Lodestar X2C 2nd light, darks issues

[arkosg]

Hello everyone,

Conditions were excellent last night, so I set up my 8" Celestron CPC 800 alt/az at f/3.3 on my back deck, 

using LodestarLive 0.10.  Moon was waning crescent, and for objects above the glow of Vancouver and 

our town, viewing was very good - so good that I stayed up late enough to catch Orion (after seeing Don's 

shots, I had to see it for myself!    ).

I didn't note exposure times - I was trying a bunch of things, including the various stacking methods, etc.  

Typically I'd set the integration time to between 15 and 30 s (this seemed to minimize jitter), and then would 

stack 3 - 10 images (based on subjects brightness, field rotation, etc.)  Often I'd just lose track of the integrations 

because I was busy trying to colour correct, but often after a few integrations the benefit was less obvious anyway.

I viewed Paul's video tutorials, and noted the efforts of other members, and tried to spend more time colour 

correcting on the fly in LL; in general, I think the images were much better corrected, though I did end up 

applying some minor additional colour correction in iPhoto (the histograms were so narrow in LL at times 

that I couldn't tell how well they were aligned - looking forward to version 0.11!)

I've put the majority of the shots in my Gallery, but here are a few of the results.  Enjoy!

- Greg A

[arkosg]

As a follow up (and cry for help!):  I can't seem to get darks to function in the way they are supposed to, 

and I'm sure it is just me doing something wrong, despite having RTFM.  Here is what I am doing:

- scope aperture is covered

- click on 'Dark Frame Acquisition' button

- click on Exposure Control tab, set time (eg. 30 s)

- Start Exposure, let it run and take 5-10 shots (eg. at bottom, Exposures reads: Dark: 5)

- use Dark Calibration tab to write file for later use

I then click on `Image Acquisition' button at bottom, set the exposure to match my dark (eg. 30 s), 

and Start Exposure.  What I end up getting (example: Double Cluster) is something very dark with 

limited dynamic range, regardless of how much I fiddle with histogram, etc. in Display Processing tab:

If I "kill" the dark frame (go back to Dark Frame Acquisition button, Exposure Control, and press reset, 

(eg. at bottom, Exposures reads: Dark: 0), then I can take my image above again and get a "proper" 

image, like the ones I've been posting, albeit with hot pixels, etc. (reloading the saved Dark Frame 

master causes the same issue as noted above, so it's reproducible):

I know I must be doing something wrong here somewhere, but I can't figure out what it is - or perhaps

I'm missing something in the LL Display Processing when a dark is being used?  Help!  

Greg A

[HiloDon]

Nice work, Greg. Take another look at Paul's second video where he works on M1. He first uses the gamma control with modify all on to brighten and move the histogram to the right. It also widens some. Then the contrast control is moved to widen the histogram. It also separates the colors more. If you don't widen it, it is very difficult to align the colors. I originally had this problem as well. Once it's widened each color can be changed to be equal in height and width with the contrast control so they can be aligned with the brightness control. The alignment controls need to be finer, but if the histogram is wider, v0.10 works fairly well.

I don't understand your dark frame problem, but if have had something like you describe happen to me a few times. What I usually do is to close the program and reopen it and start fresh. I always make sure that stacking is off when I take my darks, Don't know if that matters. Then I take my darks and just leave it. I find it's best to take the darks just before you go to your image acquisition. I like working with 10 or 15 second exposures and the stack them as needed. If you have something brighter, you can go with less exposure and still use the 10 or 15 second dark. The one thing I noticed is that more brightness correction is needed when a dark is used, but I don't think you need more exposure. Not sure I'm right on all this because I'm still on a learning curve to. But, it does seem like we're all making progress. Perhaps Paul understands better what's going on with your darks.

Anyway, keep trying. Things are looking good. Your framing and image scale are excellent, too.

Don

[Martin Meredith]

Hi Greg

Great images! Good to see the nebulosity in M45, and that's lots of details in M33 when you consider that this can be a real pain visually.

Re: darks, what you're doing seems fine to me though I haven't had the problem.

BTW On some occasions I've used (master/saved) darks from earlier sessions with reasonable results. It helps if you record the temperature as well as the duration at which the dark was collected, then select the closest one. I recently bought a cheap (£5) digital inside-outside thermometer (also measures humidity), which I intend to use to check mirror cooling, so I've started taking these measurements to see loosely how they correlate with viewing, dewing, fan usage etc.

cheers

Martin

[arkosg]

Hi all (and hopefully Paul, too!)

The weather has been grey here, but I've been trying out darks indoors with the Lodestar and a SLR lens.  I've noticed something interesting, which may account for the darks issues I've been having.  Hopefully others can verify.

What I've been doing is taking darks before imaging, usually 10 frames worth for a given exposure, and SAVING THEM to a master darks file.  For example, this past time I had 5s, 15s and 30s master dark files.  Then I would restore the dark file I wished to use, depending on what my image exposure setting was.  This led to the issues mentioned above - very dark, low dynamic range, very different than imaging without darks at all - though it did get rid of hot pixels, etc.

What I tried last night indoors was to shoot darks as normal but NOT save them, but rather just start imaging right away after.  This worked as I expected - it required slight tweaking of the displayed image relative to no darks, but it was reasonably close and it took care of the hot pixels, etc.  However, when I tried to load in the Master darks file, for the same exposure time and same number of dark frames, it went back to the "bad" behaviour.  I will have to star test this next time I'm out observing to verify that it works for astronomical objects, not just bookshelves in darkened rooms.  

Paul - could this be a bug in the Save/Restore process?  Does it apply saved Master Dark files differently than the equivalent dark frames not saved?

Cheers,

Greg A

[HiloDon]

Hi all (and hopefully Paul, too!)

The weather has been grey here, but I've been trying out darks indoors with the Lodestar and a SLR lens.  I've noticed something interesting, which may account for the darks issues I've been having.  Hopefully others can verify.

What I've been doing is taking darks before imaging, usually 10 frames worth for a given exposure, and SAVING THEM to a master darks file.  For example, this past time I had 5s, 15s and 30s master dark files.  Then I would restore the dark file I wished to use, depending on what my image exposure setting was.  This led to the issues mentioned above - very dark, low dynamic range, very different than imaging without darks at all - though it did get rid of hot pixels, etc.

What I tried last night indoors was to shoot darks as normal but NOT save them, but rather just start imaging right away after.  This worked as I expected - it required slight tweaking of the displayed image relative to no darks, but it was reasonably close and it took care of the hot pixels, etc.  However, when I tried to load in the Master darks file, for the same exposure time and same number of dark frames, it went back to the "bad" behaviour.  I will have to star test this next time I'm out observing to verify that it works for astronomical objects, not just bookshelves in darkened rooms.  

Paul - could this be a bug in the Save/Restore process?  Does it apply saved Master Dark files differently than the equivalent dark frames not saved?

Cheers,

Greg A

I think you're on to something, Greg. I think the same thing happened to me. I'm pretty sure that I didn't have that problem in the earlier versions. I use a MacBook Pro. I usually don't use the saved ones, because it's better to take the darks close to the time you're viewing. But, I remember the last time out, I tried a saved one because I used a different time and had that same problem you did. I'm wondering if it's just the color mode. I'll check it the next time out.

Don

[Paul81]

Hi Greg,

Firstly, sorry about the slow reply, I have been away this week (for work) and had no internet access!

I have read through your post and had a look through the code, and I think you may be right about a bug in the load master dark code. When you load a master dark it resets the current dark data and inputs in each pixel from the FITS file into the rolling median algorithm which is used to compute a master dark (and used to subtract from your light images).

The rolling median will probably not be quite right under this circumstance, hence the funky result! I think I need to seed the rolling median better when restoring the master dark.

Will try a few things out to today and report back!

Other than that your set of images at the start of the post are looking really good, glad the videos were of use!

Paul

[arkosg]

Thanks Paul!

I got out last night and was able to confirm that just shooting some darks & using them (without save & restore) seemed to work as expected.  

It's a complex bit of code under the hood I'm sure, so small issues are bound to come up, especially as more people jump on the LL bandwagon 

and put it through its paces in ways that you never intended!  

I have a feature-related request to add to your (growing) queue.... I think it would be awesome for those of us using the cameras for outreach to 

have an option to turn on a re-sizeable second output window that mirrors the main control window.  I've used other cams with dual outputs, 

and had one output running through a frame grabber to my computer while the other went to an external monitor for groups to view (rather than 

all peeking over my shoulder).  I've tried mirroring my MacBook output via HDMI to my external monitor and it does work, but the image is small 

since it has to display all the controls, etc.  What would be perfect is a second window I can move to the external monitor ("spanned") and maximize 

to take up the entire screen, showing the same thing as the control window.

I have no idea how hard this would be.... but I think it would be killer for outreach!  Thanks for even considering it...

Cheers!

- Greg A

[Paul81]

Hi Greg,

That should be pretty easy to add in - I have added it to the future ideas list (and will likely implement as this is similar to something I was looking at for the NSN folks). It'll probably be a while to I get around to it though! 

I can replicate your issue with the darks BTW. I tried my first though to fix it but sadly it didn't so need to look a bit harder at what is going on. There might be some knock on effects to the median stacking as it is the same rolling median estimator used for the darks and median stacked lights.

Paul

[Paul81]

I just realised when I tried to replicate the darks issue I was using the wrong master dark file... Opps...

However, using my test harness when I use the right master dark file and run a single exposure through, using a fixed set of display processing parameters I get the following:

If I try the same thing but instead of loading a master dark I use a fresh stack of the original FITS files (same display processing parameters) and then run a single exposure through I get the same:

I also looked at 9 pixel value across the image and they equal the same value in both test cases. The pixel values are also the correct value given the values in the image and the master dark (i.e. pixel = image - dark).

So it looks like it is working OK.... I will run a pixel per pixel comparison on the two images to check they are identical.

Hmmm... when I get a chance I will also repeat with some colour data  - maybe it is related to something there...

Paul

[arkosg]

Hi Paul,

Thanks for trying to run this darks issue down!  With the examples above, is that with the existing code (eg. v 0.10)?  Or after you've changed something in the code...?

Thanks for considering the dual window output idea!  I'm really enjoying using LL, and for those of us doing outreach, I think something like that would 

really help with sharing the wonderful images with others.  I realize this isn't your day job, so I appreciate that it may take a bit to get something like this 

up and running!  

Cheers,

- Greg A