Hi everybody, some weeks ago I was very like to discover a new variable star (my first one):
https://www.aavso.org/vsx/index.php?view=detail.top&oid=474676
I used Astroart to get the images, fotodif for the photometry and vstar for the period study. I really enjoyed it!
In the VSX database we may see the epoch: 02 Oct 2016 (HJD 2457663.57404)
... in HJD.
I uploaded a phase-plot with my data and with NSVS data and It looks fine. But I had to convert NSVS dates to HJD.
Till now, everything was ok...
But yesterday I used VPHOT to do the photometry since I intend to use it. And I uploaded the observations report to AID (my first uploaded observations)
I used VSTAR to get the data and to draw the phase plot... as you may see if you open it in vstar.
The phase plot is not ok!! Why? Because the epoch is in HJD format in VSX but... Vphot reported dates in JD!!! So... What should I do? I'm afraid I should edit the vphot aavso report before uploading it, just to apply an heliocentric correction to it.
CArlos
PD: I'm going to stack the images in a 10->1 ratio to improve SNR and get better uncertainty, since I'm afraid that it's oversampled. But this is not the problem now.
Hello Carlos
I don't know what the proper answer to you question is. However, if you want to get a quick view, you can open your data in Vstar, and when you select the phase plot, edit the VSX value for epoch from 2457663.57404 to xxxxx.56985 and your plot should look right. The JD to HJD correction based on the BAA ap is .00419 minutes for your object (TYC 3224-2602-1).
Cheers
Gary
Thanks Gary, I'm afraid I should modify Vphot report to HJD before uploading observations...
That's right? But this is not clear in the vphot manual...
Carlos
Hi, "solved". The "problem" is that Vstar needs JD to graph the phase plot and the VSX epoch is in HJD, so... It needs to be corrected by the user.
Is it possible to program Vstar to detect the VSX epoch format just to adjust it automatially?
Regards
This is a modified versions of my post to the VStar 6 bug report topic
Problem with HJD and JD
Good Idea to move to the VStar Forum
Is the problem you see that the VStar dialog for the phase plot shows the HJD Epoch from VSX as a JD? If that is the issue, I can't tell you if it is really aproblem or not. it may just be a formatting/ language issue in the dialog text or it may be a real issue. David will have to answer that. I did notice that it listed the epoch in the phase plot header but it is not identified as JD or HJD
I don't really have enough background to understand if there are other problems. For example, When I downloaded your observations from AID into VStar they were shown referenced to JD on the "X" axis in the raw data view, not HJD. Were the times actually converted to JD or is the reference frame just re-labeled without conversion?
Were the date stamps of your measurements originally in HJD?
Where they uploaded as HJD or JD? If uploaded as HJD are the JDs in AID actually converted from HJD to JD or are they HJDs simply mis-labeled as JD. In the AID download table attached you can see that only JD are shown, no HJD.
If the JDs in the AID file are really HJDs then there was a problem in the uploading process. If I recall in WebObs there is an option to select JD or HJD. I don't use VPhot and can't tell you what the JD assumptions or options are there.
If the Raw data JDs are really HJDs, the uploaded data needs to be corrected but as they now appear using the HJD epoch as a JD epoch is the rightthing to do as a work around VStar. You just have to change the X axis label of the raw data plot (right click in plot area and select properties then the plot tab and change the X-axis label. Also in the phase plot Title you should specify the epock as HJD in the title which you can do with properties.
Of course, the above procedure works because the HJDs have been recorded as JDs in AID and that needs to be corrected.
Now the bigger issue is if the HJDs were correctly converted to JDs in AID or if you correct the upload to specify them as HJDs and they are lrecorded that way in AID, what happens to the Epoch in the phase plot tool? Does it correctly distinguis between HJD and JD and convert if necessary or does it willy-nilly just apply the epoch listed to the data without conversion, even if the epoch and data are specified in different reference frames? It appears the later is what happens and VStar isn't "smart enough" to do the conversion. In this particular case there may actually be offsetting errors where HJDs were uploaded to AID as JDs and then VStar applied a HJD epoch as the JD epoch.
Hi again! Thanks for the VSTAR course!!!
Well, the problem, now, is that I did the photometry with VPHOT (I intend to work with AAVSO tools and transforming data). I created the AAVSO report and modified the date to be HJD and the date column as well, as stated above, to HJD (It's a pitty it cannot be done with VPHOT automatically). I attach the screenshot.
I've verified that the uploaded data to webobs is in HJD or at least the checkbox is ticked. (Sebastian Otero explained this to me). Screenshot attached.
The problem is that VSTAR download the data in JD format and when I try to do a phase plot it reads data from VSX. Period is ok and date is labeled as JD but it's the reported HJD date.
I've checked that VSTAR observations dates (screenshot attached) are in JD!!
Regards,
Carlos
OK, What I get from you screenshots is that there are not offsetting errors. You uploaded HJD dates and the AID converted them to JD when you uploade.
I am a bit confused by Screenshot 3. this appears to be an EXCEL file but the Tycho catalog number is in the Date column and the DATE appears to be in the MAG column. The catalog name and the catalog number have to be in the same field in the Extended File Format. So the NAME field would have to be TYC 3224-2606-1, not separated into separate fields for the catalog name (TYC) and a seoparate column for the number. (3224-2606-1) unless these are entries in a spreadhseet that converts into Extended File Format, but the headers seem to be correct for Extended File Format. Also it seems you are submitting your magnitudes in millimag format. A decimal point is required in Extended File Format. So in the first line MAG would be 11.762 and MAGERR would be 0.004 or just .004 but the decimal point is required.
https://www.aavso.org/aavso-extended-file-format
Brad Walter
Hi, yes Brad, I know it. The file was uploaded as you say. That's not the problem (following AAVSO format, I mean). It's only that I have to change "." with "," in Spain, and so on... Don't worry.
I used AAVSO format.
Carlos
Hi Carlos, Brad, Gary
Apologies for the delay. I've been a bit busy the last few days with work, kids' homework help, a nova.
Okay, let me begin to answer some of your questions.
While the VSX page shows the epoch as HJD, the value returned from the VSX web service VStar uses does not indicate whether it's JD or HJD. Sara and I could talk with the VSX team about this.
Yes, I could certainly write a plugin to convert between HJD at JD as a general utility.
Currently VStar handles HJD conversion of loaded observations in a way that probably makes too many assumptions and is not flexible. It should be possible to convert observations to and from HJD at will. Right now it's a one-shot. Indeed, there's some history related to this. I haven't forgotten you Sebastian!
Of most relevance, see:
As with all VStar development, sadly, progress is only as fast as the spare time I can find to spend on it. Just getting the J2000 HJD implementation written and tested took a decent amount of effort.
This is why it really helps me if people suggest priorities for tasks, e.g. by clicking the vote buttons on particular SourceForge tickets.
VStar doesn't apply any special meaning to time values (JD, HJD) other than specific HJD conversions on loaded datasets. So, if observations are in HJD and you type a time value into the epoch field for a phase plot, the assumption is that this time value is consistent with the time system of the observations (JD, HJD, BJD, MJD, ...). We have already noted here that there is missing information (specifically, HJD or JD epoch from VSX).
It would be good to have someone from HQ (perhaps Sara if she wouldn't mind) confirm what happens to time values when entered into WebObs and are marked as HJD, e.g. they are converted to JD; actually we should be able to answer that by observing what happens. My recollection is that there is a HJD field in the AID but that it's not used. You can always of course, given JD values and coordinates, convert to HJD.
What have I missed?
David
Hi All,
The following is a description of what happens on the AAVSO database side of things...
First, you should know that the aid.observations table has two separate fields for date. One for JD and one for HJD.
If a person submits a report to WebObs using the AAVSO Extended file format, they can specify with the #DATE= parameter whether the dates in their report are in HJD or JD (or a calendar date called EXCEL format).
The parameter choices are listed in the Extended File Format specs here: https://www.aavso.org/aavso-extended-file-format
If someone submits an individual CCD observation, they can also specify that it is in HJD format by checking a box and this should behave the in same way as with the file upload.
When someone uses VStar and requests data from the AID to plot, it will use the value in the JD field of the database only - that is because every observation has a JD (as originally sent to HQ or as calculated from the HJD field), while for most observations in the database, the HJD field is blank.
I hope this helps to clarify things.
Sara
Thanks for that Sara. I forgot about that rule for HJD field population. I know we talked about it in the past.
As you say, it doesn't change what VStar uses (JD).
David
"Yes, I could certainly write a plugin to convert between HJD at JD as a general utility."
Thanks David! ;)
I think this would solve most of the problems.
VSX uses HJD but actually it is not strictly like that because old epochs were imported from the GCVS which uses JD.
However, the epochs are taken from papers and nowadays most of the data are published as HJD so the epochs are actually a mix.
For long period variables it doesn't matter but for short period ones it is really meaningful as Carlos' example has shown.
Carlos' example has also shown that now researchers are able to generate, report and analyze their own data on short period stars. Carlos made all the steps: he made the photometry, he discovered a new variable, he analyzed the data, he reported the star to VSX, he reported it to the AID and he wanted to plot it with VStar.
This multitasking approach made the HJD/JD issue more evident. And it will become more and more relevant in the future so that is why having an easy to use tool to go from JD to HJD and the opposite, would be great.
Cheers,
Sebastian
No worries Sebastian.
I'll work towards flexible HJD conversion of individual loaded datasets and perhaps also write a simple plugin that converts between arbitrary, isolated JD/HJD values.
Interesting to know about the mix of JD/HJD epochs. It may still be useful to be able to present that information in VStar. Whether it's HJD or JD seems to be available in VSX records, so perhaps Sara and I could chat with Patrick about exposing that information in the web service.
David
Hi David,
Actually the information about JD or HJD is available in the AID so VStar can take that information from there.
The epochs from VSX are labelled as HJD but for stars from old publications they might be JD.
Anyway, VSX is flexible: anyone finding an epoch discrepancy may submit a revision to correct it so I don't worry too much about it. And the difference will be important only for short period stars.
Great and thanks!!