7 - Mistie Analysis and Correction
Automatic Mistie Calculations
Select (Mistie), (Automatic Mistie) calculations. Change the window to cover all data except the first breaks (500 to 3000ms). Use 9 traces for flat Alberta at line ties and compute weights. Note that SeisX will still work correctly even if some lines have only two seconds of data and other lines have three seconds. The approach we are going to take is to calculate the time shifts, phase rotation, and amplitude scalar in the first pass but only apply the time shifts. A second pass with a more detailed window will be used to do the final corrections.
Miss Tie values can be displayed as a color dot on the map. Change the map attributes (A button on the map screen) to Time Dot and make the Dot size a 1/10 of an inch. Set the Min and Max Dot values to AUTO. This will automatically compute the maximum, minimum range over the miss tie data. In this case the program has computed miss tie values from 14 to 292 milliseconds.
Display a color table on the screen with the map and select the following color scale. Note that the misties range from 0 to ~224 ms. Click on the color screen to select a different appropriate color table.
Interactive Mistie Checking
Rebuild the base map with the 3D attached. Check the computed ties at the line intersections as shown to the right. SeisX will compute the optimum time shift, phase correction and amplitude scaling to tie your data in a least squares sense. A single time shift, phase rotation and amplitude scaler will be applied to each seismic line to balance your data. Note that the Phase correction and Amplitude correction will be applied to the data but the time shift will be stored as a version number.shift text file in your project directory. It is mandatory that all seismic lines be tied with a weight of at least .001. This will require additional jump ties to those lines that are disconnected. Watch out for lines that are on top of each other. You may have to use (File) (Display) to locate lines that need to be jump tied. Change the time analysis window to 500 - 2600 milliseconds. Change the number of analysis traces to seven. Stretch out the Line Tie Analysis Display to mostly fill the screen. Notice the large time shift between lines 101 and 104 to the right.
Correct any bad ties by (Saving) the corrected seismic line tie. If you cannot tell what the correct tie is set the weight of the tie to .001 (the lowest valid weight) and set the shift to near 0. This will ensure that the least squares fit to the data will make minimum use of this tie.
Misstie 2D to 3D survey
It is generally a good idea to tie your 2D to 3D. While SeisX can tie seismic to well ties, it is not recommended. It turns out that if you try to tie a 3D survey to multiple wells you normally obtain multiple solutions.
3D is normally processed with a specific deconvolution wavelet. Hopefully this wavelet will tie all wells within the range of plus or minus 30 degrees. (If it is more than this then you have a problem!) It turns out that matching well ties is very dependent of the window selected and the filter selected. Bottom line is -- process your 3D in an optimum fashion and match all your 2D to this 3D. This is accomplished in SeisX by the 2D to 3D ties. Click on the 2D lines within the active 3D survey and manually adjust the phase, time shift and amplitude value. It is recommended to shift the phase before you adjust the time. Changing the phase will result in changing the time shift value! Don't worry too much about the phase shifts, we will correct for the time shifts first.
Further check the 2D ties on the map. Note the very large time shifts at the line intersections that need to be corrected. Line 104 and 107 have very large time shifts.
To the left you can see the obvious time shift between the two lines. The next chapter will deal with correcting these time shifts.
It is mandatory to tie your complete grid of seismic together. This will include adding jump ties for all seismic that are not connected (panel on the right)
It is a good idea to check out the line tie statistics. Change the sort field to time shift and examine the lines with the largest time shifts. Correct any obvious problems.
Automatic Grid Balancing
Select Mistie, Automatic Grid Balancing, and Calculate. If the program beeps, your seismic line grids do not tie. Examine the (Report) and connect the unconnected grids together.
View Grid Report below to the right to verify the amount of miss tie reduction has been generated. You should see a very percentage reduction ~96%
View the Residual Misties... report to see the amount of miss tie still left.
Notice that our Shifts here are all within a single time sample but the phase
errors are plus or minus 30 degrees. This is still very good. I do not think
you can see phase differences less that 30 degrees. 
Apply Only the Time Shifts
You can easily do this by renaming the computed static shift files statshift.MIG.2 to statshift.MIG.1
$ pwd D:/Projects/index/Seismic $ cd .. $ ls 101.STK.0.sgy SX200 palettes 101.hdr Seismic pre_scan.rep 101.segy.MIG.0.sgy StratSym31 project.history 102.MIG.0.sgy Synthetics segy 103.MIG.0.sgy TDCurves segy.fmt.lst 104.MIG.0.sgy TownShips.cult seisx.colormaps 105.MIG.0.sgy Wells spt.ttl 106.MIG.0.sgy __lock statshift.MIG.0 107.MIG.0.sgy asc statshift.MIG.1 Aliases culture.dat statshift.MIG.2 Devonian_Edges.cult grid.rpt statshift.STK.0 Horizons logs.default statshift.STK.1 Mistie map.pha statshift.STK.2 ProjectDefaultTD mapping2d.xdr template $ mv statshift.MIG.2 statshift.MIG.1 $ mv statshift.STK.2 statshift.STK.1
Now check the line intersection between lines 103 and 106. The seismic lines now seem to tie in time. The time shifts are now correct but the phase is out of whack (-130degrees)! If we want to correct the phase, we can now window over a much smaller interval, say 1000 to 2200ms.
It looks like we have a fairly good phase tie between 103 and our 3D. However,
we need to measure the actual values with Mistie, Interactive Line Tie
Analysis. Select a Tie Mode of 
, number of
traces to 9 and a window of 500 to 2500ms. Set the reference line to the 3D
survey. Adjust the Phase to produce a symetrical wavelet with the peak
upright. Zoom into the cross correlation by clicking on the left hand mouse
button holding the button down to select the other point. This 2D 3D tie is
displayed to the right. Note the symmetry of the cross correlation wavelet.
While the data looks okay, we measure a -36 degree phase shift between the 3D data set and out 2D lines. Here is what our Dialog box should look like
Click on a few more points, Saving the statistics for these 2D3D line ties Notice if you tie with line 107, the phase tie will be around -107degrees. We will check a few 2D ties with our 3D. You will notice that we obtain different values from different intervals in the same well and different values from different wells
Automatic Mistie Calculations -- Phase
Select (Mistie), (Automatic Mistie) calculations. Change the window to cover the target data (1000 to 2200ms). Use 9 traces for flat Alberta at line ties and compute weights. Note that SeisX will still work correctly even if some lines have only two seconds of data.
Examine the Mistie Statistics from this second run. You will notice that
the time shifts are all plus or minus one or two samples. The Phase
rotations required are not the same as in the first run.
The reason for this
is we have used a different window that covers the target horizons. We can
do this because we have corrected for the time shifts.
$ pwd D:/Projects/index/Seismic/process $ mkdir balance $ cd balance $ pwd D:/Projects/index/Seismic/process/balance $
BEFORE YOU CALCULATE, change output directory to "balance", Turn Use Line Groups to No, Select Reference line as 103 use a File Description String "Phase Balance", set process ID to 2 and press Calculate.
Finally, submit 
the batch job
to be processed. In
order to work with the phase corrected data, we neet to (Project), (Attach
SeisX'ed Files) the balanced lines (menu to the right).
Check out the magnitude of the time shifts. The maximum time shift to the left
(blue) is only 4ms!
We now have a very good time and character tie. It is important that time shifts and phase rotations be properly applied in order to interpret subtle stratigraphic plays. Be careful with phase changes. You might want to restrict phase rotations to groups of seismic lines that were acquired and processed at the same time.
Note: click on any screen picture and you will have the full size image. Press (Alt) and the left arrow and you will return to the document.
Go to the next, previous section, Table of Contents