Focusing of RGB with LRGB image

[Tommohawk]

I've just been rethinking slightly my imaging / processing techniques - mostly because I've just had a dabble at binning RGB  -  and whilst doing this I've had an unrelated thought. 

I've always been careful to focus all L, R, G and B images, but if in the processing we are going to blur the RGB to reduce noise, does it really matter if the RGB isnt precisely focused? Obviously one wouldnt mis-focus deliberately, but my filters arent quite parfocal and if I want to do an auromated overnight run, given i dont have a motor/auto focuser,  would it matter if one colour was slightly defocused?

Just wondered what folk might think of that?

 

[tooth_dr]

11 minutes ago, Tommohawk said:

I've just been rethinking slightly my imaging / processing techniques - mostly because I've just had a dabble at binning RGB  -  and whilst doing this I've had an unrelated thought. 

I've always been careful to focus all L, R, G and B images, but if in the processing we are going to blur the RGB to reduce noise, does it really matter if the RGB isnt precisely focused? Obviously one wouldnt mis-focus deliberately, but my filters arent quite parfocal and if I want to do an auromated overnight run, given i dont have a motor/auto focuser,  would it matter if one colour was slightly defocused?

Just wondered what folk might think of that?

 

Interesting as I'm working through the same process.  I think it does matter to some extent, but depends how far out the focus is.  In my recent images the RGB filters werent refocused, and the final images are ok, because the luminance was well focused.  But I did notice that the stars were large and therefore required much more work, something that I struggled with.  You can see the difference in focus below, but it still produced an image.  The green focus is definitely off, and i think the blue is combination of CA and bad focus, but less so that green.

 

 

 

 

[Tommohawk]

Excellent!

Reassuring to know it wasn't such a daft thought, and also a great demo animation you've done. It does make the point that the stars may be the biggest issue if defocused.

TBH I dont think mine is that far off, as I'm using a newt at the moment. It's certainly not 100% parfocal though and I put that down to the coma corrector. At some point I need to check this out with  the frac too, but that will probably wait until galaxies/planets done. 

Funny that your green is out of focus - you'd think being mid spectrum ish it would be better than R and B. Maybe the scope it optimized for R and B though so G suffers maybe. 

[tooth_dr]

9 minutes ago, Tommohawk said:

Excellent!

Reassuring to know it wasn't such a daft thought, and also a great demo animation you've done. It does make the point that the stars may be the biggest issue if defocused.

TBH I dont think mine is that far off, as I'm using a newt at the moment. It's certainly not 100% parfocal though and I put that down to the coma corrector. At some point I need to check this out with  the frac too, but that will probably wait until galaxies/planets done. 

Funny that your green is out of focus - you'd think being mid spectrum ish it would be better than R and B. Maybe the scope it optimized for R and B though so G suffers maybe. 

I'm hoping to offset my filter focus positions using a basic autofocuser.  I agree, I normally 'aim' to sort of to get green in focus, I think in this case I worked with red.  It's an ED80 so a basic doublet, with blue bloating.  I was surprised at the green too, but maybe something occurred locally to some of the subs, be it temp change or whatever.  If this was a final RGB image, I dont think it would be passable, but it seems to do the job as a colour layer.

[Tommohawk]

Just now, tooth_dr said:

but it seems to do the job as a colour layer.

Agreed - it's a nice final image. Cant work out what it is though? 

[tooth_dr]

8 minutes ago, Tommohawk said:

Agreed - it's a nice final image. Cant work out what it is though? 

It’s M94, must not be that good though if you don’t recognise it 😂😂

[Tommohawk]

10 minutes ago, tooth_dr said:

It’s M94, must not be that good though if you don’t recognise it 😂😂

LOL. That's more a measure of my ignorance than your processing!

Funny thing is I've never imaged it myself - I think because there's so little dark sky this time of year I tend to ease off a bit. But this year because I'm furloughed and the weathers fine I'm doing much more. I did consider doing M94 the other day, but in the end went for M106 which I'm just processing now. I'll have to put M94 on the list!

Edited May 21, 2020 by Tommohawk

[MarkAR]

I would have to say you should try to get the best focus you can on each filter. Your final image will be better for it.

[Datalord]

I have a related problem, although it's not about focus, but I think my solution to the problem is applicable for you.

My narrowband filters are HA-3nm, O3 and S2 are 8nm. The obvious issue is that 8nm lets in a lot more light and so the stars bloat a bit more than in my Ha subs. I solve it by putting the stacked S2 and O3 through a MorphologicalTransformation in PI to reduce the star sizes to the same as the Ha before I combine the images.

[Tommohawk]

On 23/05/2020 at 15:45, Datalord said:

I have a related problem, although it's not about focus, but I think my solution to the problem is applicable for you.

My narrowband filters are HA-3nm, O3 and S2 are 8nm. The obvious issue is that 8nm lets in a lot more light and so the stars bloat a bit more than in my Ha subs. I solve it by putting the stacked S2 and O3 through a MorphologicalTransformation in PI to reduce the star sizes to the same as the Ha before I combine the images.

Agreed -  ideally all channels would be well focused, but in the case where one channel is defocused or stars are larger for any other reason, fixing the star sizes before combining is the way to go.

[jimjam11]

I guess another approach would be to test it and measure the result? 
Something like: Focus on 1 filter and then take 10mins with each filter. Then focus each filter manually and repeat. You could then measure the median fwhm for each set and compare the difference?

 

With my zs73 (f5.9) it makes little difference so I tend to manually focus with lum. With my 200pds (f4.5) it does so I focus electronically and use offsets between the filters. I will probably add another eaf to my zs73 in time to fully optimise for all filters/subs.