prepare v1.4 including README and doc changes

.github/workflows/rust.yml

@@ -27,20 +27,15 @@ jobs:
       - name: Install required Linux packages for "audio-visualizer"/cpal/minifb
         run: sudo apt update && sudo apt -y install libasound2-dev libxkbcommon-dev
 
-      # build with all features/fft implementations
       - run: cargo build --all-targets
-
-      # run tests with all features/fft implementations
       - run: cargo test --all-targets
 
       # run benchmark: right now, there is no reporting or so from the results
       - run: cargo bench
 
-      # test `no_std`-build with all features/fft implementations
       - run: rustup target add thumbv7em-none-eabihf
       - run: cargo build --target thumbv7em-none-eabihf
 
-      # run examples with all features/fft implementations
       - run: cargo run --release --example mp3-samples
 
   style_checks:

CHANGELOG.md

@@ -3,7 +3,9 @@
 # 1.4.0 (2023-03-04)
 - dropped all optional FFT features (`microfft-complex`, `microfft-real`,
   `rustfft-complex`) and made `microfft::real` the default FFT implementation.
-  This is breaking but only for a small percentage of users.
+  This is breaking but only for a small percentage of users. There was no
+  benefit from that feature and `microfft::real`, the default, was already the
+  fastest for no_std as well as std targets.
 - dependency updates
 
 ## 1.3.0 (2023-03-04)

Cargo.toml

@@ -4,7 +4,7 @@ description = """
 An easy to use and fast `no_std` library (with `alloc`) to get the frequency
 spectrum of a digital signal (e.g. audio) using FFT.
 """
-version = "1.3.0"
+version = "1.4.0"
 authors = ["Philipp Schuster <[email protected]>"]
 edition = "2021"
 keywords = ["fft", "spectrum", "frequencies", "audio", "dsp"]

README.md

@@ -7,15 +7,10 @@ The **MSRV** (minimum supported Rust version) is `1.61.0`.
 ## I want to understand how FFT can be used to get a spectrum
 Please see file [/EDUCATIONAL.md](/EDUCATIONAL.md).
 
-## How to use (including `no_std`-environments)
+## How to use (including `no_std`-contexts)
 Most tips and comments are located inside the code, so please check out the
 repository on GitHub! Anyway, the most basic usage looks like this:
 
-### FFT implementation as compile time configuration via Cargo features
-By default, this crate uses the `real`-module from the great `microfft`-crate.
-It's the fastest implementation and as of version `v0.5.0` there should be no
-valid reason why you should ever change this. The multiple features are there
-mainly for educational reasons and to support me during programming/testing.
 
 ### Cargo.toml
 ```toml
@@ -59,7 +54,8 @@ fn main() {
 *Measurements taken on i7-1165G7 @ 2.80GHz (Single-threaded) with optimized build*
 
 I've tested multiple FFT implementations. Below you can find out why I decided
-to use `microfft`.
+to use `microfft::real`. It is not only the fastest, but also works in `no_std`
+contexts.
 
 | Operation                                               | Time   |
 |---------------------------------------------------------| ------:|

src/error.rs

@@ -41,8 +41,7 @@ pub enum SpectrumAnalyzerError {
     InfinityValuesNotSupported,
     /// See [`crate::limit::FrequencyLimitError`].
     InvalidFrequencyLimit(FrequencyLimitError),
-    /// The number of samples must be a power of two in order for the FFT
-    /// to be fast and, depending on the used FFT implementation, to work at all.
+    /// The number of samples must be a power of two in order for the FFT.
     SamplesLengthNotAPowerOfTwo,
     /// After applying the scaling function on a specific item, the returned value is either
     /// infinity or NaN, according to IEEE-754. This is invalid. Check