SharpCap Forums

Dr. Glover

Please consider this blog post an on-going communication as I problem solve my way into my second year of this hobby.
I have something to report that is going to take several hours of pecking out my observations on the keyboard.

My first year was devoted to understanding my OTA and mount as an imaging train along the lines of the near vertical learning curve of this hobby.
That experience has, "Washed out into a firm grasp of the basics."
Yes, that learning curve included, "Up to and including learning about Seeing and sensor image scale relationships."

To recap and provide a frame of reference:
OTA: William Optics 81mm APO, reduce to f/4.72 and focal length 378mm... Very much a wide field OTA.
NCP: Within the operational limits of NCP alignment of being down in the basement of ~0.59 arc-mins of total error for a sky watcher EQ-R6 under reasonable PHD Seeing.
Observatory: Can get on target within minutes of powering up.
Take away: I had a rough start with the QHY268c, in the interim I switched to a QHY183 mono camera and a LRGB filter set.
More importantly this rough start was also highly influenced by wishful thinking on my part that I had jumped a major gap on the learning
curve of the hobby. Very much, "Not the case!!!"

Everything in SharpCap is running reasonably well since going to the QHY183mono camera.
No more gain and exposure adjustments lagging behind. "That result in my subs not timing out."
So, this recap is, "Essentially to state everything software-wise is ideally up and running and I have very much moved beyond the learning curve basics of the hobby."

With that established background now in-place and given my first-year communications with you about:
"How, I believe your analysis of SNR is 'spot-on' in scientifically being based on the lower read-out noise of back Illuminated CMOS sensors in general
as opposed to the legacies within the hobby that ultimately determine optimal exposure sub lengths and gain level settings."
A major mouthful to say the least.

I am going to call it night and will flow up by jumping right into a Live stacking observation that relates to exposure and gain settings and the speed of my imagining train as I attempt to collect LRGB sensor data.

Sincerely,
Mark

(Please bear with me)
(I am setting the stage for something I observed in my live stacking subs.)
(Also, it helps to understand. This is my reference point on optical light.) Picking up from where I left off.
So, I have gone to broad band targets being imaged by LRGB filters.
Everything is going great.
I have selected a very good DSO target that is nearly at the limits of my sensor's imagine scale for resolving contrasted details.
The Sculptor Galaxy.

My luminance frames are 120sec, zero gain, zero offset subs at 0C.
Sharpcap's live stacking of these subs is churning out a decent stack of the depth of contrasted details in the galaxy.
Looking good for pixinsight post processing.
When I reset the histogram, " I can barely make out the central bulged of Sculptor and about two dozen stars displaying signs of full-well saturation on the 2.4-micron image scale of the sensor." Everything that have I read to date says, "That is where you want to be. In the histogram basement."
(Well, maybe a little bit more of OFFSET and gain tweeking)
(Not going to happen until I do a sensor analysis benchmark.) (More on that later.)
Here is a representative stacked image. Auto stretched with black and mid levels adjusted to taste, zoomed in at 40% and clipped off my display. That image is very much representative of the learning curve factors that come into play. (Seeing/Focus/Guiding/Sensor Settings)

But there is something odd about this stack. The stack included a sub frame that was biased by dithing.
So, naturally the issue of focus was to resolve the issue of why that frame was stacked.
In doing so, "That is looking at the RAW files being saved. In DSS it quickly became apparent the easiest way to redress this issue was to exclude these subs."
It is just a timing quirk between SharpCap and PHD.

(I am going to post this as is. Because I don't want to lose my work.
However...

Returning to Sculptor and that, "There is something odd about this stack." Notice, "Not all the stars have dithering influenced star trails."
So, yes referring back to... Why is that?

For many, star bloat in a refracting OTA imagining train, "Might naturally mean to make the assumption that bloat equates to star size."
Definitely not the case.

Under Ditchburn 1976, "Star bloat is nothing more than the 'apparent' arriving luminosities spread out over an area of the sensor in the time phase domain of the visible electromagnetic spectrum.
Essentially, the full-well sensor saturation of photon readings(excitations) are measuring the rate of the arriving photon wave fronts.
(As such the larger the star(bloat) could mean two very distinct possibilities.)

In practical terms that could mean star bloat is directly related to one of two dependent unknowns.

"Is the star saturating the sensor because the star is massive and thus more luminous but very distant or is the star closer but less intense such that there exists within these two opposing unknowns the distinct possibility either could be the case."
By thinking in these terms, actual luminosity vs distance, then WHAT explains the stacked subs that where very much less than half of the stars are showing measurable trails on dithering.

Are these stars further away? Less luminant?

This is a very interesting deep drive into sensor image scale analysis.
One that I believe can be, "Measured and quantified."

Such that an image scale interferometer can be deduced from the data.
Specifically, when coupled with known parameters about a given star.

Lets back up just a bit.

Taking my posted live stacked frames of reference, "The stars showing distinctive trails are in my opinion are either very much closer in the foreground or they are very massive and thus more luminous yet further away or any combination thereof."
Such. either way, the sensor data giggling seems to reveal an imagine scale luminosity relationship as measured by the length of the resulting trails.
(What else explains the lack of star trails for many of the stars across the entire field of view.)

Much of this hobby is devoted to reducing noise. Such that on the sensor's image scale the effects of Seeing verses vibrations on the OTA are many magnitudes above this seeing level of noise.
Yet, here we have some stars showing clear signs of sensor bias while many other stars show little trailing.

Just what does that mean? In the rush to pretty pictures, does this data discarding hide away meaningful data."

Taking this observational to a logical conclusion, "Could not data that shows this trait be used to infer proximity or mass on any given star in terms of
luminosity scales.

Getting theorical, could not deliberate vibrations on the OTA be calibrated into a relationship that sorts out distance and/or mass of a given star.
Much like focusing a lasar on a window plane can reveal sound.
(Again, going to post this as is, if only to revisit.)
Sincerely,
Mark
Mark

Hi Mark,

I'm afraid that my analysis is much simpler... every star in that image does have a trail heading up/left due to to the dithering frame being included in the stack, it's just that for the bulk of the dimmer stars the trail is so weak that it does not rise above the background noise level, so cannot be seen. Certainly if you apply an extra stretch to that image you can start to pick out lines on slightly fainter stars that are not apparent in the posted version, and I see no reason that the same does not continue down to even fainter stars.

From my point of view, the real interest is how the dithering frame got included in the stack - that's not supposed to happen, although there have been some other reports of it happening from time to time. I just put more updates into this week's SharpCap 4.0 release to try to kill that problem off for good - time will tell whether they are enough!

cheers,

Robin