Not getting the results I'd like with LHaRGB

[Mike in Rancho]

A 1.9 is in the works? So soon? Any bread crumbs to throw out this time, like the "deconvolution secret project" preview from the Elf thread?

I mean, other than das language support, that is.

Brendan I saw your post early on but actually had to drive in to the office today (goodness no!), so just got to it tonight.

I used the same tricks as before, with the vignetting preset plus a 95 on the aggressiveness. Between the AutoDev and Contrast I mashed down the shadows more than you did, but that's what I tend to do anyway. Depends on whether there is legit galactic arm detail to be reasonably pulled out of the background, I suppose, and I tend to err on the side of knocking things back.

I also nearly forgot about the Ha, so had to backtrack. As in the other dataset, you did get a bit of the Ha regions just in your non-Ha data anyway, and I just lightly accented things a bit more. I actually pulled back some controls and the detail size too, since otherwise the pink was altering some of the blue-ish inner lanes too much, IMO. Anyway, I am unsure of the true Ha power here, and I didn't want to overwhelm. Undo buffer to revert the stars to their pink-free origins again also.

Quite a bit of integration you picked up on the Ha filter too, relatively. I wonder if some galactic knot accent warrants that much? Maybe so, and it will be less noisy too.

I may have ended up a wee bit purple here? I don't think I altered anything in the bias after the star sampling, perhaps I should have increased the green bias slightly, and capped it.

Still...your "problems" are good to have....I would be ecstatic to be able to produce data like yours here. Very fun to play with.

BrendanC M101 LRGB-Ha ST8 1A.jpg (457.48 KiB) Viewed 4570 times

As to your vignetting, I wonder if you are also dealing with Newt issues as I am. Very much a work in progress on my part, trying to figure out where all my "faux IFN" is coming from, lights, flats, or both. But, I figure you follow the same routine for taking flats on each filter. The Ha appears to be vignetting free, or essentially so. The RGB filters have a bit, maybe red the most. And the Lum is the biggest culprit. So...possible that your narrow filter is cleaning up errant light, but the broadband ones do not? Not sure if that's a clue or not.

[BrendanC]

Thank you @admin and @Mike in Rancho - my two favourite people in the universe (this one, anyway...)

So Mike, again, thank you for taking the time to look at the data. I know you're finding it useful too, but it's still brilliant that you're around as a sanity check because I sometimes genuinely wonder whether I'm going insane.

I agree with you about the broadband being much more susceptible to whatever's going on, especially the luminance.

Ivo, thank you for really spending time thinking about what could be going on here. My method for flats is the same as the DSLR - an LED light panel, with flat durations calculated by APT to around 23,000 ADUs. I've consulted other people with similar setups who use this successfully (with the exat same LED light panel as it happens).

Now, thinking back, because the flats take so much longer (four times longer, precisely), I might have been lazy and left the panel sitting on the OTA while I wandered off. With the DSLR, I made a point of moving it slightly between each frame, to average out any inconsistencies in the panel. So, I'm wondering whether this is 'the problem' - luminance being the most transparent of all the filters, and really showing up panel anomalies. It's definitely worth a shot, and I'll be trying it tonight.

At the end of the day though, I don't quite understand why this (no calibration files, just AutoDev applied) ...

Honeyview_No calibs.jpg (39.94 KiB) Viewed 4564 times

... plus this (calibrated with darks - not much difference because no much amp glow, it's much more noticeable on longer exposures) ...

Honeyview_Darks.jpg (40.07 KiB) Viewed 4564 times

... and this (the master flat)...

Honeyview_Flats.jpg (17.67 KiB) Viewed 4564 times

... should give me (yikes, only three attachments allowed, hang on...)

[BrendanC]

... this:

Honeyview_M101-Luminance.jpg (43.56 KiB) Viewed 4564 times

Still, if this is really a flats issue then that's probably the best problem to have, right? Because I can do something about it quickly and fairly easily.

So tonight I'll be sticking around to rotate the panel between flats. I'll also have a think about maybe making them less bright, so that I don't have to hang around so long. As well as being lazy I'm basically impatient and probably not that bright, which is not a combination best suited to astrophotography, but I like pretty pictures too.

Again, sincerely, thanks for your help.

Cheers,
Brendan

[BrendanC]

Oh, one more thing - @Mike in Rancho, would you mind sharing your log file again please, for M101? It just seems that you consistently get better results from my data, so I want to learn how you do it!

[BrendanC]

... and now, I think I might be able to say, I AM getting the results I'd like with LHaRGB!

So last night I did an experiment: stacked my luminance lights in batches of 20, so that I could then compare them and see what was going on. While the stacks moved about a bit, because of dithering and a different registration frame each time, I could clearly see a constant underlying gradient which supports Ivo's theory.

Then last night I decided to stick around and move the light panel randomly between each flat frame. And the result was much, much easier to handle with better results. A much lighter touch, not nearly as strong a Wipe needed.

Honeyview_M101 (2).jpg (58.93 KiB) Viewed 4545 times

This is GREAT.

So, now I have a name to my pain. I don't want to have to hang around for 20+ minutes doing dodgy flats with cheapo LED screen so I've got a Lacerta flat field box on order. Hopefully this will mean nice consistent flats without the pain.

Thanks again all, this has been a major headache for me and hopefully this will be a big step forward.

Cheers, Brendan

[Mike in Rancho]

Interesting. Hopefully that holds up and continues for you! When I used my laptop screen for flats, I didn't move it but I did have several sheets of paper to dim things down, and I would rotate the paper every 5 or 10 flats subs so the grain would be going a different way.

I think things are better with the flat panel and diffusion film I now use, but it's still a work in progress especially with a Newt.

Here's that log. I really should edit it some to make it more understandable. Sometimes logs need a bit of deciphering that only comes with getting used to reading them, and even then, there are a few items that don't show up anyway. Because of that you may not get a perfect match, but read ahead before applying stuff and use the latest-in-time, as there were some undo's as I experimented.

https://drive.google.com/file/d/1Q_Wjq2 ... sp=sharing

[BrendanC]

Great, thanks again Mike.

I've only ever used a Newt, and so a lot of this is familiar to me. It's just been these darned gradients that bothered me, and the flats, as per Ivo's suggestion, are at least part of that.

I'm also looking into the possibility of a light leak in the focuser tube, but that's for another day.

Thanks again for all your help.