Messier 1 ; Short Exposure

[spare badger]

Taken late September 2022
224mc on a 10" f/4.5 newt
16800 x 1s exposures at 400 gain in 8 bit capture

Minor re-edit to better blend it into the background. Would like to see what others can get with the data.

[dense slate]

any reason you did 8 bit? Ive been doing similiar photography and always use 16. camera is 12 bit.

[spare badger]

Full well depth. At 400 gain, the 224mc has a depth of around 198e (according to a sharpcap sensor analysis I performed). This means that the maximum amount of unique individual values our camera will produce is 198 where 0 would be pure black and 198 would be pure white.

So our file only needs a bit depth that's precise enough to distinguish between these 198 values. Obviously our only options are 8 bit and 16 bit. 8 bit provides a precision of 256 distinct values while 16 bit provides a precision of 65536 distinct values. Since we're sitting at 198, there's no point in saving files at 16 bit precision when our data can't even make use of that precision.

So in practical terms, 8 bit is enough in this application. In addtion, we gain the benefit of smaller file sizes and faster processing times.

[dense slate]

ahh interesting. i normally run at 600 gain on my Uranus C, so i should be fine with 8 bit across the whole gain range then?

gain is that high so i can get enough stars to stack

[spare badger]

I have the Uranus-C as well and have ran an anlysis to find this. Any gain above 450 (rounding here for simplicity) falls bellow 256 units of precision so I'd advise using 8b above 450 gain and 16 below. So yes, I'd suggest 8b at 600 gain as our precision is only at 39e (according to analysis)

[dense slate]

alright. thanks a heap. you just saved my PC storage a lot of suffering.

[spare badger]

Np, I've only had one night with the uranus, capturing 8400 frames at full res and that produced 69gb of total data, way more than I ever got with the 224mc in my largest runs.

[dense slate]

6700 frames for me at full res 16 bit was 120gb. this will be a nice change.

[old trench]

Your sensor lineage is only 59%, this analysis is not to be taken as truth

You want minimum 70%+ for something vaguely trustable

[spare badger]

Yea but it matched close enough with what player-one has up so I was fine with it

[old trench]

I’m not convinced player one’s results are accurate

My analysis (linear to 92%) has different values

[spare badger]

what were your results?

[old trench]

https://media.discordapp.net/attachments/1057376148323057794/1063198544623849622/image.png

Stand corrected, 93.1%

I was doing this specifically to test hcg

So That’s why there’s concentrated values around 180

[spare badger]

I see, in that case I can just run the test again later to get better linearity, although I think it'll have little impact onto the decision of shooting in 8 bit at such high gain values. If the full well comes out lower, oh well, but if it comes out higher, great!

[gentle hinge]

I clipped the bg a bit, because why not 😅 my first time processing this object.

[chrome estuary]

No clippers

actually yes

[pure linden]

Here's what I got:

[latent smelt]

Gain doesn't change the number of star detectable afaik

[latent smelt]

Also bit depth is quickly gained back, especially with lot of subs

So the difference between 8 and 10 bit would be almost none

Bit depth is mostly important for long exposures

[uneven crescent]

Here you go, looks ok to me.

[deft fog]

@spare badger not 100% on colours yet so might work on that abit but heres what i got (: Super lovely data!

[deft fog]

still not sure but whateb lol

[raven violet]

went a little crazy with the sharpening on the starless image so the stars look a little funky compared to m1. But overall i like my edit, good data thanks for sharing.