From: dan otto 
To: atm_free{at}yahoogroups.com
Cc: ATM List 
Reply-To: dan otto 


--0-1203399568-1052277541=:5215
Content-Type: text/plain; charset=us-ascii


James,



I forgot to put my real name at the end of my last post.  I'm Dale Eason. 
As you know I have been working on a version of the ROBO tester written in
C++. With your permission I started with your original code.  Last summer I
re-implement your original algorithms in C++ and started testing my mirror.
 It was at that time I discovered the same behavior that you recently
described. After much thought and research on the net I discovered a
different algorithm similar to what Jerry has described and also discussed
on Michael Peck ‘s Web site.  I think I wrote you about it several months
ago.  I implemented the new algorithm played with it took many measurements
but other matters put that project on hold.  After seeing your post
yesterday I resurrected the project tonight and made measurements at
different camera settings.  The new algorithm does not seem to be affected
by the light intensity.



Here is how I do it.  Starting with the knife at close to the center zone.



1.  Reading a horizontal strip of pixels across the mirror a few pixels
high ( configurable, usually set to about 8).  I take an average of all of
the strip.

2. Move the knife out of the beam until the average stops increasing.

3. Move the knife into the beam until the average is 1/2 the brightest reading.

4.  Now I in effect fold the strip over on itself about the middle and
subtract the left side from the right side.  Where the result is zero or
very close to it is the location of the null radius.  That is the zone
radius were the left and right zone matches in intensity.  If I display the
subtracted image it shows the mirror with a black ring where the null
occurs. You can see this on Pecks’s web site at
http://pw1.netcom.com/~mpeck1/astro/autof/autof.htm.

I then record the knife positions as a zone reading.



Next I move the knife away from the mirror a little bit and start over.  I
do this from  from the COC out until the black ring is out at the outer
edge of the mirror.  That usually gives me 15 to 20 zones.  Many more than
is needed of course.  It take the computer about 10 minutes to do this.  It
also is averaging about 40 frames per measurement which helps remove
turbulence.



I haven't done a dark  or light frame or tried to compensate for camera
irregularities.  But that would probably help also.




James Lerch  wrote:Dan,   I'll take you up on your
offer!  If your could independently confirm my following observations,
perhaps we can come to understand the optical theory that must exist to
explain WHY were seeing these results: #1 Increasing Image Source
brightness results in LESS total knife edge readings (Also test for the
converse as well, all the while keeping a constant "shade of
gray" null value) #2 Lowest possible values for the "Shade of
Gray" (keeping a constant image source brightness) results in LESS
total knife edge readings.   Thanks,James Lerch
http://lerch.no-ip.com/atm (My telescope construction,testing, and coating
site)----- Original Message ----- From: dan otto To: James Lerch ;
atm_free{at}yahoogroups.com Cc: ATM List Sent: Monday, May 05, 2003 7:17
PMSubject: Re: ATM Robo-Foucault, Image intensity, and Changing knife edge
readings
James, I had similar results with my version of an automated tester.  I was
testing a f7 8" mirror.  I found that that it was not as bad if I made
a narrow slit but I was still not satisfied that the test results were
valid.  They varied with the brightness of the led or the manual camera
settings just like you described.  I put the mirror back in the scope and
let the project rest so now I don't remember all my conclusions and what I
was going to try next.   I
could resurect the project to help if you need me to try other ideas for
you. The test stand, camera, and platform are still setup.

James Lerch  wrote:

Can anyone help me explain WHY this is happening, or independently confirm
my results on a mirror of similar size and F-Ratio??




---------------------------------
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo.

---------------------------------
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo.
--0-1203399568-1052277541=:5215
Content-Type: text/html; charset=us-ascii


James,
 
I forgot to put my real name at the end of my last
post.  I'm Dale Eason.  As you know I have been working
on a version of the ROBO tester written in C++.  With your
permission I started with your original code.  Last summer I
re-implement your original algorithms in C++ and started testing my
mirror.  It was at that time I discovered the same behavior that
you recently described.  After much thought and research on the
net I discovered a different algorithm similar to what Jerry has described
and also discussed on Michael Peck ‘s
Web
site.  I think I wrote you
about it several months ago.  I implemented the new algorithm
played with it took many measurements but other matters put that project on
hold.  After seeing your post yesterday I resurrected the project
tonight and made measurements at different camera settings.  The
new algorithm does not seem to be affected by the light
intensity.
 
Here is how I do it.  Starting with the knife at
close to the center zone.
 
1.  Reading a horizontal strip of pixels across the
mirror a few pixels high ( configurable, usually set to about 8). 
I take an average of all of the
strip.
2. Move the knife out of the beam until the average stops
increasing.
3. Move the knife into the beam until the average
is 1/2 the brightest reading.
4.  Now I in effect fold the strip over on itself
about the middle and subtract the left side from the right side. 
Where the result is zero or very close to it is the location of the null
radius.  That is the zone radius were the left and right zone
matches in intensity.  If I display the subtracted image it shows
the mirror with a black ring where the null occurs. You can see this on
Pecks’s web site at http://pw1.netcom.co" target="new">http://pw1.netcom.co">http://pw1.netcom.com/~mpeck1/astro/autof/autof.htm">http://pw1.netcom.co
m/~mpeck1/astro/autof/autof.htm.

I then record the knife positions as a zone
reading.
 
Next I move the knife away from the mirror a little bit and
start over.  I do this from  from the COC out until the
black ring is out at the outer edge of the mirror.  That usually
gives me 15 to 20 zones.  Many more than is needed of
course.  It take the computer about 10 minutes to do
this.  It also is averaging about 40 frames per measurement which
helps remove turbulence.
 
I haven't done a dark  or light frame or tried to
compensate for camera irregularities.  But that would probably
help also.
 James
Lerch
<jlerch1{at}tampabay.rr.com> wrote:




Dan,  
 
I'll take you up on your
offer!  If your
could independently confirm my following observations, perhaps we can come
to understand the optical theory that must exist to explain WHY were seeing
these results:
 
#1 Increasing Image Source
brightness results in
LESS total knife edge readings (Also test for the converse as well, all
the while keeping a constant "shade of gray" null
value)
 
#2 Lowest possible values for the
"Shade of Gray"
(keeping a constant image source brightness) results in LESS total knife
edge readings.  
 
Thanks,
James Lerchhttp://lerch.no-ip.com" target="new">http://lerch.no-ip.com">http://lerch.no-ip.com/atm">http://lerch.no-ip.com
/atm (My telescope construction,testing, and coating site)

----- Original Message -----

From: dan">mailto:atmpob{at}yahoo.com">dan otto 
To: James">mailto:jlerch1{at}tampabay.rr.com">James Lerch ;
atm_free{at}">mailto:atm_free{at}yahoogroups.com">atm_free{at}
yahoogroups.com 
Cc: ATM">mailto:ATM{at}shore.net">ATM List 
Sent:
Monday, May 05, 2003 7:17 PM
Subject:
Re: ATM Robo-Foucault, Image
intensity, and Changing knife edge readings

James, I had similar results with my version of an automated
tester. 
I was testing a f7 8" mirror.  I found that that it was not
as bad if I made a narrow slit but I was still not satisfied that the test
results were valid.  They varied with the brightness of the led or
the manual camera settings just like you described.  I put the
mirror back in the scope and let the project rest so now I don't remember
all my conclusions and what I was going to try next.  
 
I could resurect the project to help if you need me to try other ideas for
you.  The test stand, camera, and platform are still
setup.James Lerch
<jlerch1{at}tampabay.rr.com>
wrote:

Can anyone help me explain WHY this is happening, or independently
confirmmy results on a mirror of similar size and
F-Ratio??


Do you Yahoo!?http://us.rd.yahoo.com/search/mailsig/*http://search
.yahoo.com">The New Yahoo! Search - Faster. Easier.
Bingo.
 Do you Yahoo!?
The">http://us.rd.yahoo.com/search/mailsig/*http://search.yahoo.com">The
New Yahoo! Search - Faster. Easier. Bingo.
--0-1203399568-1052277541=:5215--

--- BBBS/NT v4.00 MP