Wednesday, June 22, 2005

A cloudy night on Mount St Helens

There was heavy cloud obscuring the view of the Mount St Helens crater last night. Processing and enhancing the images from the Volcanocam did not reveal any signs of the incandescent glow from the lavadome.

However, simply averaging all of the volcanocam images from sunset to sunrise from last night reveals two bright spots on the images - that correspond to the two anomalous glows we have been seeing for some time now (see picture below). Further adding weight to the idea that these anomalies are caused by defective pixels in the Volcanocam camera.

I originally discounted the idea of defective pixels being the cause, due to the fact that the how spot appeared intermittently and varied in size and brightness over the course of the night and from night-to-night. However I think there is a possible explanation for this in the way the images are generated.

The Volcanocan camera is a Sanyo, Model #VCC-4594 Color CCD camera and the output is an analogue NTSC video signal - not a direct digital image from the CCD. The signal is fed to a computer where it is re-digitised and resized before being transmitted to the Volcanocam webserver. I suspect that it is the processing of the digital data to analogue video, then back to digital in the computer (and any other processing/resizing) that adds the variability to the hot pixels in the final images.

Averaging the images over the whole night, like I have done below, would help overcome any variability and we see the hotpsots for what they are - defective pixels. Of course, there's still the problem that the second hot pixel that appeared recently, did not appear on many weeks worth of images at all.

I'll go back over the archives (and produce average images like the one below) to see what happens over a longer time frame...


Anonymous Anonymous said...

Nice to know what caused these mysterious spots :D But it seems pretty early to me to already have damaged pixels on a camera that was installed only a few months ago... or am I to pessismist?

2005-06-22, 11:44 AM  
Anonymous Kevin said...

It makes alot of sense. I've seen pixels go 'off' on my cameras in the past (they all had been left on for several weeks at a time). But ive always seen them at the source, the jumping around after being redigitized makes sense.
In all cases, simply cutting all power to the camera then turning it back on cleared them. I hope the USFS people will give it a try. Hopefuly the camera isn't going bad already.

2005-06-22, 5:57 PM  
Blogger Daz said...

The current Volcanocam was installed on September 23, 2004, the same day that the mountain rumbled back to life - serendipity indeed! So its a little older than a few months, but I agree that it shouldn't be failing just yet. But then again, a few defective pixels aren't necessarily uncommon in a new camera.

I have looked back over some of the images since october 2004 and the hot spot on the front/left of the crater that appeared last week was actually present in some of the images back that far. What i don't understand if its a hot pixel is why it seem to disappear compltetely for days or possibly weeks at a time - even if the variations are due to the redigitizing of the images.

As an aside I have a little more info on the camera, it has a total pixel count of 811(H)x 508(V) pixels, with 768 x 494 effective pixels.

So even though its being converted from digital to analogue then back to digital, its also being resized to 640x480 - which is an uneven fraction of the CCD size, so there is either a slight crop occuring, or the pixel dimensions are being slightly distorted.

A slight change in focal length or position of the camera would be useful - although this has already happened a couple of times since September 2004 and the second hot spot looks to be in the same location now as then. I'll post some comparison images over the weekend to confirm this.

2005-06-23, 10:14 PM  
Anonymous Anonymous said...

I have watched the movie from:

The passing moon makes a big reflection similar to the "Mysterious".
My thought:
The moon-reflection is moving and brighter and bigger than than the "mysterious".
The outburst-glow is weak, stationary and flickering; and so is the "mysterious".
So !
- make them move the camera-angle!

2005-08-07, 11:40 AM  

Post a Comment

<< Home

Click Here