jimjam11

I dont do much planetary imaging but thought I would attempt it last night and it feels like the mono workflow is a bit of a nightmare. 1. Capture was easy. I used firecapture with it set to automatically capture 60s of r, g and b. This resulted in 3 files per run. 2. I stacked each file in autostakkert, again easy to do and it ripped through the 200 videos I captured quickly. 3. This left me with 200 mono images. What is the optimum workflow from here in terms of assessing quality, combining and ultimately processing to the best combined rgb images? I used pixinsight to combine some and registax to process but this is very time consuming and repetitive, it feels like I am missing significant opportunities to streamline?

Interesting. How does it work, I have never used it and the documentation seems sparse? If I want to connect CDC I typically connect it to my mount but it sounds like I am missing something...

What functionality do you think you are missing comparing v2 to v3? I think SGPro was the benchmark for acquisition automation software for the masses but it is very CCD centric and it hasnt evolved. The move to a subscription model was the final nail for me and I havent used it since. Before paying for subscription software check out Nina or Ekos. Both are free and much better optimised for CMOS imaging (especially Nina).

Your images look good, but they do look oversampled. 1. You could use the PI SubframeSelector process to measure FWHM on all your subs and look for outliers, but also compare them to the theoretical values vlaiv suggested. I use this on every sub I capture and record the values so I know what range to expect from certain combinations (e.g. ASI1600mm + ZS73 gives me typical stars of 3.9" with my seeing) 2. Depending on the output from SubframeSelector you might not actually be in focus. There is a skill in using a b-mask (which I dont posess!) and when I switched from b-mask focussing to autofocus I got smaller stars. Using something like the b-mask tool in Nina can help ensure you are really in focus. 3. If the ZS61 is like the ZS73 it will have a tendency to bloat brighter stars.

Adding a few more images... A 2 panel mosaic of the gamma cygni region Starless: The Cocoon nebula: A finally 10 hours of the rosette (4hr RGB, 6hr Ha):

It looks like you have the flat73 and are looking at getting the flat73r. I have both but the reducer isn’t perfect: 1. It doesn’t support full frame. 2. Star sizes are worse; my long term average fwhm is 3.9” vs 5” with the reducer. 3. Star shapes are significantly worse in the corners, spacing is very finicky (although easily adjusted). 4. I have never noticed an snr improvement, there are too many other variables at play. Having said that, it does give a slightly larger fov which can be very useful! It also feels better made and the rotator is vastly superior.

The stripes are walking noise most likely caused by not dithering as Miguel said. The inconsistent (not flat) background can have a number of causes, especially if your setup has light leaks or reflections which makes background subtraction much more difficult. I don't know anything about siril, but if I got a background like this I would he looking for samples placed too close to stars in the dark areas which are then fooling the background modelling prior to subtraction. APP can do a decent job but the nuclear option is something called a synthetic flat in PI. This will flatten the background although I don't think it will help with the walking noise: http://trappedphotons.com/blog/?p=756

Nina is excellent and very fast which is important with a cmos camera. An easy way of verifying your polar alignment is just to run the phd guiding assistant for a couple of minutes, it will give you an accurate value without you having to do anything! Ideally do this pointing south towards the celestial equator...

Or use sharpcap which is a doddle for polar alignment using your guidescope. No point bothering with a polarscope nowadays if you are imaging with any kind of computer because there are so many cheap, fast and better solutions. You can make this even faster by marking the location of your tripod legs after alignment so you can start from the same place every time.

Apologies. I am talking about sampling and what sampling rates are manageable on a star tracker. In general the star adventurer is fine with sampling of >4"/px at 30s subs without guiding or other faffing. With guiding you might be able to reach 2"/px but then weight becomes a problem. Basically stick < 200mm of focal length with a star tracker and your camera unless you like hard work and guiding!

Well spotted. The ZS73 is great but bright stars (especially those as bright as M45) bloat very quickly in my experience. This is especially noticeable when switching between my ZS73 and 200p. I would guess this is sperical aberation:

600d gets you 2.06"/px. Have you considered something like the EF200 F2.8 or Samyang 135 F2? These are super light in comparison and give a much better image scale for a portable tracker. EF200 gets you 4.4"/px and the Samyang gets you 6.5"/px both of which are much easier to handle on a star tracker.

4 panel mosaic of NGC7822 1 hr RGB, 1hr Ha per panel RGB - 20x60s Gain 76 Ha - 12x300s Gain 200

Far too heavy, and depending on your camera it will yield approx 1.8"-2.3"/px (@3.8u pixels) which is beyond its tracking accuracy.

AF makes a huge difference and most parfocal filters are not parfocal enough, so once you get your offsets worked out things become much easier and you will get sharper images. Nina is in the process (V1.11) of introducing a killer feature for people like us with temperamental sky. You can configure your sequence (e.g. 5xR,5xG,5xB) and then iterate through this group for as long as you have/desire (e.g. 10Hr). It means you can get a colour image within minutes, and progressively go deeper and deeper without paying the time penalty of full filter interleaving (1xR,1xG,1xB) which takes an age. It also means you smoth out fluctuations in seeing, transparency etc so all channels are better balanced. This was technically possible before but you needed to configure a ridiculous repeating sequence (create 5:5:5 events and copy/paste until you lose the will to live). Having said that, IMO the new sequence builder in Nina is seriously intimidating and it took me several nights to get comfortable with it. It is like no sequencer I have seen before but it has flexibility and benefits which makes the effort worthwhile...

Looking forward to m33! Any reason why you have relatively little lum compared to rgb? I normally go 1:1 with lum = combined rgb not each channel being equal.

Both are in stock @FLO and they are both very well made; I would not buy a non-WO flattener for this OTA given their availability. I have both the Flat73A and Flat73R. When I got my ZS73 there was no reducer option available but I got the reducer version once it was released. With the Flat73A I configured it to the WO backspacing (11.4mm IIRC) and never touched it again. It gave perfect stars from corner-corner. The Flat73R gave me the run-around and when I first used it my stars were terrible. WO do not publish spacing guidance for dedicated cameras and I had to find my optimum by trial-error. They specify 1.8mm for DSLR and mirrorless but with my ASI1600 I have ended up somewhere around 2.7mm. Despite this the field is not as flat as the Flat73A, but it is significantly wider (With my ASI1600 it works out as 443mm vs 346mm or 1.77"/px vs 2.26"/px). I have not sold the Flat73A yet because it definitely works with a full-frame sensor; there seems to be ambiguity around the Flat73R. Initial material implied it was full-frame compatible, but the WO website now claims it is optimised for ASP and smaller and I am not convinced it would produce an acceptably flat field for a full-frame camera.

The ASI Air Pro is an amazing device and it is rapidly evolving. However, having used it for a few months it does have some limitations which have led me to use it slightly differently. Pros: I have a Polemaster and Sharcap, but for polar alignment the ASIAir is in a different league. You click go and it takes an image which is plate-solved. You click go again and it moves the mount, plate-solves and then tells you what correction to make. Polar alignment takes me <2mins with my average time approx 70s. I can PA with the Polemaster in approx 3mins, but if I need to do the rotation calibration it is more like 6mins. Sharpcap is good when it works, but I have found it very hit/miss. The £280 polemaster is redundant. Platesolving is like witchcraft. It is based on the raspberry pi but it platesolves faster than my i7 laptop running SGPro/ASTAP. Custom objects - It has an extensive built in catalog but you can save your own objects by specifying the ra/dec. Power distribution - It has 4 12V sockets including PWM control so no need for a dew heater controller. The £260 pegasus powerbox is redundant. Autofocus - As of the current beta (Sep 2020) it can do autofocus and it appears to work well. Speed - It is blindingly fast at saving an image and moving onto the next sub. This delay can be very significant with short cmos subs and something like SGPro. Rate of evolution - ZWO are developing the software rapidly. They listen to users and it is becoming more and more capable. Fewer cables - The HEQ5 doesnt have any cable management but with the ASIAir you only need to supply it with 12v power. The only other cable coming off the OTA (assuming on-OTA mounting) is the cable to the mount. 12v power - It runs on 12v power so you can run the entire setup easily on battery using a single 12v feed up to the ASIAir. Cons: Wifi connectivity - it doesnt work so save yourself the hassle and buy the TL-WR902AC. Platesolving - It solves using the jnow epoch rather than the more standard j2000. This means you cannot plan images with something like telescopius. Mosaics - It has no mosaic capability although this is possible using telescopius and custom objects if you remember to convert the telescopius ra/dec for each panel into jnow. Also remember each goto needs to be manually invoked. Platesolving - The solve returns the rotation angle, but it is not part of the goto. You therefore need to manually ensure your rotation angle is correct. An imaging sequence is still manual; once Pa is completed you need to goto your object (it will platesolve to an accurate location), turn on the cooler, configure guiding, run autofocus and finally start autorun. Once complete it can park the mount. Gain is manually set. I typically image at gain 76 for RGB and 200 for NB. Autorun cannot currently do this so you need to autorun for RGB and then rerun for NB after changing the gain. Autofocus - I believe this is based on a single star as opposed to the entire frame. This combination of pro/con has led me to an interesting place. SGPro has way more features and is completely automated, but it is slow and cannot do PA. I bought a cheap mini-pc from amazon which runs windows/sgpro and now I use the ASIAir for power distribution + polar alignment before switching to the mini pc if I am looking to do something the ASIAir cannot currently do (such as mosaics). My concerns with this approach are: SGPro switching to a subscription model. The horrible performance of SGPro made worse by the underpowered mini-pc. This leads to significant interframe wastage which I measured this by taking 20x1s exposures. SGPro needed 4 minutes(!) to complete this take. By comparison Nina only needed 47s with an average interframe delay of 1.3s which is completely acceptable (and insignificant) once subs are >60s. Hopefully there is some useful information in here for people considering the ASIAir.

I like the idea of NGC7023 as well. If you have amassed another 240hr of integration with more in progress is there a chance you could start releasing 2+ datasets per month to keep up?

I saw this when I went searching but couldn't find it anywhere to download. I am guessing they pulled it to try and get people to upgrade to the new 'improved' model...

I dont have a huge amount of experience guiding/dithering with my SA, but if I take a sequence of subs with my SA and dont guide/dither I always get terrible walking noise. When guiding/dithering I have never noticed it but I always specify a large dither. Using something like the SA it is impossible to get perfect polar alignment, so there will always be drift in dec: If I randomly dither this in RA only I get: At this point there is no correlation because of the drift, size of dither and the number of subs? To make this work I think the dither amount needs to be large enough to eliminate/minimise the overlapping, this would need to be substantially larger than a random RA/Dec dither? The longest FL I image with on my SA is 200mm which gets me 4"/px with an ASI1600. If I specified a 5' dither quantity I believe that would equate to 75 pixels which means it should be good for plenty of subs before the correlation becomes a problem? Without RA dithering walking noise would become apparent after a handful of subs. At 50mm I think you would need to specify a dither amount of 20' to get a similar result.

Loving the free (and high quality data). I initially processed this a couple of times in typical HOO and SHO: I then blended these two images in PixelMath to produce a composite: Where I thought this became interesting was animating the three blends together; it helps to highlight different structures. The SHO version clearly shows better structure separation?

Yes I saw it in this video: It is not demonstrated but it is documented in the text. I seem to recall seeing a screenshot of the app where it is also visible. There is a section within the app (shared with the sa mini). Hopefully this means you program your capture sequence and then upload it to the SA. However, I dont think the SA can trigger an astro camera which does limit its usefulness. Dithering is so important it should be default behaviour for all of these tracking style mounts.

The new star adventurer 2 is supposed to do unguided dithering which is super promising for something so portable...

Looks good. What cable did you use for the 12v run, how long is it and what is the voltage drop like?