First off, congrats! It's a good, capable little camera. Lots of fun!
Now for the questions... and there are a few.
I set the gain. Is offset labeled as something different in SGP? I think the gain had a drop down with literally every number available up to whatever you need. Thought that was interesting as opposed to a system where you could just type in your preferred gain.
With the post-126.96.36.199 builds of SGP, for the time being you need to use the "Not Set" option in the gain by binning settings in the Camera settings of SGP. You then set the gain and offset within the ASCOM driver. Offset is a key setting with these variable gain cameras...they allow you to maximize dynamic range by minimizing the offset, rather than using one overly large offset that could work at any gain setting. However, ASCOM currently does not have support for exposing offset control, so SGP does not have any way of setting offset itself. By using the Not Set option and configuring within the driver, you can control both offset and gain appropriately.
Another ASCOM setting to be aware of is USB Limit. This is a critical setting for minimizing FPN. At a low setting (minimum is 40), FPN and glows can be much more significant:
At a high setting (maximum is 100, but should never use more than 90 to maintain stability and reliability), FPN and glows can be much lower:
You want to choose the highest USB Limit setting in the ASCOM driver that is reliable and stable. You can test this with the frame and focus exposures in SGP. If you set USB Limit too high, then SGP will simply not download the frame. You can stop the F&F exposure then notch the setting down in the ASCOM driver little by little until all F&F exposures download reliably. You want to keep that setting fixed, don't change it if you can avoid it...changing it will result in changes to the FPN, even if you restore it back to the same setting. The changes may be slight, but it can still invalidate existing calibration frames (which would otherwise be reusable for months.)
File size. Yeah... I heard they were big. 30 MB. That's huge. And I only have 1TB on my laptop, half of which is gone. I can't imagine short high gain images being efficient unless I have them save to an external drive. Of course, I see there are benefits to both. So what is better? Will an hour of medium-ish gain images at 1.5 minutes be as good as the same integration of high gain 30 second images? I know one will eat up more hard drive space. Coming from the DSLR world, its hard getting used to high gain being good, where 3200 ISO will be noisy as anything.
Think of gain as a means of optimizing the camera performance to your scope and skies. If you have a really fast scope under really bright skies, then you will need to use Gain 0. You will probably suck photons down so fast that you might not get more than 60-90 second subs, and you could easily acquire so much signal that you'll swamp read noise by 100x or more, and have 99% stacking efficiency.
On the other hand, if you have a long focal length scope at a really high image scale (say SCT f/10) under very dark skies, you could probably use the max gain setting (well, max "viable" gain, the driver range goes from 0-600, but for this particular camera, any setting over 300 is just throwing away DR for no additional benefit in read noise) without any real issues with clipping. Further, at such an image scale with minimal LP, you could probably use fairly long exposures as well.
The general idea is to find the setting that gets you the minimum read noise with no clipping or acceptable clipping, at an exposure length you can handle. Using the longest subs possible is not necessarily ideal...at lower gain settings, quantization error can be quite high. Depending on what you are doing, you may be able to swamp the quantization error quite handily with skyfog noise (i.e. LRGB), however in other cases you may have a tough time swamping the noise and quantization error at all, resulting in issues with posterization. Dithering can help a bit there, however dithering is more effective with more subs... You can see how low gain and long subs are a bit counter-intuitive with this camera. This camera DOES benefit from stacking at least 80-100 subs, and can continue to benefit from stacking 250-300 subs. With effective dithering, I've been able to see continued improvements stacking upwards of 1000 frames, however it's not really necessary to go beyond 300 in any normal DSO imaging circumstance.
If you are concerned about disk space, you could try to base your exposure lengths off of your integration time goals and sub counts. If you wish to integrate 6 hours per narrow band channel, and want no more than 80 subs, then you would need to use 270 second subs. Find the gain setting under your average skies that will allow you to use that exposure length, make sure you don't clip too much, and there ya go.
Because with variable gain you have variable dynamic range, you might want to optimize your exposures a little more than that. The main thing to check is, assuming you aren't already clipping or clipping too much, are you swamping the read noise effectively? There are several guidelines to help you determine that...the 20xRN guideline, the 3xRN^2 guideline, etc. Pick one that you feel is effective, measure some of your subs, and figure out if you are sufficiently swamping the read noise with those 270 second (or whatever) subs. If you are, and you may well be more than sufficiently swamping the noise, then you might have found your sweet spot. If you find you are clipping too much, and your exposure are more than sufficiently swamping the read noise, then you have more room to optimize...maybe use 210 second subs.
The AC adapter I purchased doesn't have an on/off switch. Turning the camera off is achieved by unplugging it from either the power source of the camera. Is this bad in anyway, or am I worrying for no reason?
The camera is actually powered off the USB 5V only. The power adapter is purely for the cooler. This camera has built in cooler throttling, and the heatsink has plenty of mass to absorb the heat load if you just turn the cooler off without a warmup period. So there should be no harm in just pulling the plug. However, it's probably more ideal to make sure the cooler is turned off in SGP first before unplugging.
One last question... when I connect my Orion SSAG, PHD within SGP will run it, correct? Or do I need to do that outside of SGP?
SGP works with PHP. You will need PHP to be running in the background...so you should just connect PHP normally.
So much to learn in so little time. Any hints, tips or help would be greatly appreciated.
With CCD cameras, you have column and cluster defects and RBI. With CMOS cameras, you have glows and FPN. Now with CCD cameras, read noise is often high enough that many of the sensor defects are hidden by the noise. With CMOS cameras, read noise these days is often so low that FPN starts out at a larger magnitude, or at the very lest it may only take stacking a dozen subs or so for FPN to grow larger in magnitude than read noise. This makes dithering a very critical part of imaging with CMOS cameras. You want to make sure you are dithering enough and dithering by large enough amounts to allow any remnant FPN in your calibrated subs to be averaged out. You want to dither at least 5 pixels each time.
Here is an example of FPN reinforcement with undithered subs (example created from dark frames to demonstrate worst case FPN):
And here is an example of dithering subs (example created from same set of dark frames, with random offsets applied using the dynamic crop tool in PI):
The scale of the dithers in this test case were 3-10 pixels. Note how the fine scale FPN (including the ray structures in the glows) have all been completely eliminated. Also note that there are some larger scale pattern structures that a 10 pixel dither was not quite large enough to average out. This is the worst case scenario, these are the effective corrolary to uncalibrated, read noise limited subs (i.e. narrow band subs from a dark site, with areas of true black sky that are read noise, and FPN, limited. With proper calibration, most of the FPN would be eliminated. Dithering will usually take care of the rest. If for some reason you still have some remnant FPN, deeper exposures to swamp more of the read noise can be effective, if an option...as a larger skyfog offset can effectively "bury" the FPN.
Depending on how many subs you plan to stack, you may be able to sparse dither. If you are stacking 80 subs, you might want to dither every frame. With 100 or more subs, you could dither every two frames. With 200 or more subs, you could dither every 3 maybe 4 subs. This applies per channel. You might acquire 100+ L subs but only 30 each for RGB. Dither L every 2 subs, but dither RGB every 1 sub.
Another key factor with CMOS cameras is the glows. Glows on the ASI1600 are not bad, and easily calibrated out. However, you must make sure you cali brate effectively. With exposures 90 seconds or shorter, you want to stack 36-49 frames per master. For exposures 120 seconds and longer, you probably only need 25 frames per master. When you calibrate, because of the potential for bias shading differentials, you should NOT use both a master bias and master dark to calibrate the lights. Use only a master dark (which will have the bias in it). Further, make SURE you are NOT scaling those darks (i.e. disable any kind of dark optimization). Dark scaling will missmatch the glows, and they will either not calibrate out, or could result in inverse calibration (the glows could become dark spots). Because you will never scale the darks, there is no reason to use a bias on them or the lights. Just use the master dark.
Here is an example of dithered darks that have also been calibrated by a 25-frame master dark:
Regarding flats. Flats will likely be fairly short. However, recent testing I did indicated that there is some quirky behavior with exposures shorter than 0.2 seconds with this camera. Beyond 0.2 seconds, the behavior of the camera normalizes:
Flats will require the bias offset to be removed, however you will want to make sure that both the flats and the bias frames are no shorter than 0.2 seconds to avoid issues with variable bias shading and incompatible offsets. Using biases of 0.2 seconds and flats of 0.2 seconds or longer, and the issues that were showing up in my recent testing disappeared.
One potential problem with this camera, and it does not occur for everyone, and may have as much to do with the scope as the camera, is microlens diffraction. In many cases it just shows up as a bit of diffraction spikes around brighter stars. For particularly bright stars, however, especially with large unobstructed apertures (i.e. large refractors), it can be a problem:
Something to be aware of. I generally just try to avoid stars this bright (this is Gamma Cass). I also think I have a bit of a worst case scenario, with a 150mm aperture f/4 refractor...bright stars with this scope can be INTENSE, a lot more intense with a 150mm aperture than say an 80mm aperture (even if both scopes were f/4, aperture counts more for stars since all the light is concentrated into a point on the sensor.) I cannot say with certainty that this is why I get more microlens diffractions than other people, it's just my working theory.