Can GraXpert fix even bad calibration? Testing the new (and free!) Astro Wunderkind!

[@CuivTheLazyGeek]

Support me on Patreon!: www.patreon.com/cuivlazygeek If you're planning on buying equipment (or anything from Amazon), it will help me if you first click the affiliate links in the description! GraXpert download page: www.graxpert.com/ What are your thoughts on this? Of note: I did try many methods in PI to remove the center bright spot (including tighter smoothing), but nothing was as good as GraXpert!

[@astrophotocologne]

Hey Cuiv, thank you for the video! We are honored that you made a video about GraXpert. Thanks also for the title "Wunderkind" very nice! I thought it was good that you emphasized the dynamic approach, which I missed a bit in others videos. Though ou have chosen a very difficult case where probably every gradient software reaches its limits. By the way, the default method is RBF and not Kriging. RBF stands for Radial Basis Function, a special kind of spline algorithm. I don't usually like it when a KZbinr shares their video under another KZbinr's video, but I'm promoting The Astroimaging Channel here, where I recently featured GraXpert extensively. There I also explain my philosophy that you have to have a strategy to get the best possible results. From my point of view, it is also important that one understands the basics of how such programs work. After that, it almost doesn't matter whether you work with DBE or GraXpert. The advantage of GraXpert is perhaps that it is a bit easier to use, and the algorithms may also be a bit more useful. Anyhow, thanks again! We hope you will use GraXpert more often now.

[@CuivTheLazyGeek]

Oh wow, hello there! I didn't even realize GraXpert was made by a content creator (and team) :). Thank you for this feedback and keep it up!!

[@astrophotocologne]

​@@CuivTheLazyGeekI had the idea, the others did the hard coding work :-)

[@TheOriginalTommo]

Finally a Cuiv video that won't make me poorer 🥳🥳🥳 Hadn't heard of this tool and I struggle with DBE so thank you very much 👍👍

[@CuivTheLazyGeek]

Hehehe it's always good to do videos about free stuff :)

[@anata5127]

This is story of my past. Fighting bad calibrations with ASI294MM and asi294mc. Always thinking “I am almost there”. Never had consistency. Lost tons of data. Finally, I have sold all ZWO items, and got QHY294M (yes, that infamous sensor) and Qhy filter wheel. All problems are gone. Now, I don’t need to do triple spinning for data processing.

[@CuivTheLazyGeek]

Wow, that's very interesting to have such a stark difference between ZWO and QHY versions of the same sensor!!

[@anata5127]

@@CuivTheLazyGeek QHY produces research grade cameras. I know people who use them on multi-thousand dollar microscopes. Such companies are always superior. By the way, ampglow with QHY 294m is noticeable only after 5 min. ZWO version was giving ampglow in 30 seconds.

[@ssbhide123]

How about Siril vs GraXpert vs PixInSight vs GradientXterminator ?

[@hottokatrazi]

Siril uses the Graxpert Code

[@sampedraglio]

​@@hottokatraziso, if I use RBF I'll obtain the same result I'll obtain by using RBF in GraXpert?

[@AndreH3d]

Siril does a great job indeed, but I don’t use others so I can’t compare

[@bruins4life]

@@hottokatraziit does, but not fully identical. GraXpert does a better job in dithering the squares used for estimating the background. As a result the output is usually superior in GraXpert

[@ssrattus]

Thanks Cuiv and GraXpert!

[@CuivTheLazyGeek]

Thank you!

[@nikaxstrophotography]

Great Video Cuiv, I've been playing around, there is also the advanced tab on the right which changes the box sizes, A tip: Use RBF and less smoothing for something this complex and you will have a very acceptable result GraXpert will become part of my workflow

[@CuivTheLazyGeek]

Thanks for the tip! Yep, seems I can play with a lot of settings :)

[@deep_space_dave]

Hey Cuiv! I really could have used this when I created my last image as my flats were not exposed long enough but of course you don't know until you integrate! Thanks for making this video, very helpful! Clear Skies!

[@CuivTheLazyGeek]

My pleasure Dave!

[@Abaunzamarco]

Hello Cuiv another great video. From what I’ve heard, RFB works best with just a few sample points, I start with this, and then add more sample points as needed. In Siril in some extreme cases, I use a very soft back ground extraction on each frame before stacking, and then use RFB. Hope this may help you. Regards and clear skies.

[@CuivTheLazyGeek]

Thanks for the tip!!

[@Abaunzamarco]

@@CuivTheLazyGeek one other thing I forgot to mention. I had a similar issue with what I thought it was bad flats (too bright) and in my desperation, I tried other software to do the calibration and stacking…and to my surprise, in Astap the calibrated frames and the stack came out perfectly fine…just don’t ask me why. Perhaps it’s worth a try 🤷🏻‍♂️

[@ASC-Explore]

I agree with Adam block on gradient removal, sometimes you have to sacrifice some nebulosity in low snr areas just so you can get thw gradients removed

[@CuivTheLazyGeek]

Yes indeed - and this is what happened here. It's quite interesting!

[@christianvulpescu1398]

This is a good quote of Adam Block. He says the low SN areas do not earn to be showed!😂 I keep that in mind, too.

[@Zealor365]

This was good. You did a great job of explaining the different methods.

[@Boekoe12]

Interesting! About 99% of my final data looks like this. Still figuring out why. But looks quite promising 😊

[@CuivTheLazyGeek]

Thank you!

[@HeavenlyBackyardAstronomy]

Thanks, Guiv. I will download it and try it out.

[@CuivTheLazyGeek]

Hope you enjoy it!

[@wesmagyar]

So usually I use starx, then save a starless copy as xisf. Then edit it in graxpert then import it back into pixinsight. This usually works

[@CuivTheLazyGeek]

I'll try that as well!

[@rtpman1953]

Very interesting approach. I've been using GraXpert for 100% of my images since I first saw it announced. However, I've usually thrown in the towel if my flats were messed up. I have played with it when I didn't have any flats and can usually get acceptable images. I may be lazier than you :-) which is why I only use DBE on really weird/complex images. Great video and concept.

[@CuivTheLazyGeek]

Thanks so much! GraXpert truly is impressive

[@astroshooter1960]

Another awesome tool for astrophotography..... great video Cuiv...👍

[@CuivTheLazyGeek]

It sure is! Thanks!

[@woody5109]

Great video Cuiv, definitely going to give it a try. You had great success in only minutes, imagine after a few days, cheers.

[@CuivTheLazyGeek]

Exactly! Plus it's free so no risk involved. Have fun!

[@drtools9226]

Nice you test Graxpert. I use it about one year and it is great. It is a community Product. Which a german youtuber pushed (astrophotocologne)

[@CuivTheLazyGeek]

Thanks for the info!

[@grrrsoft]

Hey Cuiv. Nice video as always :) I tried GraXpert a while ago, but could not get the hang of it to produce reasonable results. For me, the tools in APP seem to work best. Probably because I did not want to spend a lot of time tinkering with new tools... Maybe I'll give it another try with the things I learned in this video. Anyways, it is fun to see, that you basically took exactly the same picture I did like 4 weeks ago :) "Same" camera (QHY268c), ALP-T and framing. Just from a Bortle 4 zone instead of Tokyo, so I got away using no calibration frames at all... Seeing this, I now am even more thankful, that light pollution here is not yet as extreme. Which will probably change in the next 5-10 years I'm afraid.

[@CuivTheLazyGeek]

That's another vote for APP - I really need to try it out.. also nice coincidence haha :)

[@ciskje71]

Try also Siril with background extraction with 0.8 factor instead of 2.0

[@stevenkleinberg6805]

What about going to the background view, increasing the saturation to the max and then deselecting the Sample points on the nebula regions?

[@thevelourfog5422]

Would it be possible to separate the channels of the stacked image and run each one individually through it? I'm thinking that you could be much more specific in targeting the gradient by being more aggressive in the channels that you wouldnt expect to have nebulosity ( for example, more aggressive in blue and green channels in areas where you're expecting Ha) and therefore more control over any signal being lost. I'm still fairly new to processing so not sure if that is a possibility or not 😅

[@CuivTheLazyGeek]

I guess that could indeed be possible, something to ponder

[@samwarfelphotos]

How does this compare to the background extraction in Siril?

[@astrodug]

I've been using GraXpert instead of DBE for weeks now. It's a regular part of my workflow now.

[@CuivTheLazyGeek]

Seems many have followed the same path as you have!

[@garycampbell9709]

I will keep this in mind, great video as always ❤

[@CuivTheLazyGeek]

Yay, thank you!

[@davidwillis5396]

Hi Cuiv how are you. David from Australia here. I have a question. Do you know where I can obtain a 12 volt multie power supply for my scope & cameras, mount etc. I have a 10 inch Newtonian with a NEQ6 PRO MOUNT. Any help would be greatly appreciated. Thank you

[@DSOImager]

Do you get that bright center spot with all your pictures regardless of filters/cameras/scopes? I had a similar bright center with one of my rigs.. DBE/ABE struggled to deal with it. I almost installed graxpert when I realized I was only getting this bright center with 1 filter. It was my OSC setup, and so I used this particular lum filter almost exclusively. A cheap CLS filter I tested did not have this bright spot. I ended up ordering another lum fitler (different brand) and wow.. the bright center was gone! I put up with this horrible center "gradient" for months thinking it was a light pollution issue.. or just bad flats. Turned out to be a bad filter!

[@CuivTheLazyGeek]

Very interesting! No that center spot is pretty much unique to Hyperstar - but it usually isn't visible when I take good flats

[@DSOImager]

@@CuivTheLazyGeek Interesting! Good that flats are resolving it. :)

[@Sebastian-zf9ok]

You all should check out the latest beta including AI support for gradient removal. It works extremely precise even it's still beta.

[@bobsterndlschauer5923]

Hello Cuiv another great video.

[@CuivTheLazyGeek]

Glad you enjoyed!

[@timothysheridan5602]

Another gr8 video Cuiv Sorry...I have an unrelated question...I will be installing n.i.n.a but I am running windows 8.1 and I am not sure if it will run on my system. What version would you recommend?

[@CuivTheLazyGeek]

I think the NINA requirements are Windows 10 and later...

[@timothysheridan5602]

Thanks. I will have to update my pc

[@nickreecy4229]

I tried graxpert several months ago and i havent used abe or dbe since.

[@CuivTheLazyGeek]

That's impressive! So many seem to love GraXpert!

[@adityasoni6185]

RBF is a great algorithm as i have experienced wirh Siril

[@CuivTheLazyGeek]

RBF should get implemented in PixInsight!!

[@dumpydalekobservatory]

Great video Cuiv I tried Graxpert & it froze my pc up so for now I've stuck with SIRIL which also has the option to use RBF for the background (Whatever that is?) I might try Graxpert again though seems a pretty powerful program.

[@CuivTheLazyGeek]

I need to look up what RBF actually is :-)

[@dumpydalekobservatory]

I think the force is strong with RBF lol

[@CuivTheLazyGeek]

@@dumpydalekobservatory definitely lol

[@joshuamuller9705]

I used it since the first version and since GraXpert V1.0 I most of the time don't even take flats anymore. Too much work, for a not very good result. Could of course also be a bad workflow in capturing flats on my side. But sometimes it takes a few passes to find the perfect settings for a certain picture.

[@CuivTheLazyGeek]

No flats + GraXpert sounds like an impressive approach - I'll still take flats for now though :p

[@BarryWilliams0]

I love GraXpert and don't use DBE any more. I find the interface mich simpler and the results are great. I dont have your crazy gradients to deal with though!

[@CuivTheLazyGeek]

Many people report the same thing: having stopped using DBE in favor of GraXpert - it would be awesome if it were included as a script in PixInsight

[@Astro_Shed]

Great video Cuiv, but personally I will stick with the PI tools, did not like your outcome from the free software…, and my own experience was not that good either….

[@CuivTheLazyGeek]

To each their own!

[@--Adrian--]

You should try making a Synthetic Flat. It worked better for me than DBE or GraXpert, when I had calibration issues

[@CuivTheLazyGeek]

Never had much luck making synthetic flats, always ended up including part of the object!

[@zelodec]

Synthetic flats only work for objects like a small galaxy in the middle with nothing around it. Will never work on nebulosity.

[@Alucard400]

Nice studio Ghibli shirt

[@CuivTheLazyGeek]

Hahaha thank you!!

[@Alucard400]

@@CuivTheLazyGeek I am looking forward to seeing The Boy and the Heron

[@zhouzun]

good for lrgb

[@TheCosmicGuy0111]

Fancy

[@CuivTheLazyGeek]

Indeed!

[@سبحانخالقالكون-ش3ز]

رائع

[@CuivTheLazyGeek]

For sure

[@skywatcherca]

Mr. Cuiv - as always, great video. Sir, for us who are "newbies' and not extremely experienced, what you would suggest/advise for a basic software package to control our mounts. I am using an EQ8 (Skywatcher) and I'm not sure whether I should install Planetarium or another software package to control the mount. I don't know if a review of 'a bunch' of software packages would be of interest to your subscribers, but I would find this to be of value. Take care, my friend.

[@CuivTheLazyGeek]

For controlling the mount proper, it depends on the mount. But for the EQ8 I'd be using Green Swamp Server and NINA (I have some older videos on the channel about that!)

[@sharkking9679]

Shout out to @astrophotocologne to initiate the GraXPert Project

[@CuivTheLazyGeek]

Thanks @astrophotocologne!