Free Open-Source Audio Analyzer
TELESCOPE is a free and open-source audio analysis plugin from OVNI Audio that combines thirteen different analysis views inside a single instrument.
But the most interesting part is not simply the number of meters.
Its thirteenth module, VERDICT, takes the measurements collected by the analysis engine and converts them into readable findings. Every statement includes both the numerical measurement that triggered it and the identification of the rule responsible.
There is no AI model deciding whether your mix is good or bad.
There is no cloud analysis.
There is no telemetry.
Instead, TELESCOPE uses deterministic measurements and a fixed set of rules, so the same audio produces the same analysis every time.
OVNI Audio describes the philosophy simply:
Measurement, not taste.
TELESCOPE is available completely free for Windows and macOS in VST3 format, with Audio Unit also available on macOS.
Thirteen Lenses, One Analysis Engine
TELESCOPE organizes its analysis tools into thirteen modules, or “lenses”:
- Loudness
- Dynamics
- Spectrum
- Spectrogram
- Waterfall
- CQT
- Spiral
- Scope
- Band Correlation
- Stereo Spectrogram
- Field
- Tonal Balance
- Verdict
Each lens presents the same audio from a different analytical perspective.
The idea is not to overwhelm the user with dozens of unrelated meters, but to provide different ways of examining loudness, dynamics, frequency, pitch, stereo information and tonal balance from one consistent analysis engine.
LOUDNESS
The Loudness lens provides the measurements expected from a modern professional loudness meter.
It includes:
- Integrated LUFS
- Short-Term LUFS
- Momentary LUFS
- Loudness Range
- True Peak
- Per-channel measurements
The loudness implementation follows ITU-R BS.1770 gating.
Loudness Range uses the EBU Tech 3342 specification, while true-peak measurement follows BS.1770 Annex 2.
This makes TELESCOPE useful not only for music production but also for broadcast-oriented loudness checking.
Loudness History
TELESCOPE does not simply display the current loudness value.
It maintains a history that allows you to see how loudness evolves throughout a track.
This is particularly useful when mastering.
A single integrated LUFS number may tell you the average loudness of a complete song, but it does not reveal where a chorus becomes excessively dense or where a breakdown becomes unusually quiet.
The history view makes those changes visible.
DYNAMICS
The Dynamics lens focuses on how much dynamic space remains in the signal.
Measurements include:
- PSR
- PLR
- Short-Term loudness distribution
- Clip events
PSR — Peak to Short-Term Loudness Ratio is calculated over the most recent three seconds.
PLR — Peak to Loudness Ratio is measured from the last reset.
These values can provide useful information about how aggressively a track has been compressed or limited.
61-Bin Loudness Histogram
Dynamics also includes a 61-bin Short-Term histogram.
Instead of reducing the entire track to one dynamic value, the histogram shows how frequently different loudness levels occur.
This can reveal whether a track spends most of its time within a very narrow loudness range or moves through a much wider dynamic spectrum.
Ten-Minute Clip Timeline
Clip events are displayed across a timeline covering up to approximately ten minutes.
This makes it easier to locate clipping rather than simply receiving a warning that clipping occurred somewhere in the track.
SPECTRUM
TELESCOPE includes a detailed FFT spectrum analyzer.
FFT sizes range from:
1,024 to 32,768 points
Three window functions are available:
- Hann
- Blackman-Harris
- Kaiser
The spectrum can also be displayed using different frequency organizations.
These include:
- Free FFT
- ⅓-octave bands
- Bark bands
The ⅓-octave implementation follows ISO 266 spacing.
Accurate dBFS Calibration
Every spectral bin is displayed in dBFS relative to a full-scale sine.
TELESCOPE also applies coherent-gain correction according to the selected FFT window.
That level of detail matters because FFT windows alter measured amplitude.
Without compensation, switching window types could make the apparent level of a frequency component change even when the source signal remained identical.
SPECTROGRAM
The Spectrogram lens displays frequency energy over time.
It uses:
512 logarithmically spaced rows from 20 Hz to 20 kHz
Available history lengths include:
- 10 seconds
- 30 seconds
- 60 seconds
Level is represented using colour.
Importantly, TELESCOPE displays instantaneous power rather than a long-term averaged approximation.
This makes transient events and rapidly changing spectral content easier to identify.
WATERFALL
Waterfall uses the same underlying spectral information but presents it in depth.
Frequency and level become a three-dimensional-style landscape, with older measurements moving away from the viewer.
Interestingly, OVNI Audio does not use GPU rendering for this display.
Instead, the 2.5D visualization is drawn in software.
This keeps the visualization system independent from GPU-specific behavior.
CQT
The Constant-Q Transform lens reorganizes the frequency spectrum according to musical notes rather than equally spaced FFT frequencies.
TELESCOPE uses:
24 bins per octave starting from A0
This provides twice the resolution of a conventional twelve-note chromatic division.
For musicians, this can make spectral information easier to interpret because frequencies are displayed according to their musical relationship.
Chromagram
CQT also includes a chromagram.
Energy is grouped by pitch class, allowing the harmonic content of the music to be examined independently of octave.
This can help visualize:
- Chords
- Tonal centers
- Repeated harmonic structures
- Dominant pitch classes
Key Detection
TELESCOPE estimates musical key using the Krumhansl & Kessler profiles.
Crucially, it does not simply display a key name as if the answer were certain.
The estimated key is accompanied by:
- Confidence
- Share of analyzed time
This makes the result more transparent.
If the harmonic evidence is ambiguous, the interface communicates that uncertainty rather than presenting a false sense of precision.
SPIRAL
The Spiral lens takes the same Constant-Q information and wraps it around a circular display.
One complete rotation represents one octave.
Because equivalent notes in different octaves appear along the same radial direction, harmonic relationships become visually obvious.
For example, multiple C notes across several octaves line up along the same axis.
This provides a different way of understanding octave structure and harmonic density.
SCOPE
The Scope lens concentrates on stereo behavior.
It measures:
- Correlation
- Width
- Balance
- Mono loss
Measurements can operate over windows of:
- 100 ms
- 300 ms
- 1,000 ms
Several visual representations are available.
Lissajous Display
The familiar Lissajous representation shows the relationship between the left and right channels.
A narrow vertical pattern generally indicates more mono-compatible material.
Increasing horizontal spread suggests a wider stereo image.
Strong horizontal or unusual diagonal behavior can indicate phase relationships worth investigating.
POLAR SAMPLE
Polar Sample provides another visualization of instantaneous stereo behavior.
Instead of the traditional Cartesian oscilloscope presentation, stereo information is arranged radially.
POLAR LEVEL
Polar Level displays energy according to pan direction.
A half-circle contains one ray for each degree.
Out-of-phase information folds toward the base of the display and its percentage is shown.
This can provide a particularly intuitive representation of stereo distribution.
BAND CORRELATION
A single correlation number can hide important information.
For example, a mix could have:
- Perfectly mono bass
- Extremely wide high frequencies
and still produce a relatively ordinary global correlation value.
Band Correlation solves this by calculating stereo behavior independently across ⅓-octave frequency bands.
This lets you identify exactly where stereo problems occur.
The low end may be stable while the upper mids contain problematic phase relationships.
Instead of seeing only the average, you can see the frequency-dependent behavior.
STEREO SPECTROGRAM
Stereo Spectrogram combines frequency, time and phase information.
The display uses colour to represent phase relationships:
- Red — out of phase
- Green — wide
- White — mono
Brightness represents signal level.
If a cell contains no meaningful energy, it is displayed as black rather than being assigned a misleading phase colour.
This produces a detailed map showing where stereo width and phase relationships occur across both frequency and time.
FIELD
FIELD provides a 2.5D representation of:
Energy by pan direction × frequency
The analysis uses a:
64 × 96 grid
with decay and trail visualization.
This can make it easier to understand where frequency energy is distributed across the stereo field.
But OVNI Audio makes an important distinction.
FIELD Is Not Sound Localization
TELESCOPE does not claim to determine where a sound physically existed around the listener.
FIELD and POLAR LEVEL display energy by pan direction.
They do not use:
- HRTF
- ITD
- Azimuth reconstruction
Recovering the original spatial position of individual sources from a finished stereo mix does not have a unique solution.
OVNI Audio explicitly avoids pretending otherwise.
The FIELD interface therefore permanently displays the clarification:
Energy by pan direction — not localisation.
TONAL BALANCE
The Tonal Balance lens allows your mix to be compared with a reference track.
Both tracks are normalized according to their own integrated LUFS before the comparison is made.
This is important.
Simply comparing two raw spectra can be misleading if one track is considerably louder than the other.
TELESCOPE instead compares their tonal tilt at equivalent loudness.
The resulting delta is normalized so its total sums to zero.
This emphasizes differences in tonal distribution rather than differences caused simply by overall level.
Load Your Own Reference
Rather than forcing users to compare their music against a generic genre curve, TELESCOPE lets you load an actual reference track.
That means the reference can be chosen according to your own production objective.
For example:
- A professionally mastered techno track
- A commercial ambient release
- A reference mix from the same album
- Another version of your own master
This can be more meaningful than comparing every production against a universal target.
VERDICT
VERDICT is the feature that most clearly separates TELESCOPE from a conventional collection of meters.
It takes the measurements generated by the analysis engine and applies 21 deterministic rules.
The results are converted into readable statements.
But every statement also includes:
- The measurement
- The numerical value
- The rule ID responsible for the finding
This creates an important distinction between TELESCOPE and an AI-style mix assistant.
Rules, Not AI
VERDICT does not use:
- Artificial intelligence
- Machine learning
- Neural networks
- Cloud processing
- Remote analysis
It is essentially a transparent threshold table.
The same audio produces the same report.
There is no model interpreting the music differently from one run to another.
There is also no telemetry.
Measurement, Not Taste
OVNI Audio repeatedly emphasizes that VERDICT does not determine whether your mix is artistically good.
Its rules identify measurable conditions.
That is all.
A highly unusual experimental mix could deliberately violate several conventional technical expectations and still be exactly what the artist intended.
Conversely, receiving no warnings does not mean that a mix is finished.
TELESCOPE provides information.
The producer or engineer still makes the decision.
Every Finding Shows Its Evidence
This is arguably the strongest aspect of VERDICT.
Instead of saying:
“Your mix is too wide.”
the system can point to the relevant measurement and the rule that generated the statement.
That makes the analysis inspectable.
You can understand why something was reported rather than simply trusting an opaque recommendation engine.
Find Problems in Time
VERDICT also uses per-second history.
Its headline can indicate where the first relevant issue begins.
Instead of simply telling you that something happened somewhere during a five-minute track, TELESCOPE can direct attention toward a specific time.
That can dramatically speed up mix checking.
Offline File Analysis
TELESCOPE does not require you to play an entire track in real time.
You can drop an audio file into the plugin and analyze it offline.
The same analysis classes used for live playback are used for the file.
Results include:
- Integrated loudness
- LRA
- True peak
- 30 measured ⅓-octave bands
- Per-second history
OVNI Audio states that the offline and live results are bit-identical for these measurements.
Extremely Fast Offline Analysis
On OVNI Audio’s reference M4 system, a 60-second audio file is analyzed in approximately:
0.24 to 0.63 seconds
That makes TELESCOPE useful for quickly checking finished mixes and masters without listening through the entire track every time.
Bit-Exact Pass-Through
Perhaps the most important technical characteristic of an analysis plugin is what it does not do.
TELESCOPE does not process your audio.
It reads the left and right channels and sends the information to its own analysis thread.
Nothing is written back into the signal path.
OVNI Audio has tested this using deterministic stereo noise across:
- All 13 lenses
- Bypass on and off
- Multiple processing block sizes
- Mono sources
The result was zero mismatches.
Zero Samples of Latency
Because TELESCOPE only analyzes the signal, it introduces:
0 samples of latency
It also has:
0 samples of tail
That means it can remain inserted on a track, bus or master channel without changing timing.
Analysis Runs on a Separate Thread
The measurement system operates independently from the audio processing path.
The audio thread does not wait for the analysis to finish.
This architecture is important because visual analysis can sometimes be computationally expensive.
TELESCOPE is designed so that drawing a complex visualization does not become part of the real-time audio signal path.
Only Necessary Modules Are Computed
TELESCOPE does not calculate every possible visualization continuously.
Only the modules required by the currently selected lens are normally active.
Loudness is the main exception.
It remains active continuously so that:
- Integrated LUFS
- LRA
- Histogram
- Clip count
do not develop gaps when switching between lenses.
Performance Targets
OVNI Audio publishes explicit rendering-performance targets for the interface.
At the large interface size, each lens is tested against approximately:
- 4 ms median
- 8 ms p95
At Retina scale:
- 6 ms median
- 12 ms p95
The developer also publishes the machine load factor associated with the tests rather than presenting performance numbers without context.
True-Peak Measurement
TELESCOPE implements the 4×12 polyphase FIR described in BS.1770 Annex 2.
Below 96 kHz, true-peak measurement uses 4× FIR oversampling.
Interestingly, OVNI Audio also documents known limitations of the official method rather than hiding them.
For one EBU test signal whose actual peak is −6.0 dBTP, the specified method reads approximately:
−6.3160 dBTP
TELESCOPE reports the result produced by the standard rather than silently modifying it to appear more accurate.
Mono and Stereo Input
TELESCOPE accepts:
- Mono input
- Stereo input
and produces stereo output.
Because the output is bit-exact, the purpose of that output stage is simply to allow the signal to continue through the DAW unchanged.
Multiple Languages
The interface currently supports:
- English
- Spanish
- Portuguese
- French
- German
- Italian
English and Spanish have been reviewed.
The other translations currently use established industry terminology but are pending native-speaker review.
Missing translations fall back to English rather than displaying blank interface elements.
Three Interface Sizes
The GUI can be switched between:
- S
- M
- L
These are not simply identical graphics enlarged or reduced.
Each size uses its own layout metrics so the information can be presented appropriately for the available space.
macOS Compatibility
TELESCOPE supports:
macOS 11 or later
The macOS build is Universal, supporting:
- Apple Silicon
- Intel Macs
Available formats:
- VST3
- Audio Unit
The installer is signed with an Apple Developer ID and notarized by Apple.
Windows Compatibility
On Windows:
- Windows 10 or later
- 64-bit
- VST3
The Windows version is distributed as an unsigned ZIP.
Because of that, Windows SmartScreen may display a warning.
No AAX
TELESCOPE is not currently available in AAX format.
That means it does not load directly in Pro Tools.
Free and Open Source
TELESCOPE is completely free.
The source code is released under the:
AGPLv3 licence
OVNI Audio states that both the code and installers for TELESCOPE remain free with no download deadline.
This is worth noting because earlier OVNI releases used different promotional wording around installer availability.
For TELESCOPE, the current position is straightforward:
Free and open source, with no deadline.
Email and Download
The current TELESCOPE page asks for an email address and starts the download immediately afterward.
The mailing system uses double opt-in for OVNI Audio updates.
The developer states that subscribers can unsubscribe at any time.
Best Uses
Mixing
Keep TELESCOPE on the master bus to monitor loudness, dynamics, spectrum and stereo behavior without altering the signal.
Mastering
Check LUFS, LRA, true peak, PLR, PSR, tonal balance and phase relationships from a single interface.
Reference Comparison
Load a commercial reference and compare its tonal distribution against your own track at equal integrated loudness.
Stereo Checking
Use Scope, Band Correlation and Stereo Spectrogram to investigate width and phase behavior.
Low-End Analysis
Band Correlation can reveal whether the bass region remains mono-compatible independently of wider high-frequency material.
Finding Clipping
Use the ten-minute clip timeline to identify exactly where clipping occurs.
Harmonic Analysis
Use CQT, Chromagram and Spiral to examine notes, pitch classes and octave relationships.
Key Estimation
Obtain a key estimate accompanied by confidence information rather than an unexplained musical label.
Final Mix Checks
Use VERDICT to identify measurable conditions worth inspecting before export.
Education
Because VERDICT exposes both the number and rule behind each finding, TELESCOPE can also help producers understand why certain technical mix characteristics are being flagged.
Strengths
- Completely free
- Open source
- AGPLv3
- 13 analysis lenses
- 21 deterministic VERDICT rules
- No AI
- No neural network
- No cloud analysis
- No telemetry
- Bit-exact pass-through
- 0 samples latency
- 0 tail
- ITU-R BS.1770 loudness
- Integrated LUFS
- Short-Term LUFS
- Momentary LUFS
- LRA
- True peak
- PSR
- PLR
- 61-bin Short-Term histogram
- Ten-minute clipping timeline
- FFT up to 32,768 points
- Hann, Blackman-Harris and Kaiser windows
- ⅓-octave analysis
- Bark analysis
- Spectrogram
- Waterfall
- Constant-Q analysis
- Chromagram
- Key detection with confidence
- Spiral harmonic visualization
- Lissajous scope
- Polar Sample
- Polar Level
- Band Correlation
- Stereo Spectrogram
- FIELD visualization
- Tonal Balance
- User reference tracks
- Equal-loudness reference comparison
- Offline file analysis
- Fast offline processing
- Per-second history
- Mono and stereo input
- Six interface languages
- Three GUI sizes
- VST3
- Audio Unit
- Apple Silicon
- Intel Mac
- Windows
- Signed and notarized macOS installer
Potential Limitations
TELESCOPE deliberately refuses to do several things that other analysis products may claim to do.
It does not tell you whether your mix is artistically good.
VERDICT uses fixed deterministic rules rather than artificial intelligence, so its findings should be treated as technical observations rather than creative instructions.
FIELD does not perform genuine sound-source localization. It displays energy by pan direction but cannot reconstruct the original spatial position of individual instruments from a finished stereo mix.
TELESCOPE is also stereo-focused and does not currently provide surround or immersive multichannel analysis.
There is no speech-intelligibility metric.
Pro Tools users should also note the absence of an AAX version.
And perhaps most importantly, passing every VERDICT rule does not mean a track is finished.
A technically conventional mix can still sound uninspiring.
An unconventional mix can intentionally violate several technical conventions and sound excellent.
TELESCOPE gives you measurements.
The creative decision remains yours.
Final Verdict
TELESCOPE is one of the more interesting free metering plugins to appear recently, largely because OVNI Audio has resisted turning measurement into another opaque “AI assistant.”
Its thirteen lenses cover an impressive amount of territory.
You can measure conventional loudness and true peak, examine dynamics through PSR and PLR, inspect the spectrum, visualize frequency over time, analyze musical notes with Constant-Q processing, examine stereo correlation by frequency and compare the tonal balance of your mix against a real reference.
Then there is VERDICT.
Instead of simply presenting another collection of meters, it turns measurable conditions into readable findings.
But crucially, it shows its work.
Every statement is connected to the numerical measurement and deterministic rule that produced it.
There is no mysterious model deciding what your music should sound like.
There is no remote server listening to the track.
And there is no claim that satisfying a technical threshold somehow makes a mix good.
Add bit-exact pass-through, zero latency, offline file analysis, Windows and macOS support and an AGPLv3 open-source licence, and TELESCOPE becomes a remarkably comprehensive package for a free analyzer.
For producers and mastering engineers who want to understand what is actually happening inside a mix — rather than being told how it should sound — OVNI TELESCOPE is absolutely worth downloading.
Source : ovniaudio.com
