Photometric Colour Calibration Error
- Thread starter Scooot
- Start date
fredvanner
Well-known member
I also get these faults when accessing the Vizier APASS catalog (selecting "APASS" - which queries the APASS9 catalog). I get the same (or similar) faults on different Vizier servers, which suggests this is not a just a server problem. Running with the local (XPSD) database works fine.
I suggest you download and install the APASS DR10 local database (unlike the GAIA database it is a single quite small, quick download) and use the PI APASS process.
I suggest you download and install the APASS DR10 local database (unlike the GAIA database it is a single quite small, quick download) and use the PI APASS process.
Thank you, I’ll try that if the problem persists.
I’ll try it again tomorrow to see if it’s fixed itself from my machine too.
fredvanner
Well-known member
I'm still getting errors from all tested servers with "APASS" selected as catalog (and with several different images):
All images solve fine with "APASSDR10_XPSD" selected.
This used to work fine (before I started using APASS/XPSD).
All images solve fine with "APASSDR10_XPSD" selected.
This used to work fine (before I started using APASS/XPSD).
fredvanner
Well-known member
... the Vizier catalog search doesn't like the "-out=g'mag" query component. Has the structure of the Vizier APASS search command line changed recently? If anyone has this working, could they check the log of the APASS search line in the process console log?
Thanks Fred
That works (although I wouldn't call a 2.4 Gb file small)
i think in fact what happened to me just now is that the data for M42 was cached by PI so there was no lookup for at least one phase of PCC. there was *a* lookup in the console but somehow i accidentally cleared the console before i could copy it.
you are right - the server doesn't like the apostrophes in g'mag, r'mag, etc. it seems to work if i remove those from the URL. i do have old examples on disk where the g'mag syntax did work, so it does seem like something changed over at vizier.
i wonder if there is a discussion board or changelogs for the vizier server...
you are right - the server doesn't like the apostrophes in g'mag, r'mag, etc. it seems to work if i remove those from the URL. i do have old examples on disk where the g'mag syntax did work, so it does seem like something changed over at vizier.
i wonder if there is a discussion board or changelogs for the vizier server...
fredvanner
Well-known member
... compared with 41GB in 16 files for GAIAEDR3!
Sorry if this seems a daft question but where do I download the APASS DR10 local database from.?
fredvanner
Well-known member
The response from the Vizier search actually says:
A search with "-out.all", which retrieves all catalog fields, retrieves the g'mag, r'mag, i'mag fields with no problem.
so the field is recognised (and certainly exists in the catalog), but is unauthorized - whatever that means.
A search with "-out.all", which retrieves all catalog fields, retrieves the g'mag, r'mag, i'mag fields with no problem.
fredvanner
Well-known member
Details for downloading the XPSD databases (GAIA and APASS) are included in the 1.8.8-7 release notes.
fredvanner
Well-known member
If you use the XPSD databases this problem becomes irrelevant. Actually, it only seems to be the APASS database that has a problem.
An interesting question would be whether the PCC process has been modified to add these fields to the query. If not, then something seems to have changed in the ASU (Astronomical Server URL) handling for the Vizier interface specifically affecting the APASS catalog search.
So my understanding is that the XPSD are local databases (data that I download) - I don't really want to do that.
So yes running the script against the remote database is the part that fails.
If I manually call the URL that is trying to be called a file is downloaded with the response:
Code:
#
# VizieR Astronomical Server vizier
# Date: 2021-01-27T10:29:33 [V1.99+ (14-Oct-2013)]
# In case of problem, please report to: cds-question@unistra.fr
#
#
#Coosys J2000: eq_FK5 J2000
#INFO votable-version=1.99+ (14-Oct-2013)
#INFO -ref=VOTx7234
#INFO MaxTuples=50000
#INFO queryParameters=2
#-oc.form=D.
#-unauthorized=-source=II/336/apass9-c=148.843872 69.041281-c.r=2.390218-c.u=deg-out.form=|-out.add=_RAJ,_DEJ-out=B-V-out=Vmag-out=Bmag-out=g'mag-out=r'mag-out=i'magVmag=0..12
#
#INFO Error=No catalogue or table was specified or found.
Vizier has apparently changed its behavior regarding database field queries for the APASS DR9 catalog. As Fred has pointed out, the following fields in the query URL are not being accepted now:
-out=g'mag&-out=r'mag&-out=i'mag
while they worked fine a few days ago. Interestingly, if you submit a query from Vizier for the APASS DR9 catalog:
http://vizier.u-strasbg.fr/viz-bin/VizieR?-source=II/336
by clicking the Submit button, you get the following:
So this is obviously a regression in Vizier's implementation. I cannot fix a bug in a third party application, so there is nothing we can do but wait until they fix their own problem. If this doesn't happen, I'll try to implement a workaround, but that's something I'd prefer to avoid. These things happen when one depends on third-party products that don't behave consistently.
As for local XPSD databases, they are the only solution to all of these problems. They are still optional for PCC, but will be necessary for the next generation of astrometry based tools that we'll release. I strongly recommend using them for all astrometry and photometry applications in PixInsight.
-out=g'mag&-out=r'mag&-out=i'mag
while they worked fine a few days ago. Interestingly, if you submit a query from Vizier for the APASS DR9 catalog:
http://vizier.u-strasbg.fr/viz-bin/VizieR?-source=II/336
by clicking the Submit button, you get the following:
Code:
No Suitable Catalog or Table
The following unexpected problem occured in VizieR:
No catalog or table matching your choices.
Return to the VizieR main Page
Report:
If you can't find a solution, report to CDS and insert into your message the following details:
.errorfile=/tmp/VR19654.err (2021-01-27T10:52:57)
-4
-unauthorized=-out=g'mag
-ref=VIZ601146094cc6So this is obviously a regression in Vizier's implementation. I cannot fix a bug in a third party application, so there is nothing we can do but wait until they fix their own problem. If this doesn't happen, I'll try to implement a workaround, but that's something I'd prefer to avoid. These things happen when one depends on third-party products that don't behave consistently.
As for local XPSD databases, they are the only solution to all of these problems. They are still optional for PCC, but will be necessary for the next generation of astrometry based tools that we'll release. I strongly recommend using them for all astrometry and photometry applications in PixInsight.
fredvanner
Well-known member
I'm probably accessing a different Vizier server. I get the more specific error:
However, it is clear that this is a server interface problem. I note that there are no fields in the GAIA catalogs that have an apostrophe in their names. If it is a compatibility issue between PI and the ASU interface I wouldn't bet on it being fixed any time soon (since there is the issue of which party is "wrong"). If it is an ASU / Vizier issue with handling the ' character in field names it is completely out of the control of the PI project.
Downloading the XPSD database for APASS DR10 doesn't just bypass this problem, it is much faster, is not network dependent, and uses the (generally superior) Data Release 10 (but you can download DR9 if you prefer it).
[I seem to have "crossed in the post" with Juan - but we are saying the same thing]
Code:
#
# VizieR Astronomical Server cdsarc.u-strasbg.fr
# Date: 2021-01-27T10:42:19 [V1.99+ (14-Oct-2013)]
# In case of problem, please report to: cds-question@unistra.fr
#
#
#Coosys J2000: eq_FK5 J2000
#INFO votable-version=1.99+ (14-Oct-2013)
#INFO -ref=VOTx55762
#INFO MaxTuples=50000
#INFO queryParameters=2
#-oc.form=D.
#-unauthorized=-out=g'mag
#
#INFO Error=No catalogue or table was specified or found.Downloading the XPSD database for APASS DR10 doesn't just bypass this problem, it is much faster, is not network dependent, and uses the (generally superior) Data Release 10 (but you can download DR9 if you prefer it).
[I seem to have "crossed in the post" with Juan - but we are saying the same thing]
gilbertonp
Member
Hi, same problem here whatever alternate catalog I choose. I will try XPSD database
magnifigeek
New member
Hi
A lot of people are experiencing this problem (e.g. see also the Pixinsight for beginners Facebook group). Downloading the database has fixed it for me - but I should not have needed to.
A lot of people are experiencing this problem (e.g. see also the Pixinsight for beginners Facebook group). Downloading the database has fixed it for me - but I should not have needed to.