[daip] Velocity - Frequency conversion in AIPS

[Toby Brown]

Hi Eric,

I’m just checking back in on this again sorry. I’m trying to set the FITS standard key words CRVAL, CRPIX, CUNIT, and CDELT using the ALT* keywords in the headers.

However, which CRVAL and CRPIX correspond to ALTRVAL and ALTRPIX respectively, I can’t see where the units and channel width are defined in the alternate keywords. Are you able to point me in the right direction?

Cheers
Toby

-- 
Research Officer
Herzberg Astronomy & Astrophysics Research Centre
National Research Council of Canada
drtobybrown.com <https://drtobybrown.com/>
@drtobybrown
 <https://twitter.com/DrTobyBrown>Pronouns: He/him/his

> On 10 Feb 2021, at 11:27, egreisen <egreisen at nrao.edu> wrote:
> 
> On 2021-02-10 10:32, Toby Brown via Daip wrote:
> 
> AIPS uses the values stored in keywords ALTRVAL and ALTRPIX to convert between
> frequency and velocity.  Looking in the history recorded in the FITS file I
> see that at one point the reference pixel in the combined data set for
> frequency was changed from 1 to 32.  It is now back at 1 for some reason.
> Each channel is about 10 km/sec so the error is about 300 km/sec.  It just
> a question of the reference pixel being in error or perhaps the reference
> frequency.  The 2 observations that were combined had frequencies different
> by 1.66 MHz due to being observed many months apart.  They were 1410.82 and
> 1412.48.  The header frequency converted to pixel 32 is 1410.83 which
> corresponds to one of the 2 observation dates.  The old VLA did the Doppler
> correction properly (and tracked properly) so I would trust the alternate
> reference pixel description.  The 300 km/sec is a question of the date of
> observation which may have gotten confused when combining the 2 data sets
> to make the images.  BTW - nice image.
> 
> Eric Greisen
> 
>> Hi,
>> I’m experiencing an issue converting between frequency and velocity
>> with an AIPS produced cube I was wondering if you could help me with.
>> I should caveat this with the fact I have zero experience with AIPS
>> but plenty playing with radio cubes using python and casa etc.
>> We have a VLA 21 cm cube produced in AIPS with a frequency axis. When
>> converted to velocity in python, CASA, CARTA, and by hand following
>> either the optical or radio conventions, the cube is at a
>> systematically higher velocity than the true system velocity of the
>> galaxy. However, when the conversion from frequency to velocity is
>> done using "ALTSWTCH" in AIPS, the result has the correct velocity.
>> The VLA cube can be downloaded here [1].
>> Here I show the VLA 21 cm spectrum (orange) and ALMA CO (grey)
>> spectrum comparison. The ALMA spectrum matches published
>> multi-wavelength velocities, while the VLA 21 cm cube is offset
>> approximated +300 km/s.
>> Now I’m showing the spectrum from the VLA cube *after* converting to
>> velocity with ALTSWITCH (orange) and the same ALMA CO (grey) spectrum.
>> You can see that the offset goes away.
>> Do you know why AIPS is able to make the correct conversion between
>> velocity and frequency, while the standard conventions do not work?
>> Cheers,
>> Toby
>> --
>> Research Officer
>> Herzberg Astronomy & Astrophysics Research Centre
>> National Research Council of Canada
>> drtobybrown.com <http://drtobybrown.com/> [2]
>> @drtobybrown
>> Pronouns: He/him/his
>> Links:
>> ------
>> [1] https://www.dropbox.com/s/2yow86gqxoqjroc/NGC4383.cube.fits?dl=0 <https://www.dropbox.com/s/2yow86gqxoqjroc/NGC4383.cube.fits?dl=0>
>> [2] https://drtobybrown.com <https://drtobybrown.com/>
>> _______________________________________________
>> Daip mailing list
>> Daip at listmgr.nrao.edu <mailto:Daip at listmgr.nrao.edu>
>> https://listmgr.nrao.edu/mailman/listinfo/daip <https://listmgr.nrao.edu/mailman/listinfo/daip>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listmgr.nrao.edu/pipermail/daip/attachments/20210226/4d075513/attachment-0001.html>