In the course of evaluating the new IHMP positions after the Sep2018 FPA takedown we (mostly SCO, MS, GH) went over a lot of the aspects of this problem. Now (mid Oct2018) I'm not sure everything is settled, but I wanted to record these notes. As with past reduction stes, I am including the files and work in the local directories containing these html docs.
This file originally: $scohtm/scocodes/het_eng/Work_HET/Sep29_2018/Notes.oct01_2018
Using just 3 points: LRS2-R, LRS2-B, BIB
**** Now Computing IHMP Origin on acm ****
183.617 385.540
**** Now Computing HPFacq_600 on acm ****
141.154 252.418
**** Now Computing HPFdirect_603 on acm ****
121.155 288.528
--------------------------------------------------------
Using just 3 points: LRS2-R, LRS2-B, BIB
acam_x_origin = 183.6 ± 0.5 (units of acm pixels)
acam_y_origin = 385.5 ± 1.1 (units of acm pixels)
acam = 1.27 ± 0.25 (degrees)
Not shuffle parametres, but of interest:
acam_x_HPFACQ = 141.1 252.4 # ifu=600
acam_x_HPFside = 121.1 288.5 # ifu=610
--------------------------------------------------------
=======================================================================
Here is a note from Joe (1:30pm or so Oct1)
--------------------------------------------------------
Hi Steve, Matthew,
Here are the new updated numbers for HPF fibres, and how they were arrived at.
September 29th measurements of Central CFB (145.7,254.2) on ACAM, and October 1st measurement of Side CFB (217.9,182.2) is used to map our lab fibre position measurements to ACAM coordinates.
This gives us the ACAM coordinates of all the HPF fibers. Out of this, only HR Science fibre is relevant for discussion now.
HR Science fiber's ACAM coordinates = 125.2, 289.1
Next step, I use the ACAM October 1st pixel measurements of BIB (453.0,401.5), LRS2R (737.9,581.7) and LRS2B (735.5,212.6) along with the IHMP coordinates in fplane.txt file of BIB (3.3,-73.2), LRS2R (49.8, -151.3) and LRS2B (-49.8,-148.7), to find the transformation between ACAM pixel coordinates to IHMP coordinates.
Then this transformation is applied to all the HPF fiber's ACAM coordinates derived in the first step to obtain the coordinates in fplane.txt
600 : -34.8 10.7 # Central CFB
601 : -60.2 -17.9 # HR Sky fibre
602 : -64.3 -13.7 # HE Sky fibre (decommissioned)
603 : -25.3 16.0 # HR Science fibre 604 : -29.3 20.3 # HE Science fibre (decommissioned) 610 : -54.7 -8.4 # Side CFB
611 : -32.3 -20.0
612 : -27.6 -16.4
613 : -24.0 -11.9
614 : -65.6 14.2 615 : -61.9 18.8 616 : -57.2 22.3
As you can see the HR Science fibre position (aka. HPF Direct acquisition mode position) on ACAM mirror is completely defined by the Central CFB and Side CFB measurements from September 29th and October1st respectively.
The focal plane IHMP coordinates depend on the BIB, LRS2R, LRS2B measurements also.
@Steve: Now, if you use these newly derived IHMP coordinate values, does that give you more consistent values?
Thanking you,
-cheers
joe
--------------------------------------------------------
I make a new file (Data.ALL_oct01_joe1) and rerun in het_eng/Work_HET/Sep29_2018/run_oct01_joe1:
Part to send in email:
--------------------------------------------------------------------------
The basic input file I use (SCO Oct01,2018 2:02pm)
% cat Data.ALL_oct01_joe1
# IHMP-acm positions after Sep2018 FPA take-down
# This file = het_eng/Work_HET/Sep29_2018/S/Data.ALL_oct01_joe1
# NOTE: These acm positions from SJ Oct01,2018 (EXCEPT SJ's reported HPFACQ values)
# NOTE: 610,600 ihmp positions taken from Joe Ninan email of 1:34pmCST Oct01,2018
# NOTE: Like Joe, I use here September 29th measurement of Central CFB (145.7,254.2) on ACAM
# name X,Y_ihmp X,Y_acm X,Y_acm _prev
# data
555 3.3 -73.2 453.0 401.5 448.5 402.2
056 -49.8 -148.7 735.4 212.6 731.6 212.0
066 49.8 -151.3 737.9 581.7 734.5 582.0
600 -34.8 10.7 145.7 254.2 140.8 257.7
610 -54.7 -8.4 217.9 182.2 nan nan
Results from find_ihmp_on_acm:
acam_x_origin = 183.9 -+ 0.8 (units of acm pixels)
acam_y_origin = 384.8 -+ 0.7 (units of acm pixels)
acam = 1.35 -+ 0.10 (degrees)
Not shuffle parametres, but of interest:
acam_x_HPFACQ = 147.4 -+ 0.8 # ifu=600 predicted X_acm
acam_x_HPFACQ = 255.4 -+ 0.7 # ifu=600 predicted Y_acm
Plot: oct01_joe1.png
--------------------------------------------------------
Exact procedures to get above:
find_ihmp_on_acm Data.ALL_oct01_joe1
cat Final_Residuals.List
cat trs_rotatedet.explain
acam = 270.00-ang where ang is angle from trs_rotatedet.explain
Tue Oct 2 06:22:11 CDT 2018
============================
New values from MS (from night of 20181002 UT observing). Matt's note:
Joe,
So far it seems that they give about equal flux but if you select a position between
them you get better flux. I also remeasured the central IFU (which previously only
had a single value. I get
145.8 and 256.4
I also retried the side IFU and including my value into Steven's median I get
217.8 and 182.5
I hope that when you run these new numbers that you get a value such that the HR fiber
is close to the value I think it is at. I am pretty sure that the HR fiber should be
pretty near 125.8 and 290.7
Please let us know what you get.
Matthew
New MS 20181002 values:
IFU X,Y_acm (new) X,Y_acm(old)
600 145.8 256.4 145.7 254.2
610 217.8 182.5 217.9 182.2
% cat Data.ALL_oct02
# IHMP-acm positions after Sep2018 FPA take-down
# This file = het_eng/Work_HET/Sep29_2018/S/Data.ALL_oct02
# NOTE: These acm positions from SJ Oct01,2018 and MS 20181002
# NOTE: 610,600 ihmp positions taken from Joe Ninan email of 1:34pmCST Oct01,2018
# NOTE: Added new MS 20181002 measures toget mean acm positions for 600,610
# name X,Y_ihmp X,Y_acm X,Y_acm _prev
# data
555 3.3 -73.2 453.0 401.5 448.5 402.2
056 -49.8 -148.7 735.4 212.6 731.6 212.0
066 49.8 -151.3 737.9 581.7 734.5 582.0
600 -34.8 10.7 145.8 255.3 140.8 257.7
610 -54.7 -8.4 217.8 182.3 217.9 182.2
Later in the morning, I got the following note from Joe:
------------------------------------------------------------------------------
Hi Matthew, Steve,
Thank you very much for the improving the values.
ACAM positions of Side CFB position and Central CFB position is the most
important measurement for HPF, since they pin down everything else.
So the accuracy of those two measurement, determines everything.
Using this latest and greatest updated values (as of 2nd October 2018) :
(145.8, 256.4) for Central CFB, and (217.8, 182.5)of Side CFB
We get: HR Science fiber's ACAM coordinates = 125.6, 292.0
Like Last time, using the ACAM October 1st pixel measurements of BIB (453.0,401.5),
LRS2R (737.9,581.7) and LRS2B (735.5,212.6) along with the IHMP coordinates in
fplane.txt file of BIB (3.3,-73.2), LRS2R (49.8, -151.3) and LRS2B (-49.8,-148.7),
we get the following coordinates for fplane.txt file of all fibres
600 : -34.2 10.6 # Central CFB
601 : -60.3 -17.9 # HR Sky fibre
602 : -64.4 -13.6 # HE Sky fibre (decommissioned) 603 : -24.5 15.9 # HR Science fibre
604 : -28.5 20.3 # HE Science fibre (decommissioned)
610 : -54.7 -8.4 # Side CFB
611 : -32.1 -20.4
612 : -27.4 -16.9
613 : -23.6 -12.4 614 : -65.3 14.7 615 : -61.5 19.3 616 : -56.7 22.8
Also the new updated offset we have to make to go from Central CFB (600)
to HR fibre (603) in the regular acquisition mode is given below, ie, the
Guider offset fiducial offset commands for HR fibre changes to
dx_arcsec=-9.7, and dy_asec = -5.3
Thanking you,
-cheers
joe
------------------------------------------------------------------------------
Hence, Joe's update Oct2 HPF 600,610 ihmp positions:
600 : -34.2 10.6 # Central CFB
610 : -54.7 -8.4 # Side CFB
Hence, my final file would be:
% cat Data.ALL_oct02
# IHMP-acm positions after Sep2018 FPA take-down
# This file = het_eng/Work_HET/Sep29_2018/S/Data.ALL_oct02
# NOTE: These acm positions from SJ Oct01,2018 and MS 20181002
# NOTE: 610,600 ihmp positions taken from Joe Ninan email of 7:42am Oct02,2018
# NOTE: Added new MS 20181002 measures toget mean acm positions for 600,610
# name X,Y_ihmp X,Y_acm X,Y_acm _prev
# data
555 3.3 -73.2 453.0 401.5 448.5 402.2
056 -49.8 -148.7 735.4 212.6 731.6 212.0
066 49.8 -151.3 737.9 581.7 734.5 582.0
600 -34.2 10.6 145.8 255.3 140.8 257.7
610 -54.7 -8.4 217.8 182.3 217.9 182.2
=================================================================================
Matt:
I realize that Joe is using our recent acm measures to recalculate the
ihmp x,y positions for 600 and 610. Hence, my 5 point fit using (555,056,
066,600,610) to get the acam position of the IHMP center and the roatation
angle (named acam) should really be considered a final consistancy check.
You agree right? The last two points depend on data from first 3 points.
I'm okay with that.
What I have been thinking about the last couple days is this: how were
the x,y ihmp values determined for 555, 056, 066? They are not integral
multiples of 50", so somebody used some technique to refine these numbers.
Do you know who and how? I can think of things for the LRS2 positions, since
they each sits in it's own IFUSLOT. But I'm wondering how these numbers were
modified for the BIB?
Just thinking. I don't wnat to start a big email stream over this. If you
have an understanding of these things I would really like to discuss it some
time when you are available.
-Steve
=================================================================================
Final solution:
**** Now Computing IHMP Origin on acm ****
# data
183.968 384.553
File:
dam_x_origin = 184.0 -+ 0.9 acm pixels
acam_y_origin = 384.5 -+ 0.8 acm pixels
acam = 1.4 -+ 0.12 degrees
File:
% cat Data.ALL_oct02_a
# IHMP-acm positions after Sep2018 FPA take-down
# This file = het_eng/Work_HET/Sep29_2018/S/Data.ALL_oct02
# NOTE: These acm positions from SJ Oct01,2018 and MS 20181002
# NOTE: 610,600 ihmp positions taken from Joe Ninan email of 7:42am Oct02,2018
# NOTE: Added new MS 20181002 measures toget mean acm positions for 600,610
# name X,Y_ihmp X,Y_acm X,Y_acm _prev
# data
555 3.3 -73.2 453.0 401.5 448.5 402.2
056 -49.8 -148.7 735.4 212.6 731.6 212.0
066 49.8 -151.3 737.9 581.7 734.5 582.0
600 -34.2 10.6 145.8 255.3 140.8 257.7
610 -54.7 -8.4 217.8 182.3 217.9 182.2
Run in: /home/sco/sco/scohtm/scocodes/het_eng/Work_HET/Sep29_2018/Special_Final
=========================================================================================
Tue Oct 9 08:08:05 CDT 2018
----------------------------
There are some questions about the latest sets of numbers that describe the
placement of the [acm,ihmp] system on the sky.
There may be a way to locate the ihmp center on the acm if we "fix" the
LRS2-B and LRS2-R (and BIB?) on the acm.
Some key dates:
2017 12 09 - IFU work (1)
2018 07 27 - FPA takedown and IFU installation
2018 09 28 - FPA akedown (2)
--------------------------------------------
From recent email:
Hi Matthew, Steve,
We went back to the model of the IHMP to derive these values for the nominal
positions of BIB and LRS2 as offsets from IHMP center on sky:
BIB: X=0", Y=-72.45"
LRS2: X=+/-50.0", Y=-150.0"
I assigned the sign on Y to reflect my understanding that +Y is in the direction
of the IHMP center from the LRS2 IFUs. This calculation takes into account the
small change in plate scale with field angle.
Hope that makes sense. The fact that the LRS2 values come out as intended is
just a reflection of us doing the calculation correctly (or the same way both times!)
Cheers, Gary
--------------------------------------------
Matt and Gary:
I realized that my determination of X,Y_ihmp of last night is based on the
old ihmp x,y data (including Joe Ninan's 600,610 estimates). It occurred to
me that we can do a simple graphical derivation:
1) Compute the radius from the IHMP center to BIB, LRS2-R, LRS2-B
using the coordinate Gary sent a couple days ago.
2) Convert these radii in arcsec to units of acm pixels using ps=0.2709 "/pix
3) Use Steven J's acm X,Y to make a ds9 regions file and overplot on an acm image.
4) Find where the circles intersect. This gives the IHMP center in acm pixles.
acm plate scale = 0.2709 "/pix
RAD = radius in acm pixels to the IHMP center
From Gary H:
SJ acm X,Y X,Y_ihmp (arcsec) X,Y_ihmp (acm pix) RAD
------------ ----------------- ----------------------
056 735.4 212.6 -50.0 -150.0 -184.6 -553.7 583.7
066 737.9 581.7 +50.0 -150.0 -184.6 +553.7 583.7
555 453.0 401.5 0.0 72.456 0.0 267.6 267.6
Using the above acm cooridinates and radii I make the ds9 regions files
I make acm_fiducials.reg:
% cat acm_fiducials.reg
# Region file format: DS9 version 4.1
# Filename: 20180402T092129.0_acm_sci_proc.fits
image
circle(735.4,212.6,583.7) # color=blue width=3 font="helvetica 20 normal roman" text={LRS2-B}
circle(737.9,581.7,583.7) # color=red width=3 font="helvetica 20 normal roman" text={LRS2-R}
circle(453.0,401.5,267.6) # color=cyan width=3 font="helvetica 20 normal roman" text={BIB}
I find for the X,Y_IHMP: X,Y_ihmp = 183.6 399.8
See the attached figs. Recall the two values we had before:
MS values:
cam_x_origin = 183.0
acam_y_origin = 400.7
My values:
dam_x_origin = 184.0 -+ 0.9 acm pixels
acam_y_origin = 384.5 -+ 0.8 acm pixels
So, Matt's values agrees very well with this check.
-Steve
Note:
mgp -toff for RB,BR move primarily in tracker X
astronomer@mcs play]$ shuffle_config
Copied files: fplane.txt, shuffle.cfg
from fplane.txt:
# IFUSLOT X_FP Y_FP SPECID SPECSLOT IFUID IFUROT PLATESC
056 -50.000 -150.000 998 998 998 0.000 1.00
066 50.000 -150.000 999 999 999 0.000 1.000
555 0.700 -73.130 901 901 901 0.000 1.000
=======================================
I managed to screw up Fig3 (the direction to move BIB in ihmp system).
In the first figure (acm+ihmp_fig1.png):
I use acm_lrs2_view to get LRS2-B,R region files
* Now we see that the acm Y axis runs in the direction of IHMP (and tracker) X.
* Positive X_ihmp is towards the LRS2-R direction (or larger acm Y).
* Positive Y_ihmp is towards lower Xacm values (to the left). (Recall that
the LRS2 ifu lonong axes lie along tracker X at Yihmp=-150.)
You can see even in the above figure, that to have the BIB circle agree with the
LRS2-R and LRS2-B circles we'd have to shift it to the LEFT in the acm image. This
means we would be moving the BIB cnter towards +Yihmp. The amonut of the shift
I measure (see acm+ihmp_fig2.png) is all in acm X:
BIB dX = 453.0 - 450.3 = 2.7 acm pixels = 0.73" in the +Yihmp direction
Note: In the figure the shifted BIB circle is plotted as the thinner line with
a center to the left (more +Yihmp) of the original BIB position.
Hence, since our old BIB (X,Y)_ihmp = 0.0,-72.5, then our implied revised BIB
center in the IHMP system would
BIB (X,Y)_ihmp = 0.0 , -69.8 (arcseconds)
================================================================================
================================================================================
There were lots of notes about the re-determination of the BIB center in the
ihmp coordinate system.
--------------
Steve,
I am not sure I get your BIB position calculation. If the position of
the IHMP is 183.0 and 400.7 and the BIB is at 453.0 and 401.5 then the
distance between them is 270.07 pixels or 73.16". If the BIB is still
on the center line in X (which I am not convinced is true) then the
position would be BIB(X,Y) = 0.0,-73.14 which is pretty darn close
to my value of 0.7,-73.13.
Matthew
--------------
Matt:
Because the circles all run up and down at the IHMP center, I
really can not get a handle on a change in the X_ihmp. My point
was just that I had to push the BIB circle over to the left in the acm
image (toward +Y) about 0.73". So I thought the new Y_ihmp for
BIB would be: Y_ihmp(old) + dY = -72.5" + 0.73 = -71.8"
Looking at my email, I see I bunged that up. I don't know why I
set Y_ihmp = -69.8.
In any case, I guess I would say: 0.0" -71.8" for the new IHMP
If you have a better solution for X, then go with that. Look at my second
figure. You can move that thin cyan line up and sown (move in X_ihmp) and
not really tell whether the circle fits the LRS2-R and LRS2-B intersection
any better. I really considered this just an independent check on your
estimate of (acam_x_origin,acam_y_origin). I'm not really sure that this
graphical method is good enough to refine the BIB position in ihmp
coordinates.
Anyway, I'm pretty sure I have the direction right: moving the BIB circle about
3 pixels to the left means moving +0.8" in the positive direction in Y_ihmp.
So, the old Y_ihmp=-72.5 would become Y_ihmp=-71.8.
I'm just applying my Y shift in arcsec to Gary's value. I guess I don't recall
where the values of BIB center and IHMP center that you quote below are
coming from. Maybe we could just get together and go over this. It is hard
to cover everything by email I think.
-Steve
P.S. I am using the BIB position sent by Gary a few days ago:
From recent email:
Hi Matthew, Steve,
We went back to the model of the IHMP to derive these values for the nominal
positions of BIB and LRS2 as offsets from IHMP center on sky:
BIB: X=0", Y=-72.45"
LRS2: X=+/-50.0", Y=-150.0"
--------------
Matt:
Now I see what you did. But when I shift my BIB circle to
better match the LRS2 B,R intersection, my new BIB center
was 450,401. So using your method:
IHMP_center = 183.3 395.6
New BIB = 450.0 401.1
So, dX = 450 - 183.3 = 266.7 acm pixels = 72.2"
Hence, my new Y_ihmp = -72.2"
You used the old BIB Y_acm = 453.0
So, dX = 453.0 - 183.3 = 269.7 acm pixels = 73.1
Hence, your new Y_ihmp = -73.1
I really can not say anything about an X shift one way or the
other (as I babbled on about in my last note).
-Steve