Resolving Power in SPcAudace

[Erik Bryssinck]

Hi all,

I've been looking for a cause of the lower resolution, given to me by SPcAudace . I have therefore compared the spectra of about a year ago with the spectra of my calibration lamp that I record today. According to the software, the resolution is about 9200 ... 9800 with my LHiResIII and 2400 l/mm grating for the spectra I treat today. Last year I had at least 13,000 and higher to more than 15.000 for the same setup.
So I analyzed the calibration spectra with BASS software. On the Ne-lines I easily reach 13.400 and more, while SPcAudace shows more around 9200 in the report for BeSS database).
I suspect that somewhere a setting in SPcAudace determines on which Ne-peak the resolution is measured. I checked the FITS header and i see under keyword:
SPC_RESL = 6532.8804646999997 [angstrom] Wavelength for resolution measuremen angstrom
SPC_RESP = 9475 Power of resolution at wavelength SPC_RESL

The same spectrum in BASS gives me:
FWHM : 0.494995Å (R = 13195 @6531.65Å) 5.31805px

I've now done another test with an older spectrum. The Ne-line with BASS software gives a resolution of 12767, while SPcAudace gives about 6373 (in Bess report window) for identical spectrum and Ne-line of 6532.88 Angstrom.

Something is not right here. Any ideas ?

thanking in advance !

best regards
Erik

[Erik Bryssinck]

I just checked the same calibrationspectrum in SPcAudace and with command "spc_resolution" and i become:

FWHM=0.532806917326 A
#
La résolution pour la raie 6532.20882749 vaut : 12260
# 12260

while in the BeSS window:

RP_SPcAudace.JPG (18.26 KiB) Viewed 6885 times

best regards
Erik

[Benjamin Mauclaire]

Hi Erik,

This issue has been fixed the 29/12/2019 and with an other patch the 9/1/2020.
You may download patchs that are available in the link "In the next release and patchs" at the upper right corner the spcaudace web page: http://spcaudace.free.fr/nouveautes_next.php.

If you have all ready done the upgrade with patchs, this is an issue that has be fixed.

Benji

[Erik Bryssinck]

Hi Benji,

Thanks for the quick reply.

I put the files in the spcaudace folder and rebooted Audela. Unfortunately I don't notice any difference. In BeSS window I see still 10062 for a particular calibrated Ne-spectrum, while in BASS 13476 the resolution is measured on the same Ne line. So, i think there is still an issue to solve.

best regards,
Erik

[Benjamin Mauclaire]

Bess export windows doesn't measure the RP, it just reads it from fits header.
To check if the patchs solve your issue, you have to process spectra with the patchs intalled.
At the end of pipeline, RP and many other informations are printed in Audela's Console.
Theses infos are also sotred in fits header.
Then Bess export.

Benji

[Erik Bryssinck]

Hi Benji,

I use the pipeline for the processing and checked the values in the FITS header.

I did another test:
I took a spectrum that was taken in june 2019.
I checked the fits-header of this earlier processed spectrum (with the SPcAudace version processed in june 2019) - see screenshot below

RP_20190618.JPG (12.41 KiB) Viewed 6826 times

As you can see the R=14062, i checked the R value with BASS software as well: BASS gives me 13962 -> so far so good.

Now i reprocessed the same spectrum with the latest SPcAudace software and the new files downloaded an placed in the SPcAudace folder:
I get a R= 10069 . This is a big difference with the processing in june 2019 of the same spectrum. (see audela's console screenshot below)

RP_20200121.JPG (18.76 KiB) Viewed 6826 times

Maybe i do something wrong, but i follow the same steps as always and now i see for the same spectrum 2 different values: in june R=14062 and now R=10069.

best regards,
Erik

[Erik Bryssinck]

Hi Benji,

I do now notice that the resolution is now calculated on the 6532 Ne line instead of the 6598 line in June 2019. But this can't explain this big difference. (the 6598 line was even lower in resolution, so the difference is actually even bigger)

best regards
Erik

[Benjamin Mauclaire]

Please send me the lines profile of the calibration lamp called "lampe-calibre....fit".

[Benjamin Mauclaire]

Hi Erik,

There has been an inversion between low resolution andd hight resolution case in spc_resolution computation.
Now the issue is fixed. I found R=13569 for HD201733 and R=12688 for HD22780.
You can download the patch (spc_calibrage.tcl).

Cheers,

Benji