PixInsight Forum (historical)
PixInsight => Tutorials and Processing Examples => Topic started by: Catanonia on 2011 April 28 19:23:40
-
Because Harry is too modest to post this, I will :)
Harry has created an absolutely stunning demonstration video showing the whole process with perfect explanations, and of course made all the better for my rubbish data supplied to him of M106
The video goes through
Star Alignment
Screen Transfers
Cropping
Background Neutralisation
DBE
Colour Calibration
Histogram stretches
Satutration
Noise reduction
and loads of other tasks and gives you a pefect feel for the work flow that you might take with PI.
This is by far the best video he or anyone has made on PI as it brings everything together into a nice smooth workflow. A bit like know how to get from A to B and C to D but still not knowing what the whole town map looks like. This video delivers that. I fact I would go as far as to say that anyone starting out with PI should watch this video first and use it as a reference chapter diving into his other videos explaining the processes more in depth as you get to that section.
So I would like to publically give a huge thanks to Harry for taking time out and creating this awesome video and I urge everyone to link it in thier favorites for reference if they are starting out with PI.
So here it is
http://www.harrysastroshed.com/CATM106.HTML
Enjoy and thanks Harry, I have learnt so much in that 15 minutes and with my data too :)
-
Thanks Harry!
-
More great work from the educational master.
We salute you Sir Harry.... 8)
-
Thanks Harry, another very useful one 8) .
Remarks:
- if the goal of the Background Neutralization step is to get a well balanced STFed image, it is also possible to do this in STF by uncoupling the channels (icon on the top left), and press "A" again. After this you can couple the channels again.
- I would usually do Background Neutralization after DBE. The gradient may also distort the color balance of your background.
- Usually, after DBE, Background Neutralization is not necessary. DBE does this as a side effect.
Georg
-
Hi George
As you say there is more than one way to skin the cat :D
I tend to use Bn before DBE only when I get a badly unbalanced colour image as if I leave it all to DBE it sometimes takes more than one application of DBE , of course there is nothing wrong with doing 2 or even 3 applications of DBE if needed . also if I can DBE right I rarely use BN after DBE , but saying that if things start to drift you can use BN at any time during your process
Regards Harry
-
I concur, Georg.
I learned PI from Harry's videos initially, and the BackgroundNeut. before DBE made sense to me. First you get rid of the 'base' color, then you deal with the more complex gradients with DBE.
However, when I saw Vicent's workshop at NEAIC, I was surprised to see that he does it the other way -- DBE first, then BackgroundNeutralization if necessary. It still doesn't make a ton of logical sense to me, but I'm sure there's a good explanation.
-
Hi
What is most important is that you have a neut background before colour calibration or it does not work correctly :D
Use BN as when you feel the need it will not hurt anything :D
Harry
-
Near as I can tell DBE has BN built in and enabled by default. I've run BN after DBE but it doesn't do anything. So I generally do DBE followed by CC.
Great video Harry! A super example that shows PI isn't difficult for even sophisticated processing (wavelets).
-
i've always done DBE as the very first step of processing... i guess you can do BN first if you carefully pick the reference preview but in my case the entire frame is usually covered with an LP gradient. and like sander since DBE neutralizes the background (if "Normalize" is not checked), i usually never do BN separately. in fact i think when you do color calibration it always analyzes the background, no? so it gets done post-DBE whether you like it or not if you do a color calibration...
-
As you say there is more than one way to skin the cat :D
Hey steady now :D:D:D
Can't wait for part II of the video :)
Nudge Nudge harry :)
-
Hi Mr cat
Spotted my little joke then >:D
I will do the Lum vid on sunday :D
Harry
-
Sorry but to me, applying BN before DBE doesn't make any sense.
Cheers!
-
BTW congratulations for the video! I'm certain many people would find it extremely valuable - despite that discrepancy we're having 8)
-
Another great video tutorial, Harry. Thanks for taking the time to do this one and, indeed, all of them.
Have to agree with Rogelio that BN before DBE doesn't make sense and , as others have stated doesn't seem to be necessary after DBE is applied.
My RGB master images usually start out with significant background imbalances so I can't see the gradients that need to be removed with DBE. To deal with this I clone the image and apply BN to this clone. I now can see the gradients so I use this image to create, test and check the background DBE model. When this look correct I create an instance of these DBE settings and apply it to the original image so DBE is being applied before BN. The resulting image has a corrected background so BN isn't needed.
Steve
-
Hi Steve
I think you are way past my little old video's 8)
Yup another way to skin , oh sorry should not say that >:D
I have tried many ways with the same result , and I appear not to be damaging the data doing it my way ??? so people you pays your money
and you take your choice :-*
regards Harry
-
Hi Harry,
I find your tutorials to be very well done and informative. I never fail to learn something so I must not be "past your little old video's" yet. :D
BTW, I didn't mean to suggest that doing BN before DBE was incorrect. Just wanted to detail how I approach that step in processing. I have also done it both ways and can't clearly see a difference. Nonetheless, on the chance that faint background "real" data might be altered if BN is done before removing gradients with DBE, I skip the BN step.
Steve
-
This is for jlake...
You deal with gradients first (DBE) because they're there polluting the real data.
Once the gradients have been dealt with, the (background) data is much more reliable for a BN operation, or anything else for that matter.
-
Thanks, RBA, that does make sense. In my mind, the BN was always getting rid of the massive color wash (often red or yellow in my case), then letting DBE get rid of the more complex gradient from the chip and cooling deficiencies.
-
In my mind, the BN was always getting rid of the massive color wash (often red or yellow in my case), then letting DBE get rid of the more complex gradient from the chip and cooling deficiencies.
That's the approach I had, but the other skinning method ;D sounds quite convincing as well.
Tasos
-
Actually, if you asked me, I'd say that the best way to work with DBE is by working each channel separately, for a bunch of reasons I've said several times in this forum already.
First, each channel is going to have different gradients, even when using an OSC camera, and each may respond better to different DBE applications (samples placed at different places, etc).
While DBE applied to a RGB image processes each channel separately, it builds its background model responding to the very same set of samples and parameters.
Sometimes this may not matter much, depending on your goals and your data, but other times it may. Foe me it matters every time.
Also, I like to "read" the background model, as it often helps me "guess" whether I'm removing JUST the gradient or something else/less, then go back to rework the samples and/or parameters if I don't like what I see. RGB composite background models often look rather psychedelic and are very hard to "read".
Someone may say that this could be solved by breaking the RGB background model into its RGB components, but then, when you go back to make adjustments, you're fighting three different channels all at once. Personally, that sounds like a much more difficult task than just applying DBE to each channel separately!!
Now, when you apply DBE separately on each channel - and that's how I do it all the time - the workflow seems quite simple and intuitive:
A slight crop on each master comes first, then DBE on each channel (where I spend as much time as necessary), then star alignment, some more cropping, RGB integration, and then finally BN followed by CC.
My opinion. Skin your goats as you like :angel:
-
I told everyone said the same thing do DBE in each channel seperately along time ago.
In theory, it should be OK on RGB.
It is much better with each filter in practice. The tolerances don't have to so wide for the good filters too.
Max
-
RBA,
I know you've mentioned doing DBE on each channel separately in the past but this more detailed explanation is very useful and makes a lot of sense. Going to give it a try.
Steve
-
I'd like to hear Juan's take on this. My sense is that as long as all samples are on the background (ie. have no stars, nebulosity etc) it doesn't help to do each channel separately. The reason is that DBE already processes each channel separately. It could be that there are disadvantages to applying the same tolerance and other parameters on all 3 planes. Clearly if you process each channel separately you can tweak the settings for each channel. If your samples are well placed you can re-use the sample set assuming registered layers.
-
Hi
I have tried separate channels and combined and can not find any difference so i stick to the simple way >:D
Harry