Aramcheck

I wanted to have a go at this galaxy to see if the outer faint halo would be visible... 66 x 3min subs with ZWO ASI 2600-MCPRO, Skywatcher 200dps + EQ6 mount. 50mm finderscope as guider with ZWO120mini. From 11th May (22:43 - 02:46) with the Moon about 40deg away. Around 1am the neighbours floodlight came on when they let their dogs out. Afterwards they switched their external lights on, which remained on for the duration...

Quite pleased with the result, as I'd found that I've not been calibrating my flats correctly (incorrect dark duration).

Processed in Pixinsight.

Cheers
Ivor

@Jerry Barnes I realised that I gave your some duff info! Recently I've been having problems with my own calibration & so the advice I gave was wrong. When integrating the Normalisation should be set to Multiplicative, not "No Normalisation" as I'd previously said.

My problem turned out to be a mistake using darks taken at a longer exposure length than the flats...

Glad you got your pedastal issue resolved.
Cheers
Ivor

I've only tried it on one small galaxy image so far, but it the results were better than with my conventional Pixinsight noise reduction (MLT/ACDNR/TGVDenoise) & also better than my old copy of Topaz DeNoise.

It took a bit of restraint to avoid the 'plastic' smoothness, but I'll definitely be using NoiseXterminator in future. Plugins for Affinity Photo & Photoshop are being developed too.

Cheers
Ivor

It looks to me that the Master Flat is overcompensating. I had a similar problem with my 2600 MC Pro which arose when creating the master flat in Pixinsight. When calibrating your flats try unchecking "Enable CFA". To create the master flat under ImageIntegration try Average Combination + No Normalisation (*) & Weights set to "Don't Care (all weights = 1)". Under Pixel Rejection pick the appropriate Rejection Algorithm ("Winsorized Sigma Clippng" works well for 30 flats) & set Normalisation to "Equalize Fluxes".

I'd also suggest trying shorter exposure times for the flats/flat darks, or a dimmer light source. With my cheap LED panel I have to make it pretty dim & shoot 0.155s flats.

I've attached the unstretched histograms of your master flat & one of my own.

Cheers
Ivor
(*) my problem occured because I was used to selecting "Multiplicative" normalisation here, which worked fine when I had a DSLR...

22 hours ago, Spile said:

That looks good enough to me! A little tilt rotation error if you feel obsessive.

Thanks @Spile. I'll no doubt get obsessive & also take up @Pixies suggestion of the steel & ptfe washers. The secondary looks like it could do with a wash in any case!

I haven't checked the collimation 'in the field' yet, but the stars are now a lot better than they were.

Here's a very rough process from last night (NGC3198).

Thanks folks!
Ivor

Yesterday I spent pretty much all day going around in circles, getting good central alignment of the secondary to the focus tube, & then messing everying up at subsequent steps...

This morning however I think the penny may have dropped. Here's what I did:-

i) Adjusted the three primary mirror screws so that distance from the OTA base to mirror cell was approx equal (measured with digital caliper)

ii) Used the Concenter to position the secondary central to the focus tube. I then took a picture through the Aline collimation cap, measured the pixel co-ordinates of the focus tube edges & used Pixelmath to draw the centre lines, and photoshop to overlay concentric circles to verify the alignment.

iii) Adjusted the secondary mirror tilt in small adjustments to bring the primary into view, taking as much care as I could to avoid rotating the secondary & only loosening one screw slightly / tightening another at any one time. Once my markers on the Primary mirror ring were visible in the Aline cap, I replaced it with the Cheshire & made final adjustment to make the Primary centre dot coincide with the out of focus cheshire cross-hairs.

iv) Adjusted the primary mirror tilt to bring the centre spot reflection concentric with the primary mirror dot.

As a sense check I then took another picture through the Aline cap & added the focus tube centre / concentric circles to check. The result is far from perfect, but I think it will have to do for the time being.

I've not had much success with the cheap laser however. I bought a Bader ClickLock Reducer to help centre it (and also the Cheshire), but haven't manged to get the laser collimated sufficiently. Best I can manage is about a 0.17 deg error & whilst the laser reflection appears close to the primary mirror spot, it's not good enough...

Looks like we might be getting a bit of clear skies tonight, so I'm eager to see what difference this makes (if any!).

Cheers
Ivor

PS: Over the last few weeks I've also adjusted the focus tube, but now I'm not convinced that was needed. In the end I used an old surveying Laser Distometer and a heath-robinson jig to hold it in position & then measured the distance fom the top of the focus tube to marks I'd made on some pvc tape stuck to the OTA, at a set distance from the focus tube base. I'm not sure of the accuracy of the distometer at these short lengths, but the measurements suggest it's aligned to the OTA to within 0.2mm in either axis.

By default StarAlignment uses PSF fits. Under "Star Detection" you could try changing the 'Compute PSF fits' from "Distortion Correction only" to "Never" & ensure that the Registration Mode is "Projective Transformation"

Juan Conejero suggests this where stars aren't focused in this thread: https://pixinsight.com/forum/index.php?threads/star-alignment-fails.13598/

Cheers
Ivor

Did you do Pixinsight's StarAlignment process before integrating (stacking)?

Cheers
Ivor

A couple of the events  talks & an introduction about Leeds AS were streamed on youtube:-

Brad Gibson - "Stellar Winds: Givers and Takers of Life" https://youtu.be/Da0I1zzZv1s
Julian Pittard - "Colliding Stellar Winds: Observations and Theory" https://youtu.be/TIVBSA9QICs

An introduction to Leeds AS: https://youtu.be/tEY5x89tyWg

Cheers
Ivor

8 hours ago, iwols said:

Any thoughts?

NGC4565 looks a lot better, although I think you could squeeze a bit more out of the image. One tip is to temporarily apply a STF or Auto-boosted STF to check whether there is a colour cast in the image... You can also look at the individual RGB histograms with HT & check that they completely overlap.

I had a quick play with your image, using PixelMath to add small offsets to the RGB channels to align the RGB histograms. Also a rough adjustment of saturation using CurvesTransformation & then another PixelMath expression to flatten out some of the background.

Warren Keller's book "Inside Pixinsight" 2nd edition is worth a look, if you haven't got a copy. It's a bit outdated now, as Pixinsight keeps being improved, but is a good starting point. There's also plenty of youtube videos, albeit of varying quality & I've found Adam Block's Fundamentals worth the ££, so worth looking at his free matierial on youtube.

When processing with Pixinsight, I find it best to work on the Luminance separately to the colour. Typically I stretch the colour data using both ArcSinhStretch & MaskedStretch, and then combine with PixelMath. For the Luminance I use the a series of weak HT stretches.

I've attached a copy of my general workflow.

Cheers
Ivor

Pix_Workflow.pdf

Hi @iwols

Thanks for sharing the original subs. I'm a bit non-plussed as you've got some good images on your website, so was this a first outing with new kit, or any other change to your usual data acquisition?

Looking at the "FitsHeader" process in Pixinsight I can see that you had 3 sessions on the 24/03, 25/03 & 07/04. The 4 x Lum subs on 24/03 are much brighter than the additional 4 on 07/04 although the noise floor still looks very high to me (0.235). The Sombrero is pretty low in the sky & the images are shot through 2.5 aimasses, so it may be best trying a different (higher) target. (I'd also suggest taking a shorter subs, but more of them, depending on what your light pollution is like)

Arranging the 30 subs in date/time order & viewing with the "Blink" process (with the 'Automatic histogram transformation to all images' selected) it looks like there may have been some cloud or additional light source on some frames? Also I would have thought that you'd have got more detail in the images - although the Sombrero isn't something I can see from our North facing garden, so that's more of a hunch than anything. Another member of the Leeds AS also uses an C8 Edge HD (albeit with a ZWO2600MC PRO) & gets some cracking images. With the Atik 414ex (6.45um) & 1422mm focal length (from "FitsHeader") your resolution (0.94"/px) should be fine.

Did you take any calibration frames (flats for each filter/night) & dark frames? These will be essential to get decent images.

BTW - the posterisation effect in the Master Lum galaxy core is down to the differences between the subs on the two nights. If you just integrate the 4 from the 07/04 that will solve that problem.

I haven't managed to get a decent enough image to share yet, but will have another go...

Cheers
Ivor

I only do OSC with Pixinsight, so I could be wrong but to me the Masters don't look right. In my images after calibration the Normalised median value is typically about 0.014. In these masters the Lum = 0.446, Red = 0.149; Green = 0.182; & Blue = 0.105.

Also to me it doesn't look like the Flat Calibration is good, as the RGB channels aren't uniform brightness in the corners/edges & there's variation between RGB. i.e. the Red is darker on the right hand edge whilst the Blue/Green are opposite.

The Lum master also has some nasty pixelation (posterisation) in the centre, which I've never seen before.

Would you be able to share a link to the uncalibrated data (or at least a sample)?

Many thanks

Ivor

Thanks for all the advice. We were away for a week (without scope), so I've not had chance to look into it properly yet.

I think the Cheshire pic may have been taken off centre, but as I've made further adjustments since, I need to start from scratch. I've bought a cheap Svbony laser & am attempting to collimate that. Also looking at Astronomyshed's youtube series I think I'll invest in a length of M5 stud bar to check the focus tube angle. (pt 1 of his series: https://youtu.be/zd-fl9SEYHw )

Thanks again!

Ivor

Gosh! Many thanks! I think I just got lucky that night & hadn't thought of entering the competition until @tomato & @Clarkey suggested it (Cheers! 🍻).

Processing the data was a bit of a challenge, as I ran out of memory on my old PC and disk space on the laptop, so had to keep swapping files between the two...

Many thanks!
Ivor

I treated myself to a Concenter with my Christmas FLO vouchers. It made centering the position of my SW200dps secondary mirror to the focus tube pretty easy, but when I then look through my Aline Collimation Cap I can't get the secondary aligned perfectly with the primary mirror. Does this mean that the focus tube needs adjustment?

Attached are handheld camera snaps (a) through the concenter; (b) through the collimation cap & (c) the view through a Cheshire.

I've modified the scope with a plastic ring to hide the primary mirror clips & added a bit of masking tape to the ring to show their position during collimation. The collimation cap image has been adjusted to make them visible & you can see that the one at the top extends much further into the view than the other two. The image through the Cheshire also shows quite a bit of crud that's accumulated on the primary.

Is the collimation good enough, or should I prioritise getting the position of the 3 mirror clips equally visible, over making the secondary nice & round?

I've also attached a examples of my most recent images, prior to attempting the collimation. I think I may have other issues with data on some of the nights as the stars in the Spider/Fly are worse than in the other two. All were taken around the same time just over a week ago.

Any advice much appreciated!
Ivor

Details here:- https://britastro.org/event/baa-spring-meeting-leeds

Cheers
Ivor

Likewise - I did a quick & dirty process in Pixinsight. DBE did a fair job of removing the vignetting.

Cheers
Ivor

Street lights are exempt from the 'statutory nuisance' lighting definition under the Environment Protection Act 1990.

If Environmental Health are unwilling to help, then try your local Councilor(s). They will have more clought.

I would have thought that you should be able to get a shield fitted to the streetlight directly affecting your daughters bedroom, but the other one will be more difficult to justify.

Cheers
Ivor

I just posted this video of Asteroid (7482) 1994 PC1 in the "Widefield, Special Events & Comets" page, & a couple of folk suggested entering it here.

Couldn't capture it on the 18th, so this is from the 19th Jan when it was visible from our back garden. Skywatcher 200dps / EQ6 & a ZWO camera.  243 x 10 sec subs from Cookridge, Leeds (Bortle 8). AstroPhotographyTool had crashed shortly before the sequence so the camera unfortunately wasn't cooled... (I had park / reboot & require the target in a hurry).

Processed in Pixinsight & then combined into a video with Adobe Premiere Elements 15... To process, I created a background image using a regular Pixinsight workflow. I then applied a quick & dirty standard workflow to each aligned sub using a ProcessContainer & then used PixelMath to reinstate the asteroid back into copies of the master background.

For the inset images I applied CometAlignment to a set of unprocessed StarAligned subs. These were then processed using an ImageContainer & a rough workflow, to give the close-up views, which were then inserted into the relevant Main image frame with the InsertImage script.

Premiere Elements was then used to animate the 243 frames into the video.

Kinda glad that it's all over & done with!

Cheers
Ivor

Couldn't capture it on the 18th, so this is from the 19th Jan when it was visible from our back garden.

Skywatcher 200dps / EQ6 & a ZWO camera.  243 x 10 sec subs from Cookridge, Leeds (Bortle 8). AstroPhotographyTool had crashed shortly before the sequence so the camera unfortunately wasn't cooled...

Processed in Pixinsight & then combined into a video with Adobe Premiere Elements 15...

Cheers
Ivor

Thanks for the heads up on PhotometricMergeMosaic. Because I didn't have the camera angle set correctly, after making the registered panels, I had quite a nightmare splitting them into strips so that PMM had a vertical (or horizontal) overlap it could work with. The results were a lot better than with GMM though & I've learned a lesson!

I managed to get some more subs on the 11th, but had to ditch most of them due to dew on the secondary mirror & naff flat calibration.

1st attempt at processing attached. Wish I'd managed to get more time on each panel. From Bortle 8 skies...

Cheers
Ivor

Thanks folks. I'll have a look at PhotometricMosaic.

I've still got a few holes to fill, if skies permit, but using StarAlignment / GMM did a reasonable job in the end - though I didn't spot the pinched stars until I reapplied the Auto-stretch on the 'finished' image. The unprocessed result currently looks like this:-

Cheers
Ivor

I think I've found the problem... Although the surrounding area looks black, it isn't. Replacing these pixels with Pixelmath and trimming the image being joined has eliminated the artifact. Also using Clonestamp to mask out any stars right on the "join" edges.

i.e.  iif($T<0.0009,0,$T)

Cheers
Ivor

I'm attempting to create a mosaic of M31. Any ideas on how to prevent this artifact that appears with GradientMergeMosaic?

Thanks
Ivor

I use blackout material on a wooden stand, which I can attach to a washing line pole... That and material draped over the washing line. Also a tarp with an extendable decorating pole in another spot. The tarp can get quite noisy if there's any wind & when packing up in the early hours.

In the past I've also used a plywood sheet attached to a pole to prevent moonlight from shining on the top of the scope, but holding that in postion for a few hours gets very boring!

I'm currently thinking of positoning a mirror in the garden to reflect one really annoying light back into the neighbours face...

Cheers
Ivor