IC 342 Ha

[Rodd]

For those of you wondering if  it would be worth spending the time collecting Ha on the Hidden Galaxy, IMO....YES!  Definitely.  I did not know, and I heard this galaxy was faint and Ha was questionable, so I decided to collect the Ha for my HALRGB target and find out.  The Moon is up, so I can't collect any other data...why not see if there is enough Ha for an HaLRGB image?    You be the judge whether there is enough for you, but in my case, I definitely will be adding the Ha to the LRGB (if i ever collect it!, which I will heartily endeavor to do).  I will be adding to this stack though--This is 7.5 hours in 300 sec subs using a TOA 130 and ASI 1600 (3nm filter).  I would like to double this data or thereabouts.   I lightly processed the stack--not quite enough to make a stand alon mono--but I did remove what gradients there were and used a little noise suppression and very little star control. just to bring out the Ha regions a bit.  I'm encouraged by the presence of what appears to be a bar structure in the core.

Rodd

[Filroden]

I think that will add some great highlights to an RGB image and help lift the galaxy.

[Rodd]

3 minutes ago, Filroden said:

I think that will add some great highlights to an RGB image and help lift the galaxy.

I agree--its not the strongest signal by any stretch, but I think it will  certainly will add some color to the Ha regions.  A side benefit, at least for me, is it will be harder to overdo, since it is on the faint side.  So a subtle hand will be required....never a bad thing in AP.

Rodd

[Rodd]

Not sure you can see them, but there are bands in the image.  I checked and the bands ( a checkerboard pattern) and it in the raw subs or the calibrated subs. They are in many of the registered subs.   But not all. Why is this?  Very strange.

Rodd

[Rodd]

Please--I have a mystery...HELP!  Star alignment is creating checkerboard banding in my subs.  The banding is not present in raw or calibrated subs.  take a look  First image is calibrated sub--no pattern, second image is aligned sub--patter.  how can this be?  I have overstretched the subs so the pattern is easilly seen.  It does sshow up in the integrated stack as you can see above if you look closely.  This is only Ha, so I can get buy--but if this happens with LRGB, forget it.  Another intersting fact is I collected 31 subs of HFG1 on teh same night--this target is very close to IC 342 and these subs do not have the pattern.  What is going on?

Calibrated sub

Aligned sub

[Adam J]

So the last image is a single sub and not calibrated? What are you using to stack and align? 

[Rodd]

2 minutes ago, Adam J said:

So the last image is a single sub and not calibrated? What are you using to stack and align? 

Adam--the banded sub is an aligned sub--the unbanded sub is the same sub before alignment.  Both are calibrated obviously. I used another sub from the calibrated subs to align.  Both are calibrated.  One is before alignment one is after

[Adam J]

5 minutes ago, Rodd said:

Adam--the banded sub is an aligned sub--the unbanded sub is the same sub before alignment.  Both are calibrated obviously. I used another sub from the calibrated subs to align.  Both are calibrated.  One is before alignment one is after

What program are you using? APP? 

Normally I would say to take a look at the master flat. 

[Rodd]

Just now, Adam J said:

What program are you using? APP? 

Normally I would say to take a look at the master flat. 

Adam--its not in the calibrated from--it cant be the flat.  It happens after alignment.  It makes no sense for it to be the flat--if it was teh flat it would be in teh calibrated subs.  All alignment does is rotate the subs to match--after alignment an individual sub should look identical except for some edge artifacts

[vlaiv]

This is consequence of small angle rotation and noise being clearly visible.

Rotating image by small angle in presence of the noise produces this effect imprinted in the noise. It will not be imprinted in the signal in the image. You need hard stretch to show it.

It can be explained as aliasing of high frequency components or it can be explained as consequence of averaging noise pixel values in some regions and not in other regions - it is same thing. You end up with noise - less noise - noise - less noise pattern which shows itself as grid.

I created similar pattern in this thread:

It depends on resampling method used - bilinear resampling gives worst results.

[Rodd]

3 minutes ago, vlaiv said:

This is consequence of small angle rotation and noise being clearly visible.

Rotating image by small angle in presence of the noise produces this effect imprinted in the noise. It will not be imprinted in the signal in the image. You need hard stretch to show it.

It can be explained as aliasing of high frequency components or it can be explained as consequence of averaging noise pixel values in some regions and not in other regions - it is same thing. You end up with noise - less noise - noise - less noise pattern which shows itself as grid.

I created similar pattern in this thread:

It depends on resampling method used - bilinear resampling gives worst results.

I use PI and have NEVER zeen this.  It did not happen on subs of another target taken the same night close by this one.  I just dont see how this could happen all of a sudden.  Subs look teh same--noise levels the same as measured by SFS.  

[vlaiv]

1 minute ago, Rodd said:

I use PI and have NEVER zeen this.  It did not happen on subs of another target taken the same night close by this one.  I just dont see how this could happen all of a sudden.  Subs look teh same--noise levels the same as measured by SFS.  

This is just "redistribution" of the noise - signal will be the same, and noise in general will remain the same over the image - it will just change it's distribution. It depends on algorithm used to rotate - interpolation. Some interpolation techniques give better and some worse results with respect to this.

I'll do another example for you here and comparison of bilinear and more advanced interpolation.

Here is "base" sub - nothing but pure gaussian noise:

Here are two subs rotated by 2 degrees - one bilinear interpolation other cubic O-moms:

These are two rotated subs - no pattern is yet visible as we did not stretch the subs to show the pattern.

Now I stretched in particular way to emphasize this pattern - left one is bilinear interpolation - and it shows pattern clearly. Right one is cubic O-moms - pattern is there but to much lesser extent. Every algorithm will produce it to some level because you need to cut into higher frequencies when you work with limited sampling rate, but some algorithms handle this much better.

If you use Lanczos 3 resampling - it should have this effect down to minimum.

[Adam J]

My next guess was going to be that you are applying debrayering to a mono image. Is it rotated such that the grid is to the sensor if rotation had not taken place. 

[Rodd]

2 minutes ago, Adam J said:

My next guess was going to be that you are applying debrayering to a mono image. Is it rotated such that the grid is to the sensor if rotation had not taken place. 

Wouldnt even begin to know how to debayer--I have never done it.  No way this could be it

[Adam J]

1 minute ago, Rodd said:

Wouldnt even begin to know how to debayer--I have never done it.  No way this could be it

Look at every setting it's Always possible to accidently tick something. 

[Rodd]

Only place it could happen would be batch prepocessing.  I will recalibrate the subs and make sure.  But Vlaiv has another cuase that is different.  Wth a try

[PhotoGav]

Rodd, if you want to make the subs and calibration files available through Dropbox I would happily run it all through my stacking processes in APP and see what comes out. 

[Rodd]

2 minutes ago, PhotoGav said:

Rodd, if you want to make the subs and calibration files available through Dropbox I would happily run it all through my stacking processes in APP and see what comes out. 

Thanks Gaav--it would take way too long to upload 149 subs!  I appreciate it.  If Vlail is right and teh noise is just redistributed, I can just limit my stretch and it wont be visible.  I just find it very starnge I have never seen this in 5 years and thousands upon thousands of aligned subs.  

[Laurin Dave]

Rodd, a friend of mine had this happen the other day with Pixinsight, I referred him to Vliav’s response and he sorted it by changing the resampling back to to Mitchell Netrali (or something like that) Which he had previously used ..  he thought it was down to a recent Pi update ..

Dave

Edited November 9, 2019 by Laurin Dave

[Rodd]

1 minute ago, Laurin Dave said:

Rodd, a friend of mine had this happen the other day with Pixinsight, I referred him to Vliav’s response and he sorted it by changing the resampling back to to Mitchell Netrali (or something like that) Which he had previously used ..  he thought it was down to a recent Pi update 

Dave

Interesting--never used that algorythm.  I will see if that works.   I do not update things very often, so that probably is not it.  After Vlaiv indicated that its just noise redistribution and will not be seen if I do not over stretch, I feel better.  This is only Ha, so most of the banding would not be in image anyway--but if it happens with LRGB I will be more concerned.  Since it has never hppenedto me, I normally would not give it a second thought--but now I am concerned

Rodd

[ollypenrice]

This isn't my field but Vlaiv and Dave seem to have a handle on it. I do sometimes see atypical background artifacts in Ha data and put it down to the fact that I'm often stretching my Ha data (especially for galaxies) way, way harder than I would stretch anything else. The endgame of galaxy Ha is, after all, often little more than the addition strings of low resolution 'blobs' to spiral arms. In the way that I combine the Ha only the brightest signal will end up in the image anyway so low level artifacts will be excluded, as Rodd says early on.

I don't think we really quantify the severity of our stretches. I certainly don't do so. It might be interesting to know the extent to which a background sky value, for instance, has been stretched. Obviously it will be far, far higher in NB than broadband.

Olly

[JamesF]

The explanation for this effect is fascinating.  It would never have occurred to me.  Makes the image look as though it was taken through the net of a football goal.

James

[Rodd]

41 minutes ago, ollypenrice said:

This isn't my field but Vlaiv and Dave seem to have a handle on it. I do sometimes see atypical background artifacts in Ha data and put it down to the fact that I'm often stretching my Ha data (especially for galaxies) way, way harder than I would stretch anything else. The endgame of galaxy Ha is, after all, often little more than the addition strings of low resolution 'blobs' to spiral arms. In the way that I combine the Ha only the brightest signal will end up in the image anyway so low level artifacts will be excluded, as Rodd says early on.

I don't think we really quantify the severity of our stretches. I certainly don't do so. It might be interesting to know the extent to which a background sky value, for instance, has been stretched. Obviously it will be far, far higher in NB than broadband.

Olly

That's true--only the high signal areas will be added so it inst really a problem for Ha--but if it happens in LRGB that will ruin the image.  Since it has never happened to me before--I just don't know what to expect.  

Rodd

[Rodd]

Thanks everyone for being there.  Wish I could say as much for the Pixinsight team.  I guess I'll push on and see if it happens with the LRGB subs.

Rodd

[Laurin Dave]

Be interesting to see Rodd, maybe the sky background with lrgb will masks ..  my mates instance was with Astrodon Oiii and asi1600