Frequently Asked Questions

(Last update: 10/XI/1998)

General

Where is the TVIPS homepage ?
How can I contact TVIPS ?
My demo dongle is used up. How to proceed ?
How often do I have to enter the license code ?
What is the viewport concept ?
The accelerator keys do not work. What's wrong ?

EMMENU

What are the configuration files for ?
Why is there no dialog to change the configuration files ?
How can I print my images ?
Can I see 16bit greyvalues using the mouse cursor ?
How can I interrupt started actions ?
What file formats does EMMENU support ?
How are the old and new EM-format defined ?

Camera

How is the coordinate system defined ?
The camera does not send images. What's wrong ?
The images look bad. What's wrong ?
What is flatfielding ( flatfield correction) for ?
How does the flatfield correction work ?
Do I have to use different flatfields ?
How often do I have to acquire new flatfields ?
How do I acquire flatfields in the EMMENU ?

TEM

It takes more than 30 seconds to take an exposure. What's wrong ?
How can I test the remote control to a Philips TEM ?
The remote control does not work. What could be wrong ?
The exposure shutter does not work correctly. What's wrong ?
The EMMENU can't handle the magnification or high tension. What could be wrong ?
How to calibrate the post magnification ?

General

Where is the TVIPS homepage ?

The tvips homepage is located at http://www.tvips.com.

How can I contact TVIPS ?

Phone ++49 (0)89 850 6567
Fax ++49 (0)89 850 9488
Mail: TVIPS GmbH, Herbststr. 7, D-82131 Gauting, Germany
EMail: support@tvips.com

My demo dongle is used up. How to proceed ?

You used up the demo time of EMMENU. You have to decide now whether you want to buy EMMENU now or not. Contact TVIPS stuff.

How often do I have to enter the license code ?

You have to enter the license code just once. It is stored in the registry in the section HKEY_LOCAL_MACHINE and therefore available for all users.

What is the viewport concept ?

Every program dealing with lots of images has the problem how to display the images in a comfortable way. The most spread way is to display each image in a separate window. The big disadvantage of this concept is that it is very nasty to bring your desired image to the top if many windows are on the screen. The viewport concept behaves like a single window ( here in TCL with a draw area of 1280x1024 pixel ). This area may be divided in the so-called viewports. Within EMMENU the viewports are square and have the selectable sizes 256x256, 512x512 or 1024x1024. Whenever an image has to be displayed, it is displayed in the next viewport. If the last viewport has been used, it starts again from the beginning and overwrites already displayed data. It is important to realize that only the display area is overwritten, the image data of the previously displayed image is not touched. In the EMMENU the last displayed image has a red border whereas all the other ones have green borders.

The accelerator keys do not work. What's wrong ?

The accelerator keys ( like F5 for the screen saver ) are bound to the main window. That means that the main window MUST be the active window to enable the accelerator key handling. If any of the dialog boxes is active, this dialog box rejects this key stroke.

EMMENU

What are the configuration files for ?

The configuration files set internal EMMENU variables accordingly to the available hardware in and around the computer ( memory, camera, TEM, etc. ). Also initial EMMENU values are set here. Experienced users may edit those files to adapt the software to their wishes.

Why is there no dialog to change the configuration files ?

The contents of the configuration files varies very much between the installations, so that an easy to use dialog that covers all the possibilities and user's wishes is quite impossible. It is much more flexible ( and faster ) to change the configuration with an ascii file editor ( like notepad ). Besides that, a TVIPS stuff member will edit the configuration file during installation and adapt it both to the available hardware and the user's wishes. If the installation is performed only by the user, TVIPS will support this user with adapted configuration files, either by floppy disk or internet.

How can I print my images ?

There are several possibilities to print, depending on what you want to do.

To send the whole contents of the 1280x1024 viewport to the printer:
1. Menu File -> Print Setup ( select a connected printer, paper, format, etc. )
2. Menu File -> Print page setup ( select the position and size of the print area on the paper )
3. Menu File -> Print preview ( check the settings of 1. and 2. , if not OK, go back to 1. or 2. )
4. Menu File -> Print ( start print process )
To print a single viewport or the whole 1280x1024 screen with my own paintings included
1. In the dialog "EM -MENU V3.0" click button "Display..." and in the dialog "Display setup" click the button "Hardcopy".
2. Select the area of interest and the overlay color handling, then save the data ( in TIFF format ).
3. Start the graphic program you prefer and load the image.
4. Modify and print the image using your graphic program.
To print image data and some dialog boxes with interesting data ( e.g. linescan )
1. Rearrange your dialog boxes on your display in a way that you like. The image data of interest must not be covered by dialog boxes.
2. Click in the toolbar on the 2nd button from the left ( with the green arrow pointing out of the monitor ) and a dialog will open where you can select if you want to copy a 256x256 area, a 512x512 area, a 1024x1024 area or the whole current visible main window. If you selected a NxN area, you have now to select it, by moving the quare with the mouse. All the contents of the desied area is now copied to Window NT's paste buffer ( if e.g. a dialog box is within the selected area, this dialog is also copied into paste buffer ).
3. Start a graphic program that supports paste from external source ( e.g. PaintShopPro, MS Photo Editor, ... ) and paste this image from the paste buffer into a new image within the graphic program.
4. Modify and print the image using your graphic program.

Can I see 16bit greyvalues using the mouse cursor ?

No, clicking with the left mouse button into the display area ( and holding the left mouse key pressed while moving the mouse ) shows you the scaled currently visible greyvalue. Due to the fact that this is an open system and the user has (nearly) anytime read and write access to this screen, it is impossible to keep in mind where this displayed image comes from and what the actual greyvalue at the pixel position would be. To get information about real 16bit grey value information at a certain pixel position you will have to start LINESCAN or click on the small button labeled [>] to switch into zoom mode.

How can I interrupt started actions ?

Try to hit the SPACE key once. If an action ( e.g. acquisition of an image series ) is interruptable, it reacts on the space key. Note that it sometimes takes a while until an action may be interrupted. DON'T hit the space key several times, because if the action is interrupted there are one or more hits of the space key left in the message queue, each program has under Windows NT. As soon as the control is back to the dialog box, Windows NT will check, whether there are unhandled keys in the queue and will react on them. Under Windows NT there is no difference clicking on a button of the dialog box or to hit the space key ( if and only if the button has the focus ). I.e. hitting the space key several times will interrupt an action, restart it, interrupt it again, restart it, etc. until all hits of the space key are out of the queue.

What file formats does EMMENU support ?

EMMENU reads and writes files in the so-called EM-format, both the old one ( with a 256 byte header ) and the new one ( with a 512 byte header ).
EMMENU supports also writing ( exporting ) image data as scaled 8-bit TIFF and unscaled 16-bit TIFF.Nearly every graphics program can read the 8-bit TIFF data, while most of them will either not read 16-bit TIFF data or will display it scrumbled. The TIFF data, created by EMMENU fulfills the TIFF 6.0 specification. Both 8 and 16 bit TIFF files have a special TemData tag. This tag is a private tag ( reserved for TVIPS by ADOBE® ) and holds all the user data also available in the EM-format. Ask TVIPS stuff for further documentation if you want to evaluate this tag from your own applications.
EMMENU also supports writing unscaled 16-bit RAW data type. This data type will help you to interchange image data, if any other format will fail. Note that each data pixel is stored in an 16bit integer ( C-type SHORT, assembler-type WORD) and uses the little endian format.
Using user macros, you can program reading and writing of every file format you like.

How are the old and new EM-format defined ?

The EM-file format was subject to a major change a time ago, so the main difference is the file header. While the old EM-file format has a 256 byte header with 40 float user values, the new format has a 512 byte header and the data type of the user data is stored in 32bit integers.

Fileheader (256 byte)

Image data

Fileheader (512 byte)

Image data

Old EM-file format

New EM-file format

Here is a description of the 256 byte header of the old EM-file format with byte = 8 bit integral type, long = 32 bit integral type and float = 32 bit floating point type. All data is stored in big endian format.

Offset	data type	Number of bytes	Comment
0	byte	1	always 0
1	byte	1	0 ( means now: Old format )
2	byte	1	always 0
3	byte	1	data type: 1-byte, 2-int(16bit), 4-int(32bit), 5-float(32 bit), 8-complex(2*float)
4	long	4	number of pixels in X direction
8	long	4	number of pixels in Y direction
12	long	4	number of images in this file
16	string	80	zero terminated comment string, i.e. max. 79 chars usable
96	float	4	high tension [kV]
100	float	4	spherical abberation [mm]
104	float	4	illum. aperture [mrad]
108	float	4	electron optical magnification [x 1000]
112	float	4	post magnification [x 1]
116	float	4	focal length [mm]
120	float	4	defocus [nm]
124	float	4	astigmatism [nm]
128	float	4	astigmatism [mrad]
132	float	4	voltage of Biprism [V]
136	float	4	specimen tilt angle [mrad]
140	float	4	specimen tilt direction [mrad]
144	float	4	illum. tilt angle [mrad]
148	float	4	illum. tilt direction [mrad]
152	float	4	Mode: 0 = image, 1 = diffraction
156	float	4	energy spread [eV]
160	float	4	chromatically abberation [mm]
164	float	4	not used
168	float	4	not used
172	float	4	not used
176	float	4	CCD pixels in x and y direction
180	float	4	CCD offset in x direction
184	float	4	CCD offset in y direction
188	float	4	CCD single pixel size [m m]
192	float	4	CCD binning factor
196	float	4	CCD readout speed [kHz]
200	float	4	CCD gain ( 0 = low, 1 = high )
204	float	4	CCD sensitivity [ADU/primary electron]
208	float	4	CCD exposure time [ms]
212	float	4	flatfield correction done ( 0 = no, 1 = yes )
216	float	4	dead pixel correction done ( 0 = no, 1 = yes )
220	float	4	mean value [ADU]
224	float	4	standard deviation [ADU]
228	float	4	not used
232	float	4	not used
236	float	4	not used
240	float	4	not used
244	float	4	Minimum [ADU]
248	float	4	Maximum [ADU]
252	float	4	Quality factor for statistic calculation [%]

Here is a description of the 512 byte header of the new EM-file format with byte = 8 bit integral type, long = 32 bit integral type and float = 32 bit floating point type. All data is stored in big endian format.

Offset	data type	Number of bytes	Comment
0	byte	1	always 0
1	byte	1	1 ( means now: New format )
2	byte	1	always 0
3	byte	1	data type: 1-byte, 2-int(16bit), 4-int(32bit), 5-float(32 bit), 8-complex(2*float)
4	long	4	number of pixels in X direction
8	long	4	number of pixels in Y direction
12	long	4	number of images in this file
16	string	80	zero terminated comment string, i.e. max. 79 chars usable
96	long	4	high tension [V]
100	long	4	spherical abberation Cs [m m]
104	long	4	illum. aperture [m rad]
108	long	4	electron optical magnification [x 1]
112	long	4	post magnification [x 0.001]
116	long	4	CCD exposure time [ms]
120	long	4	CCD pixels in x and y direction
124	long	4	CCD pixel size ( incl. binning) [m m]
128	long	4	not used
132	long	4	image length [nm]
136	long	4	defocus [0.1 nm]
140	long	4	not used
144	long	4	not used
148	long	4	not used
152	long	4	not used
156	long	4	not used
160	long	4	not used
164	long	4	not used
168	long	4	spec. tilt angle [mdeg]
172	long	4	spec. tilt deg. [mdeg]
176	long	4	not used
180	long	4	not used
184	long	4	not used
188	long	4	CCD offset in x direction [pixels]
192	long	4	CCD offset in y direction [pixels]
196	long	4	not used
200	long	4	not used
204	long	4	not used
208	long	4	CCD binning factor x and y
212	long	4	not used
216	long	4	not used
220	long	4	not used
224	long	4	not used
228	long	4	CCD sensitivity [ADU/e]
232	long	4	not used
236	long	4	not used
240	long	4	not used
244	long	4	not used
248	long	4	Date (Julian date)
252	long	4	not used

If the 4 bytes at positon 256 are not $77777777, then the rest of the header is not used, i.e. normally filled with 0. If $77777777 is however at position 256 then the layout below is valid. The reason for this structure is historically. If the magic number $77777777 is available, the program should use the float data stored between the offsets 260 and 423, since they are used completely. ( All data are in big endian format )

Offset	data type	Number of bytes	Comment
256	long	4	Magic value $77777777
260	float	4	high tension [kV]
264	float	4	spherical abberation [mm]
268	float	4	illum. aperture [mrad]
272	float	4	electron optical magnification [x 1000]
276	float	4	post magnification [x 1]
280	float	4	focal length [mm]
284	float	4	defocus [nm]
288	float	4	astigmatism [nm]
292	float	4	astigmatism [mrad]
296	float	4	voltage of Biprism [V]
300	float	4	specimen tilt angle [mrad]
304	float	4	specimen tilt direction [mrad]
308	float	4	illum. tilt angle [mrad]
312	float	4	illum. tilt direction [mrad]
316	float	4	Mode: 0 = image, 1 = diffraction
320	float	4	energy spread [eV]
324	float	4	chromatically abberation [mm]
328	float	4	not used
332	float	4	not used
336	float	4	not used
340	float	4	CCD pixels in x and y direction
344	float	4	CCD offset in x direction
348	float	4	CCD offset in y direction
352	float	4	CCD single pixel size [m m]
360	float	4	CCD binning factor
364	float	4	CCD readout speed [kHz]
368	float	4	CCD gain ( 0 = low, 1 = high )
372	float	4	CCD sensitivity [ADU/primary electron]
376	float	4	CCD exposure time [ms]
380	float	4	flatfield correction done ( 0 = no, 1 = yes )
384	float	4	dead pixel correction done ( 0 = no, 1 = yes )
388	float	4	mean value [ADU]
392	float	4	standard deviation [ADU]
396	float	4	not used
400	float	4	not used
404	float	4	not used
408	float	4	not used
412	float	4	Minimum [ADU]
416	float	4	Maximum [ADU]
420	float	4	Quality factor for statistic calculation [%]

If the 4 bytes at positon 424 are not $66666666, then the rest of the header is not used, i.e. normally filled with 0. If $66666666 is however at position 424 then the layout below is valid. There are no units specified, since the values are highly microscope dependent. ( All data are in big endian format )

Offset	data type	Number of bytes	Comment
424	long	4	Magic value $66666666
428	float	4	Spotsize
432	float	4	Intensity
436	float	4	Beam shift X
440	float	4	Beam shift Y
444	float	4	Image shift X
448	float	4	Image shift Y
452	float	4	Goniometer X
456	float	4	Goniometer Y
460	float	4	Goniometer Z
464	float	4	Goniometer Alpha
468	float	4	Goniometer Beta
472	float	4	Future, must be 0
476	float	4	Future, must be 0
480	float	4	Future, must be 0
484	float	4	Future, must be 0

The bytes in the header between 488 and 511 are not yet used and should be 0

CAMERA

How is the coordinate system defined ?

The coordinate system is defined as follow (n - number of columns, m - number of lines):

(0/0) (1/0) (2/0) ... (n-1/0) (0/1) (1/1) (2/1) ... (n-1/1) (0/2) (1/2) (2/2) ... (n-1/2) ... ... ... ... (0/m-1) (1/m-1) (2/m-1) ... (n-1/m-1)

An angle between two point is defined in degree and is increasing for rotation in mathematically positive sense.

E.g.: (128/128) - (256/128): 0 degree; (128/128) - (256/64): 45 degree

The camera does not send images. What's wrong ?

Have a look in the dialog "Output".

If there is an error message 53 or 2001 visible or the message "9901 -- CCD camera error / Camera not initialized. Use CCDINIT first. ", then the contact to the camera is missing. Maybe the camera controller ( with the temperature display ) is switched off, or one of the cables between computer and the controller or between the controller and the camera head is loose or the connectors are not correct in their sockets.

Also check in the configuration file that the correct camera controller is selected.

Due to a locking in the camera board, it is necessary to reset this board. Depending on the motherboard it is sometimes enough to logout, re-login and restart TCL/EMMENU, but mostly you will have to shutdown NT, switch the computer off ( reset is sometimes not enough ) and reboot.

The images look bad. What's wrong ?

The chicken-wire is visible in spite of flatfield correction
1. The flatfields do not fit anymore -> acquire new flatfields
2. The intensity of the acquired image is very low. Maybe lot of information is lost while the darkfield is subtracted from the uncorrected image. -> Increase intensity
3. The shutter does not work correctly. -> Check EM shutter
4. A data transfer problem -> Check all the cables and connectors between PC and camera controller and between camera controller and camera.
There is always the same structure visible in the images
1. Probably flatfields have been acquired with some specimem visible. Looking at the formula below you will see, that this specimem image is "corrected" into all the other acquired images. -> Acquire correct flatfields
There is only dark current noise visible
1. The shutter did not open -> Check shutter
2. The shutter was disabled -> Check shutter control box
3. The beam blanker was switched on -> Check ALL(!) beam blankers
4. EMMENU was set to acquire only dark or bias images -> Change format to NORMAL

What is flatfielding ( flatfield correction) for ?

Flatfielding is a shading correction process and is required to make corrections for variations in the CCD's responsivity. There are several reasons making flatfielding necessary:

The hexagonal pattern ( so-called chicken-wire ) of the fiber optic will be removed.
Changes in sensitivity of the single CCD pixels will be corrected.
Dark current variations will be removed.
Shutter properties will be removed.

All those corrections can be performed due to the extremly high linearity of the CCD chip.

How does the flatfield correction work ?

There are 2 images required for the flatfield correction. A so-called "dark image" D, an exposure without any light falling on the CCD chip, and a so-called "flat image" F, an exposure where the chip is even illuminated ( and no object is projected to the chip ). The flat-corrected image C is computed out of the uncorreced image U by the formula:

C := ( U - D ) / ( F - D )

Do I have to use different flatfields ?

Yes, you need a pair flatfield/darkfield for each combination of the readout formats, camera gains, the readout speeds, the shutter types and high tensions you will use. Using the wrong flatfield will result in wrong corrections of the image data, like over~ or undercorrection. In the EMMENU the correct flatfield and darkfield are loaded automatically accordingly to the current format/gain/readout-speed/shutter/HT settings.

How often do I have to acquire new flatfields ?

It is difficult to specify a time. Some use the same flatfields for month, some acquire new flatfields daily. Due to many possible reasons (mechanically, temperature, electronical lifetime, etc.) the sensitivity of the CCD chip changes, also there might be a change in the scintillator coupling, etc. All those reasons might change the uncorrected image in such a way, that the formerly acquired flatfields will no longer correct the image as desired. This is the latest moment where you will have to acquire new flatfields.

How do I acquire flatfields in the EMMENU ?

Preparations
- Select the camera format, the gain, the readout speed and the shutter type in the "Camera format" dialog.
- Search for a "hole" in your specimem, so that there is an even illumination on the TEM's screen.
- Open the flatfield dialog box.
Pre-Acquisition
- Adapt the beam's intensity and the exposure time. To do so, click on button "Exposure (uncorrected)" and check the mean value. A "good" value for mean is about halve the maximum possible intensity ( e.g. about 2000 for a 12bit camera [ 2^12 = 4096] )
Acquisition
- Select the number of exposures. This is the number of dark and flat exposures that will be taken. The final dark image and flat image is the average of those exposures. The more exposures you take, the more you get rid of noise ( but the more time to acquire the data is required ). At least 4 exposures should be used.
- Start the acquisition
Post-Acquisition check
- Have a look at the darkfield and the flatfield and perform a visible check. Here you will see undesired effects like "shutter did not work", etc.
- Click on the button "Exposure (corrected)" ( still with the image of the hole !!!). You should get perfect noise.
- Click on the button "Power" to compute the power spectrum of the corrected image. There shouldn't be any signs of chicken-wire reflexes nor other prominent spots visible.
- If one of the checks fails go back to 2. or 3.
- Save the flatdata

TEM

It takes more than 30 seconds to take an exposure. What's wrong ?

Obviously the remote control to the TEM is not working. Therefore the EMMENU has to wait for the timeout of the remote control software. See "The remote control does not work. What could be wrong ?"

How can I test the remote control to a Philips TEM ?

Leave the EMMENU ! Open the dialog Direct TCL command (if it is not already open) and type in the following commands in that succession:

Command to be typed	Comment
TEMCLOSE	If an error occurs, ignore it. This command is just to be in a defined state.
TEMOPEN XXX 9600	Replace XXX by the name of the serial interface where the remote control cable is connected to the computer (e.g. "COM1" )
PCOMM	Now the TEM should answer its name and software version
TEMCLOSE	Clean up

The remote control does not work. What could be wrong ?

Please go through the following check list

Is the remote control cable connected to both the computer and the TEM ?
Is the cable connected to the correct connectors ?
Is the remote control enabled on the TEM side ?
(Philips TEMs only) Does PCOMM work ? (See "How can I test the remote control to a Philips TEM ?")

If there is still no remote control, shutdown Windows NT, switch the computer completely off ( just a reset might not be sufficient ), switch it on again, reboot and go through the list above. If there is still no remote control possible, ask TVIPS stuff for additional checks.

The exposure shutter does not work correctly. What's wrong ?

(Philips TEMs only): Check the detector configuration page if the detector position is set to CENTRAL.
(Philips TEMs only):If hardware spotscan is installed, check the disable-shutter-switch located on the right side of the column. Switch must be in position NORMAL.

The EMMENU can't handle the magnification or high tension. What could be wrong ?

The EMMENU only supports a fix list of high tensions and magnifications. Those high tensions have to be specified in the microscope.conf configuration file and the magnifications have to be specified in the mag.conf configuration file. Ensure that the free high tension control is switched off. Check whether the high tension or the magnification is entered correctly in the list.

How to calibrate the post magnification ?

To calibrate the post-magnification perform the following steps:

Insert specimen with known spacing.
Select a reasonable magnification.
Set this magnification in the submenu EM-SETUP.
Acquire an image.
Calculate the power spectra.
Go to the command LINESCAN
Define a linescan over the known reflex
Move the cursor to the reflex
Click on the button "Calib"
Enter the known spacing in the dialog box
Go back to the main menu.
Have a look in the USER DATA
The calibrated post-magnification will be shown.

Note: This value is only valid for the current image. To make this value permanent you have to enter that value into the microscope configuration file.