip (predecessor of micro 4/3) , I hadn't heard of a direct comparison saying the 1600 (micro 4/3) was worse in this respect.  Amp  glow and banding was my issue with the panny chip 

 if you took arbitrary more shorter exposures with the QHY268 would they become comparable? The QSI and Morovian seem to have better cooling, maybe that helps?

Just a  pity there isn't a side by side comparison. I ended up  comparing downloading images from Astrobin in comparing and tweaking similarly  in Pixinsight, QHY268 didn't look bad to me which is what got me  thinking (actually I wanted to go with CCD but the CMOS examples look really good)

I think that if you want to do narrow band mono is very much the way to go. The only time I might consider a OSC myself is for a mobile setup when I could visit dark sites.

Adam

 

Link to comment
Share on other sites

On 31/05/2020 at 09:13, Northernlight said:

I also stumbled across a very interesting real-world comparison between the QSI 683(Kaf8300) and the QSI690 (IMX814) from Sara Wager - showing her direct comparisons of her own data she found that the KAf8300 actually produced cleaner images with less noise - despite the specs of the camera specs saying otherwise. This is very interesting coming from one of the words most respected imagers.

https://www.swagastro.com/real-world-comparison---kodak--sony-chip.html

Rich.

I've always wondered whether this came from differences in the dynamic ranges of the cameras that fed into the final images.  When I look at the two comparisons I kodak chip seems more 'washed out' compared the sony chip. In this I mean that the contrast between the areas with less signal and more signal are more pronounced in the sony ccd but are noisier.  At a very broad discussion level then the lower dynamic range has fewer 'bands' in the 16 bit system to drop the signal into.  As such each band has more information but because there are less 'bands' there can be less contrast.  In comparison in a higher dynamic range camera you have more bands but each band will effectively have less signal - so the contract in regions is higher but the signal in each band is lower so hence greater contrast but higher noise.  In principle the more exposures you take the more that this remove this issue though.   Just some thoughts as to why the sony appears noisier even thought the figures suggest otherwise.

Link to comment
Share on other sites

9 hours ago, Northernlight said:

Adam, the QSI 683 doesn't need 30 mins subs to get good results - I live under bortle 6 skies and have managed some decent images.

I took all of these with 600s subs on my qsi 683 -  and they were all about 10-12  frames per channel. - and i'm no expert at processing by any means but managed to get do ok using 600s subs, well except for the last image - got a bit carried away with the processing and it's a bit overcooked and and a bit in your face - never the less - QSI683 produces some nice images with relatively small amounts of data.

All cameras can get great images.  The benefit of the lower noise cameras is that the optimum sub-lengths can be shorter because the camera noise becomes overwhelmed by other sources of noise.  Hence the benefit of the CMOS and the Sony CCDs.  It's especially true for ultranarrow band imaging.  As such burdens on the mount is lessened.  This presentation gives the best information that I've found:-

 

Link to comment
Share on other sites

1 hour ago, Whirlwind said:

I've always wondered whether this came from differences in the dynamic ranges of the cameras that fed into the final images.  When I look at the two comparisons I kodak chip seems more 'washed out' compared the sony chip. In this I mean that the contrast between the areas with less signal and more signal are more pronounced in the sony ccd but are noisier.  At a very broad discussion level then the lower dynamic range has fewer 'bands' in the 16 bit system to drop the signal into.  As such each band has more information but because there are less 'bands' there can be less contrast.  In comparison in a higher dynamic range camera you have more bands but each band will effectively have less signal - so the contract in regions is higher but the signal in each band is lower so hence greater contrast but higher noise.  In principle the more exposures you take the more that this remove this issue though.   Just some thoughts as to why the sony appears noisier even thought the figures suggest otherwise.

That effect is largely mitigated by stacking. When you stack you are averaging values so your final image is not restricted to the same number of "bands" as in the sub frames. 

Link to comment
Share on other sites

3 hours ago, Adam J said:

That effect is largely mitigated by stacking. When you stack you are averaging values so your final image is not restricted to the same number of "bands" as in the sub frames. 

Yes, but there is a balance in this, the more you stack the more you mitigate the effect - it doesn't disappear when you just stack two images for example.  In the test examples there were relatively few stacks so hence there is possibility that with extra the dynamic range the Sony CCD has there simply wasn't enough images in the stack to offset the additional bands.  This then becomes one of the benefits of a lower noise camera, you can more quickly overcome the camera noise and hence sub-exposures can be shorter, increasing the number of images in the stack and mitigating the potential issue associated with more 'bands' because of the greater dynamic range.

Link to comment
Share on other sites

1 hour ago, Whirlwind said:

Yes, but there is a balance in this, the more you stack the more you mitigate the effect - it doesn't disappear when you just stack two images for example.  In the test examples there were relatively few stacks so hence there is possibility that with extra the dynamic range the Sony CCD has there simply wasn't enough images in the stack to offset the additional bands.  This then becomes one of the benefits of a lower noise camera, you can more quickly overcome the camera noise and hence sub-exposures can be shorter, increasing the number of images in the stack and mitigating the potential issue associated with more 'bands' because of the greater dynamic range.

2 images would double the number of levels, but who on earth stacks only two frames anyway? By the time you have taken 16 frames your going to be ok.

Yes lower read noise lets you take shorter exposures and get to that point faster. But as the KAF8300 has a 16-bit A/D and the ASI1600mm pro a 12-bit A/D it takes more frames for the ASI1600mm pro to overcome quantization noise at lower gains.

Adam

Link to comment
Share on other sites

2 hours ago, Adam J said:

2 images would double the number of levels, but who on earth stacks only two frames anyway? By the time you have taken 16 frames your going to be ok.

Yes lower read noise lets you take shorter exposures and get to that point faster. But as the KAF8300 has a 16-bit A/D and the ASI1600mm pro a 12-bit A/D it takes more frames for the ASI1600mm pro to overcome quantization noise at lower gains.

Adam

In the UK?  When you are trying ultra narrowband, set 30minute sub-exposures and then the clouds roll in, maybe?

I think I've been misunderstood.  I was hypothesising why there would be a difference between a more sensitive Sony CCD and the 'less' sensitive Kodak (i.e. both 16bit).  Even where similar images were taken why there was more noise in the Sony, but looked less washed out.  Agreed that a 12 bit camera will need a lot more frames (to offset the quantization noise) but I wasn't comparing these two.  I was conjecturing that the lower dynamic range of the Kodak was the cause.  It should be a relatively simple test though to take the ideal exposures for each camera (in reality using a fast system so that camera noise was overwhelmed earlier) and then combine to see whether the noise in the Sony was still as high.

Link to comment
Share on other sites

21 minutes ago, Whirlwind said:

In the UK?  When you are trying ultra narrowband, set 30minute sub-exposures and then the clouds roll in, maybe?

I think I've been misunderstood.  I was hypothesising why there would be a difference between a more sensitive Sony CCD and the 'less' sensitive Kodak (i.e. both 16bit).  Even where similar images were taken why there was more noise in the Sony, but looked less washed out.  Agreed that a 12 bit camera will need a lot more frames (to offset the quantization noise) but I wasn't comparing these two.  I was conjecturing that the lower dynamic range of the Kodak was the cause.  It should be a relatively simple test though to take the ideal exposures for each camera (in reality using a fast system so that camera noise was overwhelmed earlier) and then combine to see whether the noise in the Sony was still as high.

Ok, I think you need to offer up some maths in support of what your saying as its not making any sense to me at the moment.

So long as your not saturating dynamic range should not cause anything to be washed out so long as you have a 16 bit a/d on top of stacking your not going to have any issues with quantization noise. Remember everthing you are viewing online is a 8-bit jpg anyway so 0-255 levels, thats your lot and both camera will be useing all those levels in all situations unless someone clipped their black point.

The uk weather might effect you choice of sensor but that has nothing to do with your initial argument.

Perhapse offer up some example images to help people understand the point you are trying to make.

Adam

 

Edited by Adam J
Link to comment
Share on other sites

I have a friend that uses both a 383( kaf8300) and a  atik 460, he does 10-15 subs on both camera's and says the most initial noise on the 383 easily calibrates out.. There's very little comparison but I'd say the 383 has the edge, not by much thou

Is the demise of the mono chip the new demise of the ccd chat.. 

Link to comment
Share on other sites

Anyway i think we have strayed a little off topic - so back to the subject at hand - i dont think anyone can predict what will happen over the next 2 years but i do think CCD Chips have had their day and will eventually all be phased out in favour of CMOS chips which will be driven largely by the consumer digital camera market.

Will Cmos Astro camera's releases be predominantly OSC vs Mono - who knows - I certainly hope not. So i will certainly be watching the camera market with great interest to see where the trends take the astro camera market

Edited by Northernlight
update
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.