Used KAF-8300 or new ASI 1600MM-Cool?

[MartinFransson]

Decisions, decisions...

As my thoughts go now, I have two options. Both should get me in the same price range (about 2000 Euros) :

1. A used CCD with a KAF-8300 chip, for example QSI683 (rarely found)

2. A brand new ASI 1600MM-Cool

With a new camera, I get support, warranty etc. and that means a lot to me. On the other hand - do I get a better camera with the KAF-8300? Is the data easier to process? The 16 bit vs 12 bit difference is not all that big as I understand it, due to the lower read noise of the ASI. The dynamic range should be as good or even better with the ASI.

Thoughts from you more experienced imagers? I have tried looking at images from both cameras and so far I have not seen as impressive photos with the ASI but it´s a newer camera and it is also very much up to processing, sky darkness etc.

Oh, and would the focuser on my scope (150PDS) handle the weight of a QSI683 with filter wheel (1.6-1.8 kg) without sagging? The ASI is a lot lighter.

[RayD]

For me the decision to stay CCD was about the size of the files, and the fact that for a given comparable completed image, I only need to process dozens of subs, as opposed to possibly (probably) hundreds.

The QSI is very good, and several excellent imagers here use them, so that should speak volumes.

Edit:  If you have the KAF 8300 in mind have you considered the Moraivian?  I know there are some on here that use and rate them, and cheaper than the QSI.

[swag72]

At this moment in time I would still stick with the KAF8300 ...... I know that this is a proven performer and I find that the file size and exposure numbers suits me and my PC!!! 

The QSI is very good and I rate them highly - But using a Moravian side by side, there's not a lot of difference between them, so a new Moravian would be worth looking at to see if it fits your price range ...... definitely worth a look and also if you find one second hand don't dismiss them.

[MartinFransson]

The thing is I found a used QSI683 today for about 1400 Euros. Wouldn't have thought about that one otherwise. 

Regarding image numbers, isn't it possible to use the ASI in the same way? Fewer but longer (5-10 min) exposures? 

[RayD]

Just now, MartinFransson said:

The thing is I found a used QSI683 today for about 1400 Euros. Wouldn't have thought about that one otherwise. 

Regarding image numbers, isn't it possible to use the ASI in the same way? Fewer but longer (5-10 min) exposures? 

That sounds like a bargain to me.

There are many threads on the ASI, which to be fair does seem like a very good performer as people get more used to the CMOS sensor and the required settings and processing, but that is the reason why I said for a given result.  My understanding is to achieve this you will need far more subs, with each one being pretty big.

[RayD]

The QSI really does look very good indeed.  I would have bought one myself recently if it wasn't for the fact that I already had a filter wheel with 2" filters.

You don't see too many of them on the second hand market, but I have already seen lots of the ASI1600's being traded.

[Davey-T]

14 minutes ago, MartinFransson said:

The thing is I found a used QSI683 today for about 1400 Euros. Wouldn't have thought about that one otherwise. 

Regarding image numbers, isn't it possible to use the ASI in the same way? Fewer but longer (5-10 min) exposures? 

I wouldn't think about it too long, I'd be round there with the money in my hand 

I've got 2 and can't fault them

Dave

[RichLD]

6 minutes ago, MartinFransson said:

The thing is I found a used QSI683 today for about 1400 Euros. Wouldn't have thought about that one otherwise. 

Regarding image numbers, isn't it possible to use the ASI in the same way? Fewer but longer (5-10 min) exposures? 

I have a QHY163m, which uses the same sensor as the ASI. I also have a QHY9, which is probably the best value KAF8300 camera around. I have had excellent results with the 163m using 600s exposures at unity gain or thereabouts - it's a really clean chip!

The extra pixels of the 163m within the same FOV do give an advantage for widefield imaging with camera lenses. I went for the QHY163m over the ASI as it has the 128mb buffer and heated sensor window.

HTH

Rich

[vlaiv]

13 minutes ago, MartinFransson said:

The thing is I found a used QSI683 today for about 1400 Euros. Wouldn't have thought about that one otherwise. 

Regarding image numbers, isn't it possible to use the ASI in the same way? Fewer but longer (5-10 min) exposures? 

It is possible, but it is sub optimal for CMOS.

It has less dynamic range, 12 vs 16 bit ADC, so you would want something like 4 times less exposure time with CMOS (not exactly 4 times less, since dynamic range is not the same as bit count for ADC, but close enough).

This will generate 4 times more frames, at higher resolution - much more data to store and process.

[MartinFransson]

2 minutes ago, RichLD said:

I have a QHY163m, which uses the same sensor as the ASI. I also have a QHY9, which is probably the best value KAF8300 camera around. I have had excellent results with the 163m using 600s exposures at unity gain or thereabouts - it's a really clean chip!

The extra pixels of the 163m within the same FOV do give an advantage for widefield imaging with camera lenses. I went for the QHY163m over the ASI as it has the 128mb buffer and heated sensor window.

HTH

Rich

So for a given amount of exposure (say 5 hours), how would you rate the total resulting image quality of the QHY163 vs the QHY9?

[RichLD]

4 minutes ago, MartinFransson said:

So for a given amount of exposure (say 5 hours), how would you rate the total resulting image quality of the QHY163 vs the QHY9?

I've only used the QHY163m for HA and have been delighted with it -  I can't make a scientific comparison between the two cameras but can only comment that I've been extremely happy with the results I've had. I appreciate that the CMOS camera has a lack of dynamic compared to the CCD but I am struggling to see a difference when processing the individual subs - perhaps the QHY9 subs have a little more contrast straight out of the camera but I believe this is moot after the extensive processing that we perform on our images.

The bias frames of the CCD are a lot noisier as are the darks (hot pixels etc.) but both calibrate very well. TBH I like each camera equally - if I only had the CMOS I wouldn't be pining for the CCD and vice versa .

Having said all that the QHY9 allegedly has inferior electronics compared to the QSI, so my comparisons may ultimately not be that useful!

Here's my second light with the QHY163m - 15 x 600s, Baader HA 7nm, -15C, G10 OS70, Nikkor ED180 @ f/3.6, processed in PI and PS.

HTH

Rich 

 

 

[Davey-T]

Bear in mind that for all the quoted bit depth and dynamic range unless you're going to make high resolution A3 prints from your images you may never see them in more than 8 bit 256 colours or grayscale.

Dave

[MartinFransson]

39 minutes ago, Davey-T said:

Bear in mind that for all the quoted bit depth and dynamic range unless you're going to make high resolution A3 prints from your images you may never see them in more than 8 bit 256 colours or grayscale.

Dave

Fact is I DO like to make big prints 

[MartinFransson]

54 minutes ago, RichLD said:

I've only used the QHY163m for HA and have been delighted with it -  I can't make a scientific comparison between the two cameras but can only comment that I've been extremely happy with the results I've had. I appreciate that the CMOS camera has a lack of dynamic compared to the CCD but I am struggling to see a difference when processing the individual subs - perhaps the QHY9 subs have a little more contrast straight out of the camera but I believe this is moot after the extensive processing that we perform on our images.

The bias frames of the CCD are a lot noisier as are the darks (hot pixels etc.) but both calibrate very well. TBH I like each camera equally - if I only had the CMOS I wouldn't be pining for the CCD and vice versa .

Having said all that the QHY9 allegedly has inferior electronics compared to the QSI, so my comparisons may ultimately not be that useful!

Here's my second light with the QHY163m - 15 x 600s, Baader HA 7nm, -15C, G10 OS70, Nikkor ED180 @ f/3.6, processed in PI and PS.

HTH

Rich 

 

 

That's really nice! But to be fair, it is an area of very high Ha-signal. I like to image the really faint stuff and that's where I'm not sure about the QHY163 or similar.

[vlaiv]

36 minutes ago, Davey-T said:

Bear in mind that for all the quoted bit depth and dynamic range unless you're going to make high resolution A3 prints from your images you may never see them in more than 8 bit 256 colours or grayscale.

Dave

Not sure if this is so, because we tend to stretch data in post processing (non linear transforms). Also, number of subs stacked increase dynamic range. We might not be able to put on a screen 100000 levels of value, but it helps distinguish between levels 20 and 100 out of 100000 and thus show all the subtle variations in nebulosity or galaxy structure at expense of compressed stars.

On a purely mathematical note, if we put aside processing difficulties (sheer amount of data), ASI1600 should have an edge over KAF 8300 based camera due to higher peak QE, and less noise. Also shorter subs mean less likelihood of dropped frames (guiding error, wind, whatever causes frame to be bad), and thus more integration time for equal time under stars. So for 5h of shooting I would say that ASI1600 will produce better result, if properly processed.

 

 

[Davey-T]

I've yet to see a side by side comparison between a 30 minute 8300 sub and 60 30 secs subs stacked from a 1600 of a not so bright target.
What puts me of also is storing the amount of data produced.

One of the main reasons people seem to be choosing the 1600 is for mounts that they can't get to perform reliably so short subs is the line of least resistance.

As mentioned there seem to be few a 1600 for sale s/hand but few QSI683s but that maybe because fewer exist to start with

Dave

[RichLD]

14 minutes ago, MartinFransson said:

That's really nice! But to be fair, it is an area of very high Ha-signal. I like to image the really faint stuff and that's where I'm not sure about the QHY163 or similar.

Thank you! I've seen some very faint nebulosity captured by CMOS cameras, let me find a few links and I'll post them if that helps. There's no denying the QSI is a fine camera though!

[MartinFransson]

I would be doing 3-10 minute exposures, not 30-60 seconds. Don't want tons of data. That's what I do with my DSLR, but without the benefits of cooling and mono.

[MartinFransson]

2 minutes ago, RichLD said:

Thank you! I've seen some very faint nebulosity captured by CMOS cameras, let me find a few links and I'll post them if that helps. There's no denying the QSI is a fine camera though!

That would be great!

Starting to suspect the QSI sale was a hoax. Might find a Moravian or similar, though.

[RayD]

39 minutes ago, MartinFransson said:

That's really nice! But to be fair, it is an area of very high Ha-signal. I like to image the really faint stuff and that's where I'm not sure about the QHY163 or similar.

I think in reality you've pretty much answered your own query here.  For very faint stuff you are going to need either very long exposures (1800s or so I would have thought) with a CCD, or lots and lots of shorter ones with a CMOS, whether that be a ZWO or other.

As Vlav points out, if you drop a sub, it's probably not such a disaster with the CMOS.

I guess all this goes back to why there are so many threads about this, as there really is no perfect solution, it's just what suits your needs and requirements best.  I feel happy to go with longer exposures needing far less space and processing for a given result, but that doesn't mean the opposite isn't a valid route.  I've seen posts with people referring to stacking taking hours with CMOS data, but I certainly wouldn't be looking at that with my data (which is predominantly of cloud if truth be known).

[RayD]

14 minutes ago, MartinFransson said:

Starting to suspect the QSI sale was a hoax. Might find a Moravian or similar, though.

That does seem a very good price for something that really doesn't come up very often at all.  As Davey says, whether you would prefer CMOS or CCD, for that price I would snatch it.

[Davey-T]

I generally do 20 / 30 min subs but  can see the attraction of short ones if you've got an EQ5 it could make it a viable imaging mount.

Dave

[MartinFransson]

I have an HEQ5 and with guiding I have no problem with 15 minute exposures. LP prevents me from going longer than 5-10 minutes with Rgb. Narrowband is another thing of course.

[MartinFransson]

Could there be a big difference in how easy they are to use? I don´t have a permanent observatory, so I need to set up and take down everything each imaging session...

[RichLD]

Here's a good example of *lots* of subs to get an image of a very faint object (SH2-224) using the QHY163m credited to David from the QHYCCD website. 62 X 8 minutes.

http://www.qhyccd.com/images/SH2-224.jpg