Command-line options

This chapter deals with all the possible command-line options one can give when calling the radmc3d code.

Main commands

In addition to the radmc3d.inp file, which contains many ‘steering’ parameters, one can (and even must) give RADMC-3D also command-line options. The most important (and compulsory) options are the ‘command’ what RADMC-3D should do. At the moment you can choose from:

mctherm: Runs RADMC-3D for computing the dust temperatures using the Monte Carlo method.
spectrum: Runs RADMC-3D for making a spectrum based on certain settings. This option requires further command-line specifications. See chapter Making images and spectra.
sed: Runs RADMC-3D for making a SED based on certain settings. This option requires further command-line specifications. Note that a SED is like a spectrum, but for continuum processes only (no lines). See chapter Making images and spectra for more details.
image: Runs RADMC-3D for making an image. This option requires further command-line specifications. See chapter Making images and spectra.
movie: Like image, but now for a series of different vantage points. Useful for making movies in one go, without having to call RADMC-3D time and again. NOTE: This command is still under development. See chapter Making images and spectra.
mcmono: (Only expect use). Runs RADMC-3D for computing the local radiation field at each location in the model. This is only useful for when you wish to couple RADMC-3D to models of chemistry or so, which need the local radiation field. See Section Special-purpose feature: Computing the local radiation field.

Example:

radmc3d mctherm

runs the RADMC-3D code for computing the dust temperatures everywhere using the Monte Carlo method.

There are also some additional commands that may be useful for diagnostics:

subbox_****: where **** is one of the following: dust_density, dust_temperature. But other quantities will follow in later versions. See Section Making a regularly-spaced datacube (‘subbox’) of AMR-based models.
linelist: Write a list of all the lines included in this model.

Additional arguments: general

Here is a list of command line options, on top of the above listed main commands (Note: We’ll try to be complete, but as the code develops we may forget to list new options here):

setthreads [for MC] The next number sets the number of OpenMP parallel threads to be used.
npix: [for images] The next number specifies the number of pixels in both x and y direction, assuming a square image.
npixx: [for images] The next number specifies the number of pixels in x direction only.
npixy: [for images] The next number specifies the number of pixels in y direction only.
nrrefine: [for images and spectra] Specifies a maximum depth of refinement of the pixels (see Section The issue of flux conservation: recursive sub-pixeling).
fluxcons: [for images and spectra] Puts nrrefine (see above) to a large value to assue flux conservation (see Section The issue of flux conservation: recursive sub-pixeling).
norefine: [for images and spectra] Puts nrrefine (see above) to 0 so that each pixel of the image corresponds only to 1 ray. This is fast but not reliable and therefore not recommended (see Section The issue of flux conservation: recursive sub-pixeling).
nofluxcons: [for images and spectra] As norefine above.
noscat: This option makes RADMC-3D ignore the dust scattering process (though not the scattering extinction!) in the images, spectra and Monte Carlo simulations. For images and spectra this means that no scattering Monte Carlo run has to be performed before each image ray tracing (see Section Scattered light in images and spectra: The ‘Scattering Monte Carlo’ computation). This can speed up the making of images or spectra enormously. This is even more so if you make images/spectra of gas lines with LTE, LVG or ESCP methods, because if no scattering Monte Carlo needs to be made, ray-tracing can be done multi-frequency for each ray, and the populations can be calculated once in each cell, and used for all frequencies. That can speed up the line rendering enormously – of course at the cost of not including dust scattering. For lines in the infrared and submillimeter, if no large grains are present, this is usually OK, because small grains (smaller than about 1 micron) have very low scattering albedos in the infrared and submillimeter.
ilambda or inu: [for images] Specify the index of the wavelength from the wavelength_micron.inp file for which a ray-trace image should be made.
color: [for images] Allows you to make multiple images (each at a different wavelength) in one go. This will make RADMC-3D read the file color_inus.inp (see Section INPUT: (minor input files)) which is a list of indices i referring to the wavelength_micron.inp file for which the images should be made. See Section Specifying custom-made sets of wavelength points for the camera for details.
loadcolor: [for images] Same as color.
loadlambda: [for images] Allows you to make multiple images (each at a different wavelength) in one go. This will make RADMC-3D read the file camera_wavelength_micron.inp to read the precise wavelength points at which you wish to make the images. In contrast to loadcolor, which only allows you to pick from the global set of wavelength used by the Monte Carlo simulation (in the file wavelength_micron.inp), with the camera_wavelength_micron.inp files you can specify any wavelength you want, and any number of them. See Section Specifying custom-made sets of wavelength points for the camera for details.
sizeau: [for images and spectra] The next number specifies the image size in model space in units of AU (=1.496E13 cm). This image size is measured from the image left to right and top to bottom. This gives always square images. This image size in au is observer distance independent. The corresponding image size in arcsec is: image size in arcsec = image size in AU / (distance in parsec).
sizepc: [for images and spectra] Same as sizeau, but now in parsec units.
zoomau: [for images and spectra] The next four numbers set the image window precisely by specifying the xleft, xright, ybottom, ytop of the image in units of AU. The zero point of the image (the direction of the 2-D image point located at (0.0,0.0) in image coordinates) stays the same (i.e. it aims toward the 3-D point in model space given by pointau or pointpc). In this way you can move the image window left or with or up or down without having to change the pointau or pointpc 3-D locations. Also for local perspective images it is different if you move the image window in the image plane, or if you actually change the direction in which you are looking (for images from infinity this is the same). Note: If you use this option without the truepix option RADMC-3D will always make square pixels by adapting npixx or npixy such that together with the zoomau image size you get approximately square pixels. Furthermore, if truezoom is not set, RADMC-3D will alleviate the remaining tiny deviation from square pixel shape by slightly (!) adapting the zoomau window to obtain exactly square pixels.
zoompc: [for images and spectra] Same as zoomau, but now the four numbers are given in units of parsec.
truepix: [for images and spectra] If with zoomau or zoompc the image window is not square then when specifying npix one gets non-square pixels. Without the truepix option RADMC-3D will adapt the npixx or npixy number, and subsequently modify the zoom window a bit such that the pixels are square. With the truepix option RADMC-3D will not change npixx nor npixy and will allow non-square pixels to form.
truezoom: [for images and spectra] If set, RADMC-3D will always assure that the exact zoom window (specified with zoomau or zoompc) will be used, i.e. if truepix is not set but truezoom is set, RADMC-3D will only (!) adapt npixx or npixy to get approximately square pixels.
pointau: [for images and spectra] The subsequent three numbers specify a 3-D location in model space toward which the camera is pointing for images and spectra. The (0,0) coordinate in the image plane corresponds by definition to a ray going right through this 3-D point.
pointpc: [for images and spectra] Same as pointau but now in units of parsec.
incl: [for images and spectra] For the case when the camera is at infinity (i.e. at a large distance so that no local perspective has to be taken into account) this inclination specifies the direction toward which the camera for images and spectra is positioned. Incl = 0 means toward the positive \(z\)-axis (in cartesian space), incl=90 means toward a position in the \(x\)-\(y\)-plane and incl=180 means toward the negative \(z\)-axis. The angle is given in degrees.
phi: [for images and spectra] Like incl, but now the remaining angle, also given in degrees. Examples: incl=90 and phi=0 means that the observer is located at infinity toward the negative \(y\) axis; incl=90 and phi=90 means that the observer is located at infinity toward the negative \(x\) axis; incl=90 and phi=180 means that the observer is located at infinity toward the positive \(y\) axis (looking back in negative \(y\) direction). Rotation of the observer around the object around the \(z\)-axis goes in clockwise direction. The starting point of this rotation is such that for incl=0 and phi=0 the \((x,y)\) in the image plane correspond to the \((x,y)\) in the 3-D space, with \(x\) pointing toward the right and \(y\) pointing upward. Examples: if we fix the position of the observer at for instance incl=0 (i.e. we look at the object from the top from the positive \(z\)-axis at infinity downward), then increasing phi means rotating the object counter-clockwise in the image plane.
posang: [for images] This rotates the camera itself around the \((0,0)\) point in the image plane.
imageunform: Write out images in binary format
imageformatted: Write out images in text form (default)
tracetau: [for images] If this option is set, then instead of ray-tracing a true image, the camera will compute the optical depth at the wavelength given by e.g. inu and puts this into an image output as if it were a true image. Can be useful for analysis of models.
tracecolumn: [for images] Like tracetau but instead of the optical depth the simple column depth is computed in \(\mathrm{g}/\mathrm{cm}^2\). NOTE: for now only the column depth of the dust.
tracenormal: [for images: Default] Only if you specified tracetau or tracecolumn before, and you are in child mode, you may sometimes want to reset to normal imaging mode.
apert or useapert: [for images/spectra] Use the image-plane aperture information from the file aperture_info.inp.
noapert: [for images/spectra] Do not use an image-plane aperture.
diag_subpix: [for images/spectra] write subpixeling_diagnostics.out file, which contains the locations of all the pixels used, including the subpixels (see Section The solution: recursive sub-pixeling).
nphot_therm: [for MC] The nr of photons for the thermal Monte Carlo simulation. But it is better to use the radmc3d.inp for this (see Section INPUT: radmc3d.inp), because then you can see afterward with which photon statistics the run was done.
nphot_scat: [for MC] The nr of photons for the scattering Monte Carlo simulation done before each image (and thus also in the spectrum). But it is better to use the radmc3d.inp for this (see Section INPUT: radmc3d.inp), because then you can see afterward with which photon statistics the run was done.
nphot_mcmono: [for MC] The nr of photons for the monochromatic Monte Carlo simulation. But it is better to use the radmc3d.inp for this (see Section INPUT: radmc3d.inp), because then you can see afterward with which photon statistics the run was done.
countwrite: [for MC] The nr of photons between ‘sign of life’ outputs in a Monte Carlo run. Default is 1000. That means that if you have nrphot=10000000 you will see ten-thousand times something like Photonnr: 19000 on your screen. Can be annoying. By adding countwrite 100000 to the command line, you will only see a message every 100000 photon packages.

Switching on/off of radiation processes

You can switch certain radiative processes on or off with the following command-line options (though often the radmc3d.inp file also allows this):

inclstar: [for images and spectra] Include stars in spectrum or images.
nostar: [for images and spectra] Do not include stars in spectrum or images. Only the circumstellar / interstellar material is imaged as if a perfect coronograph is used.
inclline: Include line emission and extinction in the ray tracing (for images and spectra).
noline: Do not include line emission and extinction in the ray tracing (for images and spectra).
incldust: Include dust emission, extinction and (unless it is switched off) dust scattering in ray tracing (for images and spectra).
nodust: Do not include dust emission, extinction and scattering in ray tracing (for images and spectra).
maxnrscat 0: (if dust is included) Do not include scattering in the images/spectra created by the camera. With maxnrscat 1 you limit the scattering in the images/spectra to single-scattering. With maxnrscat 2 to double scattering, etc. Can be useful to figure out the relative importance of single vs multiple scattering.