Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Jupiter Edge Ring Artifact

This topic has been archived. This means that you cannot reply to this topic.
37 replies to this topic

#1 tonyb

tonyb

    Viking 1

  • *****
  • topic starter
  • Posts: 706
  • Joined: 27 Jul 2005

Posted 30 January 2014 - 11:08 AM

Has anyone figured out how to minimize edge artifacts that occur in planet capture? As you can see in my processed image there is a ring on the right edge. Processed in AutoStakkert!2 and Registax. Jupiter captured at fr 30, exp 1/30 and gain 90% in IC Capture. The IC Capture histogram was at about 190 during capture. Should I increase the gain? I can post the 1GB video captured if anyone is interested in looking.

Attached Thumbnails

  • 6344905-6337503-ST3676_AS_p50_Multi_video0005-14-01-25-19-40-44CRP25.jpg


#2 ToxMan

ToxMan

    Skylab

  • *****
  • Posts: 4,154
  • Joined: 23 Jan 2011

Posted 30 January 2014 - 12:18 PM

There could be a number of reasons for the artifact, Tony.

I assume you were using the DBK31. With my DMK31 on an 8 inch SCT, I would not use high frame rates and gain settings because my images were too grainy. Typically, I would shoot at 15 frames per second. And, fill the histogram to about 50 to 60%. That minimizes the graininess. I know, reading in this forum about what others are doing, we can presume we need to have the fastest frame rate and the highest gain settings...But, there are functional limits on each camera. The DBK31 and DMK31 are slow cameras and you can't improve an image pushing the camera beyond its functional limits.

As for the ring on the limb, this could be atmospheric turbulence, collimation and focusing issues. Are you checking collimation before imaging? Take a near by star, like Pollux (Castor is a double and won't be suitable,) Capella or Aldebaran and check it before imaging the planet. During the collimation procedure, you can also gauge your seeing conditions. And lastly, during focusing, if you aren't seeing details on your live view, then something is wrong...too much atmospheric turbulence or collimation is still off.

It is possible to get some good images with the slower camera. If you care to see some images taken with my DMK31 on my 8 inch SCT, here is a Cloudy Night link: My Gallery Most of them were taken with my old set-up.

Hope this is of some help.

Paul

#3 tonyb

tonyb

    Viking 1

  • *****
  • topic starter
  • Posts: 706
  • Joined: 27 Jul 2005

Posted 30 January 2014 - 03:01 PM

Hi there - yes I'm using the DBK31. Yes, I'm checking collimation each session and adjust it to get it as best as possible. Looks like you are capturing for about 2 mins for Jupiter? Yes, atmospheric turbulence in the Pacific NW is quite high year round. I'll get you a link to my video AVI file.

Many thanks, Tony.

#4 Bird

Bird

    Soyuz

  • *****
  • Posts: 3,957
  • Joined: 07 Aug 2003

Posted 30 January 2014 - 05:57 PM

hi Tony, out of interest, would you mind trying an experiment - rotate the camera by 90 degrees or so between a pair of imaging runs and see what happens to this edge artifact, see if it stays on the bright limb of Jupiter, or if it stays on the right hand side of the image... keep the camera setup as you did for this image above so we can be sure it will show up somewhere...

cheers, Bird

#5 tonyb

tonyb

    Viking 1

  • *****
  • topic starter
  • Posts: 706
  • Joined: 27 Jul 2005

Posted 30 January 2014 - 07:08 PM

Will do! I also put the AVI up on Creative Cloud for anyone to download the 3GB AVI to play with. http://adobe.ly/1hTYZBh

#6 WarmWeatherGuy

WarmWeatherGuy

    Skylab

  • *****
  • Posts: 4,237
  • Joined: 27 Aug 2011

Posted 30 January 2014 - 10:55 PM

I played with your file and tried lots of different things. I noticed that it is a 3 minute video. Jupiter rotates 360° in 10 hours or 36° per hour. That is 1.8° in 3 minutes. I thought maybe that was the problem so I used VDub to extract out only the first third (one minute) and tried that. It didn't seem to help much, if at all.

I am sorry to report a negative result. I have been bothered by these rings and notice them on lots of other pictures. I would like to know what causes them. If you figure it out please let us know.

#7 tonyb

tonyb

    Viking 1

  • *****
  • topic starter
  • Posts: 706
  • Joined: 27 Jul 2005

Posted 31 January 2014 - 12:09 AM

Hi Steve: I've seen some posts that say reduce the FR to 15 and the gain down to 50 or 60% which I'm going to try. I also re-stacked in AutoStakkert raising the quality selection to 75% vs 50%. It is less stacked frames - around 1727 but the result seems a little better to me. Re: capture time of 3 mins. Yep, agree that it seems long, but my seeing is never that great in PNW so I don't see much difference between 1 min capture and 3 min capture.

Attached Thumbnails

  • 6346362-ST1727_AS_p32_Multi_video0005-14-01-25-19-40-44_g4_b3_ap30_CRP25.jpg


#8 troyt

troyt

    Apollo

  • -----
  • Posts: 1,006
  • Joined: 19 Nov 2008

Posted 31 January 2014 - 12:17 AM

+3 no luck

Attached Thumbnails

  • 6346372-Jup.jpg


#9 Kokatha man

Kokatha man

    Voyager 1

  • *****
  • Posts: 14,010
  • Joined: 13 Sep 2009

Posted 31 January 2014 - 12:57 AM

.....I didn't want to be the first to say this Tony but your image appeared very noisy and (possibly) indicative of way too much sharpening - which can easily bring out an edge artefact that is usually most pronounced on the bright side (which it is most definitely in your image also...) Of course there is a possibility that something else is at play here also (the gist of Anthony's suggestions) but tbh looking at Troy's results which are much smoother and don't exhibit that edge artefact I'd suggest my appraisal of the origins is fairly accurate.

Troy's processing does look like he employed some blurring around the limb peripheries but this is a fairly common practise with Jove, but I still think his results confirm the need to back off substantially on your sharpening routine.....it's easy to try and draw too much out of any single capture image so I understand your predicament - unfortunately it is the case most of the time that we all just have to wait for good seeing to produce really well-detailed images...making sure those items mentioned (collimation & focus) are well attended too for those occasions as Paul has said, and star-testing really does reveal what the seeing is like..! :)

ps: I'd process your image also but it is taking an interminably long time to download and we're off to the Mallee for the night in a few minutes...others might like to try and judging by Troy's results this image looks quite decent...I think he's being a bit modest to say "no luck." :)

pps: rotational blur isn't an issue btw!

ppps: ...but I may be incorrect as to which is the bright side..! :grin:

#10 Kokatha man

Kokatha man

    Voyager 1

  • *****
  • Posts: 14,010
  • Joined: 13 Sep 2009

Posted 31 January 2014 - 01:47 AM

...did a quick play with the avi as we haven't left yet (obviously! :) ) and there is an artefact in your image Tony - apologies for that but whatever is causing it is greatly exacerbated by over-processing nonetheless! :)

Fwiw here's the raw stack with no sharpening whatsoever and you can see the artefact...but judicious processing and Gaussian Blur on the selected region does ameliorate it significantly and a decent outcome is possible as Troy has shown.....but tbh I can't suggest any other reasons for its appearance atm and worth trying Anthony's suggestion...

Attached Thumbnails

  • 6346451-video0005-14-01-25-19-40-44RAWStack.png


#11 WarmWeatherGuy

WarmWeatherGuy

    Skylab

  • *****
  • Posts: 4,237
  • Joined: 27 Aug 2011

Posted 31 January 2014 - 02:32 AM

... here's the raw stack with no sharpening whatsoever...


Wow, your raw stack looks way sharper than mine. Did you use AS!2 or R6?

#12 Kokatha man

Kokatha man

    Voyager 1

  • *****
  • Posts: 14,010
  • Joined: 13 Sep 2009

Posted 31 January 2014 - 07:27 AM

.....1800 frames @ 60.4% cut-off Steve...I haven't used R6 for years but the MAPs sizes in AS!2 and their placement do have some effect on your outcomes - I think it's important for Tony to appreciate that the data is quite reasonable but there really shouldn't be that artefact in the raw stack...but not being familiar with the specific camera (nor for that matter the degree of collimation at the time) I can't really add much more atm.

A question Tony - obviously a colour camera but what codec was this recorded in.....being already debayered I was interested? :question:

Btw, I hope my first response wasn't taken as rude bro...I certainly didn't have any intention whatsoever of that and just wanted to be honest in my appraisal and hopefully assist you in some way! :)

#13 tonyb

tonyb

    Viking 1

  • *****
  • topic starter
  • Posts: 706
  • Joined: 27 Jul 2005

Posted 31 January 2014 - 09:47 AM

Hi there! Thanks so much for the responses. Captured raw video (not debayered) using my Imaging Source DBK 31AU03.AS color webcam and IC Capture.AS software and Y800 codec. I agree on oversharpening. I need to back off on that. I'm interested to know MAP sizes you used in AS!2, AP placement and # of APs. I'm in imaging lock down the next week or so judging by the weather forecast but I do want to try that framerate=15 and gain at 60% idea next time I image.

#14 Andy Taylor

Andy Taylor

    Twisted, but in a Good Way

  • *****
  • Posts: 3,240
  • Joined: 24 Sep 2008

Posted 31 January 2014 - 09:50 AM

I get edge artifacts like that if I forget to set gamma to minimum when capturing.

#15 Space Cowboy

Space Cowboy

    Apollo

  • -----
  • Posts: 1,466
  • Joined: 30 May 2010

Posted 31 January 2014 - 10:34 AM

I'd put good money on the slower frame rate fixing the problem.The old DBK21 had this issue when running at its maximum frame rate (60fps for the 21 version).

#16 ToxMan

ToxMan

    Skylab

  • *****
  • Posts: 4,154
  • Joined: 23 Jan 2011

Posted 31 January 2014 - 10:46 AM

Set the camera's gamma at 98 to 100, Tony. I've been working with a local friend who is doing some planetary imaging, and he had his gamma set at 50, yielding very dark images. I think that was way too low.

Paul

#17 WarmWeatherGuy

WarmWeatherGuy

    Skylab

  • *****
  • Posts: 4,237
  • Joined: 27 Aug 2011

Posted 31 January 2014 - 10:58 AM

.....1800 frames @ 60.4% cut-off Steve...I haven't used R6 for years but the MAPs sizes in AS!2 and their placement do have some effect on your outcomes

...

A question Tony - obviously a colour camera but what codec was this recorded in.....being already debayered I was interested? :question:


I'm wondering if we have the same file. The file I got had 5382 frames that were Y800 (GBRG). I'm not sure what you mean by 60.4% cut-off. 1800 is 33.4% of 5382. Also, what size APs did you use? I have so much room for improvement. :bangbangbang: Thanks for showing me that.

#18 ToxMan

ToxMan

    Skylab

  • *****
  • Posts: 4,154
  • Joined: 23 Jan 2011

Posted 31 January 2014 - 11:36 AM

Hey Darryl, do you think there is a collimation issue? I see a tinge of blue on one side and a tinge of red on the other...not as obvious in Troy's rendition.

With a color camera, I would think that would happen if there was a little "splitting" or "separation" of the light. A prism effect from collimation being slightly off? I'm brainstorming here...

Paul

#19 WarmWeatherGuy

WarmWeatherGuy

    Skylab

  • *****
  • Posts: 4,237
  • Joined: 27 Aug 2011

Posted 31 January 2014 - 11:46 AM

Hey Darryl, do you think there is a collimation issue? I see a tinge of blue on one side and a tinge of red on the other...not as obvious in Troy's rendition.

With a color camera, I would think that would happen if there was a little "splitting" or "separation" of the light. A prism effect from collimation being slightly off? I'm brainstorming here...

Paul


I have found that can be corrected with RegiStax's RGB align. You can tell the amount of tinge by the number of pixels needed to remove it. It seems to be a function of the elevation, you get more at low elevation. The red side of Jupiter is the side closest to the horizon and the blue side is opposite the red side. I have seen this red/blue tinge when Jupiter is way up there (50° or so).

#20 ToxMan

ToxMan

    Skylab

  • *****
  • Posts: 4,154
  • Joined: 23 Jan 2011

Posted 31 January 2014 - 12:42 PM

Yup, the atmosphere will definitely disperse the light, Steve. And, low elevation would account for this, too. I would guess that in Washington state, that by 10 or 11pm, Jupiter is nearing meridian. Pacific and Mountain Standard time is an hour different, now. Tony has a lower elevation at the higher latitude, too. I'm at 32 degrees latitude, and at meridian, Jupiter is almost overhead. I'd guess that Tony is near 50 degrees latitude. Twenty degrees is a big difference. That being the case, with good collimation and focus, seeing is going to be everything.

#21 tonyb

tonyb

    Viking 1

  • *****
  • topic starter
  • Posts: 706
  • Joined: 27 Jul 2005

Posted 31 January 2014 - 03:35 PM

Hi Paul: In IC Capture, my gamma was set to 100.

Many thanks, Tony.

#22 tonyb

tonyb

    Viking 1

  • *****
  • topic starter
  • Posts: 706
  • Joined: 27 Jul 2005

Posted 31 January 2014 - 03:45 PM

I chose another video capture and processed it in AS!2 and then not so aggressive with sharpening in RG6. Then processed in Photoshop CC.

Attached Thumbnails

  • 6347465-ST1201-BEST-_AS_p22_Multi_video0008-14-01-25-20sat-smhrp-CRP25.jpg


#23 Kokatha man

Kokatha man

    Voyager 1

  • *****
  • Posts: 14,010
  • Joined: 13 Sep 2009

Posted 31 January 2014 - 04:27 PM

I'm only viewing this on my field laptop Tony - but heck! this image (and the processing! :) ) is a mega-improvement!!! :waytogo: :waytogo: :waytogo:

Paul's right imo about the colour tinges on each side of my raw stack.....on your latest one Tony I think I see both r&b on the RHS but that might not be both, only b - but it is the lappy I'm viewing on...

Stuart may well be on the money with his comments - I remember having a long discussion with Asi about 60fps captures and problems therein but my memories of the DBK21AU are very faded now...and you were capturing at 30fps Tony, so maybe the larger-sensored "31" has this problem at 30fps - especially if you've seen mention of it before, so that 15fps trial would seem a very good idea to me!

Also, debayering can be accomplished in AS!2 and I believe that it could prove superior to the usual applications of said...

Fwiw I chose a MAPs size of "125" starting with one at the centre and built a semi-symmetrical grid (not sure whether that is totally relevant) out from this making sure that the outer boxes aren't too near the edge but so that they just encompass that portion of the planet's limb they are adjacent too...ie, no portion of the disk is left uncovered by the boxes...

Steve, my "1800 frames @ 60.4% cut off" meant I chose 1800 frames to stack and this means the 1800th frame was estimated @ being 60.4% of the quality of the best frames as chosen by AS!2.....I don't think Tony was switching avi's around for folks! :)

Summing up I reckon you've developed a much better regimen for processing now Tony and that 15fps trial is well worth doing..! :)

#24 MHamburg

MHamburg

    Vanguard

  • *****
  • Posts: 2,004
  • Joined: 21 Jun 2006

Posted 31 January 2014 - 06:25 PM

Kokatha man,
Are you suggesting that it better to turn off debayering in Firecapture and then debayer in AS2? Thanks as always for your expert word.
Michael

#25 astrovienna

astrovienna

    Mercury-Atlas

  • *****
  • Posts: 2,990
  • Joined: 04 Dec 2006

Posted 31 January 2014 - 11:23 PM

Hi Tony,

This is a camera artifact. You can avoid it by reducing your frame rate one setting below max. If you search in this forum a couple years ago for DMK and artifact, you'll find several detailed threads about this. In short, these cameras (and some other brands) have to be run at one setting below their maximum frame rate to avoid this artifact. So for my old DMK21AU04, that meant I had to run it at 30fps instead of 60. Your camera, with its bigger chip, probably needs to be run at 15fps instead of 30.

I remember reading some frightfully detailed explanation of the electronics issue - some kind of electronic echo - that causes this, but it was way over my head. All I cared about was that when I cut back the frame rate, the problem disappeared.

Don't confuse this with other sources of edge artifacts, especially the diffraction artifact on the bright limb of Mars. You'll probably become well acquainted with that one over the next few months. ;)

Kevin


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics