Archive for January, 2021

Seismic Data Compression

Tuesday, January 26th, 2021

An idea that has been in the back of my mind for many years is a method to compress a seismic record down to a size that it can realistically be shared by email or FTP. And I wanted to preserve all the header information, so did not want a converter, but a compressor.

The very nature of the data means that utilities like Zip and TAR do not achieve a great enough level of compression. My method can achieve quite high levels of compression (75% or more depending on parameters), while preserving headers. But it is not lossless.

For QC purposes though, it is certainly good enough, perhaps making remote QC viable. Under some conditions, it may even be usable for processing.

So far I have it working for most of the modern Sercel systems – SN408, 428, 508, SEAL, Unite, WTU and WTU3. Also for Innoseis and SmartSolo (but only for older data sets I have, if the formats have changed, I will have some more work to do). It will also work on SEGY, but typically a lower level of compression.

It is very easy to use. It consists of two applications – a Compressor which would be at the field end and a Decompressor for the Office end. Simply specify an input and an output folder and click start.

 

Here an example of a 98MB production shot compressed down to 8MB and then decompressed. (Only 1 line displayed of course)

Original File

Restored File after 91% compression

 

Here a log of some test data. The compression ratio varies depending on the data itself and the compression parameters specified. It can be optimised for each operation to achieve a suitable level of compression.

 

It is early days yet and needs a lot more testing under true production conditions. So it probably won’t go much further until I get some work.

SMLVL

Monday, January 18th, 2021

SMLVL is a ‘companion’ to SMUH3. It provides easy handling of short offset refraction as is (or was) commonly done to refine near surface velocity models.

This is one of my hobby horses I guess. I firmly believe that the near surface has been neglected in favour of fancy processing modules that do not deliver all that they promise. Near surface investigation is a cheap and valuable tool that should be widely used on all projects. (But that is just my opinion).

SMLVL, like SMUH3 has the ability to encapsulate data and results so that everything is available for future reference rather than looking for a bunch of files and documents that have been scattered to the four winds.

And like SMUH3 it is very configurable so that you can produce a display that you want instead of the rather antiseptic plots delivered by the usual contractor software.

 

 

Details here:

https://seismatters.com/SMLVL.html

 

 

SMUphole3

Saturday, January 9th, 2021

As a change from the scamdemic, some new software.

SMUH3 is now ready.

Upholes are often neglected these days. The powers that be seem to think fancy processing can compensate for not having a reliable near surface model.

Some companies still include upholes in their programs though, and we need a way to QC the results.

SMHU3 offers more than just QC though, it optionally encapsulates the data, the results and additional documents into a single file so that all information is available at a future date.

It is easy to use and quite configurable, so you get the displays you want rather than what someone else has decided looks good.

Details here:

https://seismatters.com/SMUH3.html