JS9 is the latest in the SAOimage line of astronomical image display programs developed at the Center for Astrophysics | Harvard & Smithsonian (CfA). Our previous effort, SAOimage DS9, has been the de facto standard desktop image display for more than 30 years. Now, with JS9, you can view and manipulate astronomical image data in your browser as well as on the desktop.
Please choose from the following topics:
JS9 provides astronomical image display capabilities both on the desktop and in your browser. JS9 takes its name as well as its core functionality from the de facto standard DS9 image display program for desktop use. JS9, on the other hand, can be run either on the desktop or in a browser. It also can be used without any installation at all by dragging/dropping a FITS file onto the JS9 website. In all three cases, the JS9 display consists of one or more HTML components that are added to a web page. At a minimum, the JS9 display will contain an image display canvas. It usually also will display a menubar containing several pull-down menus and a colorbar. Traditionally, the menubar is placed on top of the image canvas and the colorbar below, but web page designers (as well as individual users) have complete control over component placement. JS9 also can display an info box, a command-line window, a pan window, a zoom window, as well as other plugins. All of these plugin windows can be brought up dynamically from the View menu.
Astronomical image data usually is offered on a JS9 web page as one or more clickable URLs. When you click on a data URL, it is displayed on the JS9 image canvas. Multiple images can be loaded into a single JS9 display. The File menu allows you to switch between these images.
Once an image is loaded, you can move the mouse (or a single finger on a touch screen) to display position and pixel value information. Pressing the left mouse button and moving the mouse (or using two fingers on a touch screen) will change the contrast/bias of the displayed image, bringing image features into relief.
These mouse/touch actions are configurable, both by web page developers and by users. See the Mouse Touch plugin for more information.
Menu options allow you to change the scaling algorithm and/or color map, add geometric regions of interest, change world coordinate system (WCS) parameters, and run site-specific astronomical analysis tasks on the original FITS data.
The following menus are in the default JS9 menu bar. Of course, web designers have control over which menus are displayed (or even whether the menu bar itself is displayed). Thus, not all of the menus listed below will be found on all JS9-enabled web pages.
Some menu options have keystroke accelerators and these are shown on the right-hand side of the menu option. Thus, for example, to open a local file, you can select the File:open...:local file menu option, or press the Meta-o key combination when the mouse is inside the JS9 display.
Note that the default menubar has a style in which all of the major menus are shown at the top-level of the menubar. This is the classic style. JS9 also supports a mac style menubar, in which the File, Edit and View menus are placed on the left side of the menubar and the Help menu on the right. The other main menus are displayed as submenus within the View menu.
The default top-level JS9 menus are:
File: The File menu displays the currently loaded images, either directly in the menu, or, if there "too many" images (as defined by the JS9.globalOpts.imagesFileSubmenu variable), using an images submenu. The displayed image is marked with a starburst. To display a different image, simply choose it from the menu. The File menu also contains the following submenus and options:
Edit: allows you to copy and paste regions, positions, and pixel values to/from the clipboard:
View: displays a list of available plugin, including the following core options:
The View menu also contains the following Display Options submenus:
Zoom: The zoomIn and zoomOut menu options zoom the primary image in or out by a factor of two. The zoomToFit options calculates a zoom factor that fits the entire image into the JS9 display. You also can set the zoom factor to a pre-specified value using menu options or you can specify a floating point zoom factor in the other zoom text box. The pan to center option pans to the center of the image, while the reset pan/zoom sets the zoom to 1 and centers the display.
The align pan/zoom ... submenu allows you to change the pan and zoom of the current image to match a target image, assuming both have WCS info available. Pixel size (i.e. as specified by the FITS CDELT1 parameter) will be taken into account when zooming, but not image rotation or flip (from the WCS params). This is quicker than the wcs reproject algorithm for aligning images that are similar.
The Zoom menu also supports image flip around the x or y axes and image rotation by +/- 90 degrees. Flipping and rotation are performed on the in-memory image, with WCS information updated as needed. The rotation is faster than the more general WCS-based rotation/reprojection in the WCS menu. NB: flipping and rotation behave differently from DS9: their action is relative to the current state of the display, whereas DS9's actions are linked to the x and y axes. Thus, flipping by x twice will return you to the original orientation (whereas flipping a second time does nothing in DS9), as will four consecutive 90 degree rotations. We find this behavior to be more intuitive than the interface offered by DS9.
Scale: Each scale option specifies a different scaling algorithm to use when converting image pixel values to RGB values. (The procedure for doing this conversion is detailed in the DS9 documentation titled "How It Works".) The log scale, for example, is especially useful with X-ray data. Histogram equalization (histeq) tries to maximize contrast across the range of data values and is useful in a wide variety of cases. The current scale is marked with a starburst.
You also can change the low and high values used to clip the image data before scaling by changing the low and/or high values in the text boxes of the Data Limits section. These values are set initially with the data's min and max values or the z1,z2 values calculated by the IRAF zscale algorithm (depending on the value of the web site's JS9.imageOpts.scaleclipping parameter. The default usually is "dataminmax".). Finally, you can use the Scale Controls plugin to set the clipping values to data min/max, zscale z1/z2 or zscale z1/max, or set the clipping values using the pixel distribution plot.
Colormap: Each colormap option specifies a different distribution of RGB values to use when converting image pixel values to RGB. (The procedure for doing this conversion is detailed in the DS9 documentation titled "How It Works".) Choose the colormap you like best, or try different ones to see how data features can be brought to the fore by color and contrast/bias.
The image filters menu option provides a number of well-known image processing functions that act directly on the displayed RGB image data (not the underlying raw astronomical data, as with the Gaussian blur in the Analysis menu). Thus, when you can run one of these routines on an image, its effect will be undone if you redraw the image by changing the contrast/bias, colormap, etc. Also, since these functions act on the current RGB image, their effect is cumulative. See the JS9.FilterRGBImage() section in the JS9 Public API for more information about the individual functions.
You can enter a floating point value into the contrast and bias text boxes to set the contrast/bias. Note that the contrast value ranges from 0 to 10 and the bias ranges from 0 to 1 (following DS9 conventions).
The reset contrast/bias option restores the original contrast/bias value (which is especially useful if you changed the contrast/bias or scale and now can't see any features at all!)
The load option brings up a File menu from which you can choose a user-defined colormap to load.
The save option saves the active colormap definition for the currently displayed image to a file (useful for editing to make a new colormap).
The invert option inverts the colormap (e.g. in the gray map, this will change white to black and vice versa).
If you have three images of the same size, scale, and pointing direction (e.g. three energy cuts of a Chandra data set), you can display each one separately and assign to them one of the red, green, or blue colormaps. If you then select the rgb mode menu option, they will be displayed as one RGB composite image. The current image (as shown in the File menu) determines which of the three independent colormaps is changed by the contrast/bias manipulation. To remove an image from the RGB composite, give it a different colormap. Re-join (or add a new image) by assigning one of the RGB colormaps. Simply turn off RGB mode to display the images separately.
Regions: Astronomical regions format is described here. Choose the appropriate menu option to create a region of type annulus, box, circle, ellipse, line, point, polygon, text. (see below for details on how to use regions). One or more regions can be created and manipulated on a single image.
The load menu option will load a file of regions and display them on the current image. Regions can be specified using the a dialect of the widely-used DS9/Funtools regions format, as described here.
The list menu option will display all regions.
The save ... save all regions in one of these formats:
The copy to ... menu option will copy regions from the current image to a specified image (or to all images). If one or more regions are actively selected in the current image, they alone will be copied. Otherwise, all regions will be copied.
The remove menu option will delete all regions.
The selected regions ... submenu contains options that apply to the selected region(s), including the ability to edit, list, save, remove, or copy the selected region(s). The submenu also allows you to change the following properties of selected regions by entering text values:
The selected regions submenu also supports the ability to toggle the "source"/"background" or "include"/"exclude" tags (which is heavily used in X-ray astronomy).
The onchange submenu contains the following options:
WCS: Use this menu option to select a world coordinate system when displaying pixel positions and regions (assuming the data file supports WCS). Available systems are: FK4, FK5, ICRS, galactic, or ecliptic. The native option chooses the system stored in the data file. The image and physical systems will display only image coordinates (not WCS coordinates). Physical coordinates are tied to the original (block 1) file coordinates while image coordinates are tied to the current image data being displayed. Server-side X-ray analysis routines operating on the original file generally want physical coords rather than image coords. Browser-based plugin routines generally use image coordinates to operate on the in-memory image data.
You also can choose the units in which WCS information is displayed: degrees, radians, or sexagesimal. Image and physical coordinates are displayed in units of pixels.
The alternate wcs menu option allows you to choose an alternate world coordinate system, if one or more are defined in the current FITS file. The submenu lists the WCS version (i.e. the letter "A" - "Z", or "default") and, if present, the WCSNAME header parameter associated with this WCS.
The reproject menu option allows you to reproject the image using the WCS associated with another displayed image. The reprojected data is placed in a new raw data layer called "reproject". You can switch between layers using the View->raw data layers option. If the display wcs-aligned option is set, the reprojected image will be aligned with the wcs reference image. If you change the pan/zoom of the latter, the former will be re-aligned. (The converse is not true: changing the reprojected image does not change the wcs reference image.) You can also reproject other images in this display using the WCS info from the currently displayed image.
Analysis: JS9 supports both client-side analysis (i.e., analysis in the browser) and server-side analysis (where requests are made to a back-end server, and returned results are displayed in the browser).
Client-side analysis tasks can be written using the JS9 Plugin capability, and the Analysis menu lists the currently loaded Analysis Plugins. For example, JS9 supplies the Imexam plugin, which offers the following tasks:
JS9 supplies a few other client-side analysis tasks:
In addition, each JS9 site can define server-side analysis tasks to be run from the Analysis menu. Server-side tasks are of two types:
For example, the JS9 web page at https://js9.si.edu offers the following tasks:
If you drag and drop a FITS file onto a JS9 web page connected to a remove server (such as https://js9.si.edu), the FITS file will, of course, not be accessible on the server (it's only stored in the browser). As a result, no server-side analysis tasks will be available. But if the remote server supports proxy loading of FITS data, the Analysis menu will display an option to upload FITS to make tasks available. Select this option and your FITS file will be sent to the server so that server-side analysis can be performed.
The server-side params submenu contains the following options:
Help: Show this and other JS9 help pages. The Help menu also contains help for Plugins, including the Imexam plugins listed above.
In keeping with the needs of mobile devices, JS9 utilizes a single mouse button for user interaction. Keyboard input also is minimized. Moving the mouse (or a single finger on a tablet) over the image will display the position/value of individual pixels. Pressing and moving the mouse (or two fingers) will change the contrast and bias of the image. A double-click (or double-tap) inside a region will bring up the configuration menu for that region.
Regions are geometric shapes that can be used to mark part of an image for analysis. JS9 allows you to create the following region types: annulus, box, circle, ellipse, line, point, polygon. A text region also can be created to annotate an image. When one of these options is selected, the region is created in the center of the image. It then can be dragged to the desired location. Arrow keys also will move the region one screen pixel in the specified direction. A selected region can be deleted by pressing the Delete key (where available. On a tablet, use the delete option in the configuration menu).
Click inside a region to select it and display the resize/rotate controls around its perimeter. Grabbing a handle allows the size of the region to be adjusted or rotated. In the case of the box and ellipse regions, the width and height are controlled separately. Click outside the region to de-select (or click another region to select that one).
For a line region, the end-points serve as both resize controls (arrow icon) and stretch points (hand icon). You might have to move the mouse around a bit to switch from one to another. The control action is to resize the region in both directions (i.e., the same as other regions). The stretch action allows you to move this end-point while the other one remains fixed.
A double-click inside a region brings up a configuration menu. You can change various values associated with a region by entering new values in the appropriate text boxes and the clicking Apply.
The configuration menu has the following standard options:
In addition, regions of different type support specific configuration
parameters in this menu.
The annulus region supports:
The meta key is the Command key on a Mac and the Control key everywhere else.
Pressing the meta key disables the value/position display. It also disables execution of plugin callbacks on all mouse events. See Local Tasks for more information about plugins and browser-based analysis.
When a keyboard is available, pressing the meta key while clicking in the primary image window will pan the image so that the mouse location is in the center of the display (this panning is disabled inside a region.)
Within a polygon or a line, if you press meta and click the mouse on the polygon or line border, a new point will be added at that point. If you press the meta key and click on an existing point, it will be deleted.
You can drag and drop FITS images and binary tables, PNG images, and JPEG images onto JS9 to display the image data. All browser-based functions (scaling, colormap, pan, zoom, regions, WCS) should work as expected. You can execute server-side analysis tasks on the FITS files if the back-end server is running on your local host machine. In this case, you probably will need to set your DataPath variable in the Analysis menu to help JS9 locate the original data file.
The View:resize:set to image size option will set the display size to the size of the current image. The View:resize:set size to full window option will resize the JS9 display to the size of the browser window's window.innerWidth and window.innerHeight properties, and will scroll the display to the center of the window. (Note that the rest of the web page is still available via scrolling, including the menubar.) The View:resize:reset to original size option will set the display size back to the original size. Display resize can be turned off by setting the JS9.globalOpts.resize variable to false in the js9Prefs.json file.
Also by default, the JS9 Display can be resized using a grab handle in the lower right-hand corner of the display. This feature can be turned off setting the JS9.globalOpts.resizeHandle variable to false in the js9Prefs.json file.
See Known Issues for a discussion of limitations imposed on resize by Webkit browsers such as Chrome and Safari.