@TF3RDL could you stop endlessly spamming the thread? It' so tiring to open the thread for news and getting another reply to yourself by you. I mean, if you want to create a new thread with your suggestions, fine. But right now 90% of posts from this thread are yours, not adding anything to the discussion or bringing anything new to the component development (just your infinite requests). Other people have pointed you the same.

• Frequency bands mode (already being developed by @Crossover though I think it is not enough)
  • IIR filter bank mode: Alternative transform algorithm that doesn't rely on FFT algorithm and it has logarithmic frequency resolution
  • Adjustable number of bands per octave: I think the making BPO adjustable allow for 1/3rd octave bands analysis and 1/24th ones
• More frequency scaling options
  • Linear and Mel/Bark/ERB frequency scales in-addition to logarithmic: Frequency bands mode should also accept linear as well as perceptual/psychoacoustic scales but in a different way (this also apply to logarithmic scale but not following octave bands, so it counts as alternate log frequency bands)
  • DC (0Hz) and Nyquist (fs/2) options for lower and upper Hz ranges respectively: As if upcoming adjustable frequency range is not enough
• FFT (fast Fourier transform)
  • Non-power of two sizes like 576 samples: And also increase the max FFT size to something like 65536 samples
  • A feature to set FFT size in milliseconds instead of samples: The actual FFT size should automatically change depending on the samplerate, kinda like built-in spectrum but better
• Expanded time smoothing options
• An option to use linear and even squared amplitude scale for calculation of RMS spectrum and also, exponential and infinite (cumulative) averaging modes should be added too
• Infinite decay rate mode for all spectrums (especially the RMS spectrum)

• Spectrum analyzer and spectrogram using custom FFT: For spectrogram display and multiple channel visualization (including Mid/Side representations of course) as well as bars display on detailed FFT mode
• Frequency bands spectrum analyzer using either FFT or CQT: For what frequency bands mode (FFT mode) would look like though I would like sinc/Lanczos FFT bin interpolation over the other because it looks more like zero-padded FFT
• Analog-style spectrum analyzer and sliding DFT visualization using AudioWorklet: What would IIR filter bank mode for frequency bands mode on foo_enhanced_spectrum_analyzer look like, though this project also have SWIFT (approximate both Gammatone filter and asymmetric FFT windowing), and variable-Q mode as well

Thank you for starting your own thread.  This will not only allow you all the freedom you need to describe your own ideas, but will allow the developer of the plugin the breathing space to post his own own news and updates in his original thread, which have been sorely missing, without fear of being "bombarded."

BTW, the Interactive Mockups you linked to are interesting;  others might try the Load from Local Device option with a music file, hit play, and then play with the "Visualization Settings" to the right of the web spectrum.

One of my own "interactive mockups" has been updated: added infinite average spectrum (previously, this mockup of one of my many foo_enhanced_spectrum_analyzer feature requests were private, and now that aspect has been made into the public now as I can now update this project and others)

It will look something like this: (not so obvious when the screenshot is a static image rather than a video or GIF of it)

@Crossover, any chances of sneak preview videos (or even a YouTube video or series of it) of what you've been working on the features like IIR filter bank and infinite averaging mode I've suggested for this component?

Calibration line "improvement": Use either linear or squared domain instead of logarithmic (dB) for calculating the calibration line based on average of all FFT bins, in-addition to a feature to use actual RMS (FIFO averaging done in squared domain) or linear averaging mode for calculating the RMS spectrum

BTW, here's the video of a mockup of new linear averaging (one of three averaging domains that I've wanted to see in the new version of @Crossover's awesome audio analyzer component) and the return of crest spectrum (that was removed in foo_enhanced_spectrum_analyzer 1.9.2.0) for the upcoming Enhanced Spectrum analyzer 2.0.0.0 update:
https://www.youtube.com/watch?v=BOcHKE7VMsk

Here's the spectral descriptor lines (horizontal = Hz, vertical = dB) could possibly look like in foo_enhanced_spectrum_analyzer:

Note that monitoring fields for these spectral descriptors (e.g. spectral centroid) are obviously going to be different from this image above and it would look more like ones from foo_loudness_peakmeter

• Spectral descriptors: This is where monitoring fields are could be introduced into Enhanced Spectrum analyzer component displaying spectral features as the song plays
• Spectrum visualization with average and crest spectrum: Re-introduction of the crest spectrum from older version of foo_enhanced_spectrum_analyzer but this time it updates at higher refresh rate and are highly-detailed and also, the averaging domain can now be selected to linear, squared, and logarithmic

const c2 = 4284900, // from 2070
      z2 = 1690000, // from 1300
      f = 80,
      f2 = x ** 2,
      s = Math.sqrt((1 + (f2) / z2) / (1 + (f2) / c2)), // shelving part
      l = ((x/f) ** 2)/Math.sqrt(0.5*(1 - (x/f) ** 2) ** 2 + (x/f) ** 2) / Math.SQRT2; // lowpass part

return s*l;

• Intensity (Horizontal B in foo_musical_spectrum): Color gradient index for each band/bin are based on the current amplitude (scaled according to graph amplitude scale settings)
• dB (no equivalent setting in foo_musical_spectrum): Similar to Vertical A gradient from foo_musical_spectrum, but it is based on decibels in terms of colorstop positions
• Notes (no equivalent setting in foo_musical_spectrum): Coloring is based on band's center frequency, repeating every octave
• Scaled (Vertical C in foo_musical_spectrum): Gradient's start and end positions for each band is based on current amplitude (from the bottom to the top point, which depends on each band's current amplitude)
• Shifted (Vertical B in foo_musical_spectrum): Similar to "Scaled" gradient mode, but instead of colorstop locations being "scaled", it simply shifts differently per-band depending on current amplitude; the gradient start position is shifted depending on the amplitude as opposed to being at the bottom

• Frequency tracking: Solid color, where the actual color is gathered from the gradient at specified position defined to spectral feature's current value in Hz (including spectral peaks, not just spectral centroid, rolloff, or even spectral spread)
• Frequency indices: Horizontal gradient where colorstop position is derived from band's center frequencies rather than screen position
• Spectral features: Displays spectral slope, kurtosis, flatness, and skewness by changing colors

BTW, here's a mockup of phase correlation (bottom) and stereo balance (top) spectrum visualization:

@Larryg49

I find every mention of the fake and non-existing 2.0.0.0 and accompanying mockup pics in this thread annoying at best.

Like Sveakul said, use panels/buttons to display the analyzers that are available. You don't have to choose for one or another. They all have their merits.

Quote from: TF3RDL on 2025-02-16 18:28:37

I had to agree with you and admit that posting mockups and links to Fanon Wiki articles a lot on this thread becomes an annoyance for users and even the developer,
After I've realizing it becomes an annoyance, 

So, since you've come to this realization, it's time for you to either,
1.   Give us a working component incorporating all the wonderful features you describe, OR
 2.   Go away -- please ! 

3. instead of fb2k use mpv with showcwt filter

• Linear and squared averaging domains to beat and vibe harder than AudioSurf's [as-steep] tag and get over Erio - Error 2.0, Suno/Udio slop, and Brazilian Funk in terms of the vibe
• IIR filter bank algorithm to make Enhanced Spectrum analyzer hit and punch harder than AudioSurf 1 Ninja Mono and Audiosurf 2 Wakeboard mode thing in terms of vibe of trap, dubstep, psytrance and bass boosted songs
• Spectrogram and waterfall display (combined spectrum analyzer and spectrogram) modes to win over freeride and visualizer modes from AudioSurf in terms of vibing
• Exponential averaging mode to make foo_enhanced_spectrum_analyzer vibe harder than [as-caterp]+[as-wb100]
• Separate channels display and Mid/Side modes to make it vibe better than Double Vision characters and [as-everybodymono], [as-nogrey] and [as-allgrey] respectively
• Stereo spectrum analyzer and correlation by frequency display to vibe better and harder than [as-bankcam], [as-swind], and Pusher
• Infinite decay rate mode to make it punchier than Audiosurf 2 in terms of vibing
• Reassigned FFT to vibe harder than [as-allgrey]+[as-caterp] on hardstyle with reassigned spectrogram on Enhanced Spectrum analyzer
• Spectral centroid and other spectral features display and monitoring fields to make FESA2/ESA2 vibe and rock better than with Audiosphere mode from Audiosurf 2, even better with spectrogram display
• Make calibration line calculation source selectable to make FESA/ESA vibe better to songs than Audiosurf especially when peak as a "source" is selected

So, since you've come to this realization, it's time for you to either,
1.   Give us a working component incorporating all the wonderful features you describe, OR
 2.   Go away -- please ! 

3. instead of fb2k use mpv with showcwt filter

4. Get AudioSurf and Audiosurf 2 to feel the vibe while letting the @Crossover and not rush the upcoming big update

• LUFS bars to reel like FabFilter Pro-L2 and vibe better than watching camera zoom and track colors change in AudioSurf
• Peak/RMS meter and clipping indicators on the graph to make foo_loudness_peakmeter vibe more than tunnels in AudioSurf
• Pan meter and phase correlation metering to make Loudness Peakmeter rock better and harder than [as-bankcam] and [as-swind] tags respectively
• Option to hide graphs to make smaller compact forms of FLPM2/LPM2 an extra oomph vs. Audiosurf and Suno/Udio AI Slop content on music streaming services and Brazilian Funk/Phonk

Whatever happened to your declaration of posting your own requests/comments to your own personal thread so we don't have to look at them here?  All you are doing here is making sure that no one else, including yourself, will ever see a new version of foo_enhanced_spectrum_analyzer (or loudness_peakmeter) again.

Moderators:  surely the deliberate, graffiti-like disruption of a component thread is not tolerated here?