Can't decide whether to change from CCd to CMOS

[alexbb]

9 hours ago, ollypenrice said:

I think the real issue is the number of files if you do long integration in short subs. AstroArt is a remarkably fast stacking and calibration programme and might be a help to CMOS users.

Olly

Storage space is more an issue for me. 300 frames are ~10GB. I have to back-up quite often (I have all the data since the beginning).

Calibrating, registering and stacking with kappa sigma rejection ~300 subs takes me just a bit over a half an hour on my 6 year old i7 powered laptop. 4 cores, 8 logical CPUs.

Usually I do narrowband and for this I expose 300s. 10 hours means 120 subs. Takes me 10-15 minutes to stack 120 subs.

8 hours ago, steviemac500 said:

Olly, any thoughts on the 8 pos fw and possible vignetting?

Attached is a master flat taken through a 1.25" Ha filter and an F/4 lens:

MasterFlat_ISO0.tif

And stretched in GIMP with auto levels:

[ollypenrice]

13 hours ago, moise212 said:

Storage space is more an issue for me. 300 frames are ~10GB. I have to back-up quite often (I have all the data since the beginning).

Calibrating, registering and stacking with kappa sigma rejection ~300 subs takes me just a bit over a half an hour on my 6 year old i7 powered laptop. 4 cores, 8 logical CPUs.

Usually I do narrowband and for this I expose 300s. 10 hours means 120 subs. Takes me 10-15 minutes to stack 120 subs.

Attached is a master flat taken through a 1.25" Ha filter and an F/4 lens:

MasterFlat_ISO0.tif

And stretched in GIMP with auto levels:

What matters about this (or any) flat is not what it looks like stretched but what the actual pixel values are in its linear form. If you open it in a programme which allows you to mouse over different parts while getting a readout of the pixel value you can compare the darkest parts of the corners with the brightest parts of the centre. In our most vignetted setup the range is a hefty 17,500 to 24,500 meaning the darkest corners are only about 69% illuminated. This is far from ideal but, none the less, flats work their magic and this camera is one of the pair we often use for large scale mosaics where flattening the component panels is vital for a seamless result. I don't know at what point an imager would describe vignetting as beyond repair but I hope not to have to work with less than 69% illumination!

Olly

[alexbb]

It's around 7200 vs 5200 for the flat above. You can see in the stretched image that the filter is not centered so there's some room for improvement. I should calibrate my filterwheel perhaps.

[Thalestris24]

On 12/5/2017 at 22:37, Gina said:

You can't get more signal by using CMOS.  It still depends on total imaging time.  With CMOS you can take shorter exposures but you need more of them.  There are some advantages of the shorter exposures with CMOS - you may be able to make better use of gaps in the clouds on what would otherwise be an unusable night sky.  Also, with a wide field and fast scope (or lens) you may be able to get away without guiding.  With the ASI1600MM-C you can use the larger ZWO filter wheel with 7 or 8 filter spaces and still get the back focus required for field flatteners or SLR lenses.  But the same results in a shorter time - definitely not. 

I usually take several hundred subs in each wavelength for NB imaging with my ASI1600MM-C and they are 32MB each - that's a lot of data.  I bought a high speed games machine and added SSD drives to get enough speed for data processing to take a reasonable time.

Hi Gina

Can I ask what your typical Ha sub length might be with the 4/3 cmos?

Thanks

Louise

[Gina]

60s mostly - 30s for the brightest DSOs and maybe 120s for the very dim ones, with gain of 600.

[Thalestris24]

11 minutes ago, Gina said:

60s mostly - 30s for the brightest DSOs and maybe 120s for the very dim ones, with gain of 600.

Oh ok - that is short! I've been doing 480s with my 1100d and 3.5nm filter with the Monkeyhead, for example, is faint and barely stackable. Rosette better, but still. Mind you, my skies are poor which doesn't help.

Thanks for the info

Louise

[Gina]

Forgot to say, that is with lenses with f ratios of around f2.8.

[Thalestris24]

I'm at f5.5 - not so fast but not slow either... I think the qe of the 1100d is around 40%. Maybe it's hard to make a comparison

[Jessun]

Interesting thread.

Did anyone run a CMOS with at least 5 or 10 min subs? Would amp glow or something else destroy the image?

/Jesper

[alexbb]

28 minutes ago, Jessun said:

Interesting thread.

Did anyone run a CMOS with at least 5 or 10 min subs? Would amp glow or something else destroy the image?

/Jesper

I usually expose for 300s when shooting through narrowband filters. Nothing to complain about the camera and amp glow, I have other issues instead.

Not sure which should be the most relevant image, I'm currently working on this:

https://drive.google.com/open?id=1I6IZ47N2_FI1bPQo1EbA0RhSHNMjb0Ot
https://www.astrobin.com/317154/D/

Or perhaps another one which is not a mosaic, but still heavily stretched:

https://www.astrobin.com/294290/C/

 

I can provide in a few hours some master darks at -15C and -25C.

[alexbb]

First is a master dark of 53 subs at -15C and the second 31 subs at -25C. 300s, unity gain (139).

MasterDark_139g_300s_-15C.tif

MasterDark_139g_300s_-25C.tif

[Gina]

Yes, very interesting   I may try longer exposures when I get imaging again.  At present I'm unable to do any imaging due to gear problems and other issues but hope to get going again soon.

[RichLD]

On 11/12/2017 at 15:30, Jessun said:

Interesting thread.

Did anyone run a CMOS with at least 5 or 10 min subs? Would amp glow or something else destroy the image?

/Jesper

Hi Keeper, I routinely run 600s ha exposures with the QHY163m (G10 OS70) and have no problems with calibration.