Back to forums

Extremely noisy RGB

Processing Noise Reduction Solar System IC2118 IC Objects
7 replies•234 views
Enrique Ojeda avatar
Enrique Ojeda avatar
Hi! I have the following problem with my RGB: it's too noisy! I apply NoiseXTerminator, different parameters, and I can't reduce the noise. My RGB is 60 images per channel of 60 seconds. My camera was refrigerated at -10C. I checked if there were low clouds or something similar in my individual images and no. My sky is Bortle 2. I thought maybe I was missing Dithering but I see that I did apply it every 6 images and when checking each one I can see the movement where it was applied. I used the Antlia Triband RGB Ultra Filter in front of my RGB filters, I thought maybe that affected it but days before I took other objects, the same parameters, and they didn't present the detail that happens to me with this particular object (IC2118). My setup is: WO RedCat III, ZWO ASI 294 MM Pro, ZWO FilterWheel, Antila RGB filters, ZWO EAF, Antlia Triband RGB Ultra Filter. The software is PixInsight. I would greatly appreciate your comments. Thank you, CS!





Respectful
Miguel T. avatar
Miguel T. avatar
  1. 1 hour per channel is far too little for a faint object. Try >6 hours per filter.
  2. Consider capturing luminance to lower the integration time needed on color filters.
  3. Why two filters in front of each other? Shouldn't do that.
  4. Use your light pollution filter alone for capturing luminance.
  5. Use your regular RGB filters alone for capturing RGB.
  6. A light pollution filter in Bortle 2 is probably hurting more than helping. I am in a Bortle 9 and stopped using mine. Light pollution with LED lights has evolved way past the classic sodium light wavelengths, you're cutting real signal from the object you're capturing, and light pollution remain present in the other wavelengths.
  7. Use NoiseXTerminator moderately (~70% maximum).
    1. There is no substitute to the right amount of signal.
    2. We can tell by the orange peel effect and multiple other artifacts when it has been overdone.
    3. Keeping some noise is required to preserve detail.
Helpful
Rob Kiefer avatar
Rob Kiefer avatar
Personally, I would never use a light pollution filter like the Antlia Triband RGB Ultra Filter in Bortle 2. RGB plus LUM with enough integration time you should work very well. 
Rob
Enrique Ojeda avatar
Enrique Ojeda avatar
Rob Kiefer:
Personally, I would never use a light pollution filter like the Antlia Triband RGB Ultra Filter in Bortle 2. RGB plus LUM with enough integration time you should work very well. 
Rob




Thanks so much Rob, I'll follow your recommendations. CS!
Enrique Ojeda avatar
Enrique Ojeda avatar
Miguel T.:
  1. 1 hour per channel is far too little for a faint object. Try >6 hours per filter.
  2. Consider capturing luminance to lower the integration time needed on color filters.
  3. Why two filters in front of each other? Shouldn't do that.
  4. Use your light pollution filter alone for capturing luminance.
  5. Use your regular RGB filters alone for capturing RGB.
  6. A light pollution filter in Bortle 2 is probably hurting more than helping. I am in a Bortle 9 and stopped using mine. Light pollution with LED lights has evolved way past the classic sodium light wavelengths, you're cutting real signal from the object you're capturing, and light pollution remain present in the other wavelengths.
  7. Use NoiseXTerminator moderately (~70% maximum).
    1. There is no substitute to the right amount of signal.
    2. We can tell by the orange peel effect and multiple other artifacts when it has been overdone.
    3. Keeping some noise is required to preserve detail.



Thank you very much, Miguel. I will follow the detailed recommendations you've given me. CS!
Well Written Respectful Supportive
Brian Puhl avatar
Brian Puhl avatar
Looking at this image, and glancing at your gallery, I noticing you shoot 60 second subs….   and If i were to guess, you probably did that here as well.    60 seconds is not enough time to really soak up any signal.  

i know the 294 has some amp glow to deal with but I'd shoot for at least 3 minute subs, if not more.   You have nothing limiting you other than amp glow in this scenario.    Push those longer subs, gather more signal faster.  

Also just to echo previous comments, but you need nothing more than a UV/IR filter in dark skies such as yours.    Bortle 2 is a dream for most.   Filtering is doing nothing productive for you on broadband signals.
Helpful Insightful
Enrique Ojeda avatar
Enrique Ojeda avatar
Brian Puhl:
Looking at this image, and glancing at your gallery, I noticing you shoot 60 second subs....   and If i were to guess, you probably did that here as well.    60 seconds is not enough time to really soak up any signal.  

i know the 294 has some amp glow to deal with but I'd shoot for at least 3 minute subs, if not more.   You have nothing limiting you other than amp glow in this scenario.    Push those longer subs, gather more signal faster.  

Also just to echo previous comments, but you need nothing more than a UV/IR filter in dark skies such as yours.    Bortle 2 is a dream for most.   Filtering is doing nothing productive for you on broadband signals.




Thanks, Brian! I appreciate your feedback. Lately, I've been shooting at 60 seconds for RGB because one of my problems is that I can't get acceptable blue stars. I've noticed some improvement with that time, but as Miguel points out, this particular object—IC2118—is too dim and requires much more time. Also, as I've been told, the additional use of the RGB filter hasn't helped. I used it at the suggestion of a colleague, but I see it wasn't a good idea. Thanks to your comments, I can rest easy, and I hope to get my RGB back next week, but with more hours and my 180-second frames. I have several hours of HA to integrate, so hopefully, I'll get good results. CS!
Respectful
Brian Puhl avatar
Brian Puhl avatar
Enrique Ojeda:
Brian Puhl:
Looking at this image, and glancing at your gallery, I noticing you shoot 60 second subs....   and If i were to guess, you probably did that here as well.    60 seconds is not enough time to really soak up any signal.  

i know the 294 has some amp glow to deal with but I'd shoot for at least 3 minute subs, if not more.   You have nothing limiting you other than amp glow in this scenario.    Push those longer subs, gather more signal faster.  

Also just to echo previous comments, but you need nothing more than a UV/IR filter in dark skies such as yours.    Bortle 2 is a dream for most.   Filtering is doing nothing productive for you on broadband signals.




Thanks, Brian! I appreciate your feedback. Lately, I've been shooting at 60 seconds for RGB because one of my problems is that I can't get acceptable blue stars. I've noticed some improvement with that time, but as Miguel points out, this particular object—IC2118—is too dim and requires much more time. Also, as I've been told, the additional use of the RGB filter hasn't helped. I used it at the suggestion of a colleague, but I see it wasn't a good idea. Thanks to your comments, I can rest easy, and I hope to get my RGB back next week, but with more hours and my 180-second frames. I have several hours of HA to integrate, so hopefully, I'll get good results. CS!



For what it's worth, changing your exposure time isn't going to change the stars.    That's typically going to be an optical issue or a focus issue, ignoring guiding factors.  Misshapen blue stars are typically a sign of chromatic abberation, but I've never really known the Redcats to exhibit poor optics.   I would more than likely look into your focuser offsets, or maybe backlash. 

As of writing this I realized you are mono, and I'm going to presume you didn't stack filters did you?  The Triband filter while imaging through another filter?    That would definitely cause some poor stars and reflections.
Helpful Insightful