09-23-2020, 03:05 PM
There is no concept of "lossless" in the process of interpolation. I've never seen the term used in that context. Can you provide link(s) referencing "lossless interpolation"?
Lossless transformations are 2-way where the original data is (1) transformed and (2) recovered without any data loss or change in the data. For example data file compression or lossless audio or image file compression.
Interpolation on the other hand is a 1-way transformation that changes the original data in an imperfect way. It consists of two processes.
1. Upsampling
Insert zero-valued samples between each pair of input samples. This is called zero-stuffing. It creates a new, higher rate signal whose frequency spectrum and bandwidth is the same as the original signal but with signal distortion consisting of images of the original signal centered on multiples of the original sampling rate.
2. Low-pass filtering the result
The goal of this process is to eliminate the distortion caused by upsampling.
The end result, assuming a perfect interpolation process, would be a signal at N times the rate of the original signal which has the same frequency spectrum and bandwidth as the original signal but without any of the distortion caused by updsampling.
There is no such thing as a perfect interpolation filter and so the result will always contain some amount of distortion. Whether or not the the distortion matters depends on the quality of the interpolator routine and how the resulting signal is being used downstream.
Lossless transformations are 2-way where the original data is (1) transformed and (2) recovered without any data loss or change in the data. For example data file compression or lossless audio or image file compression.
Interpolation on the other hand is a 1-way transformation that changes the original data in an imperfect way. It consists of two processes.
1. Upsampling
Insert zero-valued samples between each pair of input samples. This is called zero-stuffing. It creates a new, higher rate signal whose frequency spectrum and bandwidth is the same as the original signal but with signal distortion consisting of images of the original signal centered on multiples of the original sampling rate.
2. Low-pass filtering the result
The goal of this process is to eliminate the distortion caused by upsampling.
The end result, assuming a perfect interpolation process, would be a signal at N times the rate of the original signal which has the same frequency spectrum and bandwidth as the original signal but without any of the distortion caused by updsampling.
There is no such thing as a perfect interpolation filter and so the result will always contain some amount of distortion. Whether or not the the distortion matters depends on the quality of the interpolator routine and how the resulting signal is being used downstream.
