First, start up Intervision as normal. Using whatever you need for Brightness, Contrast, Slit settings, Barrier Filter, Laser Light, and Laser Intensity, focus your slide as normal. Also, make sure that your zoom level is zoom: 1.00.
Once this is done, click on Intervision Acquisition Tool, pull down the Protocols Menu, move down to Load, and over to ZSeries. This will pop up a new window called ZSeries.
Once there, click on Spacing in the upper left and change the 0.500 to 0.2.
Next, change the Resolution from Medium to Low. If your particles are crappy, you can use jump average to your heart's content. (I prefer live though).
Now, click on the Time Lapse button on the Z Series box. This will open up a new popup. At the top, Stacks is how many individual data files you want. I normally choose 30, but if you want more data, feel free to choose 50 or 100 or whatever. Then, click on Seconds on the 3rd line. This will tell the Piezo how much time it has to move through the sample and then back to the top before starting over. Normally, I have found that if you use a difference of about 25 microns from start to finish, it will need about 30 seconds to properly go through the sample in time. Feel free to experiment with this as you so choose. You can tell whether you've picked an appropriate time by actually taking data (Don't do so now). I will explain this in a bit.
Next, still in the Time Lapse Setup popup box, click on Base Name and enter the name that you want the files to have. All files will start with this name and end in 0000 through 00X where X equals (# of stacks)-1. There is also an option for Hard Shutter. When this is depressed (yellow), it means that in between stacks, it will turn the shutter off to save laser power. Normally this is good for long samples where the interval period is minutes or even hours, so that bleaching of the sample does not occur. Again, feel free to experiment with this option too.
Now, press the Close button. To take data, click on the Record dot in the Intervision Acquisition Tool. To see whether or not you picked a good interval time, wait until the first data is taken. After it is taken, it will say Saving at the top right after Status. After Saving, it should give you a countdown of numbers, normally from 10 down to 0. When it reaches 0, it starts taking data again. You need this countdown to start at LEAST from 2. If you do not see this countdown, you need to stop taking data and click back on Time Series and increase the interval time. Likewise, if the numbers are too high (i.e. starting from 60 and counting down), you can reduce the interval time so that your sample is done in an ideal amount of time. In a fast moving sample, having the numbers too high makes it hard for the tracking program to accurately relate particles in various frames, so be wary. Again, this depends on what exactly you want done and whether you want to take a long data set or a short one. When data taking is done, the Status will say Live and you can close intervision, or take more data.
Transfer your data from Epoxy to Spicysquid using FTP:
spicysquid% ftp epoxy
ftp> mget yourfiles*
The mget command grabs multiple files. The prompt command tells FTP not to ask you are you sure? for each of these multiple files.
Open up IDL in the directory that you moved the files to. To pre-track your data, type in pt3d,'base_name_of_file*' where the base_name_of_file is what you entered in when you took the data.
This will take a nice amount of time depending on how much data you had and how many microns you were analyzing, so go take a break and have some lunch or a beer while Spicysquid does some work. When it's done, type in pta=catptdata('pt.base_name_of_file*'). This will combine all your pretrack data into a single file so that you can analyze it. Once that is done, write out that file using write_gdf,pta,'ptbase_name_of_file' to write out your file. Next, type in help,pta. This will bring up something that tells you the dimensions of your data in a matrix form. Normally there are two numbers, the first number being a single whole number and the second number being a 5 to 7 digit whole number. Basically, just look for any abnormalities.
To look at your data, you need to type in plotp,pta. This will open up an IDL screen that plots each and every single one of your data points. Again, look for any abnormalities--weird stuff that you don't think should be there. Now, type in plot,pta(3,*),pta(4,*),ps=3 this will give you a general plot of your data so that you can crop it. Basically, you need to get the main body of the data and remove outliers so that you won't have problems later on. Use this command ptb=eclip(pta,[3,(left_edge),(right_edge)],[4,(bottom_edge),(top_edge)]) where the y-axis is the 4 category and the x-axis is the 3 category to crop your data. For instance, one of them could look like this: ptb=eclip(pta,[3,0,4e4],[4,4,12]) This would crop everything except the region between 0 and 4e4 on the x-axis and between 4 and 12 on the y-axis leaving you a nice little bubble. There's a description of this clipping procedure at the particle tracking tutorial web page
After you have done this, type in ptb(0:1,*)=ptb(0:1,*)*0.2352 and then ptb(2,*)=ptb(2,*)*0.2 to normalize your data. If you are wondering where the two numbers came from, the 0.2352 came from when you first entered in your noran data, it is the x and y calibration (microns per pixel). (So your data probably came out like (0.2352, 0.2352, 0.2). Likewise, the 0.2 came from the final part of the original entering. It is basically the number for the SPACING that you typed in when you originally took the data.)
Once that is done, write your file out using write_gdf,ptb,'ptb_base_name_of_file'. Now, type in emgrwrap,'ptb*',/special,/pretrack. This is Eric's program to view the pair-correlation function of your data. It will take a short amount of time and should go through all the data files and create one large file that starts with a gr so that you can easily view it. To view this gr file, type in gr=read_gdf('gr.pt*') and then plot,gr(0,*),gr(1,*) and this should bring up a graph that looks identical to the final one that popped up at the end of the pair correlation; see the picture at right. The x-axis is in microns; the first peak indicates the most probable separation for nearest neighbors, and the second smaller peak is the 2nd-nearest-neighbors, etc.
If you're interested in looking at the differences in two different pair correlation plots, you can use the oplot command to compare two. First, enter in your second gr data by using gr2=read_gdf('gr.pt_name_of_file') Note: if you do DO an overplot, make sure you change the asterisk (*) in the gr=read_gdf command above to the base name of the file that you want as the first gr and use the base name of the file that you want as the SECOND gr in the gr2=read_gdf command. To overplot, simply plot the first gr as normal and then type oplot,gr2(0,*),gr2(1,*),lines=1 where 1 can be any whole number up through 10.
Next, if you want to see where the peak positions are and the height and so on, use the following command. grwrap3,'gr.pt*' and this will show you the peak and heights and other information of the data on your pair correlation. Again, if you had more than one gr, make sure you use the appropriate name of the file instead of gr.pt*.
Start by making a ptb file as described above. The main important points are that the data need to be calibrated into microns, and you need to have clipped out any bad particles. Then run two commands:
If you wish to show colored lines in IDL, use the following commands before you plot your graph:
Any comments, let Eric Weeks <weeks(at)physics.emory.edu> know.