Adam, this is great. After a year of using Pixinsight, I’m getting back to your courses, which I purchased at the same time I acquired Pixinsight. When I go through your courses now, as opposed to a year ago, I will be adding unfamiliar (i.e., new) information to familiar (i.e., what I now know about Pixinsight) rather than adding the unfamiliar to the unfamiliar as I was doing back when I got both. By the way, adding the unfamiliar to a background of the familiar is what has been said about what distinguishes good science fiction from bad science fiction.
RC is my vote for Time Person of the Year. I am sure his tools can create world peace.. But seriously, thanks for this great idea. I have some narrowband subs that RANSAC just hates.. I can't remember if I ditched them or not but it wouldn't be a waste of time looking for them. I assume BX can fix collimation issues.
This is why you should always run subframe selector on all your subs, would have been a quick catch. It’s good to know though if I do have some out of focus data that it’s salvageable.
Hi Adam great tutorial. I wonder if you could do a tutorial on multiscale gradient removal with Pix sometime. I've found a couple online but the results always seem unsatisfactory. The gradient is anihalated which is great but the object always loses a significant amount of fine detail when I follow the existing tutorials. Thanks
The problem I run into with WBPP is that some frames are not registered (in R, G, and B) but if I load them and the reference frame into StarAlignment they pass. And this is with as close as possible the same settings as in WBPP.
Contact me through my website. Include an example data set for me to download (would not need to be many files test your case. I rarely ask for this... because frankly I find this very unlikely. The same process is run in both cases. You can use the process container that WBPP generates to see for yourself. (and see any differences in parameters)
@@AdamBlock I have tried to send you a message in all 3 browsers (Chrome, FF, and Edge) but always get "Sorry, but we were not able to verify you as human" error.
@@LHM4000 This is because you are pasting text into the form. You will need to type in the form without pasting anything. When you paste, your computer is adding a character that the form doesn't understand.
@@AdamBlock No, that is not it. I was able to send a pasted text as long as I did NOT include the Dropbox link. It's probably the ampersand that it doesn't like. So you'll have received a message explaining everything but I need a way to send you the Dropbox link that contains the files.
But if bxt focuses the stars does it focus the nebula? I believe not. But correct me if untrue. And then your star aligning unfocused nebula in your data. I mean I think it's the call of the person processing. But out of focus data should be thrown out. Just my humble opinion
I was about to ask the same question- the uncorrected parts (everything except stars) are still poor data and will negatively influence the result. If this is true, then what benefit is this 'fix'?
@@craigwallace6601 @Neilssbohr No gentleman, you are mistaken. The correction that BXT does is applied to all signal (nebula as well). It just needs to be above the noise floor. BXT is a global operation in that sense. It is not just interacting with stars... however, it is easy to see the affect on stars much more clearly.
Adam, does using BXT correct only at this step prohibit or make less effective the use of BXT later in the workflow on an integrated image? I assume not, but thought I'd ask. Hello by the way! Best, Mark
Well...that is why I employed "correct only" ..which is now the standard step...and you can run BXT later if more deconvolution is desired. So the answer is no. See Russ's page on this.
Wow! It of seems that applying BXT to all the debayerd images might improve the end result. Is my intuition wrong? It would be nice if WBPP would provide a mechanism to handle this step. Thanks for another thought provoking video.
I suspect BXT on the single integrated result is better than Batched BXT. This is because each single frame is noisy compared to an integrated result. So... I think the answer is your intuition is leading you astray. Just consider... maybe there is some nebulosity that shows up in the final stacked image... but it is too noisy for BXT to do anything with in any individual frame.
How useful (do not mind the time to process this) is to run “just correct” in all subs before stacking? Can this bring benefits or is it will be corrected in the same amount in the final stacked image?
I simply couldn't do astrophotography without Russ Cromans tools. You Mr Block are a genius 👍
you sure do come up with innovative ways to exploit pix's tools Adam, bravo!
Awesome use of BlurXTerminator, normally I would probably just toss those subs with a focus miss but this a really nice fix to save data. Thanks Adam
Thanks for watching!
What a fantastic and clear explanation of a basic problem. ❤🎉
I'm really looking forward to your list of steps!
Adam, this is great. After a year of using Pixinsight, I’m getting back to your courses, which I purchased at the same time I acquired Pixinsight. When I go through your courses now, as opposed to a year ago, I will be adding unfamiliar (i.e., new) information to familiar (i.e., what I now know about Pixinsight) rather than adding the unfamiliar to the unfamiliar as I was doing back when I got both. By the way, adding the unfamiliar to a background of the familiar is what has been said about what distinguishes good science fiction from bad science fiction.
Nice! Keep at it...
RC is my vote for Time Person of the Year. I am sure his tools can create world peace.. But seriously, thanks for this great idea. I have some narrowband subs that RANSAC just hates.. I can't remember if I ditched them or not but it wouldn't be a waste of time looking for them. I assume BX can fix collimation issues.
Very nice solution 👏👏 Cheers Simon
Glad you liked it...
The new blurx saved an image of mine that couldn't use spcc properly due to low Snr. Amazing
Never happened to me something like that, but a very handy technique to solve the issue. Thanks for sharing, Adam
This is why you should always run subframe selector on all your subs, would have been a quick catch. It’s good to know though if I do have some out of focus data that it’s salvageable.
Hi Adam great tutorial. I wonder if you could do a tutorial on multiscale gradient removal with Pix sometime. I've found a couple online but the results always seem unsatisfactory. The gradient is anihalated which is great but the object always loses a significant amount of fine detail when I follow the existing tutorials. Thanks
Are you a member of my site?
Just amazing - well done!
Thanks for watching...
🎉🎉what for a great processing Idea
Thanks
Great video Adam
Thanks Terri!
The problem I run into with WBPP is that some frames are not registered (in R, G, and B) but if I load them and the reference frame into StarAlignment they pass. And this is with as close as possible the same settings as in WBPP.
Contact me through my website. Include an example data set for me to download (would not need to be many files test your case. I rarely ask for this... because frankly I find this very unlikely. The same process is run in both cases. You can use the process container that WBPP generates to see for yourself. (and see any differences in parameters)
@@AdamBlock I have tried to send you a message in all 3 browsers (Chrome, FF, and Edge) but always get "Sorry, but we were not able to verify you as human" error.
@@LHM4000 This is because you are pasting text into the form. You will need to type in the form without pasting anything. When you paste, your computer is adding a character that the form doesn't understand.
@@AdamBlock No, that is not it. I was able to send a pasted text as long as I did NOT include the Dropbox link. It's probably the ampersand that it doesn't like. So you'll have received a message explaining everything but I need a way to send you the Dropbox link that contains the files.
@AdamBlock I figured out what it was, the Once I removed that, the entire pasted text sent successfully.
But if bxt focuses the stars does it focus the nebula? I believe not. But correct me if untrue.
And then your star aligning unfocused nebula in your data. I mean I think it's the call of the person processing. But out of focus data should be thrown out. Just my humble opinion
I was about to ask the same question- the uncorrected parts (everything except stars) are still poor data and will negatively influence the result. If this is true, then what benefit is this 'fix'?
@@craigwallace6601 @Neilssbohr No gentleman, you are mistaken. The correction that BXT does is applied to all signal (nebula as well). It just needs to be above the noise floor. BXT is a global operation in that sense. It is not just interacting with stars... however, it is easy to see the affect on stars much more clearly.
Adam, does using BXT correct only at this step prohibit or make less effective the use of BXT later in the workflow on an integrated image? I assume not, but thought I'd ask.
Hello by the way!
Best,
Mark
Well...that is why I employed "correct only" ..which is now the standard step...and you can run BXT later if more deconvolution is desired. So the answer is no. See Russ's page on this.
Wow! It of seems that applying BXT to all the debayerd images might improve the end result. Is my intuition wrong? It would be nice if WBPP would provide a mechanism to handle this step. Thanks for another thought provoking video.
I suspect BXT on the single integrated result is better than Batched BXT. This is because each single frame is noisy compared to an integrated result. So... I think the answer is your intuition is leading you astray. Just consider... maybe there is some nebulosity that shows up in the final stacked image... but it is too noisy for BXT to do anything with in any individual frame.
Thanks Adam.
@@AdamBlock
How useful (do not mind the time to process this) is to run “just correct” in all subs before stacking? Can this bring benefits or is it will be corrected in the same amount in the final stacked image?
See my answer above... BXT on the final integrated image will always be better.
can you run BlurXterminator on raw data then run WPCC
This would not be a good idea in general... actually...let me just say no. Noise is an enemy of deconvolution. It is simply a non-starter for me.
Amazed that this wasn’t caught by the imager themself. Basic analysis would have saved them the headache