LRGB, RGB vs integration time

0.00 Jos Stassen ... · view views · Share link
I often read that LRGB is always superior to RGB - yup I know, many discussions on this. We had a few good nights here in the Netherlands (very rare) and I was able to capture 59x300s R, G and B, plus 95x300s L. In total over 22 hours under B7 skies, and of course my expectation was an image that is close to a Hubbles image. Not. Plus to my surprise RGB looked better than LRGB. Not what I expected. So, I did a little experiment using less frames (14xL, 10xR, 10xG and 10xB). And this shows the opposite: LRGB results in a nicer picture than RGB. To me it seems that when using a mono camera with L, R, G and B filters, LRGB can get you a good image in less time. But there is a ceiling for quality. RGB can break through that ceiling, but it takes more integration time. Integration time is an important variable when choosing LRGB over RGB. And conclusion 2: there is also a ceiling for integration time (under B7 skies). At a certain moment adding integration time does not do much for image quality - understandable as this is quadratic growth. Have a look https://www.astrobin.com/8a8i85/ - Original (nd best IMO): 14,4 hours of RGB (59xR, 59xG, 59xB) - Revision B (nice but close to C, but 19 hours extra): 22,3 hours of LRGB (95xL, 59xR, 59xG, 59xB) - Revision C (magic compared to D and just 1 hour extra) : 3,6 hours of LRGB (14xL, 10xR, 10xG, 10xB) - Revision D: 2,5 hours of RGB (10xR, 10xG, 10xB) Makes me think what is sufficient RGB integration time? What is the max LRGB integration time before RGB makes more sense? Since it rains a lot in the Netherlands, my goal is getting maximum quality in a minimum time. Doing 22 hours again on one subject is probably not going to happen anytime soon. Any ideas how to calculate or estimate this? Thx Jos Like

0.00

...

I often read that LRGB is always superior to RGB - yup I know, many discussions on this.
We had a few good nights here in the Netherlands (very rare) and I was able to capture 59x300s R, G and B, plus 95x300s L.
In total over 22 hours under B7 skies, and of course my expectation was an image that is close to a Hubbles image.

Not. Plus to my surprise RGB looked better than LRGB. Not what I expected.

So, I did a little experiment using less frames (14xL, 10xR, 10xG and 10xB). And this shows the opposite: LRGB results in a nicer picture than RGB.

To me it seems that when using a mono camera with L, R, G and B filters, LRGB can get you a good image in less time. But there is a ceiling for quality. RGB can break through that ceiling, but it takes more integration time. Integration time is an important variable when choosing LRGB over RGB.
And conclusion 2: there is also a ceiling for integration time (under B7 skies). At a certain moment adding integration time does not do much for image quality - understandable as this is quadratic growth.

Have a look https://www.astrobin.com/8a8i85/
- Original (nd best IMO): 14,4 hours of RGB (59xR, 59xG, 59xB)
- Revision B (nice but close to C, but 19 hours extra): 22,3 hours of LRGB (95xL, 59xR, 59xG, 59xB)
- Revision C (magic compared to D and just 1 hour extra) : 3,6 hours of LRGB (14xL, 10xR, 10xG, 10xB)
- Revision D: 2,5 hours of RGB (10xR, 10xG, 10xB)

Makes me think what is sufficient RGB integration time?
What is the max LRGB integration time before RGB makes more sense?
Since it rains a lot in the Netherlands, my goal is getting maximum quality in a minimum time. Doing 22 hours again on one subject is probably not going to happen anytime soon. Any ideas how to calculate or estimate this?

Thx
Jos

9.38 Quinn Groessl ... · view views · 1 like · Share link
Alternate opinion, it's processing skill. Lum is going to improve SNR and make better images. The real question is the ratio of L to RGB. I personally like 3 to 1. Like

0.00 Pierre Fleurant ... · view views · Share link
Hi, From "Inside PixInsight", Second edition by Warren A. Keller, pg. 65: ------------- When only red, green, and blue data have been collected by a monochromatic system, their masters can be integrated into a single grayscale image and used as a luminance image. This 'pseudoluminance' (Pseudo-L, Synthetic Luminance, or Synth-L) can significantly improve your RGB result, because it's comprised of the combined signal of all of the color frames. Worth reading, is a 2013 forum post by Juan Conejero where he details why this method may be preferable to acquiring luminance. Following the same procedure as for a Super Luminance, integrate the three color masters into a pseudoluminance, later applying it to the chrominance master as we'll see (Chap. 13). As with the production of the Super Luminance, Noise Evaluation will give stronger masters more weight than weaker ones. ------------ From what I understand, running RGB at higher binning (2x2) along with 1x1 binning for L is good for CCD cameras (they don't add the read noise when binning) and not worth it for CMOS cameras (they add the read noise when binning (2x2 ->>> 4 time the read noise, or added in quadrature); You get higher SNR in the RGB and higher resolution in the L. Good idea for CCD but not worth it with CMOS. Is that reasonable? Also, can anyone send out that 2013 forum post by Juan Conejero post on the PixInsight forum, I can't find it. Thanks! Like

0.00

Pierre Fleurant

...

· Share link

Hi,

From "Inside PixInsight", Second edition by Warren A. Keller, pg. 65:

-------------
When only red, green, and blue data have been collected by a monochromatic system, their masters can be integrated into a single grayscale image and used as a luminance image. This 'pseudoluminance' (Pseudo-L, Synthetic Luminance, or Synth-L) can significantly improve your RGB result, because it's comprised of the combined signal of all of the color frames. Worth reading, is a 2013 forum post by Juan Conejero where he details why this method may be preferable to acquiring luminance.

Following the same procedure as for a Super Luminance, integrate the three color masters into a pseudoluminance, later applying it to the chrominance master as we'll see (Chap. 13). As with the production of the Super Luminance, Noise Evaluation will give stronger masters more weight than weaker ones.
------------

From what I understand, running RGB at higher binning (2x2) along with 1x1 binning for L is good for CCD cameras (they don't add the read noise when binning) and not worth it for CMOS cameras (they add the read noise when binning (2x2 ->>> 4 time the read noise, or added in quadrature); You get higher SNR in the RGB and higher resolution in the L. Good idea for CCD but not worth it with CMOS.

Is that reasonable?

Also, can anyone send out that 2013 forum post by Juan Conejero post on the PixInsight forum, I can't find it.

Thanks!

8.21 Miguel T. ... · view views · 2 likes · Share link
Jos Stassen: What is the max LRGB integration time before RGB makes more sense? Never. Your assumption that RGB without L is better past a certain threshold is wrong to begin with. It's both processing skills and your Bortle 7 adding a level of difficulty to the L data. I'm in a Bortle 9 and the only times I had to toss out L data is because the complex light pollution gradients were too uncontrollable and mismatched the RGB too greatly. Edited ... Like

8.21

Miguel T.

...

· 2 likes

· Share link

Jos Stassen:
What is the max LRGB integration time before RGB makes more sense?

Never. Your assumption that RGB without L is better past a certain threshold is wrong to begin with.

It's both processing skills and your Bortle 7 adding a level of difficulty to the L data.
I'm in a Bortle 9 and the only times I had to toss out L data is because the complex light pollution gradients were too uncontrollable and mismatched the RGB too greatly.

Edited ...

0.00 Brad Hill ... · view views · 1 like · Share link
I think you’re mainly dealing with light pollution issues. I image from Bortle 4 and there’s a huge improvement when I use L no matter how long or short total integration time is. Like

9.89 andrea tasselli ... · view views · 3 likes · Share link
RGB is cleaner and nicer than LRGB as long as the minimum SNR threshold is reached. I regret having spent over 20 hours acquiring L when it would have been better spent in acquiring RGB and better still using an OSC camera. Luminance isn't worth it for bordband targets in B7 skies and over. And sometimes in B1 skies too. Edited ... Like

0.00 Jos Stassen Topic starter ... · view views · Share link
I appreciate all the suggestions. Processing skills - only using PI since Dec – so that’s a yes. Ratio L to RGB – I’ll try this too Light pollution – not much I can do here apart from learning to control this better in PI Like

0.00 ArchStarGazer ... · view views · 2 likes · Share link
Processing is very important. Before mixing with RGB, L has to be processed to show lots of detail (e.g. using HDRMT), and there should be no areas in L that are too bright. Otherwise, colours will fade in these bright areas and the LRGB image may appear less colourful than the RGB image (this may have happened in version B of your image). Like

8.64 Brian Diaz ... · view views · 3 likes · Share link
HI B7 and luminance are not friends , i recommend superLumiance for galaxies CS Brian Edited ... Like

14.24 Arun H ... · view views · 1 like · Share link
I have had much better luck with RGB than LRGB in Bortle 6. Completely lost confidence in L to do much for me. Like

26.84 John Hayes ... · view views · Share link
Jos Stassen: I often read that LRGB is always superior to RGB - yup I know, many discussions on this. We had a few good nights here in the Netherlands (very rare) and I was able to capture 59x300s R, G and B, plus 95x300s L. In total over 22 hours under B7 skies, and of course my expectation was an image that is close to a Hubbles image. Not. Plus to my surprise RGB looked better than LRGB. Not what I expected. So, I did a little experiment using less frames (14xL, 10xR, 10xG and 10xB). And this shows the opposite: LRGB results in a nicer picture than RGB. To me it seems that when using a mono camera with L, R, G and B filters, LRGB can get you a good image in less time. But there is a ceiling for quality. RGB can break through that ceiling, but it takes more integration time. Integration time is an important variable when choosing LRGB over RGB. And conclusion 2: there is also a ceiling for integration time (under B7 skies). At a certain moment adding integration time does not do much for image quality - understandable as this is quadratic growth. Have a look https://www.astrobin.com/8a8i85/ - Original (nd best IMO): 14,4 hours of RGB (59xR, 59xG, 59xB) - Revision B (nice but close to C, but 19 hours extra): 22,3 hours of LRGB (95xL, 59xR, 59xG, 59xB) - Revision C (magic compared to D and just 1 hour extra) : 3,6 hours of LRGB (14xL, 10xR, 10xG, 10xB) - Revision D: 2,5 hours of RGB (10xR, 10xG, 10xB) Makes me think what is sufficient RGB integration time? What is the max LRGB integration time before RGB makes more sense? Since it rains a lot in the Netherlands, my goal is getting maximum quality in a minimum time. Doing 22 hours again on one subject is probably not going to happen anytime soon. Any ideas how to calculate or estimate this? Thx Jos How are you combining L and RGB? That could be a big source of issues--depending on how you are processing the data. John Like

26.84

John Hayes

...

· Share link

Jos Stassen:
I often read that LRGB is always superior to RGB - yup I know, many discussions on this.
We had a few good nights here in the Netherlands (very rare) and I was able to capture 59x300s R, G and B, plus 95x300s L.
In total over 22 hours under B7 skies, and of course my expectation was an image that is close to a Hubbles image.

Not. Plus to my surprise RGB looked better than LRGB. Not what I expected.

So, I did a little experiment using less frames (14xL, 10xR, 10xG and 10xB). And this shows the opposite: LRGB results in a nicer picture than RGB.

To me it seems that when using a mono camera with L, R, G and B filters, LRGB can get you a good image in less time. But there is a ceiling for quality. RGB can break through that ceiling, but it takes more integration time. Integration time is an important variable when choosing LRGB over RGB.
And conclusion 2: there is also a ceiling for integration time (under B7 skies). At a certain moment adding integration time does not do much for image quality - understandable as this is quadratic growth.

Have a look https://www.astrobin.com/8a8i85/
- Original (nd best IMO): 14,4 hours of RGB (59xR, 59xG, 59xB)
- Revision B (nice but close to C, but 19 hours extra): 22,3 hours of LRGB (95xL, 59xR, 59xG, 59xB)
- Revision C (magic compared to D and just 1 hour extra) : 3,6 hours of LRGB (14xL, 10xR, 10xG, 10xB)
- Revision D: 2,5 hours of RGB (10xR, 10xG, 10xB)

Makes me think what is sufficient RGB integration time?
What is the max LRGB integration time before RGB makes more sense?
Since it rains a lot in the Netherlands, my goal is getting maximum quality in a minimum time. Doing 22 hours again on one subject is probably not going to happen anytime soon. Any ideas how to calculate or estimate this?

Thx
Jos

How are you combining L and RGB? That could be a big source of issues--depending on how you are processing the data.

John

0.00 Jos Stassen Topic starter ... · view views · Share link
John Hayes: Jos Stassen: I often read that LRGB is always superior to RGB - yup I know, many discussions on this. We had a few good nights here in the Netherlands (very rare) and I was able to capture 59x300s R, G and B, plus 95x300s L. In total over 22 hours under B7 skies, and of course my expectation was an image that is close to a Hubbles image. Not. Plus to my surprise RGB looked better than LRGB. Not what I expected. So, I did a little experiment using less frames (14xL, 10xR, 10xG and 10xB). And this shows the opposite: LRGB results in a nicer picture than RGB. To me it seems that when using a mono camera with L, R, G and B filters, LRGB can get you a good image in less time. But there is a ceiling for quality. RGB can break through that ceiling, but it takes more integration time. Integration time is an important variable when choosing LRGB over RGB. And conclusion 2: there is also a ceiling for integration time (under B7 skies). At a certain moment adding integration time does not do much for image quality - understandable as this is quadratic growth. Have a look https://www.astrobin.com/8a8i85/ - Original (nd best IMO): 14,4 hours of RGB (59xR, 59xG, 59xB) - Revision B (nice but close to C, but 19 hours extra): 22,3 hours of LRGB (95xL, 59xR, 59xG, 59xB) - Revision C (magic compared to D and just 1 hour extra) : 3,6 hours of LRGB (14xL, 10xR, 10xG, 10xB) - Revision D: 2,5 hours of RGB (10xR, 10xG, 10xB) Makes me think what is sufficient RGB integration time? What is the max LRGB integration time before RGB makes more sense? Since it rains a lot in the Netherlands, my goal is getting maximum quality in a minimum time. Doing 22 hours again on one subject is probably not going to happen anytime soon. Any ideas how to calculate or estimate this? Thx Jos How are you combining L and RGB? That could be a big source of issues--depending on how you are processing the data. John In PhotoShop as the luminance layer Like

0.00

Jos Stassen

Topic starter

...

· Share link

John Hayes:
Jos Stassen:
I often read that LRGB is always superior to RGB - yup I know, many discussions on this.
We had a few good nights here in the Netherlands (very rare) and I was able to capture 59x300s R, G and B, plus 95x300s L.
In total over 22 hours under B7 skies, and of course my expectation was an image that is close to a Hubbles image.

Not. Plus to my surprise RGB looked better than LRGB. Not what I expected.

So, I did a little experiment using less frames (14xL, 10xR, 10xG and 10xB). And this shows the opposite: LRGB results in a nicer picture than RGB.

To me it seems that when using a mono camera with L, R, G and B filters, LRGB can get you a good image in less time. But there is a ceiling for quality. RGB can break through that ceiling, but it takes more integration time. Integration time is an important variable when choosing LRGB over RGB.
And conclusion 2: there is also a ceiling for integration time (under B7 skies). At a certain moment adding integration time does not do much for image quality - understandable as this is quadratic growth.

Have a look https://www.astrobin.com/8a8i85/
- Original (nd best IMO): 14,4 hours of RGB (59xR, 59xG, 59xB)
- Revision B (nice but close to C, but 19 hours extra): 22,3 hours of LRGB (95xL, 59xR, 59xG, 59xB)
- Revision C (magic compared to D and just 1 hour extra) : 3,6 hours of LRGB (14xL, 10xR, 10xG, 10xB)
- Revision D: 2,5 hours of RGB (10xR, 10xG, 10xB)

Makes me think what is sufficient RGB integration time?
What is the max LRGB integration time before RGB makes more sense?
Since it rains a lot in the Netherlands, my goal is getting maximum quality in a minimum time. Doing 22 hours again on one subject is probably not going to happen anytime soon. Any ideas how to calculate or estimate this?

Thx
Jos

How are you combining L and RGB? That could be a big source of issues--depending on how you are processing the data.

John

In PhotoShop as the luminance layer

26.84 John Hayes ... · view views · 3 likes · Share link
Jos Stassen: In PhotoShop as the luminance layer Ok...that should be fine. In general, the Lum data should drive the SNR and the RGB data should drive the color noise in the LRGB combination. If you gather all of your data under similar conditions, the lum channel will have the highest signal of the 4 channels and if you devote most of your exposure time to lum, it will always have a higher SNR than the RGB data. So, something is screwy if you are seeing higher SNR from your RGB data. Either the acquisition conditions are changing or there's something going on with the processing. John Like

26.84

John Hayes

...

· 3 likes

· Share link

Jos Stassen:
In PhotoShop as the luminance layer

Ok...that should be fine. In general, the Lum data should drive the SNR and the RGB data should drive the color noise in the LRGB combination. If you gather all of your data under similar conditions, the lum channel will have the highest signal of the 4 channels and if you devote most of your exposure time to lum, it will always have a higher SNR than the RGB data. So, something is screwy if you are seeing higher SNR from your RGB data. Either the acquisition conditions are changing or there's something going on with the processing.

John

5.02 Christian Bennich ... · view views · 2 likes · Share link
I have been really happy with @Adam Block tutorials, also on LRGB. https://www.adamblockstudios.com/categories/m83-lrgb-2024 It is a subscription only video though. (Well worth the investment) The intricate details on the process and why it’s easy to “dilute” the RGB data with your L-data etc. is explained in detail. Like

1.51 Andy Crossley ... · view views · 1 like · Share link
Christian Bennich: I have been really happy with @Adam Block tutorials, also on LRGB. https://www.adamblockstudios.com/categories/m83-lrgb-2024 It is a subscription only video though. (Well worth the investment) The intricate details on the process and why it’s easy to “dilute” the RGB data with your L-data etc. is explained in detail. Came to recommend the very same thing. As another friend has also been helping me with lately. Treat the RGB layer almost like a watercolour image, providing colouring, and use the L layer to bring out sharpness/brightness/detail. Like

1.51

Andy Crossley

...

· 1 like

· Share link

Christian Bennich:
I have been really happy with @Adam Block tutorials, also on LRGB.

https://www.adamblockstudios.com/categories/m83-lrgb-2024
It is a subscription only video though. (Well worth the investment)

The intricate details on the process and why it’s easy to “dilute” the RGB data with your L-data etc. is explained in detail.

Came to recommend the very same thing.

As another friend has also been helping me with lately. Treat the RGB layer almost like a watercolour image, providing colouring, and use the L layer to bring out sharpness/brightness/detail.

0.00 Jos Stassen Topic starter ... · view views · 2 likes · Share link
Brian Diaz: HI B7 and luminance are not friends , i recommend superLumiance for galaxies CS Brian I did a super quick SuperLuminance try out, amazing! Thx Brian! Like

13.55 Bruce Donzanti ... · view views · 1 like · Share link
Jos Stassen: Brian Diaz: HI B7 and luminance are not friends , i recommend superLumiance for galaxies CS Brian I did a super quick SuperLuminance try out, amazing! Thx Brian! Completely agree, Brian. I eventually learned this the hard way after many attempts. Finally, this past year, I captured my best galaxies using a super lum under my B7 sky. Grabbing many more hours (like double the RGB data) of just lum under these sky conditions is a waste of good imaging time. Bruce Like

13.55

Bruce Donzanti

...

· 1 like

· Share link

Jos Stassen:
Brian Diaz:
HI

B7 and luminance are not friends , i recommend superLumiance for galaxies

CS
Brian

I did a super quick SuperLuminance try out, amazing! Thx Brian!

Completely agree, Brian. I eventually learned this the hard way after many attempts. Finally, this past year, I captured my best galaxies using a super lum under my B7 sky. Grabbing many more hours (like double the RGB data) of just lum under these sky conditions is a waste of good imaging time.

Bruce

1.51 Rostokko ... · view views · Share link
Are there any benefits in using a quad filter or similar to capture L in B>6 areas? Is anyone doing that? Like

9.89 andrea tasselli ... · view views · 1 like · Share link
I have tried and gives a significant benefit compared to the standard UV/IR-Cut. Like

	Distinct awarded users	Total awarded images
Image of the day
Top picks
Top pick nominations

	Image of the day	Top pick	Top pick nominations	Total submitted
Deep sky
Solar system
Extremely wide field
Star trails
Northern lights
Noctilucent clouds
Landscape
Artificial satellite

	Image of the day	Top pick	Top pick nominations	Total submitted
Backyard
Traveller
Own remote observatory
Amateur hosting facility
Public amaeteur data
Professional, scientific grade data
Mix of multiple sources
Other
Unknown

AstroBin

AstroBin

Image Index:

Contribution Index (beta):

LRGB, RGB vs integration time [Deep Sky] Acquisition techniques · Jos Stassen · ... · 18 · 735 · 2

AstroBin

AstroBin

Image Index:

Contribution Index (beta):

LRGB, RGB vs integration time [Deep Sky] Acquisition techniques · Jos Stassen · ... · 18 · 735 · 2

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(

Sorry

Share this forum post

Post editing not available

Sorry

Sorry

Sorry :-(