Reducing 'Walking Noise'

[pwhit]

I recently bought a ZWO ASI 1600 cooled OSC camera and have initially been taking unguided 20 second subs (after careful polar alignment) without any calibration frames just to get a feel for what the camera is capable of. After integrating the subs I noticed that the RGB masters suffer from 'walking noise' (i.e. streaky noise aligned from top left to bottom right). Advice on the web suggests that I can eliminate this by using autoguiding with dithering which is what I will do in the future. However, I wondered if there is any way of reducing walking noise on the images that I've already taken. Can anyone suggest the best Noise Reduction method to try with suggested settings etc.

[bulrichl]

You integrated the subframes "without any calibration frames". The "streaky noise" is caused by hot pixels. You didn't guide either, so the stars were slowly drifting during the imaging session. After registering the frames the stars match but the hot pixels not any more, they reflect the drift that occurred during the imaging session.

You can retrieve your frames by taking (at least!) some dark frames, making a MasterDark according to https://www.pixinsight.com/tutorials/master-frames/ and calibrating the light frames with the MasterDark. The next necessary step is CosmeticCorrection which takes care of the hot pixels that remained after the calibration process. Then debayer, register and integrate as usual.

Detailed information about PI's calibration process is given e.g. here: https://pixinsight.com/forum/index.php?topic=11968.0

Dithering is good advice for the next imaging session. However, I would suggest not to omit the calibration process either.

Bernd

[pwhit]

Hi Bernd

Thanks for your response. I tried using a MasterDark from my Dark Library (same exposure and temp as lights) but this didn't seem to have any noticeable effect but I didn't try Cosmetic Correction so will give that a try.

A follow-up question regarding calibration :-      I have a home-made EL panel which I use for Flats (optimum exposure determined in APT as 0.005 sec). What would you suggest regarding calibrating the flats with this camera ?  The Flat Frames look pretty clean when I stretch them so I'm wondering whether calibrating them is necessary especially given such a short exposure.

Pete

[pfile]

flats *must* be calibrated with at least a bias frame, or else the scaling calculation that's done prior to division of the light by the flat goes wrong, and then the flat overcorrects the light.

rob

[ngc1535]

A follow-up question regarding calibration :-      I have a home-made EL panel which I use for Flats (optimum exposure determined in APT as 0.005 sec). What would you suggest regarding calibrating the flats with this camera ?  The Flat Frames look pretty clean when I stretch them so I'm wondering whether calibrating them is necessary especially given such a short exposure.
Pete

Really? Most El panels flicker at some rate related to A/C current. 0.005 sure sounds like a very very short exposure.
-adam

[pfile]

wow, yeah i missed that. one way or another EL panels and LED panels flicker, from 60hz all the way up to 20khz.

i think with this type of camera you should be able to reduce the gain to force a longer exposure, right? also you can look into dimming the panel either electronically or with sheets of drafting vellum.

rob

[pwhit]

Thanks for the further tips. I determined the flats exposure time by using the 'CCD Flats Aid' in Astrophotography Tool (APT). I've also read various articles that suggest that 'scaling' of calibration frames doesn't work well for the ZWO ASI 1600 camera. So although I understand the basic concepts of calibration I'm still not sure what is the best calibration strategy for this camera, hence the reason why I've started off by doing no calibration (resulting in 'streaky' noise) and intend to gradually extend the use of calibration frames until I reach an optimum result. I understand this is a popular camera so I'm keen to find out what other users have determined to be the best strategy.

[bulrichl]

For the calibration of frames taken with a CMOS camera I suggest (as a good starting point) the approach and settings that I use with my ASI294MC Pro (see my tutorial, right down at the bottom).

This implies that you have to take dark frames, flat frames and flat-dark frames (no bias frames needed).

Bernd

[pwhit]

Hi Bernd

I read your interesting tutorial recently and did wonder at the time whether I should try your suggestions for CMOS cameras. Could you confirm that when you say a 'flat-dark' you're referring to a dark frame with exposure time being the same as for the flats (in my case 0.005 sec unless I reduce the light output from the EL panel as suggested earlier to get a longer exposure). I tend to take my flats immediately after an imaging session so I would just carry on taking similar frames but with the cap on the scope instead of using the light panel. I also tend to use Dark Frames from a prepared library with same exposure, temperature and gain settings as the light frames. Apologies for asking some basic questions.

Pete

[bulrichl]

Could you confirm that when you say a 'flat-dark' you're referring to a dark frame with exposure time being the same as for the flats (in my case 0.005 sec unless I reduce the light output from the EL panel as suggested earlier to get a longer exposure). I tend to take my flats immediately after an imaging session so I would just carry on taking similar frames but with the cap on the scope instead of using the light panel.

Yes, same exposure time, same gain and temperature as for the flat frames and cap on for the flat-darks. Adam and Rob have already addressed the possible issue of a too short exposure time for the flat frames. Some flat field panels will yield uneven illumination with too short exposure times.

I also tend to use Dark Frames from a prepared library with same exposure, temperature and gain settings as the light frames. Apologies for asking some basic questions.

All right,

Bernd

[pwhit]

My current home-made EL panel has a polycarbonate sheet over the EL light strip which diffuses the light, so I'm intending to add another layer of polycarbonate to push up the exposure times (from the current 0.005 sec). What would you recommend is the optimum exposure time I should be aiming for as this will dictate the thickness of additional polycarbonate sheeting.

[John_Gill]

Hi,

My flats are about 1 - 2 seconds.  Not sure if this is good or bad, but it seems to work for me.

Look up
John

[bulrichl]

Hi Pete and John,

the minimum exposure time depends on the frequency f of the power converter's AC output which supplies the EL foil / the LED's.

If the exposure time is less than approximately 1/f, this may result in a highly uneven illumination. So the lower limit for the exposure time of the flats should be approximately 1/f (or somewhat higher). The frequencies of different types of power converters may vary greatly. However, I guess there will be no power converter with a frequency less than 50 Hz, so exposure time greater than 0.02 s will be OK in either case (1 - 2 s definitely will be fine). If the frequency is not known you can try different exposure times and judge from the evenness of illumination, which minimum exposure time is necessary.

An alternative to inserting another layer of polycarbonate would be, as Rob suggested, to lower the gain (for the flat frames and the flat-darks frames likewise). However, dimming of the EL panel is either a provided feature of the power converter that was delivered with the EL foil / the LED's or it may not work at all. Simply decreasing the voltage is no option because that normally does not work reliably.

Bernd

[pwhit]

Hi Bernd

Thanks once again for your suggestions. I can't see any uneven illumination on my existing flat frames when stretched so I don't think I need to increase the exposure time by much. If I change the gain setting I'll probably forget to reset it when I next take some light frames so I think I'll go for the added sheet in front of the EL strip. Anyway lots of options to try !

Pete