Crystal Impact Home  
  About us Diamond Endeavour Match!  
About Diamond
Function List...
Brochure (PDF)...
Get Diamond

Order Now

Demo Version

A Diamond trial version is available free-of-charge and can be downloaded. More...
Frequently Asked Questions...
User Group...

Diamond Version 2.1: Frequently Asked Questions

Please note: The following sections refers to the "old" version 2.1 of Diamond. FAQ's for the current version 3 are available here.







  • Problems with graphics drivers

    Several users of Diamond reported bugs with new graphics adapters when Diamond is running in rendering mode and double buffering is active. The harmless version of this bug is just a black window (and blank printout page), the harmful version is a crash (protection fault of Diamond in special DLL of graphics driver). Bugs have been reported for/under:
    "Win98SE with a Nvidia TNT2 5.22",
    "ATI Rage 128 Xpert2000 AGP 99 with 32 MB",
    "NT SVP 5, Dual processor board 2x Celerons 500 MHz, graphics adapter ATI Rage 128, 16 MB, RAM 256MB">,
    "laptop computer ASUS F7400 with a graphics card (ATI RAGE LT PRO) with 4Mb video memory, AGP graphics controller with hardware 3D acceleration. I work at 1024x768 32-bit resolution",
    "Windows98 with ATI 3D Rage Pro AGP 2x"

    The rendering mode of DIAMOND uses the OpenGL library. Since this is no native part of the Windows Operating System, DIAMOND installs a generic driver, if no other is found during the installation procedure. If a new graphics board is installed, the manufacturer normally delivers new drivers, which should support special features of the graphics board. But sometimes they contain bugs, especially concerning the OpenGL functions. So please make sure, that you have installed the most recent driver version, which is often available via the Internet from the manufacturer.

  • Rendering bug when using transparency

    Diamond 2.1 has major problems displaying correctly spheres that are intersected by semitransparent planes/polyhedra faces. Often that part of the sphere that would be hidden by an opaque plane is invisible also with a semitransparent plane. Please find attached a dsf file as an example. Actually, this problem can even be seen in one of the examples in your Diamond gallery: "SURFACE ADSORPTION OF SMALL MOLECULES".
    A similar problem occurs when two semitransparent faces overlap.

    The rendering in Diamond is made with OpenGL. But unfortunately the OpenGL implementation on MS Windows does not support "alpha-buffering", which is necessary for the correct representation of transparent objects, if there is intersection etc.
    For the current version 2.x there is no patch or workaround. The 3.0 version of Diamond will have a POVRay interface, which offers more versatile rendering facilities - and correct transparency. This version will be available not before 3rd quarter of 2002.

  • Background not bright white in rendered mode

    Trying to transfer rendered pictures from Diamond into other programs a white background turns into a pale color (in most cases pink). Is there any way to overcome this problem?

    This effect can be observed with some (not all) graphics cards when using the 'HiColor' (16-bit) mode. Switching to 'TrueColor' (24-bit or 32-bit) mode solves the problem.

  • Shading edge missing for polyhedron faces

    In non-rendering mode, the option "Shading edge" does not work for polyhedra.

    The "Shading edge" option is unfortunately limited to atoms, bonds, and cell edges only but does not work for polyhedron faces.

  • Post-processing of rendered pictures

    One of the main enhancement of Diamond 2.0 in relation to 1.2 is obviously the "rendering mode". But I was a little disappointed when I discovered that this beautiful way of structural representation was only available on the screen of my PC. [.....] and it was not possible, when you use the command "copy", to work on the picture with CorelDraw! (for example) because the copy is a bitmap one. So, there is not any possibility to "ungroup" the different object. You must work in "normal" mode to obtain a suitable picture...

    The "Rendering Mode" is technically (from the developer's view) very different from the "Standard mode". In standard mode, Windows functions are used to draw objects like spheres, polygons for thick bonds etc., whereas in rendering mode, the OpenGL 3D graphics interface is used. The Windows drawing objects can be exported as 2D objects in a vector-oriented graphics format, but the 3D objects cannot! So it is unfortunately impossible to give the user the opportunity to post-process the structure picture, e.g. by ungrouping objects.

  • Stereo view seems to work not properly

    When I use stereoscopic projection, it seems that background and front are exchanged, however, I was not able to get a reasonable stereoscopic impression.

    For the inspection of a stereo picture stereoscopic glasses should be used. Those glasses separate both pictures by a system of mirrors. With some practice, a stereo picture can also be inspected without such aid. Look between both pictures and focus to an infinite point (try to squint a little bit). Doing so a third picture, which is 3-dimensional, appears between the others. But note, that the last method means, that the left and the right picture are exchanged. This may lead to curious impressions. The effect must be compensated by changing the sign of the stereo angle, which can be found in the 'Display/Structure...' dialog.



  • Problems with printer drivers

    Heavy problems with printer "HP LaserJet 2100 Series PCL 6" on MS Windows 98: When printing a rendered picture, using that driver, Diamond crashes. It also crashes when "File/Printer setup..." is called, even if this dialog is closed with "Cancel" and no printing is performed.

    The problems are obviously caused by bugs in the printer drivers or DLL's for the HP LaserJet 2100. The Diamond users report that they also have trouble in other graphics applications with that driver. It is recommended not to use Diamond together with that driver on an MS Windows 98 system.

  • PostScript printing

    Several users miss a PostScript export function in Diamond.

    Diamond has no special PostScript export function. To get a PostScript file, use the "Print" command from the File menu, where you select a PostScript printer driver and print into a file rather than print directly to a printer. If your printer driver does not support PostScript printing, check for the latest original PostScript printer driver on Adobe's website (



  • CIF read errors

    Frequently reported errors are read errors when importing from CIF files.

    Since CIF is a human-well-readable format (but rather difficult for computer programs to check all syntax pitfalls), it tempts people to manipulate data with a text editor. One important reason to mention here why Diamond does not read data from a CIF file is that the block of data is not initiated with a keyword beginning with "data_", e.g. "data_compound1".
    There is a DOS program available from the IUCr for CIF syntax checking: VCIF.EXE. (See also:

  • Import of Crystal Visualizer files from ICSD CD-ROM

    I have been unable to load data files stored using the Crystal Visualizer program provided with the ICSD CD-ROM.  I have tried both .Z and .ZP formats without any success.

    Crystal Visualizer is an old precursor of DIAMOND 1.xx and its format was supported by the older DIAMOND versions. But this format is no longer supported by DIAMOND 2.0. To visualize structures from the ICSD (Inorganic Crystal Structure Database), export them using the CRYSTIN or CIF format (use 'Export' command from 'File' menu within RETRIEVE). These formats can directly be loaded into DIAMOND and you can start to build up new structure models.

  • PDB export format and CHIME

    I tried to export a structure using the PDB export format. Then I opened this file in an other program (CHIME) and found, that ALL atoms of the atomic parameter list were shown, though I only used a part of them to build up the structure. Moreover, only the exact positions in the parameter list are shown but further atoms were missing. Is this a problem of CHIME or is there a bug in the PDB export format?

    DIAMOND treats the PDB format as it was originally intended, i.e. to save the crystal structure description including bibliography, spacegroup, unit cell dimensions and the atomic parameter list. CHIME and some other programs (e.g. RasMol) also use this format, because cartesian coordinates are used and even an original protein structure from the Brookhaven file normally describes just one complete molecule (if you ignore the spacegroup information). So in fact you tried to export the picture rather than the structure, and that's not possible at the moment. We take this as a suggestion for further improvements.

  • Import of xyz-files

    When I try to import a structure in xyz-format, the structure is not displayed correctly.

    DIAMOND has originally been designed to work with crystal structures. Hence, if you import a structure without translational symmetry (e.g. a molecule in xyz-format), the program tries to build some sort of unit cell around it. Unfortunately, the corresponding algorithm sometimes produces unexpected results due to a non-optimal setting of some internal tolerance parameter.

    Please use the following work-around:

    After loading/importing your xyz-file, select the "Add all atoms" command in the "Structure" menu. Afterwards, select "Connect atoms" (also in the "Structure" menu).

  • Broken up Diamond picture in CorelDraw 7

    Some users report that a rendered picture that has been copied in Diamond with "Edit/Copy" is pasted broken up or distorted into CorelDraw 7.

    Since the Copy/Paste procedure works correctly with other graphics applications (MS Paint, Micrografx Picture Publisher, MS Office, Adobe Illustrator, etc.), there seems to be something with CorelDraw...



  • Bug: Atom type assignments are overwritten after re-opening a DSF file

    If you overwrite Diamond's default assigment of atom types to atoms of the parameter list, by using the "Atom types" dialog, which is opened from the "Edit" menu, and choosing the "Assign" button, these assignments are overwritten the next time you open the DSF file again.
    Use the following workaround to assign atoms to different atom types alternatively:
    1. In the "Atomic parameters" dialog ("Edit" menu), assign the atoms that should be differentitated slightly different oxidation numbers, e.g. "0.0", "0.1", "0.2", etc.
    2. In the "Atom types" dialog ("Edit" menu), overwrite the default atom type symbols, e.g. from "V+0.20" to "V3". 3. Eventually check effective radii, atom type designs, and connectivity.
    These assignments will be persistent in the DSF file when saving and re-opening.

  • How to add centroids and readjust bonding for organometallic cyclopentadienyl complexes

    How can I add centroids and readjust bonding for organometallic cyclopentadienyl complexes?

    You can define a "dummy atom", e.g. by selecting the five atoms of the cp-ring, then call the command "Insert atom..." from the "Structure" menu. "How to insert an atom" is described in the online help: Highlight the "Insert atom..." command in the Structure menu, then press F1 key (the "Insert atom command" page of online help file will open), then go to "How to insert an atom".

  • Recognition and presentation of hydrogen bonds

    Up to now we have not found any option which would make a recognition and presentation of hydrogen bonds available...

    DIAMOND recognizes any bonds within crystal structures by analyzing the interatomic distances. The user has complete control over the bonding criteria via the "Connectivity" dialog. This is the key for solving your problem. You can download a simple example file (orthorhombic D2O, taken from the ICSD database) and use it for the following demonstration:

    1. First let us build up the contents of a unit cell (the water molecules) open up the file ICE.CRY with the "File/Open command".
    2. fill in the unit cell with the "Structure/Fill/unit cell" command
    3. check the connectivity with the "Structure/Connectivity" command: You will find a list of all combinations of atoms, which can be connected. As default the Deuterium-Oxygen pair is selected. You also will see a histogram of interatomic D-O-distances up to 5 Angstroms. The red bars indicate the lower and upper distance limit for D-O-bonds, and these bars can be moved with the mouse or by changing the values for 'DMin' and DMax'. Now change the upper limit to 1.5 and press OK.
    4. change the "Bond Design" of D-O-Bonds with the "Edit/Bond Types" command. Select the D-O atom pair and change the "Style" value on the "Design" page to "Thick"
    5. Now connect all atoms according to the current connectivity with the "Structure/Connect Atoms" command.
    6. complete all incomplete fragments with the "Structure/Complete fragments" command

      Now let us add Hydrogen bonds as thin fragmentated lines
    7. check the connectivity again with the "Structure/Connectivity" command. Now change the lower distance limit for D-O-Bonds to 1.5 Angstroms and the higher limit to 2.0 Angstroms.
    8. change the "Bond Design" of D-O-Bonds again with the "Edit/Bond Types" command. Select the D-O atom pair and change the "Style" value on the "Design" page to "Thin". Now increase the value for "Fragmentation". Finally make sure, that the checkbox "Update created bonds" is NOT marked, because we do not want to change the style of already created bonds.
    9. Finally connect all atoms according to the new connectivity with the "Structure/Connect Atoms" command.
  • How to select and manipulate groups of atoms?

    When someone wants to delete a whole group of atoms, e.g. all the hydrogens, he/she has to do that manually, one by one. It would be nice if when you say "delete" to have besides the options "atom(s)", "molecule" the option "group of atoms" so that you can delete at once all the hydrogens for example

    In former DIAMOND versions ( < 2.0 ) you found several menu entries to manipulate particular groups of atoms, bonds, etc. Since it is not possible to consider all combinations of groups and operations DIAMOND 2.1 follows an other strategy. The key is to switch the right info pane to a particular list of visible objects. This is done with the "List" command of the 'Display" menu. For example: "Display / List / atom types" shows a list of all atom types (i.e. combination of chemical element and oxidation state) of the current structure. Now you can select one or multiple entries of the list with the mouse. Then you open the context menu (right mouse button) and choose the command "Select atom(s)". This will select all corresponding atoms in the graphics pane. At last you should give back the focus to the graphics pane. Therefore you click on the graphics pane while pressing the <Shift> key. Now you are free to manipulate the selected atoms, e.g. delete them by typing the <Del> key.

  • Difficulties in finding the correct space group in the Collapsible tree

    For the novice crystallographer, it is VERY difficult to find the correct space group in the Collapsible Tree.  (i.e. I spent quite a while looking for the P21/A space group).  It would be helpful if I could remove the collapsible tree of space groups and alphabetize ALL SPACE GROUPS in a single list not just those in the outer shell.

    When you type in the required space group symbol and then switch to the list by pressing the <TAB> key, the program automatically jumps to the symbol, so in fact you don't have to search it.

  • 'Fill Unit Cell' command creates atoms outside the unit cell

    I think that I have found a strange error in Diamond. In structure types with very long c-axis I observed that in the 3-D drawing as well as some of the projections (where the c-axis is parallel to the paper) some of the drawn atoms are outside the unit cell.

    The 'error' that you have observed is caused by the 'epsilon value' that Diamond uses to define the borders of the unit cell, when it processes the 3D and 2D projections. The epsilon value is subtracted from the lower boundaries and added to the upper boundaries of the parallelepiped that has the crystal coordinates (0, 0, 0, 1, 1, 1). The default epsilon value is 0.01, leading to (-0.01, -0.01, -0.01, 1.01, 1.01, 1.01). The epsilon value has been introduced to catch the atoms that are (approximately) on edges or faces of the unit cell. For a large axis, say c = 200 Angstroems, the 0.01 becomes 0.01 * 200 = 2 Angstroems in reality, so that atoms farer away from the unit cell edges or faces will be created, too.
    To change the epsilon value, start Diamond, open a file, select a structure, and then - before starting with the 3D projection - use the 'Structure/Fill/super cell...' command. In the Super Cell dialog, set the Epsilon value to a smaller value (e.g. 0.001), select the first option (unit cell), and click OK. (This will fill the unit cell and save the new epsilon value in the Diamond INI file.)



  • Diamond now crashes when starting after months/years of successfull working

    Some users report that Diamond crashes when trying to start the application, although it has been working correctly for months or years.

    Obviously when closing the Diamond application the last time, it has written something "delicate" (due to a non-discovered bug) into its INI file and now crashes whenever it is started and tries to restore settings from that file. You can find the INI file "diamond.ini" in the Windows directory, i.e. usually "c:\windows" or "c:\winnt" (on MS Windows NT). Rename that file to e.g. "diamond.ini.bak" or delete it, then start Diamond. Now it should work with the default settings (like after first installation).

  • 16 bit setup not working on some Windows NT machines

    I have the problem that I can't install your product on my NT machines due to the use of a 16 bit install wizard. My question is if there's a workaround to this problem. It strikes me as being rather odd since the users who requested the use of your product were all running NT machines as well and they didn't have any problems installing the DEMO version.

    In the InstallShield knowledge base we have found the following contribution:
    ERRMSG: The Error Message "16 bit application error message - setup.exe not a valid file" Occurs on Windows NT

    I receive the following error message, "16 bit application error message - setup.exe not a valid file", on some Windows NT workstations when I run my setup.

    Windows NT 4.0 Service Pack 1 and Service Pack 2 have a bug supporting long filenames, in this case, long directory names.

    Windows NT 4.0 Service Pack 3 fixes this problem.
    Note: This issue does not occur on Windows 95 since it uses the VFAT (Virtual File Allocation Table) system, which allows filenames up to 255 characters, mixed-case characters, and spaces.

  • "Send To" converts file name to 8.3 format

    Is it possible to send Diamond an input file (using the 'Send To' right-click file option in Windows NT) and have its full name preserved? As is stands now, the long file name gets converted from its long version (i.e. up to 256 characters) to its 8.3 equivalent when the file is opened in Diamond using the 'Send To' option.

    The 8.3-name conversion when using "Send To" is caused by rather "old-fashioned" run-time library code containing in version 2.x of Diamond. It will be removed in the Diamond 3 version. (On Windows 2000, this works already correct with the current Diamond version.)

  • Crash when closing Edit Atom Types dialog

    One Diamond user reported that Diamond crashes (protection fault) under MS Windows 2000 when closing the "Edit Atom Types" dialog, regardless with OK or Cancel. The same happens with the "Edit Bond Types" dialog, but with no other dialog. Another Diamond user reported a crash, too, but with a relay of approximately 15 seconds after closing the dialog.

    We could not reproduce this error on none of our Windows 2000 computers, but the user with the relayed crash reported that the crash does not more occur after updating graphics drivers. (Compare also "Problems with graphics drivers" above.)

  • Where can I find the distances table

    In an earlier version I could get a list of all the distances (intra and inter molecular) from one atom to all other atoms in a certain distance range. I am not able to find this anymore. Is this option taken away or is it just hidden somewhere?

    The table of distances and/or angles is displayed in a pane right beneath the structure picture pane. To activate this pane or to switch from the data sheet to distances and angles view mode, choose the command "Display / Distances and angles". To change the settings in the table of distances and angles, click with the right mouse button into that table and choose the appropriate command from the context menu.