windjammer

It makes life a lot easier if you have a permanent setup. Setup and teardown is just a fag, and disturbing the image train from one set of exposures to the next just adds pointless issues that have to be sorted. Having a permanent setup, which only needs waterproof covers to be viable, means that one can work on reliability: I used to sit through 5 hours of 600 second exposures ready with the toolbox for emergency surgery. Now I can set the cams running, go to bed, sleep through (most of) it and process data the next day. And this with home made kit: its just a matter of tackling the next bottleneck that is holding you back. I'm in London so rubbish Bortles, but I have found narrowband can blast through light pollution and pretty mucky skies: not a star visible but the cameras returning acceptable exposures. SHO is worth experimenting with if you have never tried it. I also think the UK cadence of cloud and clear skies is not so bad! 4 -5 days on the trot of astronomy nights brings out my inner zombie so not too many of those is desirable. My target is one imaging project per month - data collection to 'final' product. This year has been pretty dodgy weather so far, but I have actually done more, and better, than any year to date! So: never give in, never surrender! Simon

Very impressive - I like the colour rendering. The normal blue/red mix eveyone uses doesn't really work for me, this is much better! The detail of the Witch on the Broom is very delicate. The fullsize FoV is humungous. That RASA is working v well indeed, I can't believe only an 1 hour per filter. Simon

What is that wallpaper !

Thanks, must get some new specs....

Yes, much better. I tried it for 10 mins and bought it!. The standard settings for SN,LL and ST are far too severe IMHO. So my general approach is to do most stretching in 32 bit mode (PI, AP, PS or whatever), convert to 16 bit and feed it into Tpz. Then check out the ST(standard) mode. If reduce noise is >20 - 30% to kill the noise I try LL (low light). If LL at 20-30% does not do the trick, I try the SN (severe noise) mode. The blur the noise reduction adds increases from ST to SN. So the enhance detail slider makes up for that. As usual, the trick is to avoid baking in artefacts early on which are impossible to remove later. You can also use Tpz multiple times in the processing as the issues change - usually in a more and more refined fashion as you converge to a final image solution. Simon

>>To me, it looks like a space brain..... Like something out of Mars Attacks! >>I noticed a faint bubble just below-left of the centre of the image. I looked and looked but can't see it! Simon

Its a good looking fish, splashing its way through the Milky Way spinning off stars and turbulence! I love the catfish type barbels around the 'mouth' region.

Lovely image. Quite a lot of noise baked into the nebula - Topaz Denoise AI is a great tool for dealing with it if you haven't tried it. It is a plugin for PS and there is a free trial vs to try it out. The tool is pretty much point and shoot. Simon

yep, that's not bad at all, especially to get the colour. FoV is just right as well.

Hi Here is NGC 896, IC 1795, LBN 645, the Fishhead Nebula in L(Ha)SHO narrow band. Image and annotated vs. The image is rendered with the Foraxx Palette Utility script in PixInsight. Rig and processing details at the end. Data taken in Astronomik narrowband filters: Ha (656nm), Sii (672nm) and Oiii (501nm). Total exposure time 15 hrs. Ha 1x1 bin - 22x 600s = 3.7hrs, 03-04 September, seeing 2.6", scope West side, prime focus Sii 1x1 bin - 33x 600s = 5.5hrs, 04-05 September, seeing 1.8", scope West side, prime focus Oiii 1x1 bin - 35x 600s = 5.8 hrs, 05-06 September, seeing 1.2", scope West side, prime focus Simon and annotated: ---- Master Lights: FWHM (pxl) pre BXT post BXT Ha 2.440 1.476 Sii 2.942 1.666 Oiii 3.163 2.524 ----- Plate solver: Resolution ............... 1.247 arcsec/px Focal distance ........... 750.71 mm Pixel size ............... 4.54 um Field of view ............ 56' 1.8" x 43' 58.3" Image center ............. RA: 2 28 13.183 Dec: +62 01 33.70 Image bounds: top-left .............. RA: 2 24 25.590 Dec: +62 25 19.95 top-right ............. RA: 2 32 27.835 Dec: +62 21 12.52 bottom-left ........... RA: 2 24 04.018 Dec: +61 41 26.22 bottom-right .......... RA: 2 31 54.880 Dec: +61 37 24.65 ----- Rig: Imaging scope: SW Startravel 150mm F5 Refractor, Baader Diamond Track, (2.5x Celestron Luminos 2inch imaging barlow), Atik 460EX mono Guide scope: SW Evostar 90mm F10, with guiding XY stage, ZWO 120MM camera Guiding: 2 stage PHD: high frequency guide scope (mount tracking) and low frequency OAG image train guiding (guidescope flex) Mount: Home made German Equatorial pillow block mount, permanently rooftop mounted. Spring loaded DEC axis gearing. Other gadgets: ST4 based anti vibration shutter, ST4 based PEC ----- Processing Lights: PixInsight: Lights, Darks, Flats, Biases: master dark/dark library-> masterbias-> superbias-> calibrated flats-> master flat-> calibrated lights-> cosmetic correction-> aligned lights-> master light-> BXT PixInsight: Master BXT lights-> crop-> linfit-> final master lights PixInsight: final master lights->StarNet2 starless-> LRGB Channel Combination (Ha, SHO)-> export xisf starless master. GradXpert Gradient removal:->import starless xisf-> GXPT(20pc,3sg)-> export xisf, fits Affinity Photo 32 bit image processing:-> import LSHO gxpt starless fits-> accept default stretch-> curves-> lvl(master)-> Topaz Denoise(LL, 22, 31)-> Tpzdn(ST,15,13)-> export tiff 16 bit PixInsight: import tiff16-> channel separation-> Foraxx Palette Utility Script-> export Foraxx tiff 16 Affinity Photo final tart ups:-> import Foraxx tiff 16-> cvs-> scol (50%blend)-> 2x clarity, B&C tweaks->Tpzdn(LL,3,0)-> paste in star mask layer, blend mode 'screen', B&C, White Balance adjust. ----- Processing Star Mask: PixInsight: final master lights-> RGB Channel Combination (SHO)-> export xisf GradXpert Gradient removal:->import xisf-> GXPT(20pc,3sg)-> export xisf PixInsight: import SHO gxpt master-> StarNet2 star mask-> SCNR(Mg, 0.57)-> export fits starmask Affinity Photo 32 bit image processing:-> import SHO gxpt starmask fits-> accept default stretch-> vibrance-> Sii master light star mask overlay (B&C adj, blend mode 'luminosity')-> paste Star Mask layer on top of starless final (blend mode 'screen'). -----

A lovely sense of depth to the fov, especially in the starless top right. The more heavily emphasised star layer in your final vs loses a bit that visual impression, I might dial them back a bit. But a great pic!

If you have PI then starnet2 process will create star mask for you to process stars separately (if you didn't do this already?). The nebula is quire noisy and stars have a green cast. I had a quick play Topaz denoise on the neb and Affinity Photo for hue, level, B&C on the star mask- hope you don't mind. Simon

Lots of v. interesting stuff in this discussion - I will archive it and give it a long read! I might start giving my subs the once over with an FT package to get a better feel of what is going on. What I would give my eye teeth for is two cameras, pixel dimensions differing by a factor of two, on the same scope, same night, same target and compare the subs through the wringer discussed above! And then stand back.... : ) Simon

Beautiful. I love the Foraxx palette. Bracken has this as 2.6o x 3.4o, an enormous FoV! Spectacular - how many hours went into this ? Simon

OK - FWHM/1.6 it is then! A very clear discussion. I would like to see CCD Imager's fabled sub though : ) Simon

I will have a look

>>Hope this makes sense. Yes, actually very clear. And the 1.6 - is that just the point at which the 'generalised' gaussian reaches our 10% cutoff point (I understand the FT of a gaussian is another gaussian)? And the Nyquist x2 is higher because his 'simple' case does not include noise considerations? Simon

..and BTW, that is a v nice looking bit of FFT software. Is that readily available ? Also the line profile generator ? Simon

So, if I understand you correctly, the 1.6xFWHM recommendation will capture the signal from frequency components with greater than 10% of the peak frequency amplitude. And we are talking about the spatial frequencies in the FT of the point response fn. And by 1.6xFWHM you mean: if my FWHM is 3.4 asec my sensor resolution should be 3.4/1.6 ~ 2asec per pixel. An explanation of your 1.6 factor would be very helpful. Simon

Stars looking a bit eggy in the corners ?

Well, the PixInsight FWHMEccentricity script reports the FWHM of Ha master lights over a few projects (at F5, no barlow) comes in at ~ 2.7 pixels. The PI Image Solver script reports resolution is 1.25 asec per pixel, so FWHM comes in at 2.7 * 1.25 = 3.4 asec The seeing website: https://www.meteoblue.com/en/weather/outdoorsports/seeing/london_united-kingdom_2643743 reports London UK seeing in range 1.5 to 3 asec. Usually at the higher end TBH! The Rayleigh criterion would be 0.8 asec at 500nm, so we are nowhere that. With modest enhancement parameters (no artefacts) in BXT the FWHM comes down to ~ 1.6 pixels, or 1.6 * 1.25 = 2 asec. I agree with the point that deconvolution tools need extra pixels to work with (no evidence, just a gut feel!). So my thinking is working at F5 with 1.25 asec/pxl sensor is probably about right, and a little bit higher F number might work as well if the light flux doesn't take exposure times to extreme lengths. 600s to 900s is my practical limit, just in terms of getting a reasonable number of exposures in a session. Another optical element in the image train is a bit of a downer as well. Simon

Yes - F5 and 2.5x barlow. I had to bin the camera 2x2 and even so 600s exposures. Early on this year I realised (duh) a 2.5x barlow and then binning 2x2 was just snakes and ladders, so I've reverted to F5 and 1x1 binning on later projects. You end up with the same number of pixels in the relevant object and in a wider fov, which you can always crop. I might be in the market for a 1.5x barlow tho.... Simon

A lovely image, great colours - I would definitely have a go at the noise. I scraped the image and ran it through Topaz Denoise

Actually I can't get my head round masking - its a weird world of double negatives. I'm going at it full on and all of sudden I lose the plot - no idea what the mess of layers is doing any more! Simon

I imaged the Bubble about a year ago but was never that happy with the end result - so this last week of clouds was an opportunity to re-process with new tools and a better handle on processing images. Also, in a moment of chutzpah / hubris I will probably regret: a side by side comparison of a Hubble close up of the Bubble and my humble Startravel 150 achro refractor.. : ) Astronomik narrowband Type 2c 1.25" filters: Ha (656nm), Sii (672nm) and Oiii (501nm). Total exposures 8.6hrs. Ha 2x2 bin - 16x600s (East side) + 600sx5 (West side) = 3.5hrs, 8-9 October 2022 SII 2x2 bin - 11x600s = 1.8hrs, 10-11 October 2022 OIII 2x2 bin - 20x600s = 3.3 hrs, 17-18 October 2022 Imaging scope: SW Startravel 150mm F5, 2.5x Celetron Luminos 2inch imaging barlow, Atik 460EX mono Simon and Hubble vs Startravel 150: