ZWO Bit depth question.

[r3i]

On 10/16/2016 at 10:02, ollypenrice said:

Très gentil, madame! I did much the same sum. The discrepency may be to do with active pixel area, perhpas. But why don't they just give the blessèd information since we so obviously need it!

Olly

A decent FOV calculator will give you the chip size info, e.g.: http://www.newastro.com/book_new/camera_app.html

You give it the pixel size & number of pixels per row & column and it does the maths. 

 

[ollypenrice]

15 minutes ago, r3i said:

A decent FOV calculator will give you the chip size info, e.g.: http://www.newastro.com/book_new/camera_app.html

You give it the pixel size & number of pixels per row & column and it does the maths. 

 

Thanks. Both of mine ask for mm so this is a good link. Cheers,

Olly

[Tiny Small]

I find it odd that when talking about these cameras, everyone only thinks of the ASI1600. Maybe because it has a 16Mp sensor?  The problem I see with this is, that with the pixel size, it's very similar to a DSLR sensor in terms of field of view. I have a ZWO cooled mono camera but not the 1600. I went for the 178 based on pixel size, resolution and the FOV on my 130 PDS. With this camera I get a 6Mp image (more than enough for most astro stuff) at 0.87"/p (with the 0.9x CC fitted). I would need a huge scope and a mount that could carry it to get that scale with the 1600. The 178 is also 14 bit as opposed to 12 (still wish it was 16 though).

These cameras are good but still in their infancy and as such, suffer from some slightly annoying quirks. If you read the CN threads about the 178, you will see issues that owners (including me) have found. One of the biggest bones of contention is amp glow. Un-calibrated on 300, 600 and 900 second exposures, it rages from terrible to horrific. But it calibrates out perfectly well... though that isn't always a simple process. For some reason, to calibrate 120s exposures, I have to use 300s darks. I'm sure that if I redid my darks library I could get better calibration files. Apparently the sensor resets on every start up, but only on this model. Anyway, despite it not being simple, it is relatively easy to calibrate out. On short exposures, it isn't an issue and the CN threads are full of pretty decent images gained through stacking a few hundred shorter subs.

The images, once calibrated are also pretty clean in regards to noise as stacking more shorter images really does reduce it massively. Attached is a 100% crop of a misaligned stack which shows the difference in noise between a single sub and the rest of the stack (153 x 120s lum). The image is stretched in DSS and you can see part of a single sub poking out the bottom of the frame.

 

This is before any other processing.

One thing that I think people forget about these cameras is the price. My camera was £708. How much would I have to spend to get a 6 Mp camera from Atik or Starlight Express? A lot more. These are not designed to replace or directly compete with CCDs, but they are an affordable entry into cooled mono imaging. And the coolers on these cameras are excellent. I use set point cooling at -20 C and it hits that fairly quickly and stays there. Occasionally I will see a 0.1 to 0.5 degree change as the sensor responds to a changing environment, but then it stabilises again bang on -20.

If you already have a decent CCD, there would be little to tempt you into one of these other than to gain resolution at a much lower cost. If on the other hand, you don't have a CCD and don't fancy spending a small fortune, these can be an excellent choice.

 

EDIT; One thing I forgot to mention is that the image above is at a gain of 240, which on my camera is almost halfway to maximum.

[ollypenrice]

2 hours ago, Tiny Small said:

I find it odd that when talking about these cameras, everyone only thinks of the ASI1600. Maybe because it has a 16Mp sensor?  The problem I see with this is, that with the pixel size, it's very similar to a DSLR sensor in terms of field of view. I have a ZWO cooled mono camera but not the 1600. I went for the 178 based on pixel size, resolution and the FOV on my 130 PDS. With this camera I get a 6Mp image (more than enough for most astro stuff) at 0.87"/p (with the 0.9x CC fitted). I would need a huge scope and a mount that could carry it to get that scale with the 1600. The 178 is also 14 bit as opposed to 12 (still wish it was 16 though).

These cameras are good but still in their infancy and as such, suffer from some slightly annoying quirks. If you read the CN threads about the 178, you will see issues that owners (including me) have found. One of the biggest bones of contention is amp glow. Un-calibrated on 300, 600 and 900 second exposures, it rages from terrible to horrific. But it calibrates out perfectly well... though that isn't always a simple process. For some reason, to calibrate 120s exposures, I have to use 300s darks. I'm sure that if I redid my darks library I could get better calibration files. Apparently the sensor resets on every start up, but only on this model. Anyway, despite it not being simple, it is relatively easy to calibrate out. On short exposures, it isn't an issue and the CN threads are full of pretty decent images gained through stacking a few hundred shorter subs.

The images, once calibrated are also pretty clean in regards to noise as stacking more shorter images really does reduce it massively. Attached is a 100% crop of a misaligned stack which shows the difference in noise between a single sub and the rest of the stack (153 x 120s lum). The image is stretched in DSS and you can see part of a single sub poking out the bottom of the frame.

 

This is before any other processing.

One thing that I think people forget about these cameras is the price. My camera was £708. How much would I have to spend to get a 6 Mp camera from Atik or Starlight Express? A lot more. These are not designed to replace or directly compete with CCDs, but they are an affordable entry into cooled mono imaging. And the coolers on these cameras are excellent. I use set point cooling at -20 C and it hits that fairly quickly and stays there. Occasionally I will see a 0.1 to 0.5 degree change as the sensor responds to a changing environment, but then it stabilises again bang on -20.

If you already have a decent CCD, there would be little to tempt you into one of these other than to gain resolution at a much lower cost. If on the other hand, you don't have a CCD and don't fancy spending a small fortune, these can be an excellent choice.

 

EDIT; One thing I forgot to mention is that the image above is at a gain of 240, which on my camera is almost halfway to maximum.

Helpful post. Thanks. According to my calculations the 1600 would give me 0.77"PP on our TEC 140 (True FL 1015mm.) That's not a bad place to be for galaxies on a moderate sized scope and we've imaged here successfully at 0.66"PP so it stikes me as do-able.

I think we tend to regard as a 'normal' field of view anything that we are using at the moment! I would find it a mighty wrench to come down from full frame for nebular imaging since, even now, we are often doing mosaics. But for galaxies and planetaries I could live with the 1600 or even a Sony 460/490.

I guess that with the amp glow issue you are pretty much stuck with darks for calibration. With CCD I use bias as dark and bad pixel map. On one camera a bit of amp glow does hit one corner but it's an easy cosmetic fix in Ps. This ZWO amp glow sounds more problematic and probably demabnds darks.

Olly

[Tiny Small]

4 minutes ago, ollypenrice said:

This ZWO amp glow sounds more problematic and probably demabnds darks.

Olly

The bigger problem is the sensor recalibrating... meaning a darks library can become redundant very quickly. I have no issue with having to use darks, but what I would have an issue with is having to retake them every session. Using a dedicated darks library should be one of the attractions of a cooled sensor. As it stands, now I've found (I think) useable settings and a sensor orientation that I probably won't change, I'm going to redo the darks library and see how I get on. AFAIK, the 1600 doesn't suffer from this issue, but there are plenty of users on here that will correct me if I'm wrong.

Even with this issue though, the price makes them an attractive option for a lot of people. It's amazing how much you're prepared to put up with if the price is right.

[ollypenrice]

1 minute ago, Tiny Small said:

 

Even with this issue though, the price makes them an attractive option for a lot of people. It's amazing how much you're prepared to put up with if the price is right.

Very true!

[riklaunim]

Dark current higher for CMOS in 2016? Are we talking about the same chips?

[harry page]

13 hours ago, sharkmelley said:

What makes them harder to cool than a CCD?

Is the ASI 1600 dark current really higher than CCD?  I haven't seen figures for the ASI 1600 sensor in particular and I wholeheartedly agree that for an astro-camera they really ought to be published.  However the popular KAF-8300 CCD sensor has a dark current of around 1.5e/sec at 20C whilst Sony Exmor CMOS has a dark current of around 0.15e/sec at 20C.  I don't know how the ASI 1600 sensor compares with Sony Exmor.

Mark

Hi

the CMOS internal circuit dissipates significant heat straight into the chip , making cooling more difficult .

The dark current on my sx h35 is Less than 0.08 electrons/second ( which is considered noisy )  , I do not have the numbers for the Cmos chips but I will investigate as I find the 0.15 very low indeed and if were true

cooling would hardley be required !!!!!

[harry page]

3 hours ago, riklaunim said:

Dark current higher for CMOS in 2016? Are we talking about the same chips?

Hi

we have to remember that most ccd / cmos chips are not desighned for astronomy and have problems that camera makers have to adapt to

Hence the need for cooling , flushing etc.

Harry

[harry page]

7 hours ago, Tiny Small said:

I find it odd that when talking about these cameras, everyone only thinks of the ASI1600. Maybe because it has a 16Mp sensor?  The problem I see with this is, that with the pixel size, it's very similar to a DSLR sensor in terms of field of view. I have a ZWO cooled mono camera but not the 1600. I went for the 178 based on pixel size, resolution and the FOV on my 130 PDS. With this camera I get a 6Mp image (more than enough for most astro stuff) at 0.87"/p (with the 0.9x CC fitted). I would need a huge scope and a mount that could carry it to get that scale with the 1600. The 178 is also 14 bit as opposed to 12 (still wish it was 16 though).

These cameras are good but still in their infancy and as such, suffer from some slightly annoying quirks. If you read the CN threads about the 178, you will see issues that owners (including me) have found. One of the biggest bones of contention is amp glow. Un-calibrated on 300, 600 and 900 second exposures, it rages from terrible to horrific. But it calibrates out perfectly well... though that isn't always a simple process. For some reason, to calibrate 120s exposures, I have to use 300s darks. I'm sure that if I redid my darks library I could get better calibration files. Apparently the sensor resets on every start up, but only on this model. Anyway, despite it not being simple, it is relatively easy to calibrate out. On short exposures, it isn't an issue and the CN threads are full of pretty decent images gained through stacking a few hundred shorter subs.

The images, once calibrated are also pretty clean in regards to noise as stacking more shorter images really does reduce it massively. Attached is a 100% crop of a misaligned stack which shows the difference in noise between a single sub and the rest of the stack (153 x 120s lum). The image is stretched in DSS and you can see part of a single sub poking out the bottom of the frame.

 

This is before any other processing.

One thing that I think people forget about these cameras is the price. My camera was £708. How much would I have to spend to get a 6 Mp camera from Atik or Starlight Express? A lot more. These are not designed to replace or directly compete with CCDs, but they are an affordable entry into cooled mono imaging. And the coolers on these cameras are excellent. I use set point cooling at -20 C and it hits that fairly quickly and stays there. Occasionally I will see a 0.1 to 0.5 degree change as the sensor responds to a changing environment, but then it stabilises again bang on -20.

If you already have a decent CCD, there would be little to tempt you into one of these other than to gain resolution at a much lower cost. If on the other hand, you don't have a CCD and don't fancy spending a small fortune, these can be an excellent choice.

 

EDIT; One thing I forgot to mention is that the image above is at a gain of 240, which on my camera is almost halfway to maximum.

Hi

 

I agree with a lot of what you say ( yes value for money ), these cameras are not ccd beating cameras ( Yet )  , but I can see the future is Cmos and if manufactures can overcome the limitations of the chip

the possability of very high QE etc are very exciting

 

Regards

 

Harry

[harry page]

Hi

Just for your info this is a 60 sec dark ( cooled ) from a zwo 174 , see if you can spot the amp glow

Please do not think I am anti cmos , just want people to know current limitations

Kind regards

Harry

 

[Tiny Small]

1 minute ago, harry page said:

Hi

Just for your info this is a 60 sec dark ( cooled ) from a zwo 174 , see if you can spot the amp glow

Please do not think I am anti cmos , just want people to know current limitations

Kind regards

Harry

 

The uncooled 174 is notorious amongst the CN lot for exactly that reason. It's the worst of the bunch... though for planetary it's hard to beat.

[harry page]

7 minutes ago, Tiny Small said:

The uncooled 174 is notorious amongst the CN lot for exactly that reason. It's the worst of the bunch... though for planetary it's hard to beat.

Hi

This is from the cooled camera !!

Harry

[Tiny Small]

Wow.

[Thalestris24]

The qhy8l used to have a bad amp glow problem but they updated the driver and it was much diminished thereafter.

Louise

[Tiny Small]

26 minutes ago, Thalestris24 said:

The qhy8l used to have a bad amp glow problem but they updated the driver and it was much diminished thereafter.

Louise

Hopefully ZWO will address this too.

[sharkmelley]

2 hours ago, harry page said:

Hi

the CMOS internal circuit dissipates significant heat straight into the chip , making cooling more difficult .

The dark current on my sx h35 is Less than 0.08 electrons/second ( which is considered noisy )  , I do not have the numbers for the Cmos chips but I will investigate as I find the 0.15 very low indeed and if were true

cooling would hardley be required !!!!!

Take a look at Roger Clark's review of the Canon 7DII (figure 3): http://www.clarkvision.com/reviews/evaluation-canon-7dii/  He measures the dark current to be around 0.08 e/pixel/sec at 20C (temperature taken from EXIF).  My Sony A7S gives a dark current of around 0.15-0.18 e/pixel/sec at an indoor ambient temperature of 20C after a few hours of solid imaging i.e. after the sensor has warmed up.   These latest generation CMOS sensors have very good dark current suppression.

I haven't yet seen dark current figures for the ASI 1600.

Mark

[Zakalwe]

I think the latest ZWO driver goes a long way to controlling amp glow.

Regarding bit depth, have a read of this:

http://www.cloudynights.com/topic/553142-zwo-asi1600mm-cool-performance-stats/#entry7479621

Quote

I have always been one that criticized this camera's 12 bit ADC but today I decided to (hopefully) put my worries to rest so I did some analysis comparing the high dynamic range mode and the unity gain mode. I can confidently say I am not worried anymore. I measured a dynamic range of 5700:1 dynamic range in HDR mode and much lower (expectedly) in unity gain mode. These data can be seen attached from Pixinsight. To compare, my KAF-1301E measured a dynamic range of 6999:1 dynamic range using the exact same method and my KAF-8300 measured a disappointingly low 2800:1 dynamic range. The full well of 20ke- noted in their literature was pretty spot on. I still dislike the ADC chosen but from their point of view I can see why. Of course the dynamic range of 5700:1 calls into the fact that it's a 12 bit ADC and is limited to 4095:1 which is a bummer to me.. but still much higher than the KAF-8300. Even the perfect electronics of the QSI 683 quote a 3160:1 dynamic range (my chip had 12e- of read noise rather than the 8e- in the documents from QSI). 

 

[Thalestris24]

I still think 16-bit is likely to be better! Not that I'm in the market for any more cameras myself.... 

Louise

[Andyb90]

I'm new to astro cameras having recently purchased the asi1600mm cool. I've been trying to get my head around dynamic range. I found this site which gives a simple calculation for dynamic range:

http://www.photometrics.com/resources/learningzone/dynamicrange.php

the calc given is: linear full well capacity (electrons)/read noise (electrons).

I think the article is saying a 16 bit ADC would allow for 65,536 grey levels, but to achieve that the full well capacity would have to be big enough and the read noise low enough to achieve 16 bit.

So if you had say read noise 5e- and full well capacity 22,000 electrons the dynamic range would be 4400, so 12 bit (4096 grey levels). If that is correct does it mean you would not be able to achieve full 16 bit, even with a 16 bit ADC, unless you stacked enough subs?

Andy.

 

[Tiny Small]

On ‎20‎/‎10‎/‎2016 at 15:29, Thalestris24 said:

I still think 16-bit is likely to be better! Not that I'm in the market for any more cameras myself.... 

Louise

I have a 4cmx4cm, 16bit, 16 megapixel CCD that plugs onto the back of my medium format cameras. I can state with absolute certainty that 16bit is by far superior to 12 or 14 bits. My digital back is approaching 20 years old and still gives cleaner, sharper and more detailed images than cameras like the Nikon D4x. Even with interpolation up to 50 megapixels, it is unbelievably clean and sharp. I would use it for astro but it's limited to 30s exposures and weighs about 8 pounds :/

[rwg]

On 20/10/2016 at 15:27, Zakalwe said:

I think the latest ZWO driver goes a long way to controlling amp glow.

 

 

Now that's interesting... QHY have had an amp glow control option on their Sony CMOS sensor models for a while now and it is definitely effective (In SharpCap you can turn it off manually and see the difference it makes). As someone pointed out the 174 is bad for this, the 178 has much less of a problem. Anyway, maybe ZWO have also worked out what QHY have done to reduce the amp glow?

cheers,

Robin

[rwg]

On 17/10/2016 at 20:54, harry page said:

Hi

This is from the cooled camera !!

Harry

Sadly the amp glow doesn't respond much to cooling on these sensors

Robin