Deminishing returns.....

Started 3 months ago | Discussions
Kemo-sabe Junior Member • Posts: 39
Deminishing returns.....
1

193 5' subs Bortle 9

Stacked in Sequator RNC stretched PP DxO PhotoLab4

FD 800mm f5.6l Canon 5D MKIII iso400 w Optolong L-eNhance dou-band 2" filter

DIY FD 300mm f4.0l guide-scope QHY5LII-C guide camera

iOptron CEM25P mount on tripier

adding more data doesn't appear to improve the image. Have I hit a point of diminishing returns?

Saabster Regular Member • Posts: 265
Re: Deminishing returns.....
1

Kemo-sabe wrote:

193 5' subs Bortle 9

Stacked in Sequator RNC stretched PP DxO PhotoLab4

FD 800mm f5.6l Canon 5D MKIII iso400 w Optolong L-eNhance dou-band 2" filter

DIY FD 300mm f4.0l guide-scope QHY5LII-C guide camera

iOptron CEM25P mount on tripier

adding more data doesn't appear to improve the image. Have I hit a point of diminishing returns?

With such short subs, it's possible. Why are you taking only 5 sec subs, especially since you're guiding? The L-Enhance needs longer time since it filters out so much light.

If you increase your exposure time to even 30 seconds at ISO 800 you'll be amazed at what you can get.

Did you take darks and flats? What processing did you do after stacking?

Here's a single 15sec sub from my Pentax K-1 at ISO 800 that I did some stretching in RawTherapee (a free image processing program.)

You have about 15 minutes of data. Imagine what you can do with longer exposures.

1x15sec ISO 800 Processed in RawTherapee.

Astrozoid
Astrozoid Contributing Member • Posts: 846
Re: Deminishing returns.....

Kemo-sabe wrote:

193 5' subs Bortle 9

adding more data doesn't appear to improve the image. Have I hit a point of diminishing returns?

You can always double your signal-to-noise ratio with 4x more exposure.

This is a ballpark way to look at it.

Does your image currently have the SNR you want it to have? If yes, then you don't need to shoot anymore.

If no, then look at your total exposure and ask yourself, are you prepared to spend 4x that much to double the SNR?

Now, looking at your image, you would say, you might say, wow, 772 x 5' more exposure seems like a lot for just 2x increase in SNR.

HOWEVER, if you want to record faint detail in the outer portions of the Orion Nebula, and you want to keep shooting from your current location, your only solution is to put a LOT more exposure in the image.

Of course, there may be more detail there that is just hidden because in your processing your have clipped the black point.

EDIT: seeing the previous post, it assumes you are shooting 5 second exposures. I am assuming you are shooting 5 MINUTE exposures.

If you are shooting 5 SECOND exposures, you need a LOT more total exposure.

Jerry

1llusive
1llusive Veteran Member • Posts: 3,206
Re: Deminishing returns.....

Looks about right for 16 minutes of exposure.

 1llusive's gear list:1llusive's gear list
Nikon Z6 II Sigma 100-400mm F5-6.3 Nikon Z 24-70mm F4 Nikon Z 85mm F1.8 +1 more
jasonm10 Regular Member • Posts: 173
Re: Deminishing returns.....

Saabster wrote:

Kemo-sabe wrote:

193 5' subs Bortle 9

Stacked in Sequator RNC stretched PP DxO PhotoLab4

FD 800mm f5.6l Canon 5D MKIII iso400 w Optolong L-eNhance dou-band 2" filter

DIY FD 300mm f4.0l guide-scope QHY5LII-C guide camera

iOptron CEM25P mount on tripier

adding more data doesn't appear to improve the image. Have I hit a point of diminishing returns?

With such short subs, it's possible. Why are you taking only 5 sec subs, especially since you're guiding? The L-Enhance needs longer time since it filters out so much light.

If you increase your exposure time to even 30 seconds at ISO 800 you'll be amazed at what you can get.

Did you take darks and flats? What processing did you do after stacking?

Here's a single 15sec sub from my Pentax K-1 at ISO 800 that I did some stretching in RawTherapee (a free image processing program.)

You have about 15 minutes of data. Imagine what you can do with longer exposures.

1x15sec ISO 800 Processed in RawTherapee.

Plus all the internet shooters appear to be routinely taking 500+ exposures as well. The more data to stretch the better.

Astrozoid
Astrozoid Contributing Member • Posts: 846
Re: Deminishing returns.....
1

1llusive wrote:

Looks about right for 16 minutes of exposure.

Original post said it was "193 5' subs".

That would be 193 x 5 minutes, not 5 seconds.

Unless the original poster meant " instead of '.

Two other hints maybe... 5 sec ar f/5.6 at iso400 with a strong light pollution filter would not record this much. And, 15 minutes of total exposure is not when most people would start  asking if they were at the point of diminishing returns.

So, if its seconds, then NO, he is not close to the point of diminishing returns.

If minutes, then YES, he is, and needs darker skies with his current setup, or a mono camera and narrowband filters.

Bortle 9 is brutal.

1llusive
1llusive Veteran Member • Posts: 3,206
Re: Deminishing returns.....

Astrozoid wrote:

1llusive wrote:

Looks about right for 16 minutes of exposure.

Original post said it was "193 5' subs".

That would be 193 x 5 minutes, not 5 seconds.

Unless the original poster meant " instead of '.

Two other hints maybe... 5 sec ar f/5.6 at iso400 with a strong light pollution filter would not record this much. And, 15 minutes of total exposure is not when most people would start asking if they were at the point of diminishing returns.

So, if its seconds, then NO, he is not close to the point of diminishing returns.

If minutes, then YES, he is, and needs darker skies with his current setup, or a mono camera and narrowband filters.

Bortle 9 is brutal.

I've shot a similar/better image with 17 x 1 minute exposures. I think 16 sounds about right. There is no way that is 965 minutes. Think about that.

 1llusive's gear list:1llusive's gear list
Nikon Z6 II Sigma 100-400mm F5-6.3 Nikon Z 24-70mm F4 Nikon Z 85mm F1.8 +1 more
jasonm10 Regular Member • Posts: 173
Re: Deminishing returns.....

1llusive wrote:

Astrozoid wrote:

1llusive wrote:

Looks about right for 16 minutes of exposure.

Original post said it was "193 5' subs".

That would be 193 x 5 minutes, not 5 seconds.

Unless the original poster meant " instead of '.

Two other hints maybe... 5 sec ar f/5.6 at iso400 with a strong light pollution filter would not record this much. And, 15 minutes of total exposure is not when most people would start asking if they were at the point of diminishing returns.

So, if its seconds, then NO, he is not close to the point of diminishing returns.

If minutes, then YES, he is, and needs darker skies with his current setup, or a mono camera and narrowband filters.

Bortle 9 is brutal.

I've shot a similar/better image with 17 x 1 minute exposures. I think 16 sounds about right. There is no way that is 965 minutes. Think about that.

Even with my limited experience, that doesn't look like a total 16 hour exposure time.

I could be wrong.

Astrozoid
Astrozoid Contributing Member • Posts: 846
Re: Deminishing returns.....

jasonm10 wrote:

1llusive wrote:

Astrozoid wrote:

1llusive wrote:

Looks about right for 16 minutes of exposure.

Original post said it was "193 5' subs".

That would be 193 x 5 minutes, not 5 seconds.

Unless the original poster meant " instead of '.

Two other hints maybe... 5 sec ar f/5.6 at iso400 with a strong light pollution filter would not record this much. And, 15 minutes of total exposure is not when most people would start asking if they were at the point of diminishing returns.

So, if its seconds, then NO, he is not close to the point of diminishing returns.

If minutes, then YES, he is, and needs darker skies with his current setup, or a mono camera and narrowband filters.

Bortle 9 is brutal.

I've shot a similar/better image with 17 x 1 minute exposures. I think 16 sounds about right. There is no way that is 965 minutes. Think about that.

Even with my limited experience, that doesn't look like a total 16 hour exposure time.

I could be wrong.

It could be processing inexperience.

It will be interesting to see if the original poster ever tells us.

Jerry

Astrozoid
Astrozoid Contributing Member • Posts: 846
Re: Deminishing returns.....

1llusive wrote:

I've shot a similar/better image with 17 x 1 minute exposures.

It is always so meaningless to mention a comparison like this.

First of all, IF you are correct, and it's 17 minutes total, a 1-minute sub is 12 times as long as a 5-second sub. So it's not an apples to apples comparison to start with.

A 5-second sub with a strong light-pollution filter at ISO 400 with that camera might not even be sky-noise limited. Think about that.

Second, it is completely useless information without telling us how dark your skies were, or what exact equipment you shot it with. Maybe you used a 12-inch f/2.8 scope. Maybe you shot under Bortle 1 skies. Maybe you used a camera with twice the quantum efficiency.

Plus, the op said he was guiding. Who guides 5 second exposures?

Jerry

1llusive
1llusive Veteran Member • Posts: 3,206
Re: Deminishing returns.....

Astrozoid wrote:

1llusive wrote:

I've shot a similar/better image with 17 x 1 minute exposures.

It is always so meaningless to mention a comparison like this.

First of all, IF you are correct, and it's 17 minutes total, a 1-minute sub is 12 times as long as a 5-second sub. So it's not an apples to apples comparison to start with.

A 5-second sub with a strong light-pollution filter at ISO 400 with that camera might not even be sky-noise limited. Think about that.

Second, it is completely useless information without telling us how dark your skies were, or what exact equipment you shot it with. Maybe you used a 12-inch f/2.8 scope. Maybe you shot under Bortle 1 skies. Maybe you used a camera with twice the quantum efficiency.

Jerry

Nikon D5300, 300mm f/5.6, red zone (driveway)

https://ericfrankephotos.smugmug.com/Galleries/Astrophotography/i-jDcGksG/A

 1llusive's gear list:1llusive's gear list
Nikon Z6 II Sigma 100-400mm F5-6.3 Nikon Z 24-70mm F4 Nikon Z 85mm F1.8 +1 more
1llusive
1llusive Veteran Member • Posts: 3,206
Re: Deminishing returns.....

jasonm10 wrote:

1llusive wrote:

Astrozoid wrote:

1llusive wrote:

Looks about right for 16 minutes of exposure.

Original post said it was "193 5' subs".

That would be 193 x 5 minutes, not 5 seconds.

Unless the original poster meant " instead of '.

Two other hints maybe... 5 sec ar f/5.6 at iso400 with a strong light pollution filter would not record this much. And, 15 minutes of total exposure is not when most people would start asking if they were at the point of diminishing returns.

So, if its seconds, then NO, he is not close to the point of diminishing returns.

If minutes, then YES, he is, and needs darker skies with his current setup, or a mono camera and narrowband filters.

Bortle 9 is brutal.

I've shot a similar/better image with 17 x 1 minute exposures. I think 16 sounds about right. There is no way that is 965 minutes. Think about that.

Even with my limited experience, that doesn't look like a total 16 hour exposure time.

I could be wrong.

And you would be right - I'm saying 16 minutes, not hours.

 1llusive's gear list:1llusive's gear list
Nikon Z6 II Sigma 100-400mm F5-6.3 Nikon Z 24-70mm F4 Nikon Z 85mm F1.8 +1 more
Adrien D Regular Member • Posts: 145
Re: Deminishing returns.....

With the overexposed core, 5 seconds with a filter seems low (or the post processing did that)

But in a bortle 9 sky, i can't imagine 5 minutes exposure at f/5.6 giving anything else than overexposed sky.

EricTheAstroJunkie Contributing Member • Posts: 762
Re: Deminishing returns.....

Adrien D wrote:

With the overexposed core, 5 seconds with a filter seems low (or the post processing did that)

But in a bortle 9 sky, i can't imagine 5 minutes exposure at f/5.6 giving anything else than overexposed sky.

They used the l-enhance filter and shot at ISO 400, even in bortle 8/9 with that filter on my Astrotech AT65EDQ and a D5300 I had to 10 minute exposures at ISO 400 to really see much.

 EricTheAstroJunkie's gear list:EricTheAstroJunkie's gear list
Nikon D200 Nikon D7000 Nikon D800E Nikon D5300 Sony a7S +14 more
OP Kemo-sabe Junior Member • Posts: 39
Re: Deminishing returns.....

Astrozoid, you are correct in your analysis. The images were shot in Bortle 9 sky and consist of 193 5 minute subs with a Canon FD 800mm f5.6l lens with a Canon 5D MKIII at iso400 incorporating an Optolong L-eNhance dou-band 2" drop-in filter. My guide scope consist of a DIY Canon FD 300mm f4.0l lens coupled to a QHY5LII-C guide camera. The Mount is an iOptron CEM25P with ADM side by side saddle and iPolar scope for polar alignment. I have Pegasus power distribution for the mount/equipment and dual Pegasus focus controllers for remote focusing the primary optics and guide-scope.

I am a newbie to processing astro images. It has has taken me more than 1 year to automate and dial-in the hardware software aspects of astro imaging and learn the nuances of the system.

However, astro processing is another story. I was to embarrassed by my pitiful results to post such a sad image in the light of so many awesome images that I have seen on this forum. Pride is a terrible thing.

My son suggested that I post the image to see if I could get some advice and input on what I am doing wrong in my processing workflow. Or, is this the best that one might expect from Bortle 9 skies with the optical equipment I am using or a combination of both?

I am using Clarke's methodology of in camera dark frame extraction for the subs. Using Clarke's Methodology the subs actually take 10 minutes to shoot per sub. I use Sequator and/or DSS for stacking. PS RC-Astro filter GradientXterminator to remove gradients. Then RNC stretch or PS to stretch. StarTools or DxO PhotoLab4 Post Processing.

I have been experimenting with Canon's DPP4 RAW processor to process and convert the RAW subs to TIFF prior to stacking with varying results. In an attempt to pull more fine detail from the subs.

Astrozoid
Astrozoid Contributing Member • Posts: 846
Re: Deminishing returns.....

Kemo-sabe wrote:

Astrozoid, you are correct in your analysis. The images were shot in Bortle 9 sky and consist of 193 5 minute subs with a Canon FD 800mm f5.6l lens with a Canon 5D MKIII at iso400 incorporating an Optolong L-eNhance dou-band 2" drop-in filter. My guide scope consist of a DIY Canon FD 300mm f4.0l lens coupled to a QHY5LII-C guide camera. The Mount is an iOptron CEM25P with ADM side by side saddle and iPolar scope for polar alignment. I have Pegasus power distribution for the mount/equipment and dual Pegasus focus controllers for remote focusing the primary optics and guide-scope.

I am a newbie to processing astro images. It has has taken me more than 1 year to automate and dial-in the hardware software aspects of astro imaging and learn the nuances of the system.

However, astro processing is another story. I was to embarrassed by my pitiful results to post such a sad image in the light of so many awesome images that I have seen on this forum. Pride is a terrible thing.

My son suggested that I post the image to see if I could get some advice and input on what I am doing wrong in my processing workflow. Or, is this the best that one might expect from Bortle 9 skies with the optical equipment I am using or a combination of both?

I am using Clarke's methodology of in camera dark frame extraction for the subs. Using Clarke's Methodology the subs actually take 10 minutes to shoot per sub. I use Sequator and/or DSS for stacking. PS RC-Astro filter GradientXterminator to remove gradients. Then RNC stretch or PS to stretch. StarTools or DxO PhotoLab4 Post Processing.

I have been experimenting with Canon's DPP4 RAW processor to process and convert the RAW subs to TIFF prior to stacking with varying results. In an attempt to pull more fine detail from the subs.

Post a link to the raw, linear stack and we'll see if there is more there to be brought out in processing.

Jerry

thinker Regular Member • Posts: 224
Re: Deminishing returns.....

Astrozoid wrote:

Kemo-sabe wrote:

Astrozoid, you are correct in your analysis. The images were shot in Bortle 9 sky and consist of 193 5 minute subs with a Canon FD 800mm f5.6l lens with a Canon 5D MKIII at iso400 incorporating an Optolong L-eNhance dou-band 2" drop-in filter. My guide scope consist of a DIY Canon FD 300mm f4.0l lens coupled to a QHY5LII-C guide camera. The Mount is an iOptron CEM25P with ADM side by side saddle and iPolar scope for polar alignment. I have Pegasus power distribution for the mount/equipment and dual Pegasus focus controllers for remote focusing the primary optics and guide-scope.

I am a newbie to processing astro images. It has has taken me more than 1 year to automate and dial-in the hardware software aspects of astro imaging and learn the nuances of the system.

However, astro processing is another story. I was to embarrassed by my pitiful results to post such a sad image in the light of so many awesome images that I have seen on this forum. Pride is a terrible thing.

My son suggested that I post the image to see if I could get some advice and input on what I am doing wrong in my processing workflow. Or, is this the best that one might expect from Bortle 9 skies with the optical equipment I am using or a combination of both?

I am using Clarke's methodology of in camera dark frame extraction for the subs. Using Clarke's Methodology the subs actually take 10 minutes to shoot per sub. I use Sequator and/or DSS for stacking. PS RC-Astro filter GradientXterminator to remove gradients. Then RNC stretch or PS to stretch. StarTools or DxO PhotoLab4 Post Processing.

I have been experimenting with Canon's DPP4 RAW processor to process and convert the RAW subs to TIFF prior to stacking with varying results. In an attempt to pull more fine detail from the subs.

Post a link to the raw, linear stack and we'll see if there is more there to be brought out in processing.

Jerry

Could you please comment on the use of ISO400 with 7.8 electrons read noise vs ISO3200 (2.4 electrons) or even ISO6400/ISO12800 (with his filter)?

And the use of in-camera dark frame subtraction, as a misunderstanding of Roger's "On-Sensor Dark Current Suppression Technology", besides doubling imaging time?

-- hide signature --

(Harvey) - Jane, I've been thinking...
(Jane) - Oh, do you want an aspirin?

Alen K Senior Member • Posts: 1,123
Re: Deminishing returns.....

Kemo-sabe wrote:

I am using Clarke's methodology of in camera dark frame extraction for the subs. Using Clarke's Methodology the subs actually take 10 minutes to shoot per sub. I use Sequator and/or DSS for stacking.

Wow, 32+ hours taking exposures? I give you high marks for effort but IMO you needlessly wasted at least half of that time. I would not recommend in-camera dark-frame subtraction (AKA Long-Exposure Noise Reduction or LENR) for anything but single exposures (think nightscape) or very short stacks. There are more efficient methods. And where on Roger Clark's website does he advocate using LENR? I must have missed that.

Now, maybe you have no choice but to shoot in Bortle 9 but personally I would shoot from darker skies. Heck, I live in Bortle 7 but get Bortle 4 by driving less than an hour. It's a no brainer. Use of a dual-band filter is obviously going to help in badly light-polluted skies but it's not a panacea. For one thing, it's only going to be good for emission nebulae.

As an example, here is 12 minutes at f/5.6 from Bortle 4.
https://www.cloudynights.com/gallery/image/83577-an-early-season-m42/
No dark frames. In fact, no calibration frames at all. And no filter. Yeah, it's noisy as heck but it IS only 12 minutes. Imagine what an hour would have given me.

Astrozoid
Astrozoid Contributing Member • Posts: 846
Re: Deminishing returns.....

thinker wrote:

Astrozoid wrote:

Kemo-sabe wrote:

Astrozoid, you are correct in your analysis. The images were shot in Bortle 9 sky and consist of 193 5 minute subs with a Canon FD 800mm f5.6l lens with a Canon 5D MKIII at iso400 incorporating an Optolong L-eNhance dou-band 2" drop-in filter. My guide scope consist of a DIY Canon FD 300mm f4.0l lens coupled to a QHY5LII-C guide camera. The Mount is an iOptron CEM25P with ADM side by side saddle and iPolar scope for polar alignment. I have Pegasus power distribution for the mount/equipment and dual Pegasus focus controllers for remote focusing the primary optics and guide-scope.

I am a newbie to processing astro images. It has has taken me more than 1 year to automate and dial-in the hardware software aspects of astro imaging and learn the nuances of the system.

However, astro processing is another story. I was to embarrassed by my pitiful results to post such a sad image in the light of so many awesome images that I have seen on this forum. Pride is a terrible thing.

My son suggested that I post the image to see if I could get some advice and input on what I am doing wrong in my processing workflow. Or, is this the best that one might expect from Bortle 9 skies with the optical equipment I am using or a combination of both?

I am using Clarke's methodology of in camera dark frame extraction for the subs. Using Clarke's Methodology the subs actually take 10 minutes to shoot per sub. I use Sequator and/or DSS for stacking. PS RC-Astro filter GradientXterminator to remove gradients. Then RNC stretch or PS to stretch. StarTools or DxO PhotoLab4 Post Processing.

I have been experimenting with Canon's DPP4 RAW processor to process and convert the RAW subs to TIFF prior to stacking with varying results. In an attempt to pull more fine detail from the subs.

Post a link to the raw, linear stack and we'll see if there is more there to be brought out in processing.

Jerry

Could you please comment on the use of ISO400 with 7.8 electrons read noise vs ISO3200 (2.4 electrons) or even ISO6400/ISO12800 (with his filter)?

And the use of in-camera dark frame subtraction, as a misunderstanding of Roger's "On-Sensor Dark Current Suppression Technology", besides doubling imaging time?

A lower read noise would only help you if you were shooting under truly dark skies or with a narrowband filter.

Once your exposure is sky-noise limited, and not read-noise limited, you can pretty much totally forget about read noise, it is trivial.

With a really bright sky, you don't have to worry about read noise, so use a low ISO. Just expose so the peak of the histogram on the back of the camera is 1/3 of the way over from the left. Then turn off LENR and shoot as many frames as you can. Dithering will also help tremendously.

Personally, I have never recommended LENR - except for single frames.

Photons are only recorded when the camera is open to the sky. Photons are the only thing that make up the signal half of the signal-to-noise ratio. There is no other way to improve the signal. But you could have recorded twice as many photons and doubled your signal if you had shot lights continusously instead of wasting half of your clearsky time shooting LENR in-camera darks.

So I advise to shoot darks on a cloudy night in your garage under similar temperatures. Then you can shoot darks all night long and create a really good master dark frame.

When you use LENR you get a double whammy hit because not only have you lost half the signal you could have recorded, you are adding noise in each dark that you shoot in the camera. You would add a lot less noise with a good master dark.

LENR is not the same thing as on-sensor dark suppression technology.

And, most importantly, dark suppression technology is not removing thermal signal. It just removes the artifacts, like bright pixels, from thermal (dark) signal. It does not remove the shot noise associated with the dark signal. The only way to deal with this is by collecting more photons to improve the signal half of the signal-to-noise ratio.

Jerry

OP Kemo-sabe Junior Member • Posts: 39
Re: Deminishing returns.....

Astrozoid wrote:

Kemo-sabe wrote:

Astrozoid, you are correct in your analysis. The images were shot in Bortle 9 sky and consist of 193 5 minute subs with a Canon FD 800mm f5.6l lens with a Canon 5D MKIII at iso400 incorporating an Optolong L-eNhance dou-band 2" drop-in filter. My guide scope consist of a DIY Canon FD 300mm f4.0l lens coupled to a QHY5LII-C guide camera. The Mount is an iOptron CEM25P with ADM side by side saddle and iPolar scope for polar alignment. I have Pegasus power distribution for the mount/equipment and dual Pegasus focus controllers for remote focusing the primary optics and guide-scope.

I am a newbie to processing astro images. It has has taken me more than 1 year to automate and dial-in the hardware software aspects of astro imaging and learn the nuances of the system.

However, astro processing is another story. I was to embarrassed by my pitiful results to post such a sad image in the light of so many awesome images that I have seen on this forum. Pride is a terrible thing.

My son suggested that I post the image to see if I could get some advice and input on what I am doing wrong in my processing workflow. Or, is this the best that one might expect from Bortle 9 skies with the optical equipment I am using or a combination of both?

I am using Clarke's methodology of in camera dark frame extraction for the subs. Using Clarke's Methodology the subs actually take 10 minutes to shoot per sub. I use Sequator and/or DSS for stacking. PS RC-Astro filter GradientXterminator to remove gradients. Then RNC stretch or PS to stretch. StarTools or DxO PhotoLab4 Post Processing.

I have been experimenting with Canon's DPP4 RAW processor to process and convert the RAW subs to TIFF prior to stacking with varying results. In an attempt to pull more fine detail from the subs.

Post a link to the raw, linear stack and we'll see if there is more there to be brought out in processing.

Jerry

It will take a little time for me to gather all the subs into 1 file ( used .  When I complete that task I will post a link to my Google Drive.

Keyboard shortcuts:
FForum MMy threads