Jump to content


Photo

Jup DSLR - What am I doing wrong?

  • Please log in to reply
16 replies to this topic

#1 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5459
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 29 October 2012 - 08:14 AM

A little background - I played around with a 7MP point n shoot afocal for about a year, thought I had learned Registax, seeing and focus. Still learning new stuff about Registax.
Then I got a Canon 1000d(XS) and the 1st vid I took of Saturn was better then all my point n shoots, but it was still grainy and a little blurry. I was really happy to get a decent image scale though!
Well now into the Jup season and still grainy and blurry. I see better DSLR pics and what am I missing?

I use a cheap 2x barlow with a steadypix adapter in front of the camera sensor. If I remove the barrel of the barlow I can move the lens really close to the sensor and get less mag, which gives a brighter image but not really more detail. I have learned that if I use the barlow at the "regular" distance so I have about 5000 FL (10"sct is native 2500fl), I get much better results, because I can shrink the image to 70% and it looks better than the "barlow lens closer" images at the same scale.
But at that scale the resulting image is very noisy and blurry, the colors don't look good and it's very dim.
Am I underexposed? I use EOS Mov Rec and a laptop, I set the exposure to where I can see detail and focus, but it's kinda dim, if I up the exposure I can't see detail on the planet anymore. EOS Mov Rec settings are something like ISO 1600 and 1/5th sec exposure - 21 fps.
Here is a single frame from the avi

Attached Files



#2 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5459
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 29 October 2012 - 08:14 AM

and one after all the Wavelets ALL THE WAY

Attached Files



#3 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5459
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 29 October 2012 - 08:15 AM

and one after starting the the wavelets at layer 2

Attached Files



#4 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5459
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 29 October 2012 - 08:17 AM

Now if I shrink that to 70% it doesn't look half bad, but at this scale it sucks.
Can someone tell me what I am doing wrong?

#5 bunyon

bunyon

    Gemini

  • *****
  • Posts: 3236
  • Joined: 23 Oct 2010
  • Loc: Winston-Salem, NC

Posted 29 October 2012 - 08:28 AM

I don't know DSLRs for planetary at all.

But that looks like a Jupiter captured in mediocre to poor seeing. Which happens. A lot.

Are you confident the seeing was good?

#6 m1618

m1618

    Mariner 2

  • *****
  • Posts: 280
  • Joined: 18 Sep 2012
  • Loc: Orange County - CA

Posted 29 October 2012 - 08:30 AM

Have you tried RGB align?

(I am "learning" registax...trying to fly an airplane by touching all the controls)

#7 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5459
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 29 October 2012 - 08:33 AM

I thought it was pretty good, maybe above mediocre, the planet was not jumping around, but it was going in and out of focus every couple of seconds, maybe steady for 4 or 5 seconds at a time.
Even still, I see these wonderfully smooth images in "poor" seeing...mine are never smooth or bright.

#8 EdmontonAB

EdmontonAB

    Apollo

  • *****
  • Posts: 1269
  • Joined: 30 Apr 2008
  • Loc: Sooke, BC , Canada

Posted 29 October 2012 - 12:24 PM

Try it after checking off the "Used linked wavelets" box.

#9 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5459
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 29 October 2012 - 02:31 PM

I always use the RBG align after stacking, it only seems to make a subtle difference.

I have tried using the linked wavelets but it just turns to garbage intantly, just clicking turns the image to colored pixels. I have never been able to use linked wavelets. I thought it has something to do with EOS Mov Rec using the Mjpeg wrapper, but that's just a guess, I really have no idea why.
Sample of the video here:

http://www.youtube.c...g&feature=yo...

#10 Space Cowboy

Space Cowboy

    Apollo

  • -----
  • Posts: 1041
  • Joined: 30 May 2010
  • Loc: Cheshire, UK

Posted 29 October 2012 - 03:09 PM

Your images are very good considering you are not using a planetary camera.
The seeing looks pretty good on your video clip. How many frames are you capturing? If you take a 3-4 min video capture this should provide enough frames to produce a smooth image with little noise. I suspect from your comments this may be your problem. At the focal length you are using rotational blur should not be an issue.

#11 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5459
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 29 October 2012 - 03:27 PM

Typical vids are 180 seconds, I end up with 3800 to 4000 frames, the use the best 900 for stacking.
The screenshots above I actually used the best 1300, trying to get a smoother image.

#12 Kokatha man

Kokatha man

    Fly Me to the Moon

  • *****
  • Posts: 6859
  • Joined: 13 Sep 2009
  • Loc: "cooker-ta man" downunda...

Posted 29 October 2012 - 06:14 PM

Shawn - with the preface that I've never used a dslr for planetary imaging, it appears from the video that the seeing just wasn't up to scratch.....noise is certainly NOT an issue with your capture as your examples with heavy-wavelet applications clearly demonstrate: in fact, the low noise level in them and also the single frame are quite amazing wrt lack of noise..! :waytogo: :shocked:

But the detail you've extracted would have had some (lesser) reflections in the single frame - ie, in any sort of decent seeing, instead of featureless bands you'd have seen some details in the belts - or at least much more tonal variation in them than is displayed.....so I'd be pretty safe in suggesting that the seeing wasn't that hot, and that you actually did very well with your processing in the circumstances..! :cool:

The various elements that influence seeing manifest themselves in several different ways, such that from some perspectives you can easily wonder why you don't get as much as you'd hoped for from any specific captures..... :)

#13 Kokatha man

Kokatha man

    Fly Me to the Moon

  • *****
  • Posts: 6859
  • Joined: 13 Sep 2009
  • Loc: "cooker-ta man" downunda...

Posted 29 October 2012 - 06:19 PM

ps - to cover another aspect Shawn, I suppose I should have also asked how confident you were of your focus..?

If you are confident about the focus, then poor seeing! :)

#14 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5459
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 29 October 2012 - 07:42 PM

Darryl, thanks for your input!
1st of all...Focus..I am never confident, I constantly doubt! I always fiddle with it between captures hoping one of them will be perfect. But I was FAIRLY confident as Ganymede was a round disk.
Noise...Really? Man I sure would have thought the images were noisy, but if you say so I will go with it. It just seems to have a grainy textured mushy appearance...Maybe I have been looking at too many DMK and Flea3 images...lol
So no finer detail in the video means seeing or focus, well I can live with that, at least it means it CAN get better. Now would someone kindly dispose of that wretch of a jet stream for me?
On a side note, I get MUCH smoother stacks from AS!2, but I am stuck until Emil finishes the next version. The conversion of Mjpeg wraper to uncompressed for AS!2 leaves horrible rings in the frames.
I just thought there must be something I am not doing.
AS!2 - 2000 frames, this video was shot 10 minutes prior to the one above. Aside from the rings, it's a bunch smoother and the wavelets don't give it a noisy appearance like above.

Attached Files



#15 cheapersleeper

cheapersleeper

    Gemini

  • *****
  • Posts: 3260
  • Joined: 22 Jan 2010
  • Loc: Sachse TX

Posted 29 October 2012 - 08:53 PM

I am only a beginner so take what I say with a grain of salt...

I differ with what others are saying. In this last image on my monitor it looks like RGB are badly aligned. Also, in my one month of Wavelet use, I have found it a delicate dance between Sharpen and Denoise over several Wavelet layers. Up with the Sharpen until detail is apparent, then denoise to smooth back down. From your examples it does not look as if you are doing this. I think your data may be pretty good but not great and your processing is what may be holding you back.

Regards,
Brad

This if from a recent discussion: http://www.cloudynig...at.php/Cat/0...

#16 rumples riot

rumples riot

    Mercury-Atlas

  • *****
  • Posts: 2964
  • Joined: 01 Jul 2004
  • Loc: South Australia

Posted 29 October 2012 - 10:31 PM

Seeing is most likely the problem here. Checking focus constantly is the norm for any seasoned planetary imaging. Last year Bird and I were almost having a competition to see who was focusing most. LOL.

Seeing such as you have here is most likely fast seeing which is commonly caused by jet stream interaction with the middle atmospheric layers. The image will typically look like no wobble but goes in and out of focus very quickly. Sometimes faster than others but typically around 1-3 seconds out of focus and just a second or two in focus. The faster the oscillation the faster the jet is interaction with the layer below.

Wobble is generally seen when winds about 200-500 feet up are turbulent and this can produce a flag waving look to the image but still is the jet is reasonably still quite good looking detail.

Your image also seems quite dim and in the last image you have onion rings. This is typically using a frame rate that does not allow full exposure to each frame. The subsequent stacking and sharpening will show up the lack of dynamic range in the data and produce those rings. Aim for histograms of around 75% (not sure how you will achieve that with a DSLR; no doubt someone more knowledgable on DSLR capture can help with that) to reduce the rings.

Hope this helps. Feel free to ask questions if I have written something you don't understand.

#17 shawnhar

shawnhar

    Fly Me to the Moon

  • *****
  • Posts: 5459
  • Joined: 25 Jun 2010
  • Loc: Knoxville, TN

Posted 30 October 2012 - 07:32 AM

Thanks!

Seeing such as you have here is most likely fast seeing which is commonly caused by jet stream interaction with the middle atmospheric layers. The image will typically look like no wobble but goes in and out of focus very quickly. Sometimes faster than others but typically around 1-3 seconds out of focus and just a second or two in focus. The faster the oscillation the faster the jet is interaction with the layer below.


That is it, I believe now the combination of fast seeing and a limit of 21 fps are the limiting factors here. I forgot about fast seeing.
As for the dimness, I am suprprised you are the 1st to mention it, and although the onion rings in the last image were a result of the video conversion process to uncompressed, I see what you are saying about the lack of dynamic range!

For my DSLR there are really only 2 settings, ISO and Exposure. The program controls that and displays a live view of what the sensor sees.
The ISO would be like the "gain" and the Exposure would be like exposure. Since it captures 21 fps as an output of what the chip sees you can capture 21fps of a white blow out ball or a very dim barely there ball.






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics