These questions are collected from our NITRC help forum. Questions are still being actively added to the list. If you have more questions to ask, or if you want to share your experience in dealing with your problems, please visit the forum site

http://www.nitrc.org/forum/forum.php?forum_id=1553

or send an e-mail to Hanchuan Peng or other Vaa3D developers.

Content

To produce the missing libtiff3, you can run your build command (e.g. build.linux) with "-T", which will recompile the tiff libraries.

You can build libtiff under v3d_main/common_lib/src_packages/tiff-3.8.2.1 . That is a libtiff library slightly changed by me so it can be used to read LSM files. If you do in this way, after you re-compile libtiff, you can set a path to the tiff. And it will work.

Most probably this is because you are using an already built qt library rather than building the source code yourself. ou will need to download the Qt source code from

http://qt.nokia.com/downloads/downloads#qt-lib

and build from there using the simple commands: (btw, the qt4-demos has a different path, and will not work, unfortunately).

1. first unzip qt. using "tar zxvf qt_xxxxx.tar.gz" (or similar command)
2. build qt: first enter qt uzipped folder, and run "./configure -nomake examples", and then run "make -j4" or (whatever -jk, where 'k' depends on how many processors you have on the machine) and then run "sudo make install"
3. add "qmake" of the qt to your path. For example, if you are using bash and qt4.7.3, then your qmake path should have the default like "/usr/local/Trolltech/Qt-4.7.3/bin/qmake". therefore you should just add in your .bashrc (or .bashrc_profile) file the sentence "export PATH=$PATH:/usr/local/Trolltech/Qt-4.7.3/bin". Do NOT forget to run "source .bashrc" or a similar command to update the environment variables.
4. Now QT is built, and what you need to build V3D is just one command "./build.linux -B".

Note: It has been shown that our program is most compatible with Qt-4.7.1 or Qt-4.7.3. Problems have been reported when Qt-4.8.x is used on Linux platform. Please be careful with your Qt version.

It seems to be a problem with a Lion gcc compiler. llvm g++ was set as the default compiler, which is not 100% compatible with the gcc/g++ code.

We are still doing more test on it. The simplest solution is to re-link the /usr/bin/g++ to the /usr/bin/g++-4.2 or the same folder. Suppose you are using linux/mac os, you can use the following command:

	mv /usr/bin/g++ /usr/bin/g++_old
	ln -s /usr/bin/g++-4.2 /usr/bin/g++
	mv /usr/bin/gcc /usr/bin/gcc_old
	ln -s /usr/bin/gcc-4.0 /usr/bin/gcc

Note: You may need the administrative right to change the default compilers.

Yes, your graphics card should support OpenGL. Basically any graphics card that supports OpenGL 1.4 should work well. But if OpenGL 2.0 is supported it will be super. Also the larger video memory the better.

One way to try is to load a SWC (neuron) downloadable from vaa3d.org test data page. It is just a simple surface file. If you cannot even visualize that, then you definitely have some problem with your card.

Note: We have noticed that some Intel cards do have problems with rendering.

In the Tri-view window control panel, now we add a new color manager for multi-channel images. Color manager will do on-line color blending based on user-selected colors, different blending modes, as well as on-line contrast and brightness adjusting. It also allows the look-glasses based color blending. This computation slows down the tri-view, however, it provides a lot more flexibility to visualize multi-channel images.

You can define a 3D roi using one mouse stroke in the 3d viewer. you can also define an roi in the tri-view by pressing ctrl/.wiki + left mouse key. In the tri-view window, you can also remove the just-defined corner of the ROI by pressing alt + left mouse key.

You can also use the main menu to clear the entire ROIs (for XY, YZ, and ZX views), and also copy and paste ROIs from images to each other using a standard Vaa3D ROI plugin.

A detailed instruction for tracing neurons is described in our wiki page:

Vaa3DNeuron1

You can also refer to our neuron tracing example movies to trace neurons.

you can merge multiple swc files into one using the plugins->neuron_utilies->sort neuron.

This function is provided in the latest vaa3d_tool repository. For more information regarding check-out location and usage, please see this page.

You can use the image modulation plugin. To do so, go to the main menu, select "Image/Data" -> "color channel" -> "split color channels", then you can have multiple images, each corresponding to a separate channel.

Then, you go to main menu -> "Plugins" -> "Vaa3D_PluginInterface_Demos" -> "Multi_Image_Interface" -> "Modulate 2 images..." and then in the dialog you pick two images (for these separated channels) and produce a single new image which has two channels have high intensity at the same time.

Then you go to main menu -> "Plugins" -> "image_segmentation" -> "Label_Objects" -> "Label image objects", and fill in information in the dialog and you can produce a segmentation image, which can be used for collecting statistics.

Assume you have an original grayscale / multicolor image opened, and also a segmented mask image opened, you can go to the main menu, select "File" -> "Export" -> "Export 3D ... " and then generate a pointcloud annotation file. The actual statistics has been saved in the file. You can load the file using plain text editor, or the Vaa3D Matlab file IO toolbox function "load_v3d_pointcloud_file".