Help with planetary imaging (can't get good 50 fps frames)

[Sedna]

Hi everyone, right off the bat I know everyone will tell me that I need to be doing this with a webcam. Well, I don't have a webcam yet, so I'm doing my best at planetary imaging using a DSLR. I'm shooting 50 fps and importing the video frames to Photoshop (I have Mac so I can't use these other software packages like Registax).

I cannot seem to get a good stacked image no matter what I do (prune bad frames, stack more frames, sharpen the final image, etc.). I think my frames are just bad. I'm using a Celestron 8SE with a x2 Barlow and an eyepiece projection (24 mm). I've also tried imaging without the eyepiece projection (just the Barlow) and I can't get *anything* (so for Jupiter, no clouds, no details, just a big white glowing blob).

One thing which I suspect is an issue: with my Canon 70D, I don't think I can shoot movies in raw mode. The movie options I have are ALL-I and IPB, which I believe are both compression algorithms ...

Anyway, to show you all what I'm working with, I'm posted a few example frames below (eyepiece projections). These are typical of the frames I stack. I'm also posting a "final" image after I've stacked many frames.

I know that I have a good telescope for planetary, and I know that some people do great planetary imaging with a DSLR, so I'm getting a bit frustrated . Any advice is much appreciated. Thanks!

 

[Stargazer33]

First off they are not the worst I've seen. Some of my images have been truly awful. You are right in your assumption that DSLRs are not very good for taking video files unfortunately.

It looks like either seeing was bad when you were imaging or focus is slightly off. Do you focus with a bahtinov mask?

Also it looks a little over exposed. It is better to be slightly too dark than slightly too bright. If the image is a little dark you can lighten it and bring out detail. If the image captured is too bright or "burnt out" you cannot get the detail back.

HTH

[Sedna]

Hi Bryan, thanks for your reply. I actually enhanced the contrast and brightness on the individual frames before posting here. Not sure if that's typically done before or after stacking. Since this was an eyepiece projection, I definitely don't think the original frames were too bright. I have a dumb question though. If I'm recording video at 50 fps, the sampling rate is clearly 50 Hz, so is each frame always 20 ms exposure, or can they be arbitrarily shorter than this? Sorry, my wife is the photographer, not me

My seeing conditions were good, probably only thing working against me is that Jupiter is low in the sky and setting early now. I didn't have much more luck with Saturn though, which sets much later. I can post my frames of Saturn later.

[michael8554]

For uncompressed video, have you tried APT or BYE to record a X5 LiveView MOV, which gives you a true 1:1 pixel resolution at 1104x736 ?

Michael

[vlaiv]

8 minutes ago, Sedna said:

Hi Bryan, thanks for your reply. I actually enhanced the contrast and brightness on the individual frames before posting here. Not sure if that's typically done before or after stacking. Since this was an eyepiece projection, I definitely don't think the original frames were too bright. I have a dumb question though. If I'm recording video at 50 fps, the sampling rate is clearly 50 Hz, so is each frame always 20 ms exposure, or can they be arbitrarily shorter than this? Sorry, my wife is the photographer, not me

My seeing conditions were good, probably only thing working against me is that Jupiter is low in the sky and setting early now. I didn't have much more luck with Saturn though, which sets much later. I can post my frames of Saturn later.

Exposure can be arbitrary long and only relation to FPS is that if you have longer exposure you will not be able to achieve target FPS (so longer than 20ms - frame rate drops from 50fps).

How did you stack and process your images?

You mentioned using Mac, I'm not sure what sort of software is available on Mac, but usual frame work for doing planetary imaging would be: Pipp -> AS!3 -> Registax (or AstraImage) -> PS/Gimp for final touch up. Pipp serves to calibrate movie frames, and has quite a few useful tools to pre process your movies (hence the name - Planetary Imaging Pre Processor). You can skip this step if you don't calibrate your movies.

Next would be AS!3 for stacking (Registax does it also, but I've found AS!3 much better), and there are probably couple of other pieces of software that can do planetary stacking.

Regisax is useful for wavelet sharpening, but AstraImage can do it also.

[Sedna]

37 minutes ago, michael8554 said:

For uncompressed video, have you tried APT or BYE to record a X5 LiveView MOV, which gives you a true 1:1 pixel resolution at 1104x736 ?

Michael

That sounds great, but I've never seen that in the video menu on my camera (Canon 70D). Can you give me a little more information? Thanks!

[Sedna]

32 minutes ago, vlaiv said:

Exposure can be arbitrary long and only relation to FPS is that if you have longer exposure you will not be able to achieve target FPS (so longer than 20ms - frame rate drops from 50fps).

How did you stack and process your images?

You mentioned using Mac, I'm not sure what sort of software is available on Mac, but usual frame work for doing planetary imaging would be: Pipp -> AS!3 -> Registax (or AstraImage) -> PS/Gimp for final touch up. Pipp serves to calibrate movie frames, and has quite a few useful tools to pre process your movies (hence the name - Planetary Imaging Pre Processor). You can skip this step if you don't calibrate your movies.

Next would be AS!3 for stacking (Registax does it also, but I've found AS!3 much better), and there are probably couple of other pieces of software that can do planetary stacking.

Regisax is useful for wavelet sharpening, but AstraImage can do it also.

Hi Vlaiv, thanks for your message. I stacked my frames in Photoshop (there is an option to import frames from video). I'm wondering if most people use freeware like Registax because it's freeware, or if the stacking algorithm is actually better for AP?

[Buzzard75]

19 minutes ago, Sedna said:

That sounds great, but I've never seen that in the video menu on my camera (Canon 70D). Can you give me a little more information? Thanks!

APT (Astrophotography Tool) and BYE (BackyardEOS) are capture software that you run on your computer. I use them myself. I do not believe there is a native Mac version of either.

[Sedna]

1 minute ago, Buzzard75 said:

APT (Astrophotography Tool) and BYE (BackyardEOS) are capture software that you run on your computer. I use them myself. I do not believe there is a native Mac version of either.

Got it, thanks. Do they allow for a higher sampling rate (than 50 fps)?

[Buzzard75]

16 minutes ago, Sedna said:

Got it, thanks. Do they allow for a higher sampling rate (than 50 fps)?

I think it's camera and connection dependent. I get around 30 with my 750D. As was mentioned, the software is just better at capturing uncompressed video than a DSLR is.

[michael8554]

Planetary Imagers do tend to go for high frame rates for a number of reasons including getting lots of good frames before Jupiter has rotated much.

Michael 

[vlaiv]

10 hours ago, Sedna said:

Hi Vlaiv, thanks for your message. I stacked my frames in Photoshop (there is an option to import frames from video). I'm wondering if most people use freeware like Registax because it's freeware, or if the stacking algorithm is actually better for AP?

There is quite a bit of difference in just overlapping frames and taking average and using dedicated Planetary stacker to produce planetary images from movie. It is definitively better to use later. Such programs to an extent manage to sort out deformations to each frame and do sort of "micro alignment" of features when they stack subs (they use alignment points over planet / surface for this). Result is much "cleaner" and less blurry stack.

Due to seeing and this deformation of frames (just look at video and you will see planet jump a bit around and gets stretched / squeezed - all kind of wobbles and blurring) - result of stack will always be blurry and high frequency components of the image will be attenuated (that is what blurry means in "math" terms), so you need a way of restoring those high frequencies (or sharpening). Best suited sharpening method for planetary images is wavelet based decomposition and sharpening (rather than simple unsharp mask or similar) - this is where Registax comes into play (you can also try deconvolution - both methods work in the same way - restoring attenuated frequencies, but do it in mathematically different way).

[GuLinux]

You have indeed a nice camera, but as others said, it's not exactly the best way to get planetary images. Among other advantages, a dedicated planetary camera has a massive advantage in allowing to save raw video frames (in SER formats) rather than encoded (and lossy) videos.

But apart from that, I really think one of the worst thing in your setup is the optical path. As you mentioned, you have a very good telescope, but putting in between both a barlow and an eyepiece means having too many elements, and eyepiece projection is not a very reliable technique, as it can create more troubles than anything else.

You want only the telescope, and a *good* barlow. Here you can find a decent guide on what focal length you might want to obtain (and therefore, wether you need a barlow, and with which multiplying factor): http://astronomy.tools/calculators/ccd_suitability . It's not exactly suited for planetary imaging, but it gives you an idea.

As for why you can't get an image at all without the eyepiece, I highly suspect it's just a pointing issue. If the planet is not perfectly centered in the camera field of view, you might just miss it.  If you point your object with an eyepiece, then remove the eyepiece and put on the camera, the weight difference between the two setups might shift the mount. In this case, you need to use either a flip mirror, or a (very well aligned) guide scope.

[Stub Mandrel]

15 hours ago, Sedna said:

I've also tried imaging without the eyepiece projection (just the Barlow) and I can't get *anything* (so for Jupiter, no clouds, no details, just a big white glowing blob).

Sounds like the exposure is too long.

With jupiter one strategy is to use a long exposure and get focus on the moons, then adjust to get best exposure for the planet.

My feeling is that even a cheap webcam is likely to give results as good as a DSLR on planets, because you can use something like Sharpcap for tuning focus and capture. Even the most basic dedicated planetary camera is a step up again.

[Sedna]

On 02/10/2018 at 03:10, vlaiv said:

There is quite a bit of difference in just overlapping frames and taking average and using dedicated Planetary stacker to produce planetary images from movie. It is definitively better to use later. Such programs to an extent manage to sort out deformations to each frame and do sort of "micro alignment" of features when they stack subs (they use alignment points over planet / surface for this). Result is much "cleaner" and less blurry stack.

Due to seeing and this deformation of frames (just look at video and you will see planet jump a bit around and gets stretched / squeezed - all kind of wobbles and blurring) - result of stack will always be blurry and high frequency components of the image will be attenuated (that is what blurry means in "math" terms), so you need a way of restoring those high frequencies (or sharpening). Best suited sharpening method for planetary images is wavelet based decomposition and sharpening (rather than simple unsharp mask or similar) - this is where Registax comes into play (you can also try deconvolution - both methods work in the same way - restoring attenuated frequencies, but do it in mathematically different way).

Thanks, for helpful. Now I just need to find one of these planetary stacker programs for Mac.

[Sedna]

On 02/10/2018 at 03:23, GuLinux said:

You have indeed a nice camera, but as others said, it's not exactly the best way to get planetary images. Among other advantages, a dedicated planetary camera has a massive advantage in allowing to save raw video frames (in SER formats) rather than encoded (and lossy) videos.

But apart from that, I really think one of the worst thing in your setup is the optical path. As you mentioned, you have a very good telescope, but putting in between both a barlow and an eyepiece means having too many elements, and eyepiece projection is not a very reliable technique, as it can create more troubles than anything else.

You want only the telescope, and a *good* barlow. Here you can find a decent guide on what focal length you might want to obtain (and therefore, wether you need a barlow, and with which multiplying factor): http://astronomy.tools/calculators/ccd_suitability . It's not exactly suited for planetary imaging, but it gives you an idea.

As for why you can't get an image at all without the eyepiece, I highly suspect it's just a pointing issue. If the planet is not perfectly centered in the camera field of view, you might just miss it.  If you point your object with an eyepiece, then remove the eyepiece and put on the camera, the weight difference between the two setups might shift the mount. In this case, you need to use either a flip mirror, or a (very well aligned) guide scope.

Hi GuLinux, thanks for the thorough answer. I have a very good mount (EQ mount which should be overkill for planetary) and I actually don't miss the planet without the eyepiece, I just always get an overexposed image, e.g., Jupiter is clearly there, but just a glowing white orb, no detail. I will try again with DSLR, but I hear everyone's suggestion that I need better equipment.

[Stub Mandrel]

46 minutes ago, Sedna said:

I just always get an overexposed image, e.g., Jupiter is clearly there, but just a glowing white orb, no detail.

Jupiter is very bright. I usually use about a 23ms exposure, and that's with an x3 barlow cutting the light down.