[meteorite-list] Aussie Photographs Meteor Through Telescope: NOT

From: Mr EMan <mstreman53_at_meteoritecentral.com>
Date: Sat, 29 Aug 2009 09:40:17 -0700 (PDT)
Message-ID: <145736.35153.qm_at_web55204.mail.re4.yahoo.com>

I need to revise and extend my remarks from before. This probably is a meteor in spite of our first judgments. I too piped in as an early naysayer because I was thinking in the film paradigm. I've rethought the image in the digital paradigm.

Long and boring and technical rationale:

I've looked at the photo through an image processing application where I can zoom down to pixels and believe I can account for some things which we casually dismissed before because it was not what we were used to seeing. There is some actual color data in the head of the fireball when I adjust the Gamma.

1. If we assume the bulbous tip was a terminal burst not just the end of the exposure then there need not be an trail of equal diameter all the way back up the trace on the image that is the width of the "bulb". I've looked at the trail and it appears to have a uniform width save for the tip. The wispy segmented trail is a result of a fast moving object crossing many sectors of a Charged Coupling Device (CCD) while several passes of the frame scan program are going on.(my interpretation) The color data should be present all along the trail unless the bulb is a terminal flaring representing a several-magnitude flash of energy as we see in a terminal burst.

2. The tail is not as squiggly as we first thought, but seems to be an artifact of the CCD array and how the image information is captured. Be it remembered that while film collects the entire image the entire time of the exposure, digital "timed exposure" imagery is the summing/melding of thousands of passes over the CCD sector by sector, pixel by pixel. Each pixel has to be given permission to purge information to make ready for the next pass. even at computer Hz rates this can cause a pile up of information as the data is read and written to storage. Some pixels simply will not be ready to receive and hold light data as the fireball is passing.

3. To make a timed exposure in a digital camera, the data of one pass is added to the data of all the cycles before it. The image data is also processed by several algorithms to try to accommodate a range of conditions--none of which are optimized for a high speed intensely bright object on a black background. Likewise, we have no way of determining when the meteor passed as all scan data is lost once added to the image file.

4. I won't delve into the full technical aspects of latency of signal and how the microprocessor polls the signal from each pixel on the CCD, etc. But- for an allegory we are all familiar with, think of how "wagon wheels" in old westerns appear to spin backwards on film. It looks that way owing to a difference in frame rate of the film and the actual speed of the wheel's rotation. A timing discrepancy in a digital frame gives rise to a "smeared streak" with black gaps at the spots where the data is being reset so we can get a "-==_ --==-_ -==-_ -==--_ -==_" for what would have been a continuous straight line to our eye.

5. When we look at the "mask" of a CRT TV that hides the edge of the phosphors to make them appear uniform, we see a grid of black rectangles. CCDs have a similar grid/blind grid. When an object crosses the screen horizontally-- keeping on just that row the line is straight. Cross it diagonally and it looks like a series of step downs or step ups. Add in the aberration of lens curvature, a slight internal vibration from the drive motor, process through a jpg compression algorithm and you get a "squiggly" line even at normal zoom!

All in all, in light of what I remember now about digital still-frame photography, this is a righteous shot.

Elton

--- On Sat, 8/29/09, Rob Matson <mojave_meteorites at cox.net> wrote:
> Subject: Re: [meteorite-list] Aussie Photographs Meteor Through Telescope: NOT, ... Absolutely *NOT* a meteor.--Rob
>
> http://www.iceinspace.com.au/forum/showthread.php?t=49164&highlight=meteor
Received on Sat 29 Aug 2009 12:40:17 PM PDT


Help support this free mailing list:



StumbleUpon
del.icio.us
reddit
Yahoo MyWeb