Topic:
How amplitude trace scaling works in GeoFrame.

Description:
        

Solution:
When converting data from SEGY to IESX 8-bit, a scale factor is determined to
ensure that the magnitude of the data will not exceed 127., the largest value
that can be represented with 8 bits. The data is then scaled down and the scale 
factor, that was applied, is stored in each IESX trace header. 

When retrieving this data back to 32-bit SEGY the inverse scale factor is 
applied so that the original magnitude of the data is retrieved. The problem is
that those numbers that are smaller than the scale factor were rounded off to 
the nearest integer when scaling was applied. This data is irretrievably lost. 
Scaling back to the original magnitude there is a quantization equal to the 
scale factor.

Therefore it is not possible to get 100% of the values back to 32-bit SEGY
except for the quantification mentioned above. 
For example, where the largest number in the data is 12700.0 in order to fit it 
  in 8 bits it must be scaled  down to integer 127. 
  In this case the scale factor would be 0.01. 
  A number like 12649.0 would become 126.49, or integer 126. 
  If this is scaled back up to 32 bit a value of 12600, not 12649.0, 
  is returned therefore losing 49. 
  The quantization is 100, and you lose up to one half of that.

When converting data from SEGY to IESX 16-bit, a scale factor is determined to 
ensure that the magnitude of the data will not exceed 32767. 
    (32767 is the largest value that can be represented with 16 bits.) 
Again the data is scaled down and the scale factor, that was applied, 
is stored in the IESX trace header the same way as above for 8-bit data. 



Last Modified on: 28-NOV-00