Jump to content


Photo

Strange Diagonal Fixed Noise Pattern from Flea3

  • Please log in to reply
10 replies to this topic

#1 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3059
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 13 November 2012 - 12:33 PM

After going over a month of not using my scope due to poor conditions, I tried imaging Jupiter through light fog Sunday night with marginal data at best. The f25 RGB data appeared noisier than usual so I decided to take a flat using the same setup as Jupiter, 60fps, with the gain eased down to yield a matching 65% histogram. I used a round 15" fluorescent light and held it right in front of my dew shield. The flat stack was compiled in Registax. While I have not done this before, the amount of diagonal noise seems excessive. My Flea3 uses the heavy point grey firewire cable and the laptop adapter. My computer is my company issued lenovo T400 running Windows XP. I used FireCapture 2.1 with the FlyCap2.2 to capture the data. Rotating the camera had no effect on the overall noise or direction.

Does anyone have any suggestions? This level of background noise seems unacceptable to me and I doubt it can be easily flat frame removed, but I have not tried. Unfortunately, my Jupiter data is ROI 400x400 which I neglected to set during my flat so the file sizes do not match.

Thanks in advance!

Regards,
Steve

Attached Files



#2 Bart Declercq

Bart Declercq

    Messenger

  • *****
  • Posts: 493
  • Joined: 21 Jan 2007
  • Loc: Haaltert, Belgium

Posted 13 November 2012 - 04:24 PM

Steve,

I've seen this pattern before : did you clean your chip using a cloth (and perhaps some kind of cleaning fluid?)

This sometimes causes such lines in the direction of the last "wipe".

The slight wavyness and the hint of crosshatching (wipe in the other diagonal perhaps?) suggests a non-electronic cause.

#3 ToxMan

ToxMan

    Mercury-Atlas

  • *****
  • Posts: 2766
  • Joined: 23 Jan 2011
  • Loc: Tucson, Arizona, USA

Posted 13 November 2012 - 07:58 PM

Is this mono, unfiltered, Steve? It almost looks like Moire pattern, the bane of my imaging in near-IR and R filters. By the way, Bart, I make a dry pass after cleaning DMK31 camera chip cover, never following the same routine, making direction of wipes randomly. But, the direction of the Moire pattern never changes. Always worse in high gain and frame rate settings.

Attached Files



#4 DesertRat

DesertRat

    Fly Me to the Moon

  • *****
  • Posts: 5176
  • Joined: 18 Jun 2006
  • Loc: Valley of the Sun

Posted 13 November 2012 - 07:59 PM

Does'nt look like any electronic noise I've seen before. I'll assume you just coadded the frames with no alignment. There are also faint lines here and there perpendicular to the obvious lines. I think Bart may be on to something.

I have not done this test, I have flat fields for the Flea3 I took to measure gain and noise figures. But I was not imaging at f/25, I just placed the Flea3 outfitted with a camera lens at f/4 plus filter butted right up to an ELD panel. In a test like that anything on the cover window would be blurred out. At f/25 not so.

But assuming you do get a good flat field, yes even this could be removed (I've seen far worse). There is a only a hint of a couple of dust bunnies (at f/25 they show up well) - so maybe you did clean it recently. I used flat fields long ago for ccd camera images of the Moon and Jupiter, and it did help a lot. But with video and the planet swimming around flat fields are not as needed. Hot pixel removal however is crucial.

Glenn

#5 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3059
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 13 November 2012 - 08:24 PM

Guys, thanks for the feedback. I only use dry cotton q-tips. Last time I cleaned the chip, no chemicals, in a left to right direction until the dust motes were gone, maybe 3 or 4 swipes each time with a new q-tip. I used an Astronomik Lum filter for this flat which I don't use on Jupiter. This flat was made using the Registax5 flatfield creation. I've seen this before last time out imaging the sun and just thought it was due to the settings. I'll send this data to Point Grey and see that they say about it. In the meantime, I'll take a flat avi with the same ROI as Jupiter next time out.

Regards,
Steve

#6 skywatch

skywatch

    Ranger 4

  • -----
  • Posts: 313
  • Joined: 06 Feb 2008
  • Loc: Ottawa 45.3085N 76.175W

Posted 18 November 2012 - 12:24 AM

Aliasing of the 60 Hz frame rate and the 60 Hz flashing of the florescent light? Personally though I try not to over about 45 fps because funny things can happen at the high frame rates.

-Rolf

#7 Kokatha man

Kokatha man

    Fly Me to the Moon

  • *****
  • Posts: 6780
  • Joined: 13 Sep 2009
  • Loc: "cooker-ta man" downunda...

Posted 18 November 2012 - 03:26 AM

.....I don't think my appraisal is worth much Steve, but those lines are anything but symmetrical or uniform when you look closely...and Bart's original suggestion would seem to be nearer to the money imho because of this random pattern which only looks "electronic" if you don't look at it very closely... :question: :confused:

I'm interested in your use of "dry" cleaning tips - I appreciate that the glass cover for the sensor would be/is toughened, but I wonder if using dry tips might be smearing and not actually removing any crud that accumulates..? :question:

#8 Kokatha man

Kokatha man

    Fly Me to the Moon

  • *****
  • Posts: 6780
  • Joined: 13 Sep 2009
  • Loc: "cooker-ta man" downunda...

Posted 18 November 2012 - 03:34 AM

ps - fwiw here's an enhanced image of the pattern in your shot above...as I said before, I'm really in the dark about this Steve!

Attached Files



#9 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3059
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 18 November 2012 - 10:43 AM

Rolf, Mo, thanks for the feedback. You may not believe this but last night the diagonal noise was completely gone (although some slight vertical noise is now present). You're probably thinking that I've lost my mind but I have attached some new flat data that shows the change. The only new things I tried were:
1) Slighty tightening the 2 screws that hold the firewire to the back of the camera since the heavy cable droops down and will pull out if I step on it by accident.
2) I rerouted the firewire cable to the cheap plastic adapter card that pulls into my laptop and again avoid unecesssary strain from the heavy cable.
3) I tried collecting 16bit fit data only to find that my computer can't seem to keep up with the direct write. (I have not tried .SER but not sure what to do with that format data even if I had it). After finally giving up, I toggled by to 8bit data
4) It was quite cold last night, around 23F which may have helped?

So after "borrowing" the round flourescent light from the basement, my flat light was back in business and I took flats with each channel using the same fps (80/80/45) & histogram (65%) as my actual Jupiter data and no lines to be found. I plan to go out shortly and make some test avis on the sun which is when I first noticed this issue.

In conclusion, it may have been the toggling between 8/16/8bit that turned off the diagonal noise. Point Grey has not seen type of noise before. I will post more data when I get it. It's hard for me to believe that only my Flea3 has this issue. Anyway, keep this trick in mind, it may work for you.

Please attachments

Regards,
Steve

Attached Files



#10 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3059
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 18 November 2012 - 10:49 AM

Red vs Green, 80fps same Registax sharpening. Green is much worse

Attached Files



#11 sfugardi

sfugardi

    Gemini

  • -----
  • Posts: 3059
  • Joined: 30 Sep 2006
  • Loc: CT

Posted 18 November 2012 - 10:55 AM

L-B comparison. Vertical lines in B are terrible.

Attached Files








Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics