ollypenrice

57 minutes ago, AstroRookie said:

@ollypenrice

Don't know if I could orientate the camera as you showed, I think my guide scope and camera would somewhere be in the way - I'll check. I have considered upgrading the focuser, but if that is not the problem then this is yet another "useless" upgrade, and a "30 days money back guarantee" is taking into account my location not really helpful 😉

@michael8554

Diff flex: that's interesting, as I have to say up to now I've always been imaging near the zenith and this is the first time I have been imaging at +- 40° altitude south (rosette nebula) - I'll have a look on how the cables and guide scope behave at that position

Thank you both!

AstroRookie

 

 

This change in orientation might also introduce tilt.

Olly

A small amount of matt black paint fully covering the scratch and of a soft, rounded shape, will entirely eliminate it as a cause.

13 minutes ago, AstroRookie said:

Hi Clarkey,

Maybe stupid question, but would that not be reflected in the graph of the guide software?

AstroRookie

No, the guide camera might be following the star perfectly but the imaging scope (or just its mirror) might be moving relative to the guidescope. An off axis guider removes this possibility by guiding on the imaging light cone itself. So called 'Mirror flop' has essentially the same consequences as flexure.

However, I'd put money on tilt, in your case. (Not a lot of money but a bit... ) From what I can see, the elongations all go the same way, meaning they are parallel with each other. Optical defects rarely produce this effect but tilt does. Since you see the same thing in two cameras, it is unlikely to be a tilted chip. (These are not uncommon and some cameras have chip tilt adjustment built in.) I would look for sag or slack in the focuser. Are all your attachments screw-fit rather than push fit? Screw fit is best. Are you placing your camera in this orientation, parallel with the counterweight bar? It's the best.

Tilt can spring up suddenly because some mechanical component becomes loose or a bearing breaks.

Olly

I would rather say that OIII is faint, rather than noisy. 

Russel Croman's Noise Xterminator is an order of magnitude better on astrophotos than any other NR routine I've tried. Because our RASA 8 data invites extreme stretching, I do often lift it well above the noise floor but StarXt sorts it out in an invisible or, in extreme cases, almost invisible manner. Here's a close crop pushed beyond its limit. Before Noise Xt...

After noise Xt...

Had I wanted to erase the brighter parts of the noise reduced image, feeling that the NR had damaged them, this would have been a ten second operation in Photoshop Layers but, really, are they damaged? Here I had the 'Preserve details' option in StarXt set to minimum. Alternatively, the noise reduced image could now be further sharpened.

Olly

12 minutes ago, AKB said:

NOW you tell me … !

See my post of last night’s session:

 

Ouch, and if you want to do this collision thing properly (!!!) do it with a RASA...

Phew.

Olly

I have three automated sheds based here and have developed some Golden Rules, the first of which is to accept that, if a thing can go wrong, it will. For this reason we have a 'No Possible Collision' rule. No telescope can ever stop in a position in which the roof can collide with it.  (A much more lavish remote hosting provider a few miles away has exactly the same rule and he is orders of magnitude ahead of me in IT savvy.)  A design using 'rolling roof and upper sides' makes this rule easier to comply with.

Another simple thing is to make as much of the gear inside as showerproof as possible. Despite the best efforts of the IT guys who are responsible for the sheds here, unwanted openings have happened several times. Simple shower protection of fixed items can save a lot of money.

Good UPS backup is essential but not so easy to assure because office UPS machines don't like the temperature extremes of observatories and tend to have a shortish life. You probably won't have our upper temp extremes, though.

Olly

It would be a shame to see a thread about books turn into a thread about screens!

Olly

We're running a RASA 8 quite happily on an NEQ6. This will weigh about the same as other 8 inch catadioptric systems. I've also imaged with a TEC 140 on this mount. I don't think a 6 or 8 inch SCT or RC will overload it.

However, that's not the problem. The problem in imaging at high resolution is guiding accuracy. Have a look at your guide RMS in arcseconds while using your refractor.  Fifeskies, above, gets 0.5 arcsecs 'on a good night.'  This is a good value for an EQ6 and represents the best you are likely to achieve, but not all examples of the mount manage this. Anyway, you probably know your RMS or can easily find it. Multiply it by two and this gives you a decent estimate of the best resolution in arcsecnds per pixel that your mount will support. If your RMS is 0.9" there is no point in building a system which is working at 1.0"PP. If your RMS is 0.5" then, yes, that would work.

Your local seeing is likely to impose ts own limit, as well. Where are you based? How stable is your seeing?

Olly

The first guideline is simply 'Don't.' Even quite a significant amount of dust has no discernible effect whatever.

When, eventually, it becomes necessary you can follow the FLO guidelines above.  My one caveat would be to suggest a first use of cotton wool by dabbing once and discarding. I wouldn't use any wiping action till the mirror was already very clean and I would always wipe both gently and in curved strokes since straight scratches (god forbid) will show at the eyepiece when curved ones won't.  Remove rings etc before starting, or use thin rubber gloves.

The website I normally recommend is no longer active so I took a quick look at the first few U-tubers who popped up in Google. They were not at all good. Stick with the FLO advice.

Olly

Just now, paul mc c said:

For some reason I never had a problem but I have just installed a reducer and I am now getting some of the dark darker streaks as seen in the left bottom corner, the third one is just the non stretch image.

Images are just raw shots in Asiair.

As long as the same defects appear in the lights, these flats will fix them. There is really no way of knowing till you try them.

If your flats were perfect, ie totally flat, you wouldn't need them.

However, Adam's point is correct.

Olly

If they work, they're good. If they don't, they're bad. When you take an astrophoto and apply them, do you see the defects seen in the first two flats. (ie slight vignetting, severe dark dust bunnies on the left, smaller ones near the centre and dark spots most notably near the centre?)  The small dark spots may not disappear because the source must be close to the sensor and little signal may be getting past them.

What are we looking at in the third image? I don't know this software.

Olly

13 hours ago, bomberbaz said:

Yes I have just checked and can confirm I did indeed clip it. Having reviewed what I did I figured it out. I tweaked the black levels in exposure to darken the background and this is where the clipping took place.

I have quickly redone the image from the base stack but this time I ran graXpert before any gimp tweaking and it allowed me to retain all the data without clipping.

I do think some data from a dark site would help an awful lot though.  See below

I don't do background levels by eye, I always use the Ps Colour Sampler (in the Eyedropper tools) to measure it. I feel 23 per channel is ideal but sometimes have to settle for less. Imaging with the RASA complicates this because there is far less background sky than with slower systems. The RASA reveals faint nebulosity just above the background and also dust-darkened regions which lie just below the broader background. The deeper you go, the more you will encounter this.

Olly

28 minutes ago, bomberbaz said:

I have been guilty of black clipping myself in the past but not so now. I watched a nebula processing tutorial online which basically mirrored what your saying @ollypenrice

However having just checked the image above, it seems Gimp histogram shows it is black clipped but I am sure it wasn't when I processed it. I shall have another look at the master image and see if I can get something back out of it, strange.

steve

Your image doesn't look obviously clipped, to my eye. Sometimes it can happen at the conversion to JPEG. Maybe have a look at your last 16 or 32 bit image prior to JPEG conversion for the web?

Olly

Edit: I have tested, several times, a screen grab against its original JPEG to be sure that the screen grab itself doesn't introduce any clipping and I have never found that it did. I think the screen grab does represent the original JPEG, therefore.

36 minutes ago, PeterC65 said:

That's very interesting @ollypenrice. When I do EAA I almost always black clip to darken down the background. It sounds like it would be better not doing this if my intention is to get the best possible image with post processing.

 

Certainly yes. The thing about the faint data is that, when its gone, it's gone. Once you get your eye in, you can see when an image is badly clipped. The sky is jet black and even, because the noise-brightened pixels and any faint genuine signal have been clipped out. For a proper grasp of the black point, though, you need to keep looking at the histo in levels.

It is actually easier to remove LP gradients from an unclipped image because the software has a more genuine picture of where it is. Gradient removal should be the second operation on the stack, the first being the edge-cropping of any border artifacts. If you don't crop these out they will confuse the gradient tools and give a messy histogram which is hard to interpret.

Olly

44 minutes ago, Elp said:

Don't forget your very dark skies and apertures you're using too.

Yes, so in sites of greater LP you need far longer total exposure. When using F5 and F7 instruments we typically exposed for 15 to 20 hours.

It's a slow business...

Olly

If you use a star removal software you can process the nebula and the stars separately. This would have been a holy grail five years ago. It is possible, using this method, to make stars as small and tight as you like, so much so, in fact, that I usually give mine a blur.

Stars can also be made to look tight if the image is black clipped, because the faint signal around the stars is discarded. Unfortunately so is the faint nebular detail as well - unless you are processing the stars separately. The histogram clearly indicates the clipping in both the images posted here.

It is of critical importance to get the histogram right, with no black clipping. There must be a little flat line left of the peak.

If the image is black clipped it is impossible to know much about the stars because most of what you'll see will be stellar core and the faint object nebulosity is also gone for good. So is most of the star colour, which is contained around the outer edges. It can be tempting to use the black point to cut out gradient, but don't be tempted to do this. We use DBE in Pixinsight to remove gradients but other tools work as well. Clipping is never, ever, the way to do it. Check the histogram after every processing operation, and trust it.

Olly

With an observatory based telescope, a modern F2 astrograph and a modern, cooled, CMOS astro camera, we spend between 3 and 6 hours capturing a target.  Post-processng takes at least three hours, though a large mosaic may take a week.

If there are any short cuts, I have yet to find them.

Olly

In France La Poste has had some of its services, notably larger parcel services, privatized and some not.  I dread the discovery that the privatized Chronopost are to deliver a parcel because their personal best (ie worst), so far, is nineteen consecutive orders not satisfactorily delivered to us.  Looking on the internet at customer feedback, our experiences are not unusual. When a delivery simply comes with the post lady, however, it...  just comes. If we cross paths with her on the local back roads we'll both stop and she'll gain a couple of minutes on her round by passing stuff through the car windows. Humanity!

I was picking up a parcel from a local drop-off point when a rusted, dented van pulled up, sign written with what translates as 'Dirt Cheap Van Hire.'  This, ladies and gentlemen, was the Chronopost agent in all his sub-contracted glory. 

Olly

On 24/02/2024 at 19:07, TiffsAndAstro said:

In an hour or so I'm gonna try quick polar along, 3 maybe 2 star alignment then full polar alignment. 

 

Cue clouds

Polar alignment and software star alignment are totally different things.

Polar alignent physically points the RA axis of the mount towards the north celestial pole, near Polaris.  You are aligning the real metal mount so its RA rotation axis is parallel with the earth's axis.

Star alignment is an all-software procedure in which you are synchronizing the position of the stars in the sky with their positions in the virtual planetarium contained in your  mount's software. Using a 3 star alignment means you can do this synchronization with quite a poor physical polar alignment. If you have a very good polar alignment, a one star alignment will do.

Olly

2 hours ago, R26 oldtimer said:

I see your point. There are two reasons that make me want to try RGB stars. One is that I don't like magenta SHO stars even with inverted green SCNR, but that is sort of fixed with adding HOO stars instead. The other reason is that ultra narrowband filters produce really weird diffraction spikes with my newt that don't coincide between filters.

Other than that I completely agree with you, no need for true colour stars in a false colour image.

If it's good star shapes that you want, wouldn't you get the best ones through your red filter?

Olly

Nice to see this getting some attention. As you demonstrate, it's a fine nebula.

Olly

Nice. I can't see that it makes any sense to mix a colour mapped palette like SHO with a natural colour RGB palette, but RGB stars in HOO make sense since HOO resembles RGB.

Olly

That IFN is really impressive.

Olly

A square sensor is only a disadvantage if you consider it as having one short side: if you consider it as having its short side extended to the same length as its long side, it becomes an advantage!  In truth, it's the best possible shape for any straight sided sensor. You just need a bigger square sensor...$$$$$$$$ 

Olly

Because I worship at the shrine of, 'If it ain't broke, don't fix it,' and because setting up new astrophotographic gear is usually fraught with stress and frustration, I do no astronomical comfort buying whatever. My idea of comfort is using stuff that works.

However, as for comfort buying in other areas - now you're talking!

lly

Newt experts, like Pieter Vandevelde who built this fast instrument himself, set the camera up under the OTA rather than above it. Here he is at my place:

You'll need less counterweight that way and be less likely to bump the camera when working around it.

I'd be prepared to devote initial sessions to careful collimation with the Quattro and, perhaps, consider using an artificial star to get it as close as possible on otherwise wasted cloudy spells.

Olly