Skip to main content

Get more from the data you already have collected.

Learn how you can save time and with our updated and interactive filtering in Oasis montaj. We will show you a redesigned 1D and 2D filtering workflow performing industry-standard filters and transforms.



Geoff Plastow
Senior Geophysicist at Seequent



19 min

See more on demand videos


Find out more about Seequent's mining solution

Learn more

Video Transcript

[00:00:00.618](ethereal music)

[00:00:10.120]<v ->Hello, good morning.</v>

[00:00:10.953]Good afternoon to everyone in attendance.

[00:00:14.020]Today, we’re going to be looking at redesigned filtering

[00:00:16.630]workflows that we’ve introduced into the latest version

[00:00:19.460]of Oasis montaj version 9.9.

[00:00:24.010]We’re going to be working with some real world data,

[00:00:26.280]and tackling some real world problems today.

[00:00:28.820]So I hope you enjoy today’s demonstration.

[00:00:31.580]So just a quick introduction for myself.

[00:00:35.160]My name is Geoff Plastow,

[00:00:36.420]and I’m a Senior Geophysicist here at Seequentt,

[00:00:39.110]and I’m currently based in Vancouver, British Columbia.

[00:00:43.060]As you can imagine from the title of this webinar,

[00:00:46.330]we have redesigned some of the key filtering workflows

[00:00:49.480]inside of Oasis montaj.

[00:00:51.800]This updated filtering process is now much more visual,

[00:00:55.610]much more interactive and much more intuitive.

[00:00:59.250]In turn, this is going to make you more efficient

[00:01:01.410]with your time and takes a lot of the guesswork,

[00:01:04.020]and trial and error involved in the filtering process.

[00:01:07.710]So now you could focus on understanding,

[00:01:10.100]and interpreting the data.

[00:01:13.320]So before we get too far into today’s conversation,

[00:01:16.800]perhaps I can mention why we might even want to filter

[00:01:19.500]our data to begin with.

[00:01:21.500]For those attending,

[00:01:22.810]feel free to use the chat panel,

[00:01:24.500]to type in some other reasons why you might want to filter

[00:01:27.260]your own data.

[00:01:29.580]Some of the most common reasons why we might want to filter

[00:01:32.140]our data is to accentuate or to highlight

[00:01:35.020]part of a geophysical or geological response,

[00:01:38.350]or to perhaps suppress something that we see in the data.

[00:01:42.060]For example,

[00:01:43.200]we may want to suppress systematic noise or random noise,

[00:01:47.590]and accentuate subtle geological and geophysical anomalies.

[00:01:52.280]I’m going to jump into Oasis montaj now.

[00:02:03.680]In Oasis montaj,

[00:02:05.210]part of the filtering process allows us to perform

[00:02:07.830]transforms on our data as well.

[00:02:10.410]We can transform a very simple magnetic datasets

[00:02:13.570]like we’re looking at now

[00:02:15.454]into something more interpretable,

[00:02:17.410]and perhaps a higher order deliverable.

[00:02:19.900]This is what we’re going to do today.

[00:02:22.420]So let’s have some fun and let’s get to work on an airborne

[00:02:25.150]magnetic datasets.

[00:02:28.100]So here we are in the latest version of Oasis montaj

[00:02:31.110]released in November of 2020.

[00:02:33.910]We’re going to start with the Geosoft database

[00:02:36.120]containing some airborne magnetic data.

[00:02:39.180]So on the right side of the screen,

[00:02:40.690]we see our Geosoft database, it’s a very simple database.

[00:02:44.430]We have our location information, Eastern and Northern.

[00:02:47.820]We have our radar altimeter,

[00:02:49.320]which is the sensor height above the surface of the earth.

[00:02:52.340]And we have our final magnetic data.

[00:02:55.110]And here we see a profile of our magnetic data in red.

[00:02:59.690]On the left-hand side of the screen,

[00:03:02.010]we see a map containing the same data gridded.

[00:03:07.220]The areas in red or pink represent areas of a high magnetic

[00:03:10.760]intensity and the areas that are cool or blue in color

[00:03:14.190]represent areas of low magnetic intensity.

[00:03:18.980]Now all geophysical projects have their challenges.

[00:03:22.890]And of course, this one is no exception.

[00:03:26.670]Below this image,

[00:03:27.870]we’re looking at some satellite imagery

[00:03:29.760]over the survey area,

[00:03:31.370]and we can see all of these specs and dots

[00:03:34.580]over the survey area, and they represent culture.

[00:03:37.860]They represent drill pads,

[00:03:39.360]they represent drill casings and shacks,

[00:03:41.770]and other metallic objects.

[00:03:45.300]And our airborne geophysical survey has picked up

[00:03:48.130]all of those items and we can see them in our dataset

[00:03:51.430]as these very small magnetic anomalies.

[00:03:54.950]And our first challenge that we’re going to tackle today

[00:03:57.480]is we’re going to try to separate or remove these small

[00:04:00.810]magnetic features from our magnetic dataset.

[00:04:04.230]So we’re going to use the 1D FFT filtering capability

[00:04:08.290]in Oasis montaj to do that.

[00:04:11.400]So I’m going to click 1D FFT and 1D filtering.

[00:04:17.410]So right off the bat where we can see our redesign

[00:04:20.460]1D filtering menu, this is completely resizable,

[00:04:24.800]and I’m going to make it full screen

[00:04:26.150]so everyone can see it clearly.

[00:04:28.860]The 1D FFT filtering is divided into three panels.

[00:04:33.370]At the top, we have our power spectrum,

[00:04:36.460]which we can look at wave length or wave number.

[00:04:40.320]And as a quick reminder,

[00:04:41.760]the power spectrum is a 48 transform of our magnetic data.

[00:04:46.470]And we’re looking at it in the energy space.

[00:04:49.150]So we’re looking at wave length or wave number

[00:04:52.280]on the horizontal scale.

[00:04:53.640]And we’re looking at energy on the top scale,

[00:04:56.790]on the vertical scale.

[00:04:59.110]The center panel gives me a nice preview

[00:05:02.400]of our magnetic data.

[00:05:05.400]And at the bottom here, we have our filter design panel.

[00:05:11.000]So I’m going to select the database channel

[00:05:13.350]that we want to filter.

[00:05:14.410]We’re going to filter our magnetic data,

[00:05:16.820]and we’re going to provide an output channel.

[00:05:19.400]I’m going to call this our low pass filter data.

[00:05:23.230]And here I can design the first filter that we can apply.

[00:05:26.950]I’m going to apply a low pass filter,

[00:05:29.920]and here I can adjust the filter design properties.

[00:05:34.600]And these are in our ground units where it’s around meters.

[00:05:37.750]So I can apply,

[00:05:38.670]let’s say a 200 meter filter and right off the bat,

[00:05:43.280]we can see our power spectrum window updates.

[00:05:47.610]In blue, we see our filter response.

[00:05:50.950]We can see our original data here in the frequency domain

[00:05:54.880]of our original spectrum.

[00:05:56.730]And we see the filtered spectrum as well.

[00:05:59.750]We can also see a preview of the filtered response.

[00:06:04.020]I’m just going to make it a little bit larger.

[00:06:07.520]So you can see here that we’ve made a little dent

[00:06:09.780]in these cultural anomalies, but not much.

[00:06:12.960]So we can simply just increase this number automatically.

[00:06:17.640]And I hit enter and the filter updates,

[00:06:20.970]the filter of response updates in real time,

[00:06:23.350]and we see a preview of our filtered response right away.

[00:06:27.780]So this takes a lot of guesswork out of designing a filter.

[00:06:31.080]We can see it in the frequency domain.

[00:06:32.950]We can see a live preview of our data,

[00:06:35.620]and we can automatically adjust and fine tune things.

[00:06:38.870]That’s how we find the ideal filter settings.

[00:06:41.800]Once we’re ready, we can click okay.

[00:06:43.980]And we can apply this filter to our database.

[00:06:47.670]I’m just going to display the low pass filter data

[00:06:53.370]on top of our original dataset,

[00:06:56.120]so we can examine it a little bit more clearly.

[00:06:59.240]So we can see,

[00:07:00.073]we did a pretty good job of removing these small cultural

[00:07:03.480]anomalies, at the same time,

[00:07:06.060]doing our best to maintain the geologic response

[00:07:08.950]of the project area.

[00:07:11.600]The next thing I’m going to do is I’m just going to create

[00:07:13.560]a quick grid of our low pass filter data.

[00:07:20.210]So now we have an updated image of our culture removed

[00:07:24.880]magnetic data.

[00:07:27.150]So let’s go ahead and check out some of the 2D filtering

[00:07:30.400]capabilities inside of Oasis montaj.

[00:07:34.740]As you may have noticed,

[00:07:35.800]we’ve renamed our mag map extension 2D filtering.

[00:07:40.080]We’ve done this to address the more broadly applicable

[00:07:43.160]2D filtering capabilities of the extension.

[00:07:47.070]So let’s go ahead and start filtering our magnetic data.

[00:07:54.290]Our input here is going to be our low-pass grid

[00:07:57.380]that we just created, and we’re going,

[00:08:00.100]the first filter we’re going to apply is going to be a first

[00:08:03.130]vertical derivative.

[00:08:05.250]This is a very common geophysical filter that is applied

[00:08:08.430]to accentuate near surface information,

[00:08:11.800]and to accentuate lineaments contacts and faults,

[00:08:14.607]and the dataset.

[00:08:17.160]I’m going to provide the name for the output grid,

[00:08:20.160]and a name for our filter design file.

[00:08:23.040]Once we’re ready, I can click, edit or create a filter.

[00:08:31.720]We’re met with a new redesigned 2D filtering window.

[00:08:36.520]Again, this is completely customizable and can be resized.

[00:08:40.500]I’m going to go ahead and make it full screen.

[00:08:44.120]Inside of this, we can resize and rescale our windows.

[00:08:49.150]Let’s just talk a bit about what we’re looking at.

[00:08:52.670]In the top right-hand corner,

[00:08:54.220]we can see a grid of our magnetic data.

[00:08:56.820]We can see our coordinates and we can also see the value

[00:08:59.640]of our data.

[00:09:00.720]So here, we’re looking at nanoteslas.

[00:09:04.490]On the left-hand side here,

[00:09:05.950]again, we see a power spectrum of our input data.

[00:09:09.710]So again,

[00:09:10.970]this is the power spectrum or the Fourier transform

[00:09:14.470]of our input survey data.

[00:09:16.160]And we’re looking at it in the frequency domain.

[00:09:19.270]Like we can switch between wave length and wave number,

[00:09:22.060]if you’re comfortable in working on either of those


[00:09:26.460]We can see our filter design here in blue.

[00:09:29.770]We can see our original input power spectrum in black,

[00:09:34.180]and the filtered spectrum here in red.

[00:09:38.370]If we want, we can resize this window,

[00:09:41.100]if we want to look at a specific part of the power spectrum.

[00:09:46.380]And again, we can resize these as we need.

[00:09:50.010]What’s great is we have a huge option,

[00:09:52.500]a huge list of filters that we can use to accentuate,

[00:09:55.940]and transform our data.

[00:09:57.770]The first filter I’m using is a regional residual separation


[00:10:02.870]Here, we can assign the cutoff wavelength.

[00:10:06.130]If we want, we can just simply just click and scroll,

[00:10:09.180]and you’ll see that the filter is automatically updated

[00:10:12.120]in the power spectrum.

[00:10:13.640]And the result,

[00:10:15.200]the preview of the filter is updated in the bottom left-hand


[00:10:20.010]This is so useful,

[00:10:21.550]and it takes out so much trial and error

[00:10:24.140]out of our filtering process.

[00:10:26.530]So we can simply adjust the filtering parameters

[00:10:29.450]to accentuate the geophysical or geological features

[00:10:33.280]that we’re interested in.

[00:10:35.180]We can quickly switch between low pass, high pass,

[00:10:38.660]and we can also stack our filters.

[00:10:40.990]We can apply multiple filters at once, which is very common.

[00:10:45.500]So I’m first performing a regional filter on our dataset,

[00:10:49.510]and now I’m performing a derivative

[00:10:51.540]in the vertical set direction.

[00:10:54.000]If I wanted to see what the other directions look like,

[00:10:56.180]I can simply flip between them and I can get an automatic

[00:11:00.510]preview of what those results look like.

[00:11:04.720]What’s nice is if I’m stacking my filters,

[00:11:07.270]I can see the previewed result of all of them combined,

[00:11:10.920]or I can see the individual results of the regional filter

[00:11:14.570]or just the derivative filter as they’re applied.

[00:11:18.240]Once I’m happy with the design of my filter,

[00:11:21.210]I can click, okay.

[00:11:23.620]And then, okay to create my filtered 1VD grid.

[00:11:29.990]And you can see from this result,

[00:11:31.430]we’ve accentuated some of the faults in this area,

[00:11:34.780]as well as some of the intrusive bodies

[00:11:37.000]that are sticking out.

[00:11:39.320]So let’s jump back in and run another filter.

[00:11:44.360]The next filter I’m going to run

[00:11:46.200]is a reduce to pole transform.

[00:11:50.720]So I’m just going to rename my grid here,

[00:11:54.520]and I’m going to edit my filter file.

[00:12:01.200]And again, we’re going to jump into our 2D filtering.

[00:12:05.200]So in case you didn’t know,

[00:12:06.690]reduction-to-pole or RTP is a transform,

[00:12:10.500]which removes the dependence of magnetic inclination

[00:12:13.230]from our magnetic data.

[00:12:16.280]RTP or reduction-to-pole is a transform that aims

[00:12:19.560]to remove any assymetry or distortion

[00:12:22.630]caused by magnetic inclination.

[00:12:25.230]This more or less results in anomalies

[00:12:28.250]above their magnetic sources.

[00:12:30.720]This is a pretty common transformant magnetic processing,

[00:12:34.120]and really eases some of the processes involved

[00:12:37.610]in interpretation.

[00:12:40.060]So I can go ahead and again, make this a full screen.

[00:12:44.400]We can see our power spectrum here on the left-hand side.

[00:12:47.420]We can see our original and our output grid.

[00:12:50.690]So to do an RTP transform,

[00:12:52.870]I can simply select this from our dropdown list.

[00:12:57.430]What’s great is all I have to do is enter

[00:12:59.460]in the survey date.

[00:13:01.870]So this is the date that the data was acquired.

[00:13:04.180]I can click calculate,

[00:13:05.750]and it will automatically calculate the inclination,

[00:13:08.990]and declination for this project area.

[00:13:11.820]And I can click, okay.

[00:13:13.580]And I can perform our reduction-to-pole

[00:13:16.230]on our magnetic data.

[00:13:18.810]So you can see that the 2D filtering,

[00:13:21.290]or redesigned 2D filtering extension really,

[00:13:25.620]really assist interpretation by allowing us to preview

[00:13:28.890]the result before re-filtering.

[00:13:30.630]And it takes a lot of the guesswork out of the process.

[00:13:35.070]What I’d like to show you next is our new matched filtering

[00:13:40.180]inside of our 2D filtering extension.

[00:13:44.580]So what is matched filtering?

[00:13:46.750]Matched filtering is a separation filter,

[00:13:49.680]or layer filtering process,

[00:13:51.810]which is a new addition to the latest version

[00:13:53.870]of Oasis montaj.

[00:13:56.100]Matched filtering gives us our first glimpse,

[00:13:59.300]or insight as to how our magnetic structure changes

[00:14:02.790]from near surface or short wavelength, to depth,

[00:14:06.860]long wavelength.

[00:14:08.450]It provides us some quick insights into the relationship

[00:14:11.430]between wavelength and equivalent source depth

[00:14:14.430]in our magnetic data.

[00:14:16.610]The aim of matched filtering is an attempt

[00:14:18.590]to isolate magnetic sources to an ideal layer

[00:14:22.050]at an equivalent depth.

[00:14:24.620]So in the matched filtering menu,

[00:14:27.220]we can see a similar window that we saw

[00:14:29.490]in our 2D filtering extension.

[00:14:31.630]We have our input magnetic data,

[00:14:33.600]which is the RTP grid that we just created.

[00:14:37.220]And we can see our power spectrum on the left-hand side.

[00:14:41.640]Again, we have the black line,

[00:14:43.820]which represents the original survey data,

[00:14:48.140]and we have our long wavelengths at left-hand side

[00:14:53.100]of the screen.

[00:14:53.933]This represents the deeper structure or deeper sources

[00:14:57.030]in our dataset.

[00:14:58.800]And on the right-hand side,

[00:15:00.040]we have our high-frequency content,

[00:15:02.540]or our near surface structure.

[00:15:05.410]And we can see a number of filters that have been applied

[00:15:09.580]to our data set.

[00:15:11.290]So I’m just going to zoom in to part of the spectrum.

[00:15:13.770]So we can little bit see our filters

[00:15:16.640]a little bit more clearly.

[00:15:18.690]So the first filter that we apply is a low pass filter.

[00:15:23.280]The last filter that we apply here is a high pass filter.

[00:15:27.570]And then for each of the equivalent depths

[00:15:30.200]that we are going to try to extract from our dataset,

[00:15:33.460]we provide a tapered band pass filter,

[00:15:37.100]so we can extract up to four equivalent depth.

[00:15:40.620]In this case,

[00:15:41.453]I have three and we can see the preview of them

[00:15:45.110]on the bottom right-hand corner of the screen.

[00:15:47.670]I can preview equivalent source depths at five kilometers.

[00:15:52.850]Here we are at two and a half kilometers,

[00:15:55.800]and then 1000 meters.

[00:15:58.680]And you’ll see, as I move here,

[00:16:01.060]the updated spectrum that we’re looking at is updated

[00:16:04.150]in the power spectrum window.

[00:16:09.060]I’m just going to resize this.

[00:16:11.600]If we want,

[00:16:12.433]we can interactively select an equivalent source depth

[00:16:15.580]from our window.

[00:16:16.800]I can click on this edit button and I can pick the depth

[00:16:20.010]on the spectrum.

[00:16:21.540]And it’s going to actually provide me some information

[00:16:25.480]about what depth we were going to pick.

[00:16:28.350]So if I wanted to extract something at around

[00:16:31.990]the 700 meter equivalent source depth layer,

[00:16:37.270]I can go ahead and just click that,

[00:16:39.410]and it’ll automatically update itself here.

[00:16:42.610]Once I’m ready to perform the separation filter,

[00:16:45.170]I can click, okay.

[00:16:46.880]And it will extract my matched grids.

[00:16:56.110]So we have our four filtered grids.

[00:16:58.270]Our first grade represents all of our high-frequency

[00:17:01.680]near surface structure.

[00:17:02.980]You can see it’s really extracting these near surface

[00:17:05.380]liniments and faults.

[00:17:08.240]Again, now here’s our one kilometer or sorry,

[00:17:10.930]our about 700 meter equivalent layer,

[00:17:15.200]our two and a half kilometer equivalent layer,

[00:17:18.020]and our five kilometer layer,

[00:17:19.670]which is really looking at our, some of our long wavelength,

[00:17:22.720]deep seated structures.

[00:17:26.350]So matched filtering makes a useful contribution,

[00:17:30.420]and enhancements to our magnetic data.

[00:17:32.617]And this example,

[00:17:33.810]it allows us to approximate and separate deep seated

[00:17:36.960]structure from near surface faults

[00:17:39.530]in our magnetic dataset.

[00:17:41.840]Matched filtering is incredibly useful in Oasis montaj,

[00:17:46.070]and part of our new addition to our 2D filtering toolkit.

[00:17:50.100]It provides fast qualitative understanding

[00:17:52.790]of the equivalent layers and the equivalent wavelengths.

[00:17:56.240]It provides us insights into the separation of deep seated

[00:18:00.160]magnetic structure and near surface magnetic responses.

[00:18:04.860]In all of our redesigned filtering workflows,

[00:18:07.710]they’ve become much more visual and intuitive to use.

[00:18:11.410]Instead of a lot of trial and error or blindly filtering

[00:18:14.040]your data,

[00:18:14.873]we can now see automatic previews of our filtered data.

[00:18:17.870]This is going to save us time and effort,

[00:18:19.830]and allows us to focus on bigger problems related

[00:18:22.500]to our exploration projects.

[00:18:28.270]So this concludes my demonstration this morning.

[00:18:31.620]Thank you very much for your time.

[00:18:34.069](ethereal music)