• Survey Mode = refraction
• Set your line number
• Set your geophone interval at the desired spacing.
• Pick a number (default is 1000 which might cause problems!) for the coordinate for your first shot point. Later shotpoints can be on either side of that. It seems (9/2002) that the number you choose might have to be on the scale of the depths/distances you are investigating. We used 1000 and the SIPQC software is crashing before printing the cross section. With the 24-channel upgrade (9/2003) it seems the simple way to set up is to lay out the geophones, call geophone location #1 something like 100 and then measure the shotpoints from there. SIPQC is easily confused - keep it simple, don't flip the polarity of the line, just move your shotpoints around on the positive parts of your coordinate system.
• Make sure the specified geophone interval is correct, and enter the x-coordinate of geophone number one. The interpretation software will let you do up to seven shots with the same spread of geophones. For each shotpoint, the x-coordinate of the geophones stays the same and you just put in the new x-coordinate of successive shotpoints (figure). Make sure that the nearest geophone to the shot is labeled #1 on the display; otherwise the software gets confused (change with "-" sign).
• Ignore shot interval - that's for reflection surveys.
• At this point, the cross section display at the top of the screen where the asterisk (*) marks the shot point should make sense. For successive shotpoints, leave the x-coordinates of the geophones alone and change the x-coordinate of the shotpoint appropriately; watch the diagram and think about the distances shown.

3. ACQUISITION menu

• Select default but note that you may have to change "sample interval" and "record length." Sample interval and record length vary sympathetically. The sampling frequency is {1/Sample interval}. Thus a sample interval of 500 ms (0.5 s) yields a sampling frequency of 2 samples/second (2 Hz). Use the fastest sampling for short surveys in fast materials. Record length is just that - how long will you record data for? How long is your line? If you have a 24 meter line in 1,000 m/s material then you will need 24 ms to get the first arrival; set it appropriately.
• Some of the software (automatic picks) tends to work a little better with a small, negative delay (-10).
• Leave acquisition filters "out"
• Stack Mode = autostack

4. FILE menu

• Use the directory named "Class"
• Set your file name (which will be a number)- SMARTSEIS will increment this one at a time when you SAVE files.
• Record all pertinent information about your refraction line and this file in your field notes (where is it?, who is with you? which way does it run?, sketch a map, etc.)
• Autosave = OFF (or each stack will get saved)

5. DISPLAY menu - you can experiment with these with data in memory

• Type: clipped and shaded works well for refraction
• AGC = fixed gain
• Time Scale changes the number of data samples per pixel
• Print Time Scale = compress by two or four to save paper!
• Trace Size: you'll come back to this after stacking some data. Units are +/- 3 Db. A 3 Db increase is a 41% increase in gain, a 3 Db decrease is a 29% decrease. Thus two steps is double (or one half) the original value.
• Filters: Some experimentation here can sometimes help you in picking first breaks. Try a low cut if you have lots of low frequency noise from distal geophones as those gains are usually sufficiently high to amplify background vibrations from distant traffic and the like.

II Collecting Data - Select DO_SURVEY on the main menu

Collect initial data and figure out the required gains for various geophones:

• Select "Noise Display", jump up and down and watch for inactive geophones, adjust the noise with up/down keys to low background noise.
• Select "Trace Display"
• Smack the ground once or twice, you should start to see some signals amidst the noise.

USUALLY you have to mess (considerably) with the DISPLAY TRACE SIZE (see above) to get a meaningful looking screen of data. The instrument defaults to the last set of gains/channel that you used. It is not uncommon to have too low of a gain for the first few phones, thereby missing the first arrival.

Stack up a reasonable amount of data. Signal/noise will increase by about the square root of the number of blows, 12-15 stacks is almost always plenty. Fool around with the trace size characteristics to get good looking first breaks.

• Select "Inspect Arrivals" and the Smartseis will pick the first breaks and give you a chance to adjust them. Adjust the picks of first breaks and save them, the file will have a .BPK" extension.

III Interpret your data - Select ANSWERS on the main menu

• Adjust Picks lets you do just that
• Solve Refraction uses the line geometry information you entered under Geometry above. The line geometry must be correct or your answers will be wild - so be really careful when setting up the survey. Note that the line geometry menu lists the current geometry settings, not necessarily those of the file you are looking at.
  

When you select Solve Refraction a box appears that lets you select the files you want to include in the analysis. Highlight and select (press the period key) the appropriate breakpoint files (.BPK) then press ENTER.

Using the cursor keys, assign each breakpoint (in T-x space) to layer 1 through 6. If you want to omit a point, assign it to layer 0. Press ENTER and the Smartseis will save another file, .LPK, and SIPQC will whir and grind, then eventually print your results or give you an error message. Assumptions include less than 8 shotpoints, level ground, geophones in a straight line, and that you selected reasonable first breaks and layers.

If SIPQC likes what you have done (velocities increase down section, geophones increase away from source, etc.) it will make a nice interpretive plot showing a shaded cross section. If it doesn't, it didn't like something you did or picked. I like to start with a small geophone interval (1-2 meters) close to source (2 meters or so) and then collect data and step the source back from there. This way, I can see what is developing as I collect data and lines.

• Linear Velocity - lets you calculate the velocity of a linear event in T-x space.

IV Put The SMARTSEIS Away - keep everything clean

• Wipe dirt off the geophones and place them in the bottom of the box
• DO NOT COIL THE CABLES; stack the cables in the box
• Put the battery back on the charger using the 2 amp charge rate.

V Some past problems:

• The first arrival changes phase across the screen - just like in the pictures in the manual. This is wrong and the problem is caused by too low of display gain on the first phones; thus you miss the first and see the second arrival. Later phones, with higher gains, show the first arrival making it look like the polarity is changing across the screen. Often it takes considerable gain to see the first arrival - experiment, turn them up to high, figure it out in each new situation. Rob Huggins at EG&G confirmed that the manual pictures were "wrong".
• The directions, throughout the manual and on the screen, are not always super-intuitive. You usually will have to play around with things to figure it out. The key is to think through each menu as you go from left to right, top to bottom.
• To run SIPQC, select "inspect arrivals" under DISPLAY, and pick first arrivals. Go to solve refraction, and the SmartSeis should put up a T-X display. Hit the "." key, then for each t-x point on layer #1 hit #1, for those on layer two, hit #2, up to six layers. Mark any point that you wish to exclude from analysis with a "0". Hit the return key when you are ready.
• A common problem is that the previous user did not know what she/he was doing. In this case all the parameters will be screwy and you have to figure it out, step by step.