Preserving details in difficult Iris

[hixx]

Hi Arrowspace90,
interesting discussion...
Yes you definately need flats for better Wipe output. However, though flats removed the vignetting in my shots, I still had to struggle with similar noise in the background when starting off.
I use an unmodified and uncooled camera. And with light pollution there is a lot of LP shot noise. Wipe will remove the LP, but not its noise component!
The only way I found to bring this LP noise down is more integration time. Try double the time (double amount of lights) and also increase the number of darks and flats and darkflats (or bias). I use to take 3-4 hrs on my f/5 refractor.
With this you'll get a much more even background.
The strategy you asked for bringing noise down: I found plenty of options in Star Tools. Check Guy's User Notes or the Unofficial manual...
- In Wipe : increase Dark Anomaly Filter until no further change.
- In AutoDev: increase Ignore Fine Detail until no further change, then reduce Outside ROI influence, then decrease Shadow Linearity (until darkening of faint details is starting to show)
- In Contrast: Decrease Dark Anomaly Headroom
- In Life: Use Isolate preset to push down background. Also helps with busy starfields.
- In Flux: there is also a Noise reduction Algorithm - helps just a tad
using all these steps (I am sure there are some more) you might be able to bring noise down just as much to have Denoise or Denoise 2 work the rest successfully.
As said, starting with a high quality stack is actually the better choice, at least for faint detail objects.
clear skies,
jochen

[arrowspace90]

This flat illustrates where I'm at.

The histogram indicates to me that all the light is in the middle. So just in between dark and light. Shouldn't this work? If not, why not?

[hixx]

1) BTW Which stacker do you use?
2) This flatframe seems ok to me.... You'll now need to create a Masterflat.
3) The Masterflat is created using a number of these flats plus either Bias frames or Darkflats
4) You should use at least 20 to 30 frames of each type.

Does that help?
jochen

[arrowspace90]

Well I am going to run up the white flag and concede defeat on this target.
Last night I re-shot Iris Nebula mostly to test out better flats. I have no doubts whatsoever that the Star Tools Creator would take one look at this and wish me future luck in learning how to acquire flats.
The vignetting of the raw stack is beyond obvious of course. This despite carefully previewing the flat histograms and moving the light curve right to the center, exactly like the photo I uploaded previously.
It worked out to .06 second exposures, though I took 30 more at .07 and 30 more at .08 that I haven't tried.
The result was about the same as my previous effort.
I took 140 light frames at 45 seconds each and the only culprit I can think of was that I did have the Celestron LP filter installed.
I tried using AutoDev after wipe, as is recommended, and which has brought me recent nice images, but it failed and the image was full of noise.
I resorted to the ManualDev Digital Development, which I know the Creator regards as hamfisted. That at least gave me an ok result, but it's not in line with my goal of a proper post process.
Perhaps there is something inherent in the RASA 8 that makes it prone to vignetting, I will make inquiries. I even scrutinized the dew shield to make sure it wasn't canted into the light train. But it is apparent, at least to me if no one else, that I am doing what people say to do with flats, and when viewed by experienced people, my flats appear "ok". Perhaps the LP filter tips the balance too far, but I am in Bortle 6 1/2.
I use APP for flats. No stretch. I don't expect answers from anyone, so I am wondering if I should next time just use the IR cut filter, then use the LP tool in APP before bringing the stack to ST. This is not something the Creator would sanction. I will have ups and downs as I gain experience. I do plan to stick with ST. Comments are welcome.
RASA 8.
ASI533
Celestron LP filter
Stacked in APP no stretch

[hixx]

this is a hard one no doubt but don't give up that easily
1) You did not mention you created DarkFlats or Bias frames
this is very important because otherwise you`ll introduce noise into the stack: The darks are correcting the lights for noise floor, but the bias noise of the Flats also need to be subtracted.
I use Darkflats - just do another 30 flats, but with lens covered. It is basically Darks with the same parameters as the Flats.

2) I am using APP as well to stack, here`s my settings different from APP default:
- in 2) CALIBRATE: unselect adaptive pedestal/reduce amp Glow !!!
- in 5) NORMALIZE: select Method: none
- in 5) NORMALIZE: unselect neutralize background !!!
- in 6) INTEGRATE: keep local normalization correction and MultibandBlending disabled

3) I had noticed APP forum articles saying Flat and Dark calibration was not working for some folks (it works for me though). You may try one stack with flats and one without. Do they show the same amount of Vignetting? If you do Calibrate with flats and your stack still has vignetting upon opening up with AutoDev in StarTools, I'd suggest to contact Mabula from APP.

hope this helps...

[arrowspace90]

Ha, not giving up! Ok, so my ASI533 OSC has zero amp glow. I'm of the potentially erroneous impression that it does not require darks.

However, I have in the past few days begun to shoot Bias frames after my flats. And as previously noted, I have improved my flats by placing the histogram light curve for them right in the middle.

The last 2 emission nebulae I have shot subsequently have worked great! But for them, there's a different filter on there, the L-eNhance instead of the Celestron RASA 8 LP filter. I'm not ready to blame my problems on that filter, but I may shoot the Iris again for a 3rd time, next time using a UV/IR cut filter instead.
Coupled with improved flats and bias frames, I will find out if any of these things made my raw data so noisy.

I would point out that my poor camera is trying its best in very sultry Texas summer conditions, only able to cool to 0 C with a 100% effort. However, these conditions did not seem to bother it when shooting a big, bright emission nebula.

[arrowspace90]

Now shooting darks also. It's not that hard and if it helps I am up for it.

[hixx]

Maybe one more chance:
if you did'nt crop already, check if croping some 10-20pixels at the borders would help.
this purple "glow" only seems very odd. maybe a combination of filter use and stacking artefacts let Wipe kick back on the corners for the red channel only.
If you do apply flats and still see this amount of "vignetting glow" or whatever you might call this, that would be my next step.

If you have the chance, try dithering when shooting . this may be performed by shooting software, hardware solutions like MGEN or ASIAIR or even manual.
it is an big help against hot/cold pixels, dark noise patterns etc.

clear skies

[arrowspace90]

Well, I am trying some new tactics, hopefully starting tonight.

I am going to greatly increase the integration time on Iris. I have already worked to understand and acquire flats that are by the book. I am also going to shoot darks and dark flats. ZWO says I don't need Bias frames with their camera.

My scope is very fast at F/2, but the dark nebulosity around Iris is a special challenge. I set up my tripod in the EXACT same spot each night, with precise leg placing circles drawn on my driveway. I will re image Iris on successive nights, weather permitting, until I have sufficient data to make this target look right and do it justice.

Perhaps I can even go back and lump in data that I have already taken, but not sure. I will use the same exposure time, gain, sensor temp, and even camera rotation each time until I get this like I want it, and like I have seen others get it. My scope is 8 inches, not 8 feet, but I will try to compensate with integration time and good calibration frames.

[hixx]

Hi arrowspace,
there is one more option for you to further reduce noise in your masterflat: in Astropixelprocessor, under 2) CALIBRATE / MasterFlat, in the "blur" dropdown, choose 2x 3x3 median or 3x 3x3 median. This will smoothen the Masterflat by taking the median of a 3x3 square, making the MasterFlat almost noise free hence ultimately reduce noise of the final stack. You can get away with such blur if you do not have sharp edged dust stains in your picture. That option has "no blur" as default, so you'll need to select it each time You create a MasterFlat. I just stumbled over it the other day and thought it might end your Iris nebula the edge.
clear skies!