NGC7000 + IC 5070 in Ha/OIII/OIII (OSC)

[SteveL]

I`ve been waiting for a clear moonless sky for many months now, just so I could give my latest filter a proper test. I purchased a Hutech IDAS LPS-V4 Nebula filter earlier this year (link to filter here) to see if a One Shot Color CCD could do a decent job of recording Ha and OIII at the same time. The theory was that as the OIII wavelength falls almost exactly half way between the green and blue pixels (at a lower sensitivity though), I should get be able to stack the 2 green channels and the blue channel all as one large stack, and therefore grab 3 times the exposure of OIII compared to the single red pixel collection Ha data. Well, it kind of worked. The Ha data is nearly sharp as previous sessions with a 13nm Ha filter, and certainly usable, but not as good as my current Baader 7nm Ha filter. The OIII data (even after stacking) seems a bit washed out, and I believe this is because of the wide passband around the OIII wavelength (50nm wide at the very top of the peak), allowing in additional noise that a real narrowband filter would block.

Even so, as an Ha/OIII/OIII combined image, taken from a moderately light polluted area in one single session without having to change filters, I would say it was worth the time to do this test. The real question will be what happens to the OIII data when the moon fills the skies, as OIII tends to be washed out from moonglow far more than Ha data.

Mount: EQ6 via EQMOD

OTA: Borg 60 @ f/3.8

Guiding: SW ED80 + SX Lodestar + Maxim

Imaging: Starlight Xpress M25C + MaximDL, 16 x 900s, Hutech LPS-V4 Nebula filter

Orchestrated: CCD Commander

Stacked: DeepSkyStacker (Red in one stack, then G1+G2+B in seperate stack)

Post Process: PSCS2 + PixInsight

(Click on image for larger version)

or the Steve Cannistra BiColor method to create a synthic green channel (link)

(Click on image for larger version)

[Catanonia]

wow, still a cracking image mate and getting me really thinking about OII and S filters.

Love the framing, the colour works well if a bit pink, but fantastic image.

[themos]

Nice to see this experiment as I've thought about it too. Not quite sure why you do separate stacks though.

[SteveL]

Not quite sure why you do separate stacks though.

My thoughts on this are long and rambling, but the short version goes something like this.

OIII is quite dim, and the usual rule-of-thumb is to capture 2 or 3 times the data for OIII as for Ha.

The OIII falls right between green and blue on the OSC bayermatrix, and checking the subs shows that the difference between blue and green through a OIII filter is almost negligable, so I extracted the 2 green pixels and the 1 blue pixel into seperate image sequences, annd treated them all as one single stack. This way I am stacking 16x900s of Ha (Red), but stacking 48x900s of OIII (G1+G2+

If you stack as a normal bayer encoded RGGB image, The striong Ha (red) uis being stacked but that fine because it has a stronmg signal. The the blue pixels are only being stacked using 16x900s, and the 2 green pixels are being averaged together per frame and then stacked, so you lose the ability of rejecting certain pixels using a sigma clip stacking routine (the averaging removing the outliers you want to clip).

The idea was to use the 2 green and 1 blue pixels in a one shot color CCD as a 3 times multiplier when grabbing OIII. I do think the wide passband of the Hutech has diluted the OIII signal somewhat, allowing noise and light pollution in, and will be trying again one night with a narrower passband OIII filter to see if my theory is right.

[themos]

So if you were to use average stacking then there would be no difference as all GGB pixels would be treated the same whether you stacked with the reds or not, have I got it right?

Processing with the reds means an unavoidable averaging of the GG somewhere along the line and you want to avoid that.

[SteveL]

I treated these as 4 sets of mono sequences, so the reds (Ha) got processed seperately, allowing me to process the G1+G2+B (OIII) as one large stack in three seperate groups in DSS to create one single output image. If I processed it as a normal OSC RGB image, the green and blue channels would get treated as normal, averaging the greens and spreading the effective OIII data captured across two channels. Well, thats how I think it works.... more subs and more data means better signal to noise, and that was the reason for getting the Hutech LPS-V4 filter originally. Whether it would be better to do two sessions with two decent narrowband filters and then combine with the same 4 seperate sets of mono sequences will be a project for another day (or two)

[Catanonia]

Confused

LOL

[adamsp123]

You may have lost me Steve but the result is superb!

[SteveL]

I nearly lost the plot while waiting for my home-brew perl script to extract the channels from all the lights/bias/flats, then watched as DSS worked its way through the three seperate calibration and light groups, only for my house to have a momentary power brownout and my desktop PC reset! aaargghh!!!

[PortableAstronomer]

Very interesting, I hadn't heard about that V4 filter before.

Sounds like the ideal filter for my OSC setup too, and your use of the bayer matrix with this filter and the way the channels are stacked, makes good sense to me.

Trouble for me is my LP is of the bright white colour (fluorescent tubes) and leaks a lot of light into the OIII area ; The sky background gets saturated even using my Baader OIII filter with long exposures.

It looks like the V4 filter could save a lot of imaging time over separate line filters, if your environment isn't too urban.

[Tim]

Just out of interest Steve, have you stacked/processed the data as an RGB? Just curious to see the difference, and how the filter performs if used as I guess it was intended?

Be nice to see the result on a mono camera too.

[martin_h]

Its funny you have done this - I was just loking at the routines in Nebulosity 2 and there is a stacking/reconstruction routine for "Nebula" filters using a OSC camera.

[Tim]

Ah, thought I had seen that somewhere!

[Tim]

Just had a good look at the data Steve, the stacked as normal version. The star colour looks great, lots of colour in them right to the edge. In the HaO3O3 version they end up a bit pinky.

I wonder how the filter would cope with bad light pollution, the results look much better than with my CLS.

Here's a quick process, but it didn't take too kindly to being shrunk to 1200px for the SGL album.

Thanks for sending the file