drizzle makes my stars worse

[young talon]

so I have the same data set twice, one with 2x drizzle and the other one without.
When I applied blurx to both and run starnet the version without blurx didn't have any stars in it but the data set with the drizzle showed alot of stars.

why is this and can I fix this?

[autumn elbow]

this is because 2x drizzle effectively upsamples your image, which turns already bad quality stars into little discs, which arent registered as stars anymore. this isnt drizzle's issue, its your optics and pixel scale.

i would recommend not using drizzle 2x, and rather use drizzle 1x

[young talon]

is 1x drizzle possible in dss? I have never seen it there before

[autumn elbow]

oh probably not, dss drizzle is not real drizzle

can you afford to get pixinsight?

[autumn elbow]

yeah i think wbpp is standard

as long as its the 1.8.9-2 or higher u can drizzle i believe

[young talon]

eh im on 1.8

Im thinking about buying pix but its just so dammn expensive

360 bucks

[iron vapor]

Honestly I wouldn't, Siril really is becoming better and better. On dev versions is already available:
spectrophotometric calibartion
better plate solve
curves
true drizzle integration (from 1x to 2x, with decimal values between them)
astrometric registration for widefield
graxpert integration

And it's still faster than pix

All that is missing now is good decon (vikram's coming) and a few minor features

[abstract mist]

You won’t regret it. Even if it is expensive. 👍🏽

[abstract mist]

Can you describe in detail why you call SIRIL drizzle “true drizzle?”

[iron vapor]

Siril used to have a "drizzle" option that was actually just upscaling the image 2x, with bilinear algo, which is not really drizzling. They now have the "authentic" drizzle (like in Pix) that can actually improve resolution and make the noise profile much nicer (especially for osc data, with bayer integration), so I call it true drizzle

[abstract mist]

So SIRIL implementation stops at 2x?

[iron vapor]

For now yes but it's still in development and it shouldn't be hard to make the limit higher

[abstract mist]

Hmm interesting. I stopped using SIRIL two years ago. Haven’t kept on what they’re doing

[iron vapor]

Tho you rarely need beyond 2x drizzling... barely any resolution improvement at the cost of a lot more noise

[iron vapor]

They're doing a lot of great work! latest dev version really is amazing

[abstract mist]

Ya I mean there is alot for why I won’t use it. But I get it’s a decent free option.

[iron vapor]

What are this reasons exactly?

Honestly, except for the absence of support for deepsnr and blurx (and masks but they're coming in the future too) there's no best choice between siril and pix
Siril is even better than pix for some applications (lucky imaging ie) and it's free

Imo it's not a decent free option, it's a serious competitor to pix price not considered

[silk drum]

and opensource

[silk drum]

no, 3x

[abstract mist]

SIRIL so far just doesn’t have the depth of pix. The fact that I can work with pixel math on multiple open images at the same time, and save my workstation while I go is huge. That’s just two reasons so far. But like I said I haven’t used it for two years, I’m going off of what I hear about it. And what friends tell me who use it.

[iron vapor]

Ah I see, from my experience with both it's just a matter of different workflows tho
I can have different instances of siril opened at the same time to make live changes to pixelmath for example. I personally prefer working with siril by far, but the work flow is really different so I guess it depends on the user's preferences

Pix is still better at post processing that's for sure, but siril is the king of pre processing and is getting better and better at post process :)

The price of pix is so freaking high, to me it's a no brainer. I could buy a narrowband filter set with that money! But to each their own I guess

[abstract mist]

Um well technically you’re correct, but I would argue you wouldn’t want that NB set.

[silk drum]

Siril is growing fast 😉

[iron vapor]

Why? Some of Baader's sho are available for almost half the price of pix and have good quality (while they're not the best obvisouly)

[abstract mist]

Typically with filters you get what you pay for. Antlias, and chromas aren’t gonna be near the pix price point. Tbf you can get cheaper filters and be totally fine it’s more luck of the draw at the point. Like quality control isn’t as high. Currently I’m imaging with 7nm ZWO filters.

[abstract mist]

After thinking about it, I’d say it like this: the flexibility and depth that pix has, has informed my acquisition. It’s made me a better astrophotographer because of the different stacking algos the different drizzle kernels, the pixel math implementation etc having the flexibility to run different tests and play with stuff is huge for just starting out. obvs, I think pix should be thought of necessary gear. Obvs I think pix is the end all software but I understand the other side of the aisle to some degree. I totally get not everyone gets into the hobby with thousands of dollars ready to spend. And there is something to be said for open source stuff, I like the idea behind it. And having software to chase like Pix makes SIRIL better for it. I think that SIRIL took awhile to implement SPCC. And their Starnet2 implementation was behind pix as well as GHS stretching. So SIRIL is better because pix is around. Pix isn’t better because SIRIL is around if that makes sense.

Now think of the new MultiScale Gradient removal process. It’s impossible to do with SIRIL. Of course not every dataset needs it, but when you do need it it’s nice to be able to do.

SIRIL will keep getting better but always slower than pix.

[iron vapor]

I think that sums up this conversation pretty well 🙂
Pix stays better than Siril (for now, since the dev of siril in the last few years has been much quicker than the slow evolution of Pix, Pix still staying ahead)
Pix is expensive, so whether it's a good thing to actually buy it depends on one's budget and needs in hardware
Which is the best, considering only the tools that are available for both, depends on the user's usual workflow. You seem to prefer pix's workflow, I prefer Siril's (I've tested both), to each their own!

Just a little note as I'm not sure you're aware of it:
Siril offers lots of drizzling kernels, registration algos (including comet registration, 2 pass alignement, features alignement, astrometric registration...), stacking algos (no norm, norm, fast norm, sigma rejection, student rejection, pix's rejection algo who's name I forgot...) and a really good pixel math integration with almost all of Pix's functions+variables, really good rgb composition that features narrowband integration, on the fly registration, cc and linear match

I'll also add that siril is faster to much fester at pre processing, and in my experience the overall post processing workflow is also faster and less teadious but that last part is personal

[proper meteor]

you actually can go further with scripts

[silk drum]

SPCC was implemented in less than 3 months. The point is we do not release stable version very often.

[abstract mist]

Fair enough. It makes sense. With recent conversations I feel like it would be best to check out SIRIL again to form better current opinions.

[proper meteor]

if you say siril 3 times

mr siril appears

its just like beetlejuice

[abstract mist]

That’s sounding like some nice stuff. I think for me where that would disconnect is the inability to work on multiple images/copy of images in one instance of pix. Like I could make an SHO and HSO image and do some pixel math to combine parts of those into a new color palette in the same instance. But sounds like SIRIL is making some nice headway. I have been intrigued by something I’ve heard called “sirilpy” (I’m sorry if that’s wrong) but I think it incorporates the python library used by some professionals for pre-processing stuff.

[nimble pollen]

Masking

[iron vapor]

It's already planned

[iron vapor]

Altough I think pix users can't use siril easily the same way siril users struggle with pix, because of the work flow difference I've talked to you about

[iron vapor]

Fair enough, that probably won't happen in siril anytime soon but I personally am happy with siril +photoshop 😆

Honestly I just hope siril will become a serious alternative to pix on all levels, pix is just too expensive and there's no middle ground...

[abstract mist]

Pix is cheap relative to Astro hobby as a whole. That’s why I think it should be considered as “gear” then it’s not so expensive. Also, I downloaded the latest stable version and couldnt find the SPCC. Apparently it’s only in the dev version? Eh, look I’m not here to try and convert you. I appreciate the convo! Clear Skies!

[iron vapor]

Yep no worries, me neither ;D It's always nice to talk when we have different opinions

And yeah all the latest additions are on the dev builds

[iron vapor]

I agree that pix is cheap relative to the rest (albeit lots of people struggle to find the money for a 1000€ starter setup so pix is just out of the equation) but it's more of a moral issue: I don't think anything justifies such a high price for a software that will be used by amateurs. Astro gear's price is justified because there's physical costs but software are all virtual so the price can't be as high... especially for an amateur use! If it was 100€ ok, 200€ perhaps (and yet I think that would only be justified for multiuse softwares) but 360€...

Some people may agree with this price, I do not but that's only my opinion

[abstract mist]

One rebuttal to that is that Pix is used by Joseph Depasquale to produce the JWST images. So at least as can be reasonably inferred pix is being used by "professionals"

But I understand the point. 👍

[iron vapor]

That's absolutely true! What I think would be more appropriate is a business model with different licenses, like lots of other apps btw:
One license for non commercial use at less than 200€ and for commercial use (wether it be nasa scientists or people who sell their astrophotos) a more expensive licence

[abstract mist]

ya know, thats not a bad idea

[iron vapor]

It feels good to have thoughtful debates... When everyone on the internet is just insulting each other for futile maters

[abstract mist]

but I think at some point the price reflects what comes with Pix and its depth. So that professionals can use it like amateurs.

[abstract mist]

[iron vapor]

I'm gonna have to leave you tho, first reunion of my college's astro club

[iron vapor]

Fair enough

[abstract mist]

Ah, have a good time! CS!

[iron vapor]

At least there's a free and open source alternative that, while not as good, is already a great option !

Guess what: there's not much clouds so we'll make a few observations !