The *r.fits file is the so-called ramp file. It is a 4096x4096xNREADS image cube, where NREADS is the number of reads taken during the integration (one read being done every 5.571s). One read produces a 4096x4096 image of about 34Mb. The file structure is simple - only the Primary Array, no Extensions - but the ramp file being a stack of individual 4096x4096 images, it easily becomes quite big, up to a 4 Gb for the longest exposure, the exposure time (1200s) being limited by the available RAM on the detector system.

File (odometer)o.fits

From the signal recorded in the ramp file and increasing from one read to the next, the flux ADU/s is calculated for each pixel and stored in the *o.fits file. All *o.files have the same size, 235Mb.

HDU #	Name	Type	Description
1		Primary Header
2	Slope	Image	Slope in ADU/s from ramp linear fit
3	Intercept	Image	ADU for NREADS=0 from ramp linear fit
4	Errslope	Image	Error estimate on slope
5	Count	Image	Number of reads involved in slope measurement (before SATURATE=60000 is reached)

File (odometer)g.fits

The guiding file is an image cube of the 32x32 guiding images recorded during the science exposure. It contains typically thousand of guiding images, the guiding loop working up to 50Hz.

HDU #	Name	Type	Description
1		Primary Header
2	IMAGE	Image	Image cube
3	METADATA	ASCII Table	16 columns x number of rows, one row per image

The ASCII table mostly contains engineering data, only GD_XOFF and GD_YOFF may be of interest for the analysis of the image sequence. The table has the following format:

Col #	Name	Units	Format	Description
1	UNIXTIME	s	D14.3	Unix timestamp expressed as seconds since January 1st, 1970 GMT
2	SEQNUM		I8	Incrementing sequence number for the frame. Sequence is reset to 0 when the guide host is rebooted
3	ISUERR_X	mrad	D7.2	ISU raw X error before PID adjustment
4	ISUERR_Y	mrad	D7.2	ISU raw Y error before PID adjustment
5	GD_XOFF	pixel	D7.2	Guide star offset in X from the null position
6	GD_YOFF	pixel	D7.2	Guide star offset in Y from the null position
7	SMRAD_X	mrad	D7.2	X position sent to the ISU - Note the X & Y axis is rotated by 45 degrees from the focal plane
8	SMRAD_Y	mrad	D7.2	Y position sent to the ISU
9	SMRAD_X	mrad	D7.2	X position read from the ISU
10	SMRAD_Y	mrad	D7.2	Y position read from the ISU
11	RISUAS_X	arcsec	D9.4	Translated (rotated) X position read from the ISU
12	RISUAS_Y	arcsec	D9.4	Translated (rotated) Y position read from the ISU
13	TCS_X	arcsec	D7.2	Last correction in RA sent to the TCS
14	TCS_Y	arcsec	D7.2	Last correction in DEC sent to the TCS
15	SEEING	arcsec	D7.2	Seeing estimate
16	STARMAG	mag	D7.2	Magnitude estimate

2. Reduced Data

This section describes the file structure of the SPIRou science products.

File (odometer)e.fits

2D extracted spectra that use the instrument profile and order localization and performs optimal extraction. There is one extracted spectrum for channels A, B, combined AB, and C; these spectra are respectively saved into different extensions of the e files, each with its proper header. The Blaze function is not removed from these spectra, so that the flux distribution along each order is unchanged. Each extension contains 49 orders of 4088 pixels.

HDU #	Name	Type	Description
1		Primary Header
2	FluxAB	Image	Flux for fibers AB combined
3	FluxA	Image	Flux for fiber B
4	FluxB	Image	Flux for fiber A
5	FluxC	Image	Flux for fiber C
6	WaveAB	Image	Wavelength vector for fibers AB combined
7	WaveA	Image	Wavelength vector for fiber A
8	WaveB	Image	Wavelength vector for fiber B
9	WaveC	Image	Wavelength vector for fiber C
10	BlazeAB	Image	Blaze function for fibers AB combined
11	BlazeA	Image	Blaze function for fiber A
12	BlazeB	Image	Blaze function for fiber B
13	BlazeC	Image	Blaze function for fiber C

File (odometer)s.fits

FITS table containing the 1D extracted and rebinned spectra, for each channel (AB, A, B, C). The spectrum is corrected from the Blaze function, which provides a first-order continuum subtraction, and overlapping parts in the contiguous orders are merged. There are two versions of the 1D file: they both have wavelength units from 965 to 2500 nm, but one has the flux rebinned in regular wavelength bins, while the other has a constant velocity bin. The table for each channel contains about 300,000 pixels (NAXIS2). There is also a telluric corrected version (only for the AB fiber).

HDU #	Name	Type	Description
1		Primary Header
2	UniformWavelength	Binary Table	Flux rebinned in regular wavelength bins
3	UniformVelocity	Binary Table	Flux rebinned with a constant velocity bin

Both UniformVelocity and UniformWavelength binary tables are in the following format:

Col #	Name	Units	Format	Description
1	Wave	nm	D	Wavelength vector
2	FluxAB	Relative Flux	D	Flux for fibers AB combined
3	FluxErrAB	Relative Flux	D	Error on FluxAB
4	FluxA	Relative Flux	D	Flux for fiber A
5	FluxErrA	Relative Flux	D	Error on FluxA
6	FluxB	Relative Flux	D	Flux for fiber B
7	FluxErrB	Relative Flux	D	Error on FluxB
8	FluxC	Relative Flux	D	Flux for fiber C
9	FluxErrC	Relative Flux	D	Error on FluxC
10*	FluxABTelluCorrected	Relative Flux	D	Telluric corrected flux for fibers AB combined
11*	FluxErrABTelluCorrected	Relative Flux	D	Error on FluxABTelluCorrected

Format D = Double-precision floating-point

*Columns 10 and 11 will only be present for telluric corrected exposures.

File (odometer)t.fits

2D spectra in the same format as the e spectra, after the correction of telluric lines has been applied. This correction uses the library of empirical spectra of fast-rotating A stars used as telluric standards and the PCA method to optimally remove the contribution of the Earth atmosphere. Parts of the spectrum are replaced by NaNs because the telluric absorption in this location is too intense and correction residuals will be too large. The correction is applied only on the combined AB channel, so there is only one Flux extension of 49 orders times 4088 pixels. This telluric-corrected spectrum is used for the RV measurement.

HDU #	Name	Type	Description
1		Primary Header
2	FluxAB	Image	The flux for the AB science channel
3	WaveAB	Image	The wavelength vector for the AB science channel
4	BlazeAB	Image	The Blaze function for the AB science channel
5	Recon	Image	The spectrum of the Earth atmosphere which has been used in the correction

File (odometer)v.fits

FITS table containing the radial velocity of the star extracted from the CCF. One cross-correlation mask is used per default, corresponding to an M3 spectral type, so it is not optimized for other types of stars. The velocity range of [-100,100] km/s is searched and a step of 1 km/s is used.

HDU #	Name	Type
1		Primary Header
2	CCF	Binary Table

CCF table is in the following format:

Col #	Name	Units	Format	Description
1	Velocity	km/s	D	Radial velocity step
2	Order0		D	Cross-correlation calculated for the individual order
3	Order1		D	Cross-correlation calculated for the individual order
...
50	Order48		D	Cross-correlation calculated for the individual order
51	Combined		D	Weighted mean of all orders, on which the velocity is modeled and measured

Format D = Double-precision floating-point

File (odometer)p.fits

Polarimetric products are only available for observations taken in polarimetric mode, from the combination of 4 consecutive exposures. Each extension contains 49 orders of 4088 pixels in image format.

HDU #	Name	Type	Description
1		Primary Header
2	Pol	Image	The polarized spectrum in the required Stokes configuration
3	PolErr	Image	The error on the polarized spectrum
4	StokesI	Image	The combined Stokes I (intensity) spectrum
5	StokesIErr	Image	The error on the Stokes I spectrum
6	Null1	Image	One null spectrum used to check the polarized signal (see Donati et al 1997)
7	Null2	Image	The other null spectrum used to check the polarized signal
8	WaveAB	Image	The wavelength vector for the AB science channel
9	BlazeAB	Image	The Blaze function for AB (useful for Stokes I)

astropy examples:

Below are examples of accessing data in each product in Python using astropy.io.fits. Note that starting with DRS 0.6, the SNR fits keyword changes from SNRxx to EXTSN0xx for order xx, so EXTSN034 for DRS order 34 for example.

from astropy.io import fits

e = fits.open('1234567e.fits')
# looking at science data channel (AB)
snr_order34 = e[1].header['SNR34']
flux_order34 = e[1].data[34]
wavelength_order34 = e[5].data[34]

t = fits.open('1234567t.fits')
tellu_corr_flux_order34 = t[1].data[34]

p = fits.open('1234567p.fits')
polarization_order34 = p[1].data[34]

v = fits.open('1234567v.fits')
rv_steps = v[1].data.field(0)
ccf_order34 = v[1].data.field(35)
ccf_combined = v[1].data.field(50)
drift_corrected_rv = v[1].header['CCFRVC']

s = fits.open('1234567s.fits')
# using uniform wavelength bins
wavelength_combined = s[1].data.field(0)
flux_combined = s[1].data.field(1)
tellu_corr_flux_combined = s[1].data.field(9)

IDL examples:

Accessing data needs MRDFITS from idlastro.gsfc.nasa.gov. Note that starting with DRS 0.6, the SNR fits keyword changes from SNRxx to EXTSN0xx for order xx, so EXTSN034 for DRS order 34 for example.

; e files
fluxAB=MRDFITS('1234567e.fits',1,header1)    ; FluxAB is the first Image Array 
print,'EXTNAME= ',SXPAR(header1, 'EXTNAME')  ; extension name
print,'OBSDATE= ',SXPAR(header1, 'OBSDATE')
print,'SNR34= ',SXPAR(header1, 'SNR34')      ; SNR for order34  - keyword goes from SNR0 to SNR48, or EXTSN001 to 048 if DRV version >=0.6
waveAB=MRDFITS('1234567e.fits',5,header5)    ; WaveAB is the fifth Image Array
print,'EXTNAME: ',SXPAR(header5, 'EXTNAME')
order=34                                     ;spectral order to be displayed, orders range from 0 to 48
plot,waveAB[*,order],fluxAB[*,order],xtitle='Wavelength (nm)',ytitle='FluxAB'

; s files
s=MRDFITS('1234567s.fits',1,header1)
print,'EXTNAME= ',SXPAR(header1, 'EXTNAME') ; First binary table 
plot,s.Wave,s.FluxAB,xtitle='Wavelength (nm), uniform wavelength sampling',ytitle='FluxAB '
oplot,s.Wave,s.FluxABTelluCorrected         ; Overplot the telluric-corrected spectrum
sv=MRDFITS('1234567s.fits',2,header2)
print,'EXTNAME= ',SXPAR(header2, 'EXTNAME') ; Second binary table
plot,sv.Wave,sv.FluxAB,xtitle='Wavelength (nm), uniform velocity sampling',ytitle='FluxAB '
oplot,sv.Wave,sv.FluxABTelluCorrected       ; Overplot the telluric-corrected spectrum