How to photograph Jupiter?

Started Apr 25, 2016 | Discussions
Juggernaut122
Juggernaut122 Contributing Member • Posts: 698
How to photograph Jupiter?

Hi guys,

Can someone guide me to the proper technique to take good pictures of Jupiter?

I find if I do long exposures (slower than about 1/5 second), then the planet gets over-exposed and shows no detail.  If I do fast enough images to avoid over-exposure - say  1/15 of a second, then the image shows detail but it is very dark.  Increasing the ISO just seems to cause more over-exposure again.

I seem to be missing some fundamental technique to get an image with detail which is bright but not over-exposed.

Do I need a special filter?
Or must this be done with stacking software to get these results?

Thanks!
Juggernaut

 Juggernaut122's gear list:Juggernaut122's gear list
Olympus Tough TG-4 Canon G7 X II Nikon D7200 Nikon AF Nikkor 50mm f/1.8D Sigma 17-50mm F2.8 EX DC OS HSM +1 more
DJ13 New Member • Posts: 17
Re: How to photograph Jupiter?

Are you using the Nikon in your profile with a telescope? I'm trying to do pretty much the same thing with my D5200 and Celestron 130SLT.  The best results I got a few years ago was to stack video frames with Registax.  Here is my best results from back then using a very basic eyepiece camera (Celestron Neximage).  I'm hoping the Nikon can do better but could also use some tips; such as should video be recorded at maximum resolution or lower?

 DJ13's gear list:DJ13's gear list
Nikon D5200 Nikon AF-S DX Nikkor 35mm F1.8G Tamron AF 55-200mm F/4-5.6 Di II LD Macro Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II
Juggernaut122
OP Juggernaut122 Contributing Member • Posts: 698
Re: How to photograph Jupiter?

Hi DJ13,

Yes, I am taking photos with the D5100 (but with an goto Dobsonian instead of an equilateral mount - so not a great setup).

I find that taking them afocally using the f/1.8 50 mm I have works better than focally directly through the scope without a lens, but the results are still pretty poor.
I live in a heavily light polluted area.

Last night, I think for the first time I imaged the red dot (I think it is in the right part of the top band, but there also seems to be something in the middle of the bottom band - but I actually suspect that my barlow has a defect in it). I don't see the moons at all in the photos.

For reasons I cannot explain, I have been told that lower resolution works better when stacking, but I have not tried using lower resolution yet - I am actually going to stack Jupiter for the first time (hopefully today).

 Juggernaut122's gear list:Juggernaut122's gear list
Olympus Tough TG-4 Canon G7 X II Nikon D7200 Nikon AF Nikkor 50mm f/1.8D Sigma 17-50mm F2.8 EX DC OS HSM +1 more
Juggernaut122
OP Juggernaut122 Contributing Member • Posts: 698
Re: How to photograph Jupiter?
3

DJ13 wrote:

Ok. Looks like I just improved my image quality noticeably.

This is made from a 1 minute movie on full resolution (afocally through the 35 mm f/1.8). Changed to avi with PIPP and then stacked with registax 6.

Biggest difference was when I adjusted the gamma in the final image - all of a sudden all of the hidden details came out and showed up in color. I adjusted the wavelets before I adjusted the gamma (I think).

My order of adjustment was:

1) RGB align estimate

2) adjusted the wavelets

3) adjusted the gamma (biggest difference)

4) denoised total and the "bright side" - all completely to the right (removed the ring that had formed around the planet).

5) went back to wavelets and maxed them all out to the right.

(I will also say that I really have no idea what I am doing - just trial and error based on a video I saw on  youtube about how to change the settings)

 Juggernaut122's gear list:Juggernaut122's gear list
Olympus Tough TG-4 Canon G7 X II Nikon D7200 Nikon AF Nikkor 50mm f/1.8D Sigma 17-50mm F2.8 EX DC OS HSM +1 more
DJ13 New Member • Posts: 17
Re: How to photograph Jupiter?

Nice! I'll have to try PIPP again, I was using virtualDUB to convert my 1 min video (640x480) to a 1GB avi file which Registax has trouble with.

 DJ13's gear list:DJ13's gear list
Nikon D5200 Nikon AF-S DX Nikkor 35mm F1.8G Tamron AF 55-200mm F/4-5.6 Di II LD Macro Nikon AF-S DX Nikkor 18-55mm f/3.5-5.6G VR II
fzrTom Senior Member • Posts: 2,285
Re: How to photograph Jupiter?

Juggernaut122 wrote:

DJ13 wrote:

Ok. Looks like I just improved my image quality noticeably.

This is made from a 1 minute movie on full resolution (afocally through the 35 mm f/1.8). Changed to avi with PIPP and then stacked with registax 6.

Biggest difference was when I adjusted the gamma in the final image - all of a sudden all of the hidden details came out and showed up in color. I adjusted the wavelets before I adjusted the gamma (I think).

My order of adjustment was:

1) RGB align estimate

2) adjusted the wavelets

3) adjusted the gamma (biggest difference)

4) denoised total and the "bright side" - all completely to the right (removed the ring that had formed around the planet).

5) went back to wavelets and maxed them all out to the right.

(I will also say that I really have no idea what I am doing - just trial and error based on a video I saw on youtube about how to change the settings)

Did you try to select RESAMPLING Lanczos 2 ?

-- hide signature --

Franck

 fzrTom's gear list:fzrTom's gear list
Sony a7S Sony 50mm F1.4 Tamron SP 70-300mm F/4-5.6 Di USD Tamron SP 24-70mm F2.8 Di VC USD Tamron SP 150-600mm F/5-6.3 Di VC USD +3 more
Juggernaut122
OP Juggernaut122 Contributing Member • Posts: 698
Re: How to photograph Jupiter?

fzrTom wrote:

Did you try to select RESAMPLING Lanczos 2 ?

Hi Franck,

I don't see that option in Registax 6.

Where can I find it, and what does it do?

Thanks!
Juggernaut

 Juggernaut122's gear list:Juggernaut122's gear list
Olympus Tough TG-4 Canon G7 X II Nikon D7200 Nikon AF Nikkor 50mm f/1.8D Sigma 17-50mm F2.8 EX DC OS HSM +1 more
Waecky Regular Member • Posts: 180
Re: How to photograph Jupiter?

Juggernaut122 wrote:

Hi guys,

Can someone guide me to the proper technique to take good pictures of Jupiter?

I find if I do long exposures (slower than about 1/5 second), then the planet gets over-exposed and shows no dSoletail. If I do fast enough images to avoid over-exposure - say 1/15 of a second, then the image shows detail but it is very dark. Increasing the ISO just seems to cause more over-exposure again.

I seem to be missing some fundamental technique to get an image with detail which is bright but not over-exposed.

Do I need a special filter?
Or must this be done with stacking software to get these results?

Thanks!
Juggernaut

There's the Solar System Imaging Forum on cn or Christophe Pellier's Planetary Astronomy website with lots of info about this topic.

To check out when the GRS (Great Red Spot) is culminating you may use calsky.

Shoot as many pictures as possible during about 2.5 minutes, which means movie mode. For stacking and selecting the best subs you may also use Autostakkert.

For best results, consider using a dedicated planet cam ;- )

-- hide signature --

Peter

 Waecky's gear list:Waecky's gear list
Canon EOS 500D Canon EOS 6D Tamron AF 18-270mm F/3.5-6.3 Di II VC LD Aspherical (IF) MACRO Canon EF 16-35mm F2.8L USM Canon EF 24-70mm f/2.8L USM +3 more
fzrTom Senior Member • Posts: 2,285
Re: How to photograph Jupiter?

Juggernaut122 wrote:

fzrTom wrote:

Did you try to select RESAMPLING Lanczos 2 ?

Hi Franck,

I don't see that option in Registax 6.

Where can I find it, and what does it do?

Thanks!
Juggernaut

The resampling checkbox in the stack menu

One of my friend has a C8 with a dedicated camera and with this option checked the results are amazing. So as your photo is already nice, maybe you can increase again a bit the IQ. What does it do ? It simply multiply the size by 2 and allow to have more resolution.

-- hide signature --

Franck

 fzrTom's gear list:fzrTom's gear list
Sony a7S Sony 50mm F1.4 Tamron SP 70-300mm F/4-5.6 Di USD Tamron SP 24-70mm F2.8 Di VC USD Tamron SP 150-600mm F/5-6.3 Di VC USD +3 more
Allien
Allien Contributing Member • Posts: 516
Re: How to photograph Jupiter?

Juggernaut122 wrote:

DJ13 wrote:

Ok. Looks like I just improved my image quality noticeably.

This is made from a 1 minute movie on full resolution (afocally through the 35 mm f/1.8). Changed to avi with PIPP and then stacked with registax 6.

Biggest difference was when I adjusted the gamma in the final image - all of a sudden all of the hidden details came out and showed up in color. I adjusted the wavelets before I adjusted the gamma (I think).

My order of adjustment was:

1) RGB align estimate

2) adjusted the wavelets

3) adjusted the gamma (biggest difference)

4) denoised total and the "bright side" - all completely to the right (removed the ring that had formed around the planet).

5) went back to wavelets and maxed them all out to the right.

(I will also say that I really have no idea what I am doing - just trial and error based on a video I saw on youtube about how to change the settings)

This is a pretty good result really.

Earlier you mentioned or were asking about "lower resolution" cameras working better for planetary images, that is not accurate per se. What helps out is the small pixel size on some of the video cameras compared to the pixel size of a DSLR camera back in the day.

Now that we have 20 megapixel DSLR cameras that  are commonplace, chances are your D5100 has as small, or smaller pixels than the typically recommended "planet cams"

Whether or not you can shoot video with your particular DSLR at that resolution is another story, but single frames no problem.

The problem with video on some DSLR cameras is that it uses the whole sensor and essentially down-samples the resolution of the sensor.  I think the most successful approach with the DSLR are those who are able to channel their zoomed in "live view" image on the back of their camera to their computer to record a video. I assume these are Cannon cameras, maybe you can do it with Nikon too, I don't know.

Any how your afocal approach has merit, as I suspected. I have often considered trying this myself one time.  It's fiddly but at the end of the day the goal is to get the planet to cover as many pixels as possible.

Oh and one more thing, I find that typically the darker images of Jupiter to be preferable as they look the best IMO once they are stretched out in post processing. If you can see the moons, you will have over exposed the details, unless of course you have a plan to do some sort of HDR stack in Photoshop with different exposures.

Cheers.

Astrozoid
Astrozoid Contributing Member • Posts: 649
Re: How to photograph Jupiter?
4

Light pollution has nothing to do with high resolution planetary imaging. That's all about the seeing.

You need:

1. Good focus.

2. Excellent collimation.

3. Scope mirror completely cooled down.

4. Optimum sampling, which means matching the resolution that the seeing allows to the resolution that your scope is capable of, to the pixel size in your camera. You want to shoot around f/20 for mediocre seeing, and f/30 for outstanding excellent seeing.

You don't just want to blow the image up to get Jupiter as large as possible. Once you go past the point of optimum sampling, you are just getting empty magnification that doesn't record any more detail, and just makes your exposures longer for no reason.

5. "Lucky Imaging" Video recording at 1:1 pixel resolution. This means finding a way to record a small section of the sensor usually by recording Live View. You might be able to get closest with BackYardNikon. Recording video with the "high-definition" normal built-in video will trash the resolution when it downsamples the original resolution of the sensor.

You can get 1:1 pixel resolution if you shoot single still frames, but one frame is not going to look good. You need to stack hundreds or thousands of frames.

Adjust the ISO and exposure so the planet is correctly exposed in the video. This is by trial and error.

On some older Nikons, you can't control the exposure that Live View presents, it just tries to autoexpose the image, and if you can't change the brightness of Live View. With a small bright planet on a large black background, it's going to overexpose Jupiter, and basically you can't do anything about it if you can't adjust the Live View exposure, so you are basically screwed for high-res planetary imaging with some Nikons.

6. Use AutoStakkert! 2 to align and stack about 3 minutes of video. It's much better than RegiStax for stacking. AutoStakkert! 2 will also do some basic sharpening which I find works well. If you want to see if you can get a little bit more, then use RegiStax wavelet sharpening to the stack out of AutoStakkert!

I have a book on high-resolution planetary imaging with DSLRs if you are interested. You can find it on my home page at www.astropix.com

Here's what you can do with a DSLR and Lucky Video imaging with a DSLR:

Jerry

RustierOne
RustierOne Veteran Member • Posts: 4,338
Re: How to photograph Jupiter?

Astrozoid wrote:

Light pollution has nothing to do with high resolution planetary imaging. That's all about the seeing.

You need:

1. Good focus.

2. Excellent collimation.

3. Scope mirror completely cooled down.

4. Optimum sampling, which means matching the resolution that the seeing allows to the resolution that your scope is capable of, to the pixel size in your camera. You want to shoot around f/20 for mediocre seeing, and f/30 for outstanding excellent seeing.

You don't just want to blow the image up to get Jupiter as large as possible. Once you go past the point of optimum sampling, you are just getting empty magnification that doesn't record any more detail, and just makes your exposures longer for no reason.

5. "Lucky Imaging" Video recording at 1:1 pixel resolution. This means finding a way to record a small section of the sensor usually by recording Live View. You might be able to get closest with BackYardNikon. Recording video with the "high-definition" normal built-in video will trash the resolution when it downsamples the original resolution of the sensor.

You can get 1:1 pixel resolution if you shoot single still frames, but one frame is not going to look good. You need to stack hundreds or thousands of frames.

Adjust the ISO and exposure so the planet is correctly exposed in the video. This is by trial and error.

On some older Nikons, you can't control the exposure that Live View presents, it just tries to autoexpose the image, and if you can't change the brightness of Live View. With a small bright planet on a large black background, it's going to overexpose Jupiter, and basically you can't do anything about it if you can't adjust the Live View exposure, so you are basically screwed for high-res planetary imaging with some Nikons.

6. Use AutoStakkert! 2 to align and stack about 3 minutes of video. It's much better than RegiStax for stacking. AutoStakkert! 2 will also do some basic sharpening which I find works well. If you want to see if you can get a little bit more, then use RegiStax wavelet sharpening to the stack out of AutoStakkert!

I have a book on high-resolution planetary imaging with DSLRs if you are interested. You can find it on my home page at www.astropix.com

Here's what you can do with a DSLR and Lucky Video imaging with a DSLR:

Great advice, Jerry. I'm looking forward to learning how to use Autostakkert!2. Since I haven't yet received my new ZWO ASI290MC color camera, I've just been using Registax for Moon imaging. I'm also hoping to get the ZWO atmospheric dispersion corrector in time for capturing Saturn and Mars next month.

-- hide signature --

Best Regards,
Russ

 RustierOne's gear list:RustierOne's gear list
Sony Alpha NEX-5N Sony Alpha NEX-7 Sigma 19mm F2.8 EX DN Samyang 8mm F2.8 UMC Fisheye Sony E 35mm F1.8 OSS +4 more
fzrTom Senior Member • Posts: 2,285
Re: How to photograph Jupiter?

Amazing photo !

Thanks for the advice.

-- hide signature --

Franck

 fzrTom's gear list:fzrTom's gear list
Sony a7S Sony 50mm F1.4 Tamron SP 70-300mm F/4-5.6 Di USD Tamron SP 24-70mm F2.8 Di VC USD Tamron SP 150-600mm F/5-6.3 Di VC USD +3 more
rnclark Veteran Member • Posts: 3,937
Re: How to photograph Jupiter?

Astrozoid wrote:

4. Optimum sampling, which means matching the resolution that the seeing allows to the resolution that your scope is capable of, to the pixel size in your camera. You want to shoot around f/20 for mediocre seeing, and f/30 for outstanding excellent seeing.

You don't just want to blow the image up to get Jupiter as large as possible. Once you go past the point of optimum sampling, you are just getting empty magnification that doesn't record any more detail, and just makes your exposures longer for no reason.

Hi Jerry,

Optimum sampling is 2 to 3 times Dawes limit. Dawes limit is close to 0% MTF. What f-ratio that comes out to is dependent on the pixel size of the sensor. With today's smal pixels, lower f-numbers may be used. I posted the following here some time ago:

Example, 8-inch telescope (20.3 cm diameter) with a canon 7D2 with 4.09 micron pixels (0.00409 mm) and 2000 mm focal length:

First compute plate scale:

plate scale = 206265 * pixel size / focal length

Pixel size and focal length should be in the same units. 206265 is the number of arc-seconds in one radian.

plate scale = 206265 * 0.00409 / 2000 = 0.42 arc-seconds.

Dawes limit is 4.56/D, D=aperture diameter in inches, or 11.6/D, D in cm.

8-inch, 20.3 cm telescope = 11.6/20.3 = 0.57 arc-seconds.

We really want sampling at 0.57/2 to 0.57/3 = 0.28 to 0.19 arc-second.

At 2000 mm we have 0.42 arc second.

To get to 0.28 we need 0.42/0.28 = 1.5 times more focal length, or 1.5*2000 = 3000 mm (f/15).

To get to 0.19 we need 0.42/0.19 = 2.2 times more focal length, or 2.2*2000 = 4400 mm (f/22).

If you turn on 1080p video that uses the full frame, then 5472 pixel width of the 7d2 sensor gets down sampled to 1920 pixels, so video pixels become 5472/1920 = 2.85 times larger, and focal lengths need to be increased 2.85x to reach the same sampling, or focal lengths of 8550 mm (f/42) to 12540 mm f/61).

So it all comes down to pixel size.

I'll add one more to your list:

7. Excellent seeing. Where I live (Colorado) the turbulence of airflow over the mountains prevents obtaining images like your Jupiter image 99.999% of the time. Seeing as good as 2-arc-seconds is very unusual.  On Mauna Kea, Hawaii, I have seen 0.05 arc-second seeing.

Superb Jupiter image! What telescope?

Roger

Here's what you can do with a DSLR and Lucky Video imaging with a DSLR:

Astrozoid
Astrozoid Contributing Member • Posts: 649
Re: How to photograph Jupiter?
1

rnclark wrote:

Astrozoid wrote:

4. Optimum sampling, which means matching the resolution that the seeing allows to the resolution that your scope is capable of, to the pixel size in your camera. You want to shoot around f/20 for mediocre seeing, and f/30 for outstanding excellent seeing.

You don't just want to blow the image up to get Jupiter as large as possible. Once you go past the point of optimum sampling, you are just getting empty magnification that doesn't record any more detail, and just makes your exposures longer for no reason.

Hi Jerry,

Optimum sampling is 2 to 3 times Dawes limit.

Superb Jupiter image! What telescope?

Roger

Hi Roger,

I have the formula and a calculator in my book for optimum sampling for high-resolution planetary imaging that incorporates everything including pixel size.

For the highest resolution work, with excellent seeing and scopes large enough to resolve down to sub-arcsecond seeing, you really need more than just the 2x or 3x Nyquist, you really need more like 3.5x because you need to do the calculation based on the diagonal of the pixel.

This Jupiter image was shot with a C11 Edge working at f/20, and Canon 550D with Live View captured with EOS Movie Record in 6 out of 10 seeing.

Theoretically, I should have been at about f/26, but I don't usually get seeing good enough for that and I didn't have an easy way to change the Barlow spacing that night.

Here's a video animation of Jupiter rotating that I shot recently with my 130mm reftactor when the seeing wasn't very good:

https://www.youtube.com/watch?v=XQqz1fgix_s

I tried to embed the YouTube video here with the DPReview insert video tool, but it didn't seem to work.

Jerry

Astrozoid
Astrozoid Contributing Member • Posts: 649
Re: How to photograph Jupiter?
2

rnclark wrote:

If you turn on 1080p video that uses the full frame, then 5472 pixel width of the 7d2 sensor gets down sampled to 1920 pixels, so video pixels become 5472/1920 = 2.85 times larger, and focal lengths need to be increased 2.85x to reach the same sampling, or focal lengths of 8550 mm (f/42) to 12540 mm f/61).

Hi Roger,

I strongly advise against trying to use 1080p "high-definition" video because that absolutely trashes the quality when it goes from full sensor resolution to 1080p resolution.

There is some question about how 1080p video is arrived at. There's speculation that it's a combination of decimation and resampling. That's what trashes the resolution.

You need to get 1:1 pixel resolution off the sensor to record your video for high-resolution planetary imaging. Usually you do this by capturing Live View at 5x with something like BackYardEOS or EOS Movie Record or AstroPhotography Tool, but the absolute best way with a DSLR is to use a Canon 550D, 60D or 60Da and use the special high-resolution planetary mode called "640 x 480 Movie Crop Mode" which gives exactly 1:1 pixel resolution at 60 frames per second, and you can record to the memory card in the camera, you don't even need an external computer, like you do to capture Live View at 5x.

I think the Sony A7s shoot video at 1:1 pixel resolution by simple cropping of the original sensor resolution, but none of the Canons or Nikon's do in their native video modes except as mentioned in the paragraph above.

Also, if you have to push the focal ratio up to f/61, your exposures are going to be ridiculously long. And that is totally unnecessary if you just shoot the video at 1:1 pixel resolution and not at 1080p.

Jerry

rnclark Veteran Member • Posts: 3,937
Re: How to photograph Jupiter?

Astrozoid wrote:

rnclark wrote:

If you turn on 1080p video that uses the full frame, then 5472 pixel width of the 7d2 sensor gets down sampled to 1920 pixels, so video pixels become 5472/1920 = 2.85 times larger, and focal lengths need to be increased 2.85x to reach the same sampling, or focal lengths of 8550 mm (f/42) to 12540 mm f/61).

Hi Roger,

I strongly advise against trying to use 1080p "high-definition" video because that absolutely trashes the quality when it goes from full sensor resolution to 1080p resolution.

There is some question about how 1080p video is arrived at. There's speculation that it's a combination of decimation and resampling. That's what trashes the resolution.

You need to get 1:1 pixel resolution off the sensor to record your video for high-resolution planetary imaging. Usually you do this by capturing Live View at 5x with something like BackYardEOS or EOS Movie Record or AstroPhotography Tool, but the absolute best way with a DSLR is to use a Canon 550D, 60D or 60Da and use the special high-resolution planetary mode called "640 x 480 Movie Crop Mode" which gives exactly 1:1 pixel resolution at 60 frames per second, and you can record to the memory card in the camera, you don't even need an external computer, like you do to capture Live View at 5x.

I think the Sony A7s shoot video at 1:1 pixel resolution by simple cropping of the original sensor resolution, but none of the Canons or Nikon's do in their native video modes except as mentioned in the paragraph above.

Also, if you have to push the focal ratio up to f/61, your exposures are going to be ridiculously long. And that is totally unnecessary if you just shoot the video at 1:1 pixel resolution and not at 1080p.

Jerry

Hi Jerry,

Seems to me it depends a lot on the video processor and what it does to down sample the sensor.  I just checked my 7D2 on a Moon video I did last year and comparing the same physical area on the Moon, I get within 1% of the same S/N whether video or still.

Also, shooting at 1:1 means one is susceptible to the errors in Bayer deconvolution and effects from the blur filter.  Seems like a 2x2 bin would be better for resolution than 1:1.

Regardless of f-ratio, say f/20 at 1:1 pixel or f/60 at 3:1 pixel (3x3 bin), you get the same number of photons per output pixel with the same exposure time.

But I agree, it is certainly a lot easier to get to f/20 than f/60.

Roger

fzrTom Senior Member • Posts: 2,285
Thank you Jerry !

Thank you Jerry !

Really thanks a lot :

Well I didnt know Autostakkert and I have just tested it tonight ... amazing.

In comparison with registax I think that it is more difficult to use and it doesnt have all the nice toolbox that Registax has BUT it is faster and the stacking is AMAZING (so accurate !). So it can be used with Registax (Stacking with autostakkert and PP with Registax and/or GIMP/Photoshop).

Here 1 raw : Sony A58 + Tamron 150-600 ... hum nice but ...

Here the best I could get from Registax and Gimp (size 2x) ; whouaou, I was so happy with this result

Here my first try with Autostakkert : the GRS is clearly sharper, and the stacking is so accurate ! With Registax I begin to see some details, with Autostakkert they are clearly visible. This is incredible.

An animated GIF to show the différence : first image = Autostakkert and second = Registax. As you can see the Registax photo looks blured in comparison with the Autostakkert one. I was so happy with my first PP but this one is clearly better.

-- hide signature --

Franck

 fzrTom's gear list:fzrTom's gear list
Sony a7S Sony 50mm F1.4 Tamron SP 70-300mm F/4-5.6 Di USD Tamron SP 24-70mm F2.8 Di VC USD Tamron SP 150-600mm F/5-6.3 Di VC USD +3 more
Astrozoid
Astrozoid Contributing Member • Posts: 649
Re: How to photograph Jupiter?

rnclark wrote:

Seems to me it depends a lot on the video processor and what it does to down sample the sensor. I just checked my 7D2 on a Moon video I did last year and comparing the same physical area on the Moon, I get within 1% of the same S/N whether video or still.

Hi Roger,

You got within 1% of the resolution from a 1080p video as from a full-frame still? That is hard to believe.

I've done extensive tests on this and the resolution in 1080p video is absolute trash compared to a full-res still.

But, S/N is not the same thing as resolution is it?

Also, shooting at 1:1 means one is susceptible to the errors in Bayer deconvolution and effects from the blur filter. Seems like a 2x2 bin would be better for resolution than 1:1.

It's not true 2x2 binning, it's downsampling ,or decimation and downsampling. I'm pretty sure decimation is involved, and I don't think you can recover lost resolution from that. But you are the planetary scientist so I could be wrong.

If the camera actually took the 4 pixel Bayer group and created one super pixel, like DeepSkyStacker, that might be good if you don't mind shooting at f/60, but if the camera is throwing away every other line it seems like it would have to debayer each individual pixel first, so I don't see how software binning is going to help with debayerization errors.

Jerry

mac01
mac01 Contributing Member • Posts: 730
Re: How to photograph Jupiter?
1

Astrozoid wrote:

Light pollution has nothing to do with high resolution planetary imaging. That's all about the seeing.

You need:

1. Good focus.

2. Excellent collimation.

3. Scope mirror completely cooled down.

4. Optimum sampling, which means matching the resolution that the seeing allows to the resolution that your scope is capable of, to the pixel size in your camera. You want to shoot around f/20 for mediocre seeing, and f/30 for outstanding excellent seeing.

You don't just want to blow the image up to get Jupiter as large as possible. Once you go past the point of optimum sampling, you are just getting empty magnification that doesn't record any more detail, and just makes your exposures longer for no reason.

5. "Lucky Imaging" Video recording at 1:1 pixel resolution. This means finding a way to record a small section of the sensor usually by recording Live View. You might be able to get closest with BackYardNikon. Recording video with the "high-definition" normal built-in video will trash the resolution when it downsamples the original resolution of the sensor.

You can get 1:1 pixel resolution if you shoot single still frames, but one frame is not going to look good. You need to stack hundreds or thousands of frames.

Adjust the ISO and exposure so the planet is correctly exposed in the video. This is by trial and error.

On some older Nikons, you can't control the exposure that Live View presents, it just tries to autoexpose the image, and if you can't change the brightness of Live View. With a small bright planet on a large black background, it's going to overexpose Jupiter, and basically you can't do anything about it if you can't adjust the Live View exposure, so you are basically screwed for high-res planetary imaging with some Nikons.

6. Use AutoStakkert! 2 to align and stack about 3 minutes of video. It's much better than RegiStax for stacking. AutoStakkert! 2 will also do some basic sharpening which I find works well. If you want to see if you can get a little bit more, then use RegiStax wavelet sharpening to the stack out of AutoStakkert!

I have a book on high-resolution planetary imaging with DSLRs if you are interested. You can find it on my home page at www.astropix.com

Here's what you can do with a DSLR and Lucky Video imaging with a DSLR:

Jerry

Hey Jerry, thanks for all the info and all I can say is wow...

I'm switching gears from NB deep sky to planetary and trying to build some skill for the upcoming Mars and Saturn imaging sessions next month with a little Orion ccd and C8. Do you use a stepper motor for focusing with a barlow or is focusing planetary done manually?

Thanks again

Chris

 mac01's gear list:mac01's gear list
Olympus E-620 Olympus PEN E-PL5 Olympus Zuiko Digital ED 50mm 1:2.0 Macro Olympus Zuiko Digital ED 14-42mm 1:3.5-5.6 Panasonic Lumix G 20mm F1.7 ASPH +5 more
Keyboard shortcuts:
FForum MMy threads