Strange LodestarLive Focus/Framing usage....

[arkosg]

Hi everyone,

I was  working with my Lodestar X2C in conjunction with a Hyperstar on my 8" Edge HD CPC (alt/az) setup the other night, and came across something a little unusual and intriguing.  I was using 1s integrations in the Focus/Framing preview mode to compose objects/check GOTO's, and at some point realized that in many cases I was getting some very nice, sharp, and clearly defined objects in B&W this way; in fact, some of the views of objects (particularly globs and similar things) seemed almost better defined than in the colour "image" window views.  I was even able to view some extended nebulae (eg. M20, M8) in this way.

That got me to thinking:  is it possible to do B & W imaging in this way with the colour X2C?  IHow does Lodestar Live handle the camera/data in Focus/Framing mode?  I didn't think to try a screenshot, and didn't try longer integrations to see more detail, but I'm guessing it's using image data pre-debayering?   Would it be possible to output this view in some way?

When using my more typical non-Hyperstar setup, this has never really jumped out at me, but even with a waxing gibbous moon, the Lodestar X2C and Hyperstar combo were giving really decent, very crisp looking single shot near-real-time views in B & W - from a colour X2!

Cheers,

- Greg A

[Dom543]

Greg,

Have you tried to uncheck the One-Shot-Color checkbox and compare images obtained that way to those in the focusing tab?

I believe that when that checkbox is unchecked, the debayering part of the image processing is bypassed. Most likely debayering is not part of the program branch either that feeds to the focusing screen. On bright white objects, like stars, including globs, this gives higher resolution than the OSC mode. This is probably beneficial for focusing.

I say white because white light penetrates the micro filters of all pixels. So all pixels generate signal.

I experimented with unchecking the OSC box, when using an H-alpha filter. I got a very fine vertical pin stripe pattern. I believe that the reason was that H-alpha light didn't generate any signal on the Cyan and Green pixels that are located above one another and filling the odd numbered columns of the Bayer matrix. So the odd columns ended up blank. In the absence of debayering, they were not filled in with some interpolated values. But this is just my best guess explanation of what I saw, I am not an expert.

Clear Skies!

--Dom

[Paul81]

[sorry its taken an age to reply...]

In the focus and framing mode the CYMG bayer matrix is only converted to luminance (i.e. not RGB) and all display processing is performed in luminance only (so a third less data to process). This was originally to save processing power as its expected to use fast exposures and some folks CPUs may not keep up when sub 1s exposures are used.

Un-checking the OSC box won't give the same result, as although you will get a B&W image, it will have the bayer pattern over it.

If its useful, I can very easily (to the next version) add a filter option which makes it do the same thing in the imaging mode?

[arkosg]

Paul,

Yep, unchecking OSC does NOT give you the same result at all.   The focus/framing mode seems quite useful for those with colour cams - the image seems crisper and more detailed, somehow, at a given exposure than the colour images do.  If one could do this, with the added ability to stack as in the imaging mode, it seems like it would *almost* be like having a B&W and colour cam in one, no?  Or am I misunderstanding the handling/production of luminance in this mode? (I know the filters would cause some loss of signal, but one would not have to deal with the resolution loss/interpolation of the colour portions, right?)

Here are a few shots in the focus/framing mode, all 1 sec exposures, taken last night; and this was a night with smoke, a 3rd quarter moon, and poor transparency.  The first night I played with this, I was seeing much fainter objects more easily.... so perhaps if it isn't a real PITA to implement, it might be useful for those with only the X2C, especially when using it on objects with little colour... ?

Cheers,

- Greg A

[Paul81]

Hi Greg,

I don't think this would be too hard to implement, so I'll have a look at getting it into the next release (I am making mods in the affected areas any way for EAA Hubble palette).

In terms of the colour camera more like the mono, I'd need to think on that one. You still have to interpolate each pixel based upon its neighbours, as the luminance is calculated from CYMG. It might be more crisper as the luminance I only calculate from the immediate neighbours. For RGB it is a bilinear interpolation over a wider grid as this seemed to reproduce better colours. There are better algorithms than bilinear, at some point I might try some other options out to see the effects.

Paul

[arkosg]

Thanks Paul.  I know it's an oddball request, so I appreciate you even considering it.  I honestly don't know how "useful" it may be without more playing with it, but it looked promising enough to perhaps give it a try - as long as it doesn't become a PITA for you.  ; - )

Cheers,

- Greg A