Jump to content

windjammer

Members
  • Posts

    277
  • Joined

  • Last visited

Everything posted by windjammer

  1. Absolutely loads in this pic, but background not suppressed enough? Here is a quick hack, hope you don't mind!
  2. nicely resolved stars right to the centre and v good star colours. V nice. The background galaxies come in at around at 400 Mlyrs - good shapes and some colour. Amazing for only 1hr. My only comment would be to sharpen up the star edges if possible - a bit soft imho, otherwise vg!
  3. Nice detail on the counter-rotating dust cloud
  4. very nice. I like the subdued colours.
  5. You could also crop the frame and cut the offending glows. Just do a close up image. Seems a shame to waste such good material.
  6. mmm, I see what you mean. I've had the same trouble - never a really clear night without a high haze (or dew on the optics) and therefore uneven patches of crud popping up, as here. I don't know what software you have in the toolbox, but I would have a go with PI's new gradient correction process with the small scale removal knob turned up to full. Also graxpert's new AI gradient removal model is worth a try (its a free tool if you don't have it). I would have the stars removed before letting rip. I would also have a go with a curves tool to suppress the background as far as possible - you never know, the real object might just be bright enough to get away with it. In any event the blobs at 1, 6 and 9 you could probably convincingly paint out. I DID NOT just say that.
  7. Nicely done! Is this an HOS ? My own version of this in HSO was so lurid and orange I moved over to SHO for everything since... I might revisit my data Two comments if you like: There is quite a lot of noise lurking so I had a play with Topaz Denoise (hope you don't mind). Also I would go for stars a little crisper, but thats just me.
  8. >>but not without bringing out the bright background halos So is this because stars have not been extracted beforehand or the haloes left behind after extraction ? Haloes after extraction can usually be (ahem) cosmetically edited out.
  9. Its a while since I've posted, here is my effort at M81 - an RGHb mixed broad and narrow band image with excess Ha emission. Pics are a close in crop, the full FoV, and an annotated FoV. Holmberg IX, a satellite dwarf galaxy of M81 is just seen centre left as a vague blur, stars unresolved. Checking out the master lights, the G channel sort of shows some star points in the right position, but nothing made it into PI's Starnet star mask. More exposures might well have got Holmberg IX over the line, but the weather did not cooperate in March and April. Does it ever? The most distant galaxy catalogued is NGC28505, centre right of M81, coming in at 148Mpc (480 Mlyrs). It even has some colour. I've included info on the catalogued galaxies further down this post and links to the NED database where entries exist for those interested (OK, nerdy but fun). As usual, mixed feelings about the image. I have been thrashing it for weeks and conclude 12 hours of data with a 6inch achro is not really enough. The whirling dervish look of M81, fizzing across the cosmos with hair on fire works quite well. Some of the galaxy detail is a bit overcooked, but I like the excess Ha emission overlay, and the stars aren't bad, though a bit more colour in them would be nice. The double star pair TYC4383-1127 and -2044 just below and left of M81came out well without splurging (technical term) together. Exposure and processing details at the end. Simon ---- Close in crop: Full frame: Annotated image: ---- Galaxy object info: PGC HYPERLEDA I catalog of galaxies (Paturel+, 2003) (983,261 galaxies) Name RA(deg) Dec(deg) Mpc Mlrs PGC2728721 147.68 69.23 PGC2730379 148.07 69.44 PGC28505 148.29 69.00 148 482.48 PGC28529 148.45 68.97 3.685 12.01 PGC28630 148.89 69.07 3.675 11.98 PGC28757 149.38 69.05 4.408 14.37 Holmberg IX PGC2724146 149.48 68.82 PGC2726822 149.63 69.03 PGC28848 149.80 69.26 19.16 62.46 PGC28630 148.89 69.07 3.675 11.98 M81, NGC3031 https://ned.ipac.caltech.edu/byname?objname=PGC28505&hconst=67.8&omegam=0.308&omegav=0.692&wmap=4&corr_z=1 https://ned.ipac.caltech.edu/byname?objname=PGC28529&hconst=67.8&omegam=0.308&omegav=0.692&wmap=4&corr_z=1 https://ned.ipac.caltech.edu/byname?objname=PGC28630&hconst=67.8&omegam=0.308&omegav=0.692&wmap=4&corr_z=1 https://ned.ipac.caltech.edu/byname?objname=PGC28757&hconst=67.8&omegam=0.308&omegav=0.692&wmap=4&corr_z=1 https://ned.ipac.caltech.edu/byname?objname=PGC28848&hconst=67.8&omegam=0.308&omegav=0.692&wmap=4&corr_z=1 https://ned.ipac.caltech.edu/byname?objname=PGC28630&hconst=67.8&omegam=0.308&omegav=0.692&wmap=4&corr_z=1 --- Data taken in Astronomik broadband and narrowband filters: R, G, Hb (486nm) and Ha (656nm). Ha-RGHb total exposure time 12.3 hrs. R 1x1 bin - 24x 120s= 0.8hrs, 3-4 March 2024, seeing 1.3", scope West side, prime focus R 1x1 bin - 60x 120s= 2.0hrs, 3-4 March 2024, seeing 1.3", scope East side, prime focus G 1x1 bin - 30x 120s = 1.0hrs, 30-31 March 2024, seeing 1.4", scope West side, prime focus G 1x1 bin - 82x 120s = 2.7hrs, 30-31 March 2024, seeing 1.4", scope East side, prime focus Hb 1x1 bin - 12x 600s = 2.0hrs. 9-10 April 2024, seeing 2.0", scope East side, prime focus Ha 1x1 bin - 23x 600s = 3.8hrs, 14-15 April 2024, seeing 1.6", scope East side, prime focus ---- Master Lights: FWHM (pxl) pre BXT post BXT R 3.06 1.77 G 2.48 1.65 Hb 3.93 2.23 Ha 3.66 1.49 ----- Plate solver: Resolution ..............1.248 arcsec/px Focal distance ....... 750.39 mm Pixel size ............... 4.54 um Field of view .......... 55' 37.0" x 43' 39.4" Image center ......... RA: 9 55 32.466 Dec: +69 03 32.93 Image bounds: top-left .............. RA: 10 01 11.387 Dec: +69 22 03.61 top-right ............ RA: 9 50 42.343 Dec: +69 27 48.91 bottom-left ........ RA: 10 00 11.617 Dec: +68 38 46.40 bottom-right ...... RA: 9 50 02.906 Dec: +68 44 20.05 ----- 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 PixInsight: Master lights-> crop-> Gradient Correction-> BXT-> linfit-> final master lights PixInsight: final master lights-> RGB Channel Combination (RGHb)-> StarNet2 starless-> export starless fits32 RGHb master, export star mask tiff16. Affinity Photo 32 bit image processing: import starless fits32-> reject default stretch-> curves-> Topaz Denoise(ST0,81)-> curves-> Tpzdn(LL0,84)-> curves (B layer)->Tpzdn(LL3,0)-> curves-> levels (B layer)-> vibrance/saturation/selective colour-> export RGHb tiff16 PixInsight: import RGHb tiff16-> gradient correction-> export RGHb tiff16 Affinity Photo: import RGHb tiff16-> paste in star mask layer, blend mode 'screen' -> paste in Ha mask layer, blend mode 'screen' ----- Processing Star Mask: Affinity Photo: import star mask tiff16-> curves-> invert selection of brightest stars-> curves(invert selection)-> add to invert selection of brightest stars-> curves(invert selection)-> manual white balance-> vibrance/saturation-> brightness /contrast-> paste Star Mask layer on top of starless final (blend mode 'screen') ----- Processing Ha Mask - continuum subtraction: PixInsight: import R and Ha master lights-> pixel maths: create H2 = (k*Ha-R)/(k-1) for a range of k values, 2.0 a good starting point. Select a k which picks out excess emission without too much background-> k=1.5 in this case-> -> pixel maths: create H3 = H2 - K*med(H2) for a range of K values, 1.0 a good starting point. med(H2) function is the median value of the H2 pixel intensity distribution. Select a K which removes any surplus continuum emission in H2 as far as possible-> K = 1.02 in this case-> export H3 fits32 as Ha mask Affinity Photo: import fits32 Ha mask-> accept default stretch-> select target area-> curves (stretch)-> invert selection-> curves (suppress)-> curves (cosmetic suppress)-> convert to RGB16-> levels(suppress G,B channels) -> gausian blur (1.0 pixel)-> paste Ha Mask layer on top of image final (blend mode 'screen') -----
  10. You could always try a barlow as an experiment: keep the bin2 but get to 0.5"/pxl. I went through that exercise on my 150/F5 refractor but reverted to prime focus. Just not enough photons/skies/optics. Vlad (shhh!) also made convincing arguments. I might revisit though..... Simon
  11. I think the trick is to choose the right gb radius. The fov of this pic is so huge that nebulae of interest are tiny compared with the swathes of pollution! But overall, the method is indeed one of last resort!
  12. I don't use the PI scripts, so for each filter I would do separately masterbias, superbias, master dark, calibrated flats, master flat, calibrated lights, cosmetic correction. At this point I would choose one of the calibrated/cosmetic corrected lights from whatever filter to be the alignment reference for everything else. Next step would be star alignment with that reference on each stack separately. Final step is image integration combining the 6.5 and 3nm stacks for each narrowband into the master light for each narrowband. Simples ? Simon
  13. Hi I tried gradexpert, (my goto tool for this - it is free and vg normally), but was roundly defeated! Fiddled with DBE in PI but also butt kicked. So I tried old school: take an image layer and gaussian blur so all stars are blurred out and gone (gaussian radius about 20-30 pixels on your posted jpeg image). Subtract the gb image from the original. A bit of levels and repeat. This is what came out... hope this is useful ? Simon original - quick and dirty crop for square, suggest you add a black background to your mosaic to square it up first round of gaussian blur and subtract second round of gb sub and some curves:
  14. Thanks very much for the link - I will give it a whirl Simon
  15. >>My usual method is to use red continuum subtraction, Ah, do you have any further info on that ? I have been looking for a method of reliably adding in Ha for absolutely ages! Simon
  16. Here is a quick thrash - hope you don't mind
  17. The curse of the blue haloes! I have the same with astronomik blue - if you try the astrodon I would be interested in your report of any improvement. I sometimes wonder if the high cloud you mention (which seems to be almost permanent nowadays) also scatters the blue ? My last RGB outing I replaced the blue with a narrowband Hb filter and was pleased with the end result, so worth a try if you have a mind. As a last absolute cosmetic cheat, I find using the affinity photo inpaint tool works well if you don't know it: after star extraction use the inpaint tool to erase the haloes from the background. There are only a half dozen or so objectionables so not too much work, though you might have to repeat a few times as stretching proceeds. If some of the blighters persist in the star mask they can be edited out and replaced with less stretched versions pasted in over the top. Just don't tell anyone! Nothing to see here etc. Is the filter also giving the off-centre haloes, which seem to change orientation across the frame ? A very good looking pic by the way!
  18. My understanding is that CMOS sensors are more noisy than CCDs, at least in their earlier incarnations. Hence the more and shorter exposures strategy for cmos, but I am not sure if that still holds. The state of the art moves on. My CCD cam is an Atik 460, a very low noise, linear light bucket also getting long in the tooth. Full well is around 17k e- so at 16 bit is well sampled. There is no gain control required - everything can be done in software after the event. I run at -10C summer and winter, it will run at a lower temperature but imho the difference is marginal. The trade-off between number of exposures and length of exposure given limited observing time is one point on which I am undecided. In practice I choose exposure length so that the light pollution floor is just apparent (no pixels with zero count on the histogram) and accept saturation of the highlights (the brighter stars in the fov, which can be fixed in processing). And the more exposures the better of course. Having a library of darks (renewed annually) is pretty much essential, but biases I take every time - the time required is so short that one might as well as not. The point regarding the improvement in sensors is well made. The last generation of CCDs was very good indeed, so your KAF8300 should be competitive. The main improvement with CMOS is number of pixels, sensor area and high speed USB3 readout. And cost. Image processing s/w has improved massively, sophisticated algos available at the click of a mouse - IMHO that has been the real revolution: most of my investment has been in that area the last couple of years. Good luck with your new cam! Simon
  19. It is a spectacular pic. I love it. Stars are great edge to edge, no haloes, subtle colours. I would be really pleased with this. To be totally picky, arguably a red tinge to the dark clouds esp away from the centre of the frame. But are they always grey/black ? And possibly star colours also a bit red tinged. Is that what you had in mind when you said it isn't quite right ? Well done even if you don't like it !
  20. Hi everyone Here is an RGB broadband image of NGC891. I started it in November but the clouds rolled in for the next two months. A couple of clear nights this January and it is finished (sort of), not without a struggle. A few weeks under covers and the 'scope gets cranky with the damp: every plug and socket has to be reseated to chase off the gremlins. And the forgetful operator retrained... For the Blue filter I used a narrowband Hb 486nm filter, 8.5nm FWHM. The doublet objective in the refractor never does a brilliant job on broadband blue exposures, but on this subject were unusually bad for some reason: enormous star haloes beyond fixing. So I tried an Hb filter I have had for ages but never really used in anger, and it worked so well I think it will be standard procedure now. So strictly speaking the pic is an RGHb. I have kept the stars quite prominent and pulled up the background so you can see there are loads of galaxies lurking in the murk, beyond those catalogued in the annotated pic. All those small, faint and hazy blobs that look a bit elliptical or spirally are probably galaxies in the distance. The most distant galaxy catalogued that I could find listed in the NED database is PGC9101 (right and slightly below centre) that clocks in at a heroic 189Mpc or 600 Mlyrs. The other 4-digit galaxies are around 90Mpc. The 7-digit galaxies turned up nothing on Google unfortunately. Next time I might have the wit to include the galaxy cluster Abell 347 in the same frame. Thanks to a posting on SGL it turns out there is a quasar! QSO J0222+4221 (NGC 891 X4 or CX0 J02224+4221) captured as well - red shift z =1.81, equating to a Hubble look back distance of 8.5 billion lyrs. Mind boggling. A personal distance record! Location pic for the QSO is layered onto the snip attached. Exposure and gory processing info at the end. QSO info: https://www.deepskycorner.ch/obj/ngc891.en.php https://stargazerslounge.com/topic/285853-another-quasar/ Annotated: QSO finder: --- Data taken in Astronomik broadband and narrowband filters:R, G, Hb (486nm). RGHb total exposure time 12.3 hrs. Also data taken B and Ha (656nm), though unused. R 1x1 bin - 98x 120s= 3.3hrs, 6-7 November 2023, seeing 1.2", scope West side, prime focus G 1x1 bin - 97x 120s = 3.2hrs, 7-8 November 2023, seeing 1.1", scope West side, prime focus B 1x1 bin - 31x 120s = 1.0hrs, 11-12 November 2023, seeing 1.5", scope West side, prime focus Ha 1x1 bin - 9x 600s = 1.5hrs, 14-15 November 2023, seeing 1.0", scope West side, prime focus B 1x1 bin - 125x 120s = 4.2hrs, 15-16 January 2024, seeing 2.5", scope East side, prime focus Ha 1x1 bin - 14x 600s = 2.3hrs, 16-17 January 2024, seeing 1.75", scope East side, prime focus Hb 1x1 bin - 24x 600 = 4hrs, 18-19 January 2024, seeing 1.1", scope East side, prime focus Hb 1x1 bin - 11x 600s = 1.8hrs, 19-20 January 2024, seeing 1.6", scope East side, prime focus ---- Master Lights: FWHM (pxl) pre BXT post BXT R 3.284 1.918 G 3.282 2.032 Hb 4.239 2.959 B 3.089 - Ha 3.239 2.034 ----- Plate solver: Resolution ............... 0.624 arcsec/px Focal distance ........... 760.38 mm Pixel size ............... 2.30 um Field of view ............ 53' 10.7" x 40' 22.0" Image center ............. RA: 2 22 33.736 Dec: +42 21 03.50 Image bounds: top-left .............. RA: 2 20 09.853 Dec: +42 41 17.75 top-right ............. RA: 2 24 59.116 Dec: +42 40 57.89 bottom-left ........... RA: 2 20 09.974 Dec: +42 00 56.89 bottom-right .......... RA: 2 24 56.196 Dec: +42 00 36.45 ----- 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-> RGB Channel Combination (RGHb)-> export xisf starless master. GradXpert Gradient removal: import starless xisf-> GXPT(20pc,3sg)-> export xisf, fits32 Affinity Photo 32 bit image processing: import GXPT starless fits32-> accept default stretch-> curves-> lvl(B)-> curves-> paint (cosmetic star haloes)->Topaz Denoise(SN, 14, 0)-> Tpzdn(ST,22,18)-> export tiff16 GradXpert Gradient removal: import tiff16-> GXPT(20pc,3sg)-> export tiff16 Affinity Photo: import tiff16-> curves-> export tiff16 Gigapixel: import tiff16-> 2x upscale high fidelity-> export tiff16 Affinity Photo: import tiff16-> Tpzdn(LL, 0, 38)-> selective colour-> white balance-> brightness /contrast Affinity Photo final tart ups: paste in star mask layer, blend mode 'screen' ----- Processing Star Mask: PixInsight: final master lights-> RGB Channel Combination (RGHb)-> StarNet2 star mask-> export xisf starmask GradXpert Gradient removal: import starmask xisf-> GXPT(20pc,3sg)-> export tiff16 Affinity Photo: import tiff16-> curves-> levels(RGB)-> vibrance-> manual white balance-> Gaussian Blur(0.1)> levels-> curves-> brightness /contrast-> manual white balance-> brightness /contrast-> paste Star Mask layer on top of starless final (blend mode 'screen') -----
  21. A great pic - I might tweak it for more dramatic. impact. A matter of taste!
  22. Yes, much better. Still a lot of noise and grain when you zoom in. I ran your last through Topaz : ) - which I think makes an improved version (ignore the star reduction, I should have extracted stars before the denoise and put them back in, but you get the general idea...)
  23. Hi O The only issue I have with your data is the star haloes - there are a lot of them! and they largely defeated me. Load these two images (xstars and +stars) attached into the PI process Blink (0.3s blink interval) and you will see a lot of the nebula circular blobs are associated with the brighter stars. The images are basically your data after PI's Starnet2 (RGB combos with and without stars) with an automatic Histogram/STF. So pretty much not my meddling! Are the haloes a filter issue ? Might be worth investigating further. It could be StarXterminator would do a better job than Starnet2 and get rid of them - unfortunately my SXT trial sub has expired so I can't tell. I take your point re can't see what one is looking at any more. One reason why I like the PI Foraxx Palette utility script is that it is an algorithm. Press the button and it does the job with no user judgement required. No sliders. The amazing thing is that you can feed it quite different colour balanced images and it manages to always more or less produce the same answer. Which adds to confidence. Simon
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.