►
From YouTube: SJAA Astro Imaging Special Interest Group 2020 10 20 StarNet++ and Pixinsight, Planetary imaging
Description
We're very happy to have our member Francesco Meschia showcasing some techniques that he used to clean up and bring out more detail in his recent Gecko Nebula Image. Francesco will use StarNet++ and Pixinsight to correct a dust mote while still in the linear stage and also will be showing us how he enhanced the nebula while keeping the stars under control. Glenn Newell will also be sharing his experiences imaging and processing Mars during the amazing opposition that we've all been enjoying so much!
Check out Francesco's Gecko Nebula on his Astrobin page at:
https://www.astrobin.com/9nlqwd/H/?nc=user
Also member Glenn Newell talks about planetary imaging
Glenn's astrobin: https://www.astrobin.com/users/dts350z/
A
And
I
also
know
that
it
would
be
would
be
a
challenge
to
apply
the
usual
star
reduction
or
start
the
emphasis
techniques,
because,
in
addition
to
a
large
number
of
smaller
stars,
there
are
some
big
guys
here
and
those
typically
are
bad
customers
for
for
star
reduction
and
start
the
emphasis,
and
so
let's
see
what
I
did
to
to
fix
those
two
issues
and
obtain
a
nice
image
in
the
end.
So
I'm
switching
now
to
a
few
slides.
A
So
I
use
the
star
net
for
both
of
those
issues
and
the
star
net
is
a
is
essentially
an
application
of
ai,
artificial
intelligence
to
astrophotography.
This
is
a
quote
from
the
project.
Page
starting
plus
plus,
is
a
neural
network
that
can
remove
stars
from
images.
Now,
I'm
not
a
machine
learning
guy.
So
I
I
barely
know
what
it
means.
Nikola
can
probably
talk
much
more
than
I,
and
then
I
can
about
this,
but
suffice
to
say
that
this
is
a
program
that
has
been
around
for
two
years
about
two
years.
A
Initially,
it
was
a
standalone
tool
that
you
had
to
run
from
the
command
line.
Few
people
were
using
it,
then
it
got
integrated
in
the
pixel
site
and
it
arrived
on
the
mac
about
one
year
ago,
thanks
to
rob
who
built
the
the
first
star:
net
module
inside
pixel
site
and
most
recently.
Finally,
the
the
play
disaster.
Collaboration
has
included
it
in
pixinsight11811.
A
Sorry,
one
eight,
eight,
six,
okay,
I
got
it
so
what
does
a
star
net
plus
plus
do
so,
if
you,
if
you're,
not
already
familiar
it,
does
this
it
goes
from
here
to
here.
How
does
it
do
that?
Well,
somebody
taught
this
neural
network
how
to
how
an
image,
what
is
the
difference
between
an
image
with
stars
and
an
image
without
stars
and
the
this
neural
network
learned
and
so
can
apply
the
same
process
to
almost
an
image
that
you
can
throw
at
it.
A
I
personally
like
to
see
both,
but
I
like
to
be
in
control
of
the
mix
I'd
like
to
decide
how
much
I
sealed
the
nebula
and
much
I
see
the
stars
and,
as
I
learned,
how
to
integrate
the
star
net
plus
plus
in
my
pixel
size
workflow.
I
learned
that
there
are
cases
in
which
the
starless
version
of
an
image
is
a
very
useful
intermediate
step
and
I'm
going
to
present
here
these
two
cases.
The
first
one
is
the
correction
of
dust
mods.
A
Now?
You
can
use
some
painting
process
like
the
clone
clone
painting
in
pixel
side,
but
the
problem
is:
how
do
you
do
that
without
going
over
the
stars
and
by
copying
the
background
from
another
region,
you
can
mask
the
star,
but
creating
a
good
star
mask
is
already
a
task,
and
also,
even
even
when
you
put
a
star
mask
if
you
copy
clone
from
one
region
of
to
another,
the
source
region
has
stars.
You're,
just
gonna
paint
the
stars
onto
the
destination
just
by
excluding
the
the
other
stars.
A
A
Then
we
can
paint
over
the
defect
and
in
this
case
it's
easy
because
the
stars
are
not
there.
So
you're.
You
can
just
find
a
region
of
the
background
which
is
similar
in
illuminance
to
the
one
that
you
to
where
you
have
the
defect
and
just
clone
over,
maybe
using
some,
not
100
opacity.
So
you
can
graduate
your
rear
effect.
A
Then
you
add
back
the
stars
and
you
obtain
the
the
equivalent
of
the
original
of
the
image
that
you
started
from.
But
without
the
defect
that
you
just
corrected,
yes,
easier
said
than
done,
and
the
first
problem
is
that
star,
netplus
plus
only
works
on
stretched
images,
but
it
would
be
awesome
if
this
kind
of
correction
could
be
done
in
the
linear
stage,
because
it
would,
it
would
help
us
in
the
in
the
process.
A
A
How
do
we
do
that?
Well
in
the
way
I
do
it.
I
start
with
the
by
copying
the
settings
for
an
auto
sdf
string
transfer
function
into
the
histogram
transformation
tool,
like
everybody
does
when
you
wanted
to
do
an
initial
initial
stretch,
but
be
very
careful
tweak
it
so
that
you
clip
as
little
as
possible.
As
you
can
see
here,
I'm
clipping
only
227
pixels
out
of
over
6
million,
so
I'm
probably
still
okay
and
when
you
have,
when
you're
settled
on
the
right
transformation,
be
sure
to
write
down
or
copy
from
the
history
window.
A
These
two
parameters,
the
shadows
setting
and
the
midi
tones
settings
which
are
shown
here
in
the
window
and
here
in
the
source
code
for
the
transformation.
This
will
come
handy
later,
because
at
this
point
we
just
apply
all
the
corrections
that
we
want
and
then
we,
the
way
I
do
it
is
by
creating
a
first.
I
use
a
starter
to
create
this
a
star
net
version
and
the
difference
between
the
original
image,
which
I
call
temp
here
and
the
star
net
version.
A
A
So
the
way
I
do
it
is
that
to
recombine
the
two,
the
the
two
images,
the
the
one
that
I
painted
over
and
the
star
difference,
using
some
pixel
math.
That
says
okay,
add
the
two
but
then
also
add
sorry
sub.
Add
button
algebraically.
So
it's
going
to
be
subtract,
the
difference
between
what
I
painted
and
the
original.
A
If,
if
there's
actually
something
there,
if
the
level
of
the
difference
I.e,
the
stars
are
more
than
0.02,
and
this
is
arbitrary.
You
can
you
can
do
by
trial
and
error.
You
can
find
the
the
value
most
suitable
for
you
and
the.
Finally,
you
reverse
this.
The
stretch
transformation
using
this
pixel
math
expression,
which
I
didn't
invent.
A
I
found
this
on
a
I
believe
in
the
synthetic
form,
and
you
can
see
this
expression
is
being
fed
with
the
shadows,
level
and
mid-tones
level
that
I
noted
down
in
the
previous
step
here
now
enough
with
this
talking
and
I'm
going
to
show
you
how
it
works
in
in
peaks
inside.
So
this
is
the
the
image
that
I
wanted,
so
I
I
cloned
it
and
I
cloned
it
in
into
this
this
image
gear.
A
A
A
B
A
A
A
It's
basically
me
painting
over
the
defect.
Now
it's
questionable
whether
you
should
paint
over
an
image.
Of
course,
once
you
do
it,
you
have
given
up
any
pretense
of
this
image
being
a
fatal
representation
of
the
sky.
There's
no
more
science
here
this
is.
We
entered
the
realm
of
the
artistic
interpretation,
so
I
I
apologize
if,
if
somebody.
A
A
When
applied,
it
gives
a
more
fit
and
more
a
representation
of
the
stars
with
in
their
brightness
values,
which
is
more,
is
a
closer
to
what
they
were
before
micro
reaction.
So
essentially
I
corrected
the
the
trough
represented
by
the
dust
mode
without
touching
the
stars
or
almost
without
touching
the
stars
at
this
point,
this
this
temp
image
is
still
stretched.
A
I
have
to
reverse
the
stretching
transformation
to
put
it
here
and
I
do
it
directly
onto
the
this
integration
image,
one
my
main
one,
using
a
pixel
map
using
the
same
expression
that
I
showed
you
and
when
I
do
it,
the
dust
mod
is
gone
and
the
image
is
linear,
and
so
I
have
achieved
the
first.
The
first
result
that
I
wanted,
which
is
to
do
the
correction
still
in
the
while
still
in
the
linear
stage.
If
you
look
at
the
histogram
of
this
image,
the
histogram
is
all
compressed
towards
the
left.
A
A
If
I
go
back
and
forth
in
history
undo
and
redo,
you
see,
the
histogram
does
not
change,
which
means
that
I'm
not
clipping
any
anything
important
in
this
image,
which
is
again
one
of
the
things
that
I
wanted
to
achieve.
My
correction
has
not
has
altered
only
the
part
of
the
image
where
I
wanted
to
intervene
and
has
left
everything
else
unaltered,
so
that
once
I
once
have
done
that,
I
can
continue
going
through
the
history
of
this
application.
A
A
So
my
goal
is
to
stretch
the
image
so
that
in
the
end
I
preserve
the
stars.
I
preserve
good
stars
and-
and
I
don't
end
up
with
bloated
and
washed
out
stars-
it
would
be
so
nice
if
we
could
stretch
the
nebula
separately
from
the
stars
so
that
we
do
the
best
for
each
of
the
two
aspects.
The
two
sides
of
the
image
well
star
net
plus
plus
allows
you
to
do
exactly
that.
So
what
I'm
proposing
is
to
start
with
a
very
gentle
stretch
in
this
case
I'm
using
arcsinh.
A
That
gives
you
with
the
goal
of
giving
you
the
stars
that
you,
like
don't
think
of
the
nebula
focus
on
having
the
presence
of
the
stars
that
you
will
be
comfortable
with
in
the
final
image.
So
this
is
what
this.
The
initial
stretch
that
I
wanted,
the
nebula
is
almost
invisible,
but
if,
when
you
apply
sternum
plus
plus
you
get
one
image
with
only
the
nebula
and
one
image
with
only
the
stars,
which
is
essentially
the
difference
between
this
and
this.
A
At
that
point,
you
can
process
the
starters
version.
The
way
you
like
optimizing,
the
visibility
of
the
nebula
and
when
you
combine
the
starless
and
processed
image
with
the
original
star
field,
you're
going
to
obtain
an
image
in
which
the
stars
have
the
the
same
level
of
presence
that
they
had
where
you
started,
but
the
the
nebula
is
much
more
visible.
A
A
A
A
A
A
You
know
if
I
hadn't
stretched
anything
they
would
just
recombine.
They
would
be
one
the
mold
of
the
where
the
other
was
taken
from,
but
after
the
stretching
done
this
is
this
no
longer
happens.
The
two,
the
artifacts
in
the
starburst
in
the
starfield
version
and
the
status
version
no
longer
cancel
each
other,
and
you
have
to
do
something
to
avoid
that
and
the
way
I
do
it
is
to
create
the
star
mask,
which
essentially
picks
starnet
already
gave
me.
A
I
just
needed
to
binarize
it
and
apply
a
little
bit
of
morphological,
transform
to
isolate
only
those
artifacts,
because
the
artifacts
are
under
the
stars,
and
so,
if
you
create
a
mask
ventilate
the
stars,
you
have
effectively
isolated
the
artifacts
and
then
I
smooth
them
using
a
little
bit
of
mmt
multiscale
medium
transform
by
reducing
the
detail
over
some
of
the
of
the
layers
keep
in
mind.
This
is
only
effect.
This
is
only
going
to
affect
the
regions
where
the
stars
were
where
the
artifacts
are.
A
It
requires
some
trial
and
error,
but
in
the
end
you
can
smooth
them
out
pretty
good.
You
don't
want
them
to
be
super
smooth.
You
want
them
to
have
a
the
same
grain
or
grainness
as
the
the
rest
of
the
age,
at
least
that's
what
works
for
me
once
this
is
done.
You
can
recombine
the
two
images
and
the
effect
will
be
pretty
clear.
It's
pretty
pleasant,
so
let
me
go
back
to
pics
inside
and
show
you
what
I've
done
here.
A
So
the
first
thing
I
do
is
an
initial
morphological
transformation
and
it
creates
this
kind
of
result.
Still
very,
very
gentle.
The
nebula
starts
to
be
visible,
but
it's
not
yet
really
clear.
Then
I
do
a
more
a
curves
transformation
to
make
it
more
more
evident,
and
at
this
point
I
go
and
inspect
the
artifacts.
B
A
A
A
A
Let
me
show
you
what
the
result-
and
this
is
the
result
at
this
point-
the
only
other
adjustments
that
I
did.
I
did
some
sncr
to
remove
green.
I
had
my
original
image
was
rather
green,
so
this
is
possible
that
there's
some
green
left
behind
and
then
I
saturate
the
stars
and
that's
it
at
this
point
I
have
an
image
in
which
the
stars
are
nice
and
visible,
but
they
are
not
overwhelming
the
nebula.
Actually,
maybe
the
nebulizer,
I
mean
the
stars.
A
So
if,
if
a
possible
critic
that
would
be,
that
would
be
it,
but
there's
something
else
that
I
want
to
talk
to
you
about,
which
is
not
that
apparent
in
this
in
this
image.
So
let
me
go
back
to
the
slides,
so
the
final
consideration
is
whether
we,
this
image,
is
still
a
true
representation.
We
throw
in
multiple
multiple
quotes,
or
we
have
crossed
over
into
artistic
interpretation,
so
consider
that
when
we
summon
the
stretch,
the
starless
image
with
the
unprocessed
star
field,
you
end
up
with
the
stars
riding
on
the
nebula
pedestal.
A
Given
that
we
have
stretched
and
a
bit
of
pedestal,
we
it's,
it
is
no
longer
what
it
used
to
be.
This
means
that
the
the
stars
that
happen
to
sit
in
the
nebula
and
that
being
brighter
than
the
stars
that
are
just
sitting
on
the
background
sky,
we
have
effectively
altered
the
brightness
relationship
among
the
stars.
A
In
in
theory,
we
could
apply
a
technique
similar
to
the
one
that
I
told
you
about
when
I
corrected
the
dust
molds
and
I
can
go
and
and
subtract
the
corrections
that
I
did,
but
that
I
did
by
stretching.
But
this
creates
another
interesting
problem,
so
this
would
maybe
restore
the
brightness
relationship
among
stars,
but
it
would
completely
alter
the
relationship
between
the
nebulae
and
the
stars
that
are
circling
close
to
them.
A
So
my
the
way
I
I
came
to
peace
with
this,
this
problem
is
that
I
believe
that
once
we
take
the
start
and
plus
plus
path,
we
leave
the
path
of
having
a
faithful
representation
of
the
brightness
and
the
relationships
at
the
point
that
there
is
no
right
way.
It
boils
down
to
what
we
want
to
represent
with
our
image.
A
A
A
Let's
say
the
the
image
is
still
hints
at
a
relationship
between
the
brightness
of
the
nebula
and
the
stars:
it's
not
as
extreme
as
it
was.
I
corrected
it
to
make
it
more
visible,
and
I
did
it
because
I
wanted
to
make
the
nebula
accessible
to
the
to
the
viewer,
and
I
I'm
open
to
sacrificing
a
little
bit
of
the
brightness
with
the
correct
brightness
relationships
for
that.
A
A
A
B
B
C
A
A
A
F
A
A
A
E
A
A
A
A
E
C
G
A
I
just
add
them
together:
yeah,
I
don't
do
a
maxing
you
could
you
could
do
the
maximum?
I
suppose
it's
a
the
issue.
Is
that
the
when
you,
if
you
start
with
the
with
the
stars
that
are
super
imposed
on
on
a
galaxy
in
this
case
the
when
you
look
at
the
the
starlets
versus
the
sorry,
the
starfield
version
that
since
that
starnet
has
produced
those
stars,
the
the
the
amplitude
of
the
peak
would
be
depressed
by
the
same
amount
that
has
ended
up
in
the
starless
version.
A
C
H
C
A
There's
a
something
that
I
I
didn't
mention
in
the
when
we
were
looking
at
the
gecko
images
that
as
long
as
you're
in
the
linear
domain,
you
can
always
do
a.
I
mean
when
you
do
a
subtraction
or
a
sum.
Those
are
perfectly
legitimate
operation
because
you
are
summing
or
subtracting
essentially
numbers
number
of
the
photons
of
electrons
in
the
in
the
pixel
wells
after
the
image
is
stretched
the
by
subtracting
or
adding
you
are
no
longer
adding
linear
data.
So
in
theory
those
are
should
not
be
legitimate
operations.
A
A
Probably
exponential
transformation
that
was
applied
for
the
stretch
then
do
the
sum
and
then
sorry
apply
it
in
reverse.
Do
the
sum
of
subtraction
and
then
apply
it
forward,
but
it's
in
the
end.
I
I
never
saw
any
much
difference
earlier
at
the
levels
of
brightness
that
we
have.
The
stars
are
pretty
close
to
the
saturation
point.
A
It's
not.
The
difference
is
not
the
difference
between
a
linear
version,
a
locally
linear
version
and
the
the
actual
linear
original
linear
data
is
not
that
much
meaning
that
you're
using
a
linear
approximation
in
a
in
a
small
in
a
small
portion
of
the
histogram,
and
that's
probably
okay,
but
again
if
conor
hero
would
see
this,
he
would
say.
No,
that's
that's
wrong.
C
H
A
H
Well,
technically,
star
net
is
actually
inventing
data,
because
the
neural
network
actually
generates
the
image
from
random
noise
right.
It
actually
takes
the
original
image
and
it
takes
a
random
noise
and
from
the
random
noise.
It
creates
another
image
and
that's
why
we
see
the
artifacts
on
the
stairless
nebula
right,
yeah
yeah
actually
invents
another
nebula
from
random.
A
C
C
Yeah
I
mean
I
agree
that
I
mean
you
guys
all
know
about
this
documentary
school
manifesto
that
juan
wrote
right
I
mean
years
ago
that
caused
a
lot
of
stir
and
he,
you
know
different
form
threads.
He
still
comes
back
to
it
exactly
what
francisco's
saying
about
inventing
data
or
painting
and
all
that
stuff,
and
so
people
get
angry.
C
Well,
it's
not
so
much
about
science
and
as
far
as
I
can
understand,
but
it's
really
about,
like
you,
know
the
invention
of
data
so
like
like,
for
instance,
what
eric
coles
was
showing
like
just
grabbing
a
part
of
the
image
with
the
lasso
and
changing
the
colors
like
juan's
head
would
explode.
I
mean
you
just
that's
that's
right
out
right.
I
think
he
he
said
many
times
that
like
well.
C
If
you
you
know,
if
you
come
up
with
some
mask
for
the
image
it's
derived
from
the
luminance
or
that
has
some
relationship
to
the
image
and
then
you
apply
some
operation.
You
know
somewhat
globally
through
that
mask,
then
you
know
that's
okay,
but
I
don't
think
he's
ever
really
said.
It's
quote.
Unquote:
science.
I
mean
you're
right
the
minute.
C
H
C
A
C
G
G
G
G
And
in
contrast
again
to
taking
pictures
of
nebulae
or
galaxy
you're,
gonna
use
extremely
short
exposures.
The
the
shorter
the
better,
because
you're
doing
what's
called
lucky
imaging
where
you're
trying
to
just
catch
an
instant
of
good,
seeing
over
your
whole
image
or
over
part
of
your
image.
Maybe
and
you're
also
going
to
be
sort
of
grossly
over
sampling,
with
lots
of
magnification,
which
leads
to
a
high
focal
ratio
and
so
you're
using
a
barlow
versus
the
native
resolution
of
your
scope
or
using
a
focal
reducer.
G
G
G
G
So
it's
a
positive
lens
out
there
beyond
your
your
focuser
between
the
focuser
and
the
camera,
and
that
can
give
you
magnification
and
they
come
in.
You
know
two
three
four
and
some
of
them
are
are
adjustable
and,
as
I
said,
it's
it's
all
about
the
the
frame
rate.
So
you
want
to
look
for
a
camera
that
can
deliver
the
fastest
frame
rate
and
it
that
also
has
to
do
with
the
you
know.
You
use
just
a
fraction
of
the
sensor
area
or
it's
called
a
region
of
of
interest.
G
So
this
asi
174,
if
you
just
look
at
the
the
zwo
cameras,
it's
the
one
that
has
the
fastest
frame
rate
and
bruce
and
I
have
been
achieving
around
150
156
frames
per
second,
when
we're
using
a
small
region
of
interest
with
that
that
camera
and
then
another
rule
of
thumb.
You
know
I
talked
for
a
second
about
the
the
d
rotation.
G
If
you
don't
want
to
go
through
that
step,
you
just
need
to
limit
the
the
time
frame
that
you
capture
your
data
over,
so
someone's
posted
here
some
rules
of
thumbs
for
the
different
planets
based
on
their
rate
of
rotation.
Now
this
has
some
effect
of
where
they
are
in
in
their
orbit
in
relation
to
to
earth,
but
so
these
again
general
rules
of
thumb.
So
for
mars
you
know
if
you're
going
to
be
taking
videos
per
filter-
or
you
know
maybe
five
of
any
one
filter
or
what?
What
have
you?
G
You
know,
each
of
those
videos
should
not
be
longer
than
five
minutes
and
the
whole
time
that
you
spend
taking
those
multiple
videos
should
not
be
more
than
than
15
minutes,
and
then
they
have
jupiter.
Apparently,
moves
much
faster
in
terms
of
the
the
features
and
then
saturn
is
somewhat
in
in
between
there.
So
those
are
some
some
rules
of
thumb.
G
Okay,
you're
going
to
be
using,
you
know.
Instead
of
neenah
or
sharp
cat,
I
mean
nina
or
voyager
or
sgp,
or
I
don't
know,
if
I'm
sure
high
will
chime
in
here.
If
k
stars
does
does
planetary
or
not,
but
rather
than
using
those
deep
space,
long
exposure
programs
you're
going
to
be
using
something
like
sharp
cap
or
or
fire
capture,
and
I
actually
ended
up
using
using
both
for
different
features
that
they
had
and
then
go
ahead.
No.
G
I
don't
think
so:
okay
for
stacking,
the
the
popular
one,
these
days
is
auto
stacker,
it's
free
and
then
one
that
used
to
be
popular
for
stacking
and
now
is
just
used
for
some
later
processing.
Stat
steps
is
registax,
which
is
also
free
and
that's
used
mainly
for
the
wavelet
sharpening
and
then
in
the
particular
workflow.
I
found
that
that
I
used.
G
So
that's
interesting
and
then
there's
a
pipp
which
can
do
you
know
overlaps
with
some
of
these
and
some
people
use
it
just
as
a
as
a
preprocessor.
They
run
the
videos
through
that
and
do
some
operations
first
before
they
go
into
the
into
the
stacking
and
then
photoshop
or
what
have
you
for
for
doing
some
graphics
type
stuff,
the
particular
workflow
for
mars
that
I
came
across.
G
So
I
stumbled
on
this
london,
astronomer
guy
and
that's
the
name
of
his
website
and
you
can
see
his
work
there
on
the
right,
not
too
shabby
pretty
amazing
stuff
and
in
fact
even
his
daughter
was
one
of
the
winners
of
the
young
astrophotographer
of
the
year
award
this
year
with
something
she
did
on
the
sun.
I
believe
so,
actually,
that's
a
good
family
of
people
to
to
follow.
So
he
has
a
general
overview
of
planetary
imaging,
which
is,
you
know,
super
more
in
depth.
G
Then
I'm
not
even
really
going
to
go
into
it
here,
because
you
can
just
go,
read
it
and
then
he
has
specifics
on
mars
and
the
moon,
and
I
noticed
today
it
looks
like
he's
added
one
for
venus
as
well.
So
you
know
you
can
pick
your
expert,
but
this
is
one
one
expert
that
has
a
pretty
detailed
and
step-by-step
instructions
on
the
whole
workflow
of
image,
capturing
and
processing.
G
You
know
it
involved
lots
of
3d,
printing
of
of
mounting
pieces
and
adapter
pieces.
You
know
as
prototypes,
ideally
you'd
replace
them
with
with
metal
at
some
point,
but
and
lots
of
or
not
lots
of,
two
different
arduinos
for
doing.
I
made
a
focuser
for
the
the
club,
one
of
the
clubs
mac
and
I
made
automation
for
the
flip
mirror
so
that
I
could
operate
the
flip
mirror
from
in
the
house
where
I
normally
operate
the
the
telescope,
and
you
know
I
went
through
two
different
cameras.
G
I
had
a
a
camera
that
I
actually
won
at
as
a
door
prize
at
the
last
aic,
and
I
I
don't
know
I
I
feel
like
I
got
what
I
paid
for,
because
it
really
hasn't
done
anything
it
just.
It
doesn't
work
very
well
for
me,
so
I
switched
to
the
it's
a
one
shot.
Color
was
why
I
was
trying
to
use
it,
but
I
switched
to
a
filter,
wheel
and
the
asi
174,
and
then
I
did
end
up.
G
You
know
I
started
on
my
12-inch
rc
and
then
I
went
to
the
the
mac
cavs.
You
know
the,
even
though
the
aperture
is
so
much
smaller.
That
matt
kaz
is
supposed
to
be
a
good
optical
design
for
planetary,
but
then
I
ended
up
with
bruce's
encouragement
going
back
to
the
to
my
12-inch
rc
and
cranking
up
the
magnification
either
either
further,
and
I
spent
lots
of
time
you
know
coming
to
focus,
and
maybe
I
over
complicated
things
by
including
this
flip
mirror.
G
I
guess
I'm
I'm
getting
a
little
bit
ahead
of
myself
here
and
and
one
of
my
learnings
was
you
know,
even
if
I
I
got
everything
in
focus
during
the
daytime,
on
the
farthest
terrestrial
object
that
I
that
I
have,
which
is
trees.
That
are,
you
know
at
least
two
three
blocks
away
then
turns
out
at
night
for
stars
and
planets.
That's
I.
G
G
You
can
really
tell
that
it's
a
globe
and
it's
got
the
little
shine
on
it
there,
like
you,
know
this
sun
reflecting
off
a
globe
versus
kind
of
a
flat,
even
though
it
might
have
more
detail,
and
then
you
know
the
last
night
this
was
you
might
think
this
one
was
the
first
night,
but
it's
actually
the
last
night.
I
guess
the
scene
was
just
really
crap,
but
that's
what
happened
on
the
last
night,
so
okay
keyboard
focus
here.
G
G
G
F
G
G
G
B
G
G
G
B
G
Okay,
so
this
is
just
showing
the
the
flip
mirror
moving
there.
If
you
look
right
there,
you
can
see
the
knob
turning
back
and
forth,
so
that
was
convenient
to
be
able
to
operate
that
from
from
in
the
house,
and
this
is
before
this
is
see,
there's
that
atec
gp
on
the
imaging
end,
and
I
had
the
asi
174
on
the
on
the
flip-
mirror
side
of
things.
The
mirror
side
of
things.
But
after
this
I
switched
switch
cameras.
G
Yeah,
so
this
is
the
there's
a
lot
fewer
pieces
of
plastic
in
here
now
that
I've
replaced
some
of
them
with
with
metal,
as
things
have
have
shipped,
but
you
can
see
just
on
the
back
end
of
the
flip
mirror,
which
is
the
square
cube.
Here
is
the
barlow,
and
this
is
an
astrophysics
barlow,
that's
somewhat
adjustable
by
a
like
a
focal
reducer
by
adjusting
the
amount
of
space
between.
G
G
G
Okay,
so
yeah
so
adjusting
this.
These
spacers
here
would
adjust
the
amount
of
magnification
and
then
you
could
add
or
remove
spacers
here
between
the
flip
mirror
and
the
focuser
to
come
to
focus
but
like
a
focal
reducer.
When
you
change
this,
the
focal
plane
shifts
and
it's
supposed
to
shift
out
as
you
increase
magnification
as
a
focal
reducer,
would
shift
in
as
you
increase
focal
reduction,
but
I
think
that
it
shifted
out
less
fast
than
the
amount
of
extension.
G
G
B
G
G
So
the
apparent
size
of
mars
is
published
both
in
stellarium,
for
instance,
and
also
it's
been
in
articles
because
of
mars
being
at
opposition,
so
yeah,
and
then,
if
I
find
the
right
piece
of
paper,
you
know
post-it
note
on
my
desk
here
somewhere,
it's
going
to
say,
but
from
memory
it
was
you
know
it's
something
like
26
arc
seconds
or
something
I
don't
know
anyway.
That's
that's
how
it
that's
how
it
worked
out
so
yeah.
G
I
was
able
to
figure
out
exactly
what
it
was,
but
I
didn't
put
that
in
the
presentation,
so
some
some
lessons
that
I
learned
along
the
way
plan
for
plenty
of
time
just
to
find
mars
and
get
it
on
your
sensor
that
can
take.
You
know
the
first
time
you
do
it.
It
might
take
you
like
an
hour
or
more
just
to
just
to
do
that.
Simple
little
thing.
G
G
So
it
really
wasn't
as
effective
as
I
thought
it
should
be
because
you
know
basically,
I
shot
laser
through
there
and
you
know
it's
not.
The
laser
is
not
coming
to
the
the
center
in
both
arms
of
the
the
focal
reducer
or
even,
if
you,
you
know,
shoot
a
laser
down
the
optical
path
of
the
scope.
It's
not
even
centered
on
the
on
the
sensor,
so
lots
of
lots
of
sag
issues.
So
I
went
about
replacing
all
the
prototyped
plastic
parts
with
with
metal
and
you
know
try
to
just
the
normal
stuff.
G
And
you
know
once
I
got
things
you
know
stiff
and
rigid
and
and
dialed
in
then
I
pretty
much
could
go
directly
from.
I
was
using
a
a
small
refractor
mounted
on
the
top
of
my
rc
to
you
know
I
would
plate
solve
on
the
mars
and
then
go
on
to
the
directly
to
the
planetary
camera
and
do
a
spiral
search
and
find
it,
and
I
so
I
didn't
in
the
end.
G
I
wasn't
really
using
the
the
flip
mirror
at
all,
so
I
did
find
that
speaking
of
spiral
search,
I
did
find
that
the
spiral
search
in
eq
mod
is
pretty
horrible.
If
you
look
at
it
on
a
small
scale
like
around
a
planet
like
mars,
I
was
very
jerky
and
uncontrolled,
and
I
found
that
the
one
in
sharp
cap
worked
much
better
and
I
I
set
it
at
16
x
and
that
seemed
to
be
a
good
compromise
between
whoops
there.
G
It
went-
and
you
know
taking
forever
one
thing
that
I
did
struggle
with
and
maybe
reading
the
manual
or
something
would
would
help.
But
as
soon
as
I
let
off
the
button
in
sharpcap
that
spiral
search
button,
the
the
scope
would
stop
tracking.
So
I
had
to
like
as
soon
as
I
let
off
the
button,
then
I
had
to
quickly
go
to
eq
mod
and
press
sidereal
tracking
rate
so
that
it
didn't,
you
know,
fly
out
of
the
frame,
the
other
direction
to
the
west.
So.
G
And
I
didn't
see
you
know,
I
ended
up
using
fire
capture
for
the
actual
image
capturing,
but
I
didn't
see
anything
like
a
spiral
search.
I
only
saw
guiding
controls
and
fire
capture
for
mounts,
so
I
ended
up
using
sharp
cap
in
the
beginning
of
the
night
to
get
on
mars
and
then
switching
to
fire
capture
to
actually
take
the
data.
G
Another
thing
I
learned
was
that
you
know
at
this
magnification
you're
really
going
to
feel
all
of
the
different
backlashes
in
your
mechanical
components.
You
know
you
really
the
focuser,
you
know.
Even
you
know,
moonlight
should
be
a
pretty
good
focuser
right,
but
you
can.
You
can
see
the
the
effects
of
the
backlash
both
in
in
moving
the
mount
and
in
moving
the
focuser
and
everything's
just
very
hair
hair
trigger.
G
In
the
beginning,
I
was
just
trying
to
kind
of
hand
jam
everything
and
assuming
that
it
was
the
roi
was
always
in
the
center,
and
then
I
learned
it
was
actually
moving
up
into
the
upper
left
hand
corner,
and
then
I
learned
you
could
move
it
around,
which
was
way
better,
and
you
know
one
of
the
the
last
things
that
that
happened
was.
I
noticed
that
once
I'm
pointed
right
up
at
the
at
the
meridian,
so
almost
almost
straight
up,
then
stuff
would
be
falling
onto
this
on
into
the
optical
path.
G
He
said
you
know.
Well,
I
stack
fifty
percent
of
my
my
frames
and
I'm
thinking.
Well,
you
know
if
fifty
percent
is
is
good,
then
wouldn't
one
percent
or
ten
percent
be
even
sharper,
but
what
I
found
out
was
that
yes,
1
or
10
of
the
frame
is
are
higher
quality,
but
then
you
can't
apply
as
much
sharpening.
So
if
with
with
more
depth
to
your
stack,
then
they
tolerate
the
stack
it
tolerates
more
sharpening
before
you
get
artifacts.
G
G
G
C
G
G
I
think
it's
been.
Maybe
two
three
nights
back.
You
know.
I
said:
okay,
I'm
done
with
this
planetary
stuff.
What's
up
in
the
sky,
and
I
put
my
my
80
millimeter
refractor
on
the
soul
nebula
and
the
hydrogen
alpha.
The
first
sub
that
came
through
was
just
like
you
know:
tax
sharp,
I'm
just
like
yeah,
I'm.
G
G
B
G
Yeah,
I
saw
some
posts,
people
hinting
about
having
seen
stuff
what
I
was
doing
was
you
know,
starting
at
the
bottom
and
working
my
way
up
on
the
on
the
gaussian
setting
and
never
going
like.
If,
if
the
bottom
one
was
five,
then
the
next
one
would
be
four
and
then
three
and
then
two
and
then
one
right.
So
I
never
and
maybe
there's
no.
G
You
know
mathematical
reason
why
that
would
be,
but
that
seemed
to
be
the
way
to
avoid
artifacts,
but
I
definitely
saw
some
posts
about
people
talking
about
using
the
noise
reduction
part
of
the
gaussian
tool
and
making
kind
of
a
the
guy
said
it
made
a
v
shape
or
something
I
didn't
really
get,
but
so
it
I
think
it
it
needs
more
looking
into,
but
the
the
london
astronomer,
as
you
probably
know,
bruce,
because
I
think
he
read
through
all
that
stuff
too
doesn't
didn't
mention
any
any.
That
level
of
detail.
G
B
B
I
eventually
was
able
to
come
to
focus
with
the
explore
scientific
3x
expand,
expander
tele
expander,
it's
called,
which
is
similar
to
a
powermate,
but
I
you
know,
I
don't
know
if
it
was
the
spacing
between
that
and
the
back
of
the
edge
or
that
and
the
image
plane
the
focal
plane
of
the
camera.
But
the
image
didn't
look
as
clear
as
it
should
be,
considering
what
the
scene
looked
like
that
night.
G
I
think
you
know
the
technique,
is
you
defocus
it
one
way
or
the
other
until
it's
obviously
like?
Oh,
it's
obviously
getting
worse,
and
then
you
take
note
of
that.
And
then
you
go
the
other
direction
until
it's
obviously
getting
worse
and
then
you
split
the
difference
and
I
think,
and
then
you
just
wait
for
the
for
the
you
know
the
scene
to
to
oscillate
around
that
point
and
you
pray
for
the
lucky
imaging
to
to
pull
something
out.
G
Yeah,
you
should
have
a
wide
critical
focus
zone.
Although
you
would
swear
you
know,
you
could
sit
there
and
and
there's
you
know,
focusing
aids
in
in
sharp
cap
anyway,
you
know
you
could
swear
that
you
were
making
progress
and
then
I'd
always
press
the
button
one
more
time
and
it
would
just
be
horrible.
G
You
know
it's
like
it's
getting
better,
it's
getting
better,
it's
gonna!
Even
oh,
and
then
you
got
the
backlash
and
it's
I
never
could
could
run
on
it
any
better.
So,
but
you
know
using
that
other
technique
I
talked
about
was
was
what
I
used
on
the
last
night,
which
had
the
horrible
results.
So
I
don't
know
you
guys
can,
can
try
it
and
and
help
educate
us.
B
D
D
G
B
G
D
D
D
I
have,
on
occasion,
use
1100
power
from
my
backyard
in
san
jose
to
look
at
the
moon
and
seeing
all
the
little
craters
in
place,
for
instance,
in
those
kind
of
conditions.
Also,
the
deep
twilight
will
help
with
the
atmosphere
dispersion
issue
because
it
will
sort
of
wipe
it
out.
So
you
don't
have
to
worry
about
it
too
much
so
as
mars
progresses
and
gets
up
about
30
degrees.
At
that
time
of
the
of
the
night,
I
think
you
might
have
some
good
luck
with
that.
G
B
One
thing
I
wanted
to
add:
I
had
to
ask
you
and
I
don't
think
we
really
followed
through
on
figuring
it
out.
Maybe
you
knew,
but
we
did.
The
direction
of
the
conversation
went
a
different
way,
but
if
you're
using
an
adc
and
you're
using
a
one-shot
color
camera,
you
can,
you
can
just
see
it
do
its
correction,
but
I
was
wondering
how
you
would
do
it
if
you're
using
a
monochrome.
B
B
G
That
I
was
hoping
that
there
would
be
something
that
you
know
the
the
the
red
filter
and
the
especially
the
infrared.
The
infrared
filter
seemed
to
have
pretty
crisp
edges
and
the
red
almost
as
crisp,
but
the
blue
had
you
know,
flares
to
the
upper
left
and
the
in
the
lower
right
always,
and
you
could
see
it
as
it
goes
in
and
out
of
focus.
There's
always
this
limb
of
fuzziness,
and
I
was
hoping
that
some
optical
component
would
make
that
be
more
like
the
ir
and
the
in
the
red
filter.
E
D
G
B
G
Instead
of
luminance
and
there's
one
line
in
in
the
guys,
london
astronomer,
I'm
sorry,
I'm
really
bad
with
name
roger
roger
hutcheson,
and
he
he
just
has
sort
of
this
parenthetical
statement
about
now.
There's
some
you
know
argument
in
the
forums
about
whether
or
not
you
should
use
ir
for
luminance,
but
but
I
use
it
so
you
know
I
figured
look
at
his
results.
So
that's
that's
what
I
went
with
since
I
had
one,
and
it
definitely
did
was
the
best.