egg|nomz|egg changed the topic of #kspacademia to: https://gist.github.com/pdn4kd/164b9b85435d87afbec0c3a7e69d3e6d | Dogs are cats. Spiders are cat interferometers. | Космизм сегодня! | Document well, for tomorrow you may get mauled by a ネコバス. | <UmbralRaptor> egg|nomz|egg: generally if your eyes are dewing over, that's not the weather. | <ferram4> I shall beat my problems to death with an engineer.
<B787_300>
!choose go to sleep now | try to stay up for 3 more hours
<Qboid>
B787_300: Your options are: go to sleep now, try to stay up for 3 more hours. My choice: go to sleep now
<B787_300>
UmbralRaptop: also the person who did the naming at IAD needs to be drawn and quartered
<B787_300>
Terminal naming that is
<UmbralRaptop>
They're just letters with the 20 middle ones missing, yes?
<egg|laptop|egg>
iximeow: aha, 2-by-2 binning seems to significantly improve star detection
<egg|laptop|egg>
(2-by-2 soft binning behind the debayer, to stab noise basically)
<iximeow>
ooo that's a good idea
<B787_300>
UmbralRaptop: a and b gates should not be in the same building... where the fuck did the x come from for the gates right next to the main building and where in the world are the T gates... I can understand missing letters
<B787_300>
egg|laptop|egg: naturally you are capturing more light from the star as you are grabbing more of the airy disk
<B787_300>
Although you do lose spatial resolution
<egg|laptop|egg>
B787_300: I mean, it's just a soft binning, ideally the stupid alignment algorithm would do that and align the unbinned images, but at least that way it aligns *something*
* UmbralRaptop
wonders what egg's pixel scale is.
<egg|laptop|egg>
also a moth got in >_<
<iximeow>
spacemoth!
<iximeow>
!wpn spacemoth
* Qboid
gives spacemoth a Stern-Gerlach neutrino-like tardigrade
<egg|laptop|egg>
(also a cat, but that's nice it warms the bed)
<B787_300>
UmbralRaptop: but seriously during my layover I wanted a burger and fries so went to the 5 guys (at one end of the terminal A/B bulding) and my flight left from the other end... silly me thought oh it is just one building and won't be that long of a walk. Then when I was almost there I noticed that I was in the middle of the B gates and had to rush eating to make my boarding time
<B787_300>
egg|laptop|egg: is it humane to use a cat as a dew shield or dew remover?
<egg|laptop|egg>
the cat might object, also you might not like the hairs on the primary
<B787_300>
UmbralRaptop: oh also I wanted soda with ice innit and free refils
<B787_300>
I mean if you have a cat 8n rhe same room as a telescope I think hairs on ever exposed surface is a given
<UmbralRaptop>
Yeah, getting water on the cat would probably annoy it. Unless you mean having the cat lie on the OTA?
<egg|laptop|egg>
B787_300: the cat is not on the opticks though
<UmbralRaptop>
B787_300:… is the soda fountain even customer accessible?
<egg|laptop|egg>
(well, one cat *is* the opticks)
<B787_300>
UmbralRaptop: no but you can ask and get it for free unlike in Europe
<UmbralRaptop>
hm
<UmbralRaptop>
egg|laptop|egg: have you considered radioisotope heat sources?
<B787_300>
he might need them if he takes his telescope back to Zurich with him during the winter break from school
<iximeow>
/win 22
<iximeow>
oop
<bofh>
!wpn egg|laptop|egg
* Qboid
gives egg|laptop|egg an inductive duck which vaguely resembles a theory of mind
<bofh>
00:19:37 < egg|laptop|egg> also a moth got in >_<
<bofh>
yeah, those are a thing.
<bofh>
B787_300: that reminds me, I kinda wanted to see how the Five Guys at Gare du Nord was like... but decided to have something more French like a reasonable human being instead. :P
<B787_300>
if you want an expensive good burger and more fries than you know what to do with... it is great
<bofh>
so, like, this is *me*, there is no such thing as "more fries than I know what to do with" :p
<B787_300>
i am still trying to wrap my head around why there is an Iran Air office on the Champs-Élysées like who is flying there for fun
<B787_300>
bofh: i dont think you understand... they give you the cup of fries then about the same amount just floating loose in the bag
<bofh>
B787_300: I don't think you understand, that is still not enough for me most times :P
<bofh>
like I can and will go thru arbitrary amounts of fries if I have enough salt.
<B787_300>
oh at least in the US there is PLENTY of salt on the fries
<egg|laptop|egg>
bofh yeah, those are a thing. << less so than when you were there, it's rainier
<bofh>
B787_300: I don't think you understand, my daily salt intake is like 2x the FDA recommended upper limit
<bofh>
(idiopathic hypoaldosteronism, both my BP and my blood Sodium levels are normal, and I'm prone to hyponatremia given excessive heat)
<bofh>
egg|laptop|egg: heh.
<bofh>
egg|laptop|egg: like, there were unusually little rains when I was in town iirc
* UmbralRaptop
finally gets around to reading SCP-280-JP and SCP-161-JP.
<UmbralRaptop>
It's really clever how the authors made up some terms for 161-JP (eg: sleeping)
awang has joined #kspacademia
* egg|laptop|egg
pokes M31
<egg|laptop|egg>
got most of it aligned, but the image is very meh
* egg|laptop|egg
zzz
<bofh>
egg|laptop|egg: let me poke at it, also yes double-double, also I should prolly get more than, like, an hour or two of sleep a night more regularly.
<egg|laptop|egg>
bofh: yes you should sleep
<bofh>
like I have an eggstremely good ability to remain functioning given truly absurd amounts of sleep dep, but even that has limits. it's *really hard* to get to them, but the past month has been crazy enough to pull it off. :p
egg|laptop|egg has quit [Ping timeout: 190 seconds]
* UmbralRaptop
wishes that one could just drink coffee instead of sleeping. =(
<UmbralRaptop>
Also eating. >_>
<bofh>
but food is tasty,
<UmbralRaptop>
Well, yes.
egg|cell|egg has quit [Ping timeout: 180 seconds]
egg|phone|egg has joined #kspacademia
pthag has joined #kspacademia
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
<kmath>
<eggleroy> @whitequark @sigfig Good luck e.g. figuring out that “Bahnrechnung” is “orbital computation” without knowing German… https://t.co/u0iDyfxkmb
egg|laptop|egg has quit [Ping timeout: 183 seconds]
egg|laptop|egg has joined #kspacademia
<egg|laptop|egg>
!wpn Ellied
* Qboid
gives Ellied a Понтрягин sunflower
<egg|laptop|egg>
!wpn bofh
* Qboid
gives bofh an unamused kimchi
<UmbralRaptop>
!wpn iximeow and iximeow's comet
* Qboid
gives iximeow and iximeow's comet a monomorphic underfull glaive with a toast attachment
egg|laptop|egg has quit [Ping timeout: 183 seconds]
egg|laptop|egg has joined #kspacademia
<egg|laptop|egg>
UmbralRaptop: wait, iximeow found a comet?
<awang>
UmbralRaptop: No antivirus installed beyond whatever comes with Windows
<UmbralRaptop>
hrm
<awang>
I seem to attract an unfortunate amount of system breakage
egg|phone|egg has joined #kspacademia
egg|laptop|egg has quit [Ping timeout: 183 seconds]
egg|cell|egg has joined #kspacademia
egg|phone|egg has quit [Read error: Connection reset by peer]
egg|laptop|egg has joined #kspacademia
* awang
hochos software that takes two minutes to list 7 databases from a server
UmbralRaptor has joined #kspacademia
UmbralRaptop has quit [Ping timeout: 183 seconds]
<egg|laptop|egg>
!wpn UmbralRaptor
* Qboid
gives UmbralRaptor a terbium quill
* UmbralRaptor
𓌪 a random town in Sweden
<egg|laptop|egg>
UmbralRaptor: what is the point of bias frames? isn't their signal already in the darks?
<egg|laptop|egg>
cc iximeow
* UmbralRaptor
tries to recall.
pthag has quit [Remote host closed the connection]
<UmbralRaptor>
egg|laptop|egg: I *think* it's to get the 0-point of each pixel, while darks are for dark current specifically. Unless you intend to do high precision photometry, darks can cover the bias frames.
<egg|laptop|egg>
doing 360 biases is quick, 360 darks seems like a pain, esp. if you are doing actually-long eggsposures rather than my tiny 10 s lights
<iximeow>
you should be able to reuse bias and darks, right? so that should be a one-off per camera?
<egg|laptop|egg>
yeah but it's bias(gain, temperature) but dark(eggsposure, gain, temperature)
<egg|laptop|egg>
so darks take longer and may have to be more diverse
<iximeow>
indeed
* iximeow
still needs to actually build out a library of calibration frames and stop hacking at images
<egg|laptop|egg>
"Unless you intend to do high precision photometry, darks can cover the bias frames." << yes but have you considered: cows
<iximeow>
high precision cowmetry
<iximeow>
telescope cowlimation
<UmbralRaptor>
Ideally you'd do them for every image, but the time constant for imaging systems is long enough that nightly is okay.
<UmbralRaptor>
egg|laptop|egg: are you setting up a laser guide cow?
<iximeow>
them == darks?
<egg|laptop|egg>
nightly darks? D:
<UmbralRaptor>
yes.
<egg|laptop|egg>
why how wat
<iximeow>
bias and darks are a function of the sensor (as well as gain, temperature, and eggsposure), right? if you maintain the same settings the sensor won't drift that much on a daily basis would it?
<iximeow>
(ofc depending on how precise you want to be..)
<UmbralRaptor>
yeah
<bofh>
that reminds me, I need to compute bias frames for my phone camera as well.
* egg|laptop|egg
might have given bofh that silly idea :-p
<UmbralRaptor>
Also some sort of sensitivity change from the sensor being whacked by cosmic rays.
<egg|laptop|egg>
!wpn -add:wpn ray
<Qboid>
egg|laptop|egg: Weapon already added!
<egg|laptop|egg>
!wpn -add:adj cosmic
<Qboid>
egg|laptop|egg: Adjective already added!
<UmbralRaptor>
bofh: Hah!
<egg|laptop|egg>
!wpn the sensor
* Qboid
gives the sensor a stochastic meow
<iximeow>
!wpn UmbralRaptor's camera sensor
* Qboid
gives UmbralRaptor's camera sensor an oxidized divergence
<egg|laptop|egg>
UmbralRaptor: bofh: a problem might be the unknown and variable temperature of the telephone
<bofh>
egg|laptop|egg: the sensor is right above the battery, so the battery thermistor is to first order a reasonable approximation of the sensor temperature
<bofh>
UmbralRaptor: literally any, I don't care, just be consistent? ¯\_(ツ)_/¯
<UmbralRaptor>
Ah, thanks
<iximeow>
", ", whatever you feel like
<egg|laptop|egg>
bofh: consistent is one thing, but does the function that chooses the grammatical gender have to be computable,
<bofh>
egg|laptop|egg: that is not a requirement, no,
<bofh>
Gender: sum of dyadic fractions with Busy-Beaver exponent,
<egg|laptop|egg>
bofh: I mean pretty sure that's how French works to start with
<egg|laptop|egg>
bofh: sum is feminine, so is fraction, beaver is masculine (both unrelated words for it, except the one that resembles beaver is often a river name and in that case it's feminine), exponent is masculine (but exponential is feminine of course)
* UmbralRaptor
pokes eg: Mars and Jupiter with sticks.
<egg|laptop|egg>
UmbralRaptor: yeah Mars and Jupiter are basically like the beaver-rivers
<bofh>
rofl
<egg|laptop|egg>
you say "the beaver" but it doesn't have the gender of beaver because it comes from "the river 'beaver'" which has the gender of river
<egg|laptop|egg>
there are *a lot* of rivers named after beavers
<bofh>
One would imagine the animal to be named after the hydronym, not vice-versa, but ¯\_(ツ)_/¯
<UmbralRaptor>
Proposal to call beavers river rats?
<UmbralRaptor>
Uh, the advice seems generally good?
<egg|laptop|egg>
UmbralRaptor: so where does that leave me with my 360 bias frames and 120 10 s darks that I just took
<egg|laptop|egg>
:-p
<egg|laptop|egg>
bofh: exponential is feminine << this is basically the same process, it's initially an adjective, the exponential function (fonction exponentielle, it has the feminine form where the masculine would be exponentiel), so it keeps the gender of function (and the feminine form) when it becomes its own substantive
<UmbralRaptor>
egg|laptop|egg: in a position to create masters of both with beaten down read noise?
<egg|laptop|egg>
well, not quite the same process, since when it's an adjective it doesn't have gender, whereas the beavers do
<egg|laptop|egg>
UmbralRaptor: slightly confused as to how
<bofh>
egg|laptop|egg: *okay* this is somewhat sensible now
<UmbralRaptor>
np.exp(beaver)
<bofh>
somewhat.
<egg|laptop|egg>
bofh: also a fun thing is foreign (e.g. technical) terminology because you tend to give it the gender of the translation, which depends on the translation you pick, so it's quickly a right mess
* UmbralRaptor
tries to recall how one does the adding when part of a Gaussian might not exist because you can't get a negative number of electrons.
<egg|laptop|egg>
e.g. I tend to intuitively think of a heap as being feminine, (probably because of pile or something like that? not sure), but both translations (fr tas, fr-CA monceau) are actually masculine
<egg|laptop|egg>
a stack is unambiguously "pile" and feminine
<UmbralRaptor>
bofh: for the purposes of this, p-type semiconductors don't eggsist.
<UmbralRaptor>
I guess the easy approach is if you have software that says "make master dark" or the like?
<egg|laptop|egg>
UmbralRaptor: median or mean or mean of median half of image set,
<egg|laptop|egg>
(and with or without leveling image means
<egg|laptop|egg>
)
<egg|laptop|egg>
probably very much *not* leveling image means now that I think of it, pretty sure that's why my eggsisting dark was so screwy
<UmbralRaptor>
Going to guess median
<bofh>
UmbralRaptor: so like, chemical potential *can* be negative...
<egg|laptop|egg>
help I cannot looks at FITS manipulation software without being reminded of котяchrome whenever I see monochrome
<egg|laptop|egg>
s/looks/look
<Qboid>
egg|laptop|egg meant to say: help I cannot look at FITS manipulation software without being reminded of котяchrome whenever I see monochrome
* UmbralRaptor
pokes Gibbs in the free energy
<egg|laptop|egg>
UmbralRaptor: but if I subtract the master bias from the dark frames and from the light frames and then the master dark from the light frames, what did the bias do O_o
<bofh>
Uhh... yeah, what. That seems strange.
<bofh>
like for that to do nothing subtraction would have to fail to be associative (which *is* the case for floating-point, but images are integer...)
<egg|laptop|egg>
UmbralRaptor: if the dark is computed by mean, this does strictly: calibrated light frame = raw light frame - master bias - Σ(dark_i - master bias) / n_darks = raw light frame - master bias - Σ(dark_i) / n_darks + master bias
<UmbralRaptor>
presumably you can rescale one or the other separately?
<egg|laptop|egg>
bofh: UmbralRaptor: maybe if the darks are computed by median it matters?
<egg|laptop|egg>
UmbralRaptor: rescaling sounds like it adds more variables to the elephant
<egg|laptop|egg>
which seems bad if you want to get photometry out of it
<egg|laptop|egg>
bofh: UmbralRaptor: wait, even with medians? median_i(dark_i - master bias) is going to be median_i(dark_i) - master bias right?
<egg|laptop|egg>
bofh: I think the images are integer (though FITS works in mysterious ways), but in any case if you run into cancellation issues you are not carrying enough precision for your image manipulation
<bofh>
yeah even with medians what you said should hold, yeah.
<egg|laptop|egg>
bofh: UmbralRaptor: okay so what the hell is bias for then O_o
<bofh>
no idea!
<egg|laptop|egg>
I'm beginning to think that this SIPS comment actually makes sense and that bias is useless >_>
<bofh>
likewise...
<egg|laptop|egg>
(and that otoh that means that you need to take a shittonne of darks to not amplify read noise from your dark, especially for short exposures)
<egg|laptop|egg>
eggsposures
awang has joined #kspacademia
<egg|laptop|egg>
I should sleep, I need to wake up early tomorrow
<egg|laptop|egg>
going back to zurich
<bofh>
enjoy~
<egg|laptop|egg>
UmbralRaptor: hm, I'm not sure median is ideal, it tends to lose hot pixels that actually appear on means of disjoint sets
<egg|laptop|egg>
hm, on a sufficiently large set it works