---
Title: Turntable and test record – AudioLab
---

# Turntable and test record

The task **Turntable → Test record** measures a turntable, its tonearm and cartridge from a test record. AudioLab plays nothing: you play the tracks of the record, in any order, and each one is recognised as it plays – by its frequency, its level and the channel it is cut in. Two records are known by name: the **Ortofon Test Record** and **Analogue Productions – The Ultimate Analogue Test LP**.

![The two test records AudioLab knows: the Ortofon Test Record and Analogue Productions' The Ultimate Analogue Test LP, leaning against a turntable](figures/test_records.jpg)

## Quick start

**To align the turntable** step by step – which tracks to play, what to turn and which way – use the **Setup** view: [Turntable setup, step by step](turntable-setup.html). What follows is the measurement itself.

**No separate measuring interface?** A USB sound card with its own phono input can do the whole thing on its own, with no ASIO driver at all: [Turntable through a USB sound card](step-turntable-usb-card.html).

**You need**

- a test record: the **Ortofon Test Record** or **The Ultimate Analogue Test LP**;
- a **phono preamplifier** – the turntable's signal is too weak for the interface and needs the RIAA correction. An amplifier with a phono input works too, if it has a line output on the back (often labelled *REC OUT*); the turntable then goes to its phono input and the line output to the interface;
- two cables from the preamplifier's RCA outputs to the interface's line inputs (on an M4: RCA → 6.3 mm jack, TS).

**Connect**

Turntable → phono preamplifier, with the turntable's ground wire on the preamplifier's GND terminal. Preamplifier output **L → IN 3**, **R → IN 4** (the M4's rear line inputs). Nothing goes to the interface's outputs: AudioLab plays nothing here, the record does.

**Set**

1. **CONNECTIONS**: your audio interface, *Wiring* **Stereo, references first**, *Sample rate* **96 kHz**.
2. **MEASUREMENT** → Turntable → **Test record**.
3. **ANALYSIS** → *Test record*: the record on the platter.

Run the laptop on battery and keep the preamplifier away from the turntable and its mains transformer: hum makes every result worse.

**Measure**

Press **RUN** and play the record – a whole side, or the tracks you need. Each track turns green in the strip as it is recognised. Details list every result with **OK** or **check**.

**Adjust, in this order** – one thing at a time, then play the tracks that show it again:

| Details say *check* at | What to do | Tracks to play again |
| --- | --- | --- |
| Tracking ability, R worse | **Tracking force**: balance the arm to zero (it floats level), turn the dial to 0, then set the cartridge maker's force | Ortofon 9–14 |
| Tracking ability, both channels | the force at the upper end of the cartridge's range | Ortofon 9–14 |
| Tracking ability, L worse; anti-skating: one channel passes 3 % first | **anti-skating**: the same number as the force; L worse – turn it down, R worse – up, until both channels rise together | AP B1, Ortofon 9–14 |
| Crosstalk L→R against R→L differ by more than 2 dB | **azimuth** – only after force and anti-skating; view **Azimuth**, turn the headshell a little | Ortofon 5–8, AP A2–A3 |
| Speed | the turntable's pitch adjustment | Ortofon 5–8, AP A10 |

**Leave as it is** – force, anti-skating and azimuth do not change these: an arm resonance outside 8–12 Hz (only another headshell or cartridge moves it), treble falling above about 10 kHz (the cartridge with the capacitance of its cable and preamplifier input), rumble (bearing and motor), and a short pulse pointing down in the square wave (a preamplifier that inverts polarity – harmless).

The rest of this page explains each result, both records track by track, and a turntable measured before and after alignment.

## In detail

**Connect:** turntable → **phono preamplifier** (RIAA) → the device-output inputs of the wiring (*Stereo*: L → IN 2, R → IN 4; *Stereo, references first*: IN 3 / IN 4, on an M4 the rear line inputs); the turntable's ground wire to the preamplifier's terminal. An interface has no phono input: a cartridge gives about 5 mV and needs the RIAA curve and a 47 kΩ load. A receiver or amplifier with a phono input will do – take its tape or record output. Use **96 kHz**: the Ortofon sweeps reach 50 kHz.

The preamplifier is part of what is measured. Its own RIAA deviation adds to the response from the record: measure it first ([Phono preamplifier](phono.html)). Crosstalk, azimuth, speed, tracking and anti-skating hardly depend on it, as long as it does not hum.

## Measuring

1. Task **Test record**, ANALYSIS → *Test record*: the record on the platter. Laptop **on battery** (a desktop computer through a [USB isolator](phono.html#a-desktop-computer-a-usb-isolator)), the phono stage away from the turntable and its cables away from the stage's transformer: a charger or a transformer's field adds hum the cartridge picks up as well ([Phono preamplifier](phono.html#what-a-phono-stage-suffers-from-and-how-it-looks)).
2. **RUN**, or **SINGLE**: it stops by itself once every track it recognises has been heard and the groove after the last one is quiet.
3. Play the tracks in any order – the whole record side after side, or only the track you need. The strip above Details shows them: **green** heard, **grey** not yet. **Click a track** before you play it to tell AudioLab which one it is (it turns **blue** until heard): the next tone, sweep, ramp, IMD signal, noise or quiet groove heard is taken as that track, whatever it looks like – for playing one track again after an adjustment, or where two tracks look alike.
4. The **View** key above the chart switches between Setup (the guided alignment, [step by step](turntable-setup.html)), Response, Crosstalk, Tracking, Resonance, Azimuth and Waveform; over the chart is written which track a view still needs. Details list the results with **OK** or **check**, a line per track heard and everything as heard – the same as **Copy** and the **Report**. They take at most about a third of the room above the bar and scroll within it; while scrolled to their end they follow the new lines.

A track played again replaces the one before: after an adjustment, play it once more.

## What each record gives

The tracks are named as on the record's sleeve: the Ortofon record 1–15, the same on both sides (play either), the Analogue Productions record per side, A1–A10 and B1–B6.

| Result | Ortofon Test Record | Analogue Productions |
| --- | --- | --- |
| Speed | 1 kHz, tracks 5–8 | 3150 Hz, A10 |
| Crosstalk and azimuth | **curve** from the one-channel sweeps 1–4 (cut 800 Hz – 50 kHz); 1 kHz from 5–8 | **1 kHz only**, A2 and A3 |
| Channel balance | L of the last L-only tone (5 or 7) against R of the last R-only tone (6 or 8) | A2 and A3 |
| Response | 800 Hz – 20 kHz (sweeps cut at constant velocity: judged against RIAA) | 20 Hz – 1 kHz (A7), 1 – 16 kHz (A6) |
| Tracking ability | 315 Hz at 50 … 100 µm, tracks 9–14 | – |
| Anti-skating | from the tracking tracks | 315 Hz ramp 0 … +12 dB, B1 |
| Arm and cartridge resonance | – | vertical sweep 1 kHz – 10 Hz, B5 |
| Polarity, shape of the edges | square wave 2.7 ms (370 Hz), duty cycle 3:7, track 15 | – |
| Rumble | – | silent groove, B6: unweighted and A-weighted against the 1 kHz reference |
| Wow and flutter | – | 3150 Hz, A10: weighted (IEC 60386) as in **Speed & flutter**, from as much of the tone as the capture holds |
| IMD | – | 60 Hz + 4 kHz 4:1, A9: sidebands around 4 kHz, computed as the SMPTE method (which itself prescribes 60 Hz + 7 kHz) |
| Pink noise | – | B2 and B3, lateral and vertical by the phase of L and R |

**Crosstalk of the Analogue Productions record is two points.** All its sweeps are cut in both channels; the only tracks in one channel are the 1 kHz tones of A2 and A3, a point each way. A crosstalk curve over the band – and with it how an azimuth error grows in the treble – comes only from the Ortofon sweeps.

## The views

| View | Shows |
| --- | --- |
| **Setup** | the alignment step by step over the chart: what is off, which way to turn it, which tracks to play again ([Turntable setup, step by step](turntable-setup.html)) |
| **Response** | each sweep against what it should read: the RIAA curve for a sweep cut at constant velocity (Ortofon), flat for one cut with the RIAA pre-emphasis (Analogue Productions) |
| **Crosstalk** | the leak of a one-channel sweep or tone into the other channel, both ways (L → R, R → L) |
| **Tracking** | THD of each channel over the tracking tracks (µm) or over the anti-skating ramp (dB above its start), with the 3 % line |
| **Resonance** | the vertical sweep relative to 1 kHz: the resonance of tonearm and cartridge as a peak |
| **Azimuth** | for turning the headshell: L→R − R→L, one point per play of all the 1 kHz one-channel tracks (Ortofon 5–8, AP A2 and A3), aimed at 0 dB within ±2 dB, with what to do next written above |
| **Waveform** | the square wave (Ortofon track 15): two periods of L and R over each other, each period averaged over the whole track so the surface noise drops out; the rising edge at 0 ms |

## Adjusting the azimuth

The azimuth is the stylus leaning left or right, seen from the front. It cannot be seen in the signal directly, but in the crosstalk: a leaning stylus leaks more one way than the other. It is right when the left channel leaks into the right as much as the right into the left.

**How the Azimuth view works.** Each **point** is one **play of the one-channel 1 kHz tracks**: Ortofon 5–8 (L, R, L, R), Analogue Productions A2 and A3. AudioLab reads how much of each leaks into the other channel over the whole track, averages each direction over the play – **L→R** and **R→L**, written over the point – and plots their difference, **L→R − R→L**. The horizontal axis counts the plays, not time or frequency. The green line at **0 dB** is the right azimuth, the green band **±2 dB** is OK. The aim: after each turn of the headshell a new point, closer to the band.

**How to measure:**

1. First set the tracking force, the anti-skating and the cartridge's alignment (overhang): the azimuth comes last.
2. Task **Test record**, view **Azimuth**, **RUN**.
3. Play **tracks 5 to 8** through, from the start of 5 to the end of 8. Analogue Productions: **A2** and **A3**. A point appears once the last of them has played; until then the line above the chart says how many are heard.
4. Within ±2 dB: done (*Azimuth OK*). Outside it: lift the arm, loosen the cartridge or headshell screws a little and turn it by a fraction of a degree – either way, the first time.
5. Play the tracks again, all of them: a new point. The line above the chart compares it with the play before:
   - *smaller than the play before* – the turn went the right way: go on, in smaller steps as it nears zero;
   - *larger than the play before* – turn the other way;
   - *the sign changed* – past zero: back a little.
6. Tighten the screws and play the tracks once more: tightening can move it.

**How far to go.** A play of all the tracks repeats within a few tenths of a decibel; a single pair of tracks may read a little differently, so the Azimuth view takes the whole play. Within ±2 dB the azimuth is right; nearer to zero needs a finer turn of the headshell, or a thin shim under one side of the cartridge. Then look at the **Crosstalk** view: if the L→R and R→L curves cross over the band, the azimuth is as good as it gets – a turn would gain in one band what it loses in the other.

The bar shows on **L** the crosstalk L → R of the track playing, on **R** R → L with the signed difference of the last play, each with its best value.

## Results and guide values

| Result | Good | What to do |
| --- | --- | --- |
| Speed | within ±0.3 % | adjust the speed (pitch trimmer, belt, motor pulley) |
| Crosstalk at 1 kHz | 20 dB or more each way | – |
| Crosstalk L → R against R → L | equal within 2 dB | **azimuth**: turn the headshell or cartridge a little, play all the one-channel tracks again |
| Channel balance at 1 kHz | within ±0.5 dB | cartridge, wiring, preamplifier |
| Response against the record's curve | within the record's tolerance, ±1.5 dB | cartridge load (capacitance, resistance), tracking force, a worn stylus |
| Tracking ability | the cartridge's datasheet value; L and R alike | R first or both early: tracking force; L first: the anti-skating too high; one channel worse on every track: the anti-skating off |
| Anti-skating ramp: 3 % THD passed | L and R within 1 dB | adjust the anti-skating, play the ramp again |
| Arm and cartridge resonance | 8 – 12 Hz | below 8 Hz warps and footfalls excite it, above 12 Hz it reaches the bass: a lighter or heavier headshell or cartridge |
| Polarity (square wave) | L and R in phase: the traces lie over each other | out of phase: the two wires of one channel are swapped – at the headshell, in the arm or in the cable |

The guide values are AudioLab's working figures, not a standard; the record's response tolerance is its own (Ortofon: ±1.5 dB, 800 Hz – 20 kHz). Which way to turn the anti-skating the **Setup** view says ([Turntable setup, step by step](turntable-setup.html)); play the tracks again after each adjustment and aim for both channels together, as late as possible.

## Example: a turntable with an MM cartridge

*A real device, measured as described: what the measurement shows and how to read it.*

A direct-drive turntable with its MM cartridge, a tube phono stage (within ±0.8 dB of RIAA to 20 kHz), the Ortofon record played through before the alignment, with the counterweight at 3 g but the arm not balanced to zero: speed +0.01 %, channel balance −0.20 dB, crosstalk at 1 kHz −32 / −31 dB, both channels tracking 100 µm. The section *Tracking force in the measurement* below sets the same turntable after the alignment (arm balanced to zero, 3 g, anti-skating 3) beside a force set too low.

![Response view: the four Ortofon sweeps against the RIAA curve, L and R from −0.7 dB at 800 Hz rising to +1.2 to +1.6 dB from 2 kHz to 9 kHz, R peaking at +2.3 dB near 10.5 kHz, then both falling steeply, R to −7 dB at 16.5 kHz and L to −9 dB at 17.5 kHz](figures/record_response.jpg)

Within −0.7 to +1.6 dB from 800 Hz to 9 kHz, then a peak of +2.3 dB (R) near 10.5 kHz and a steep fall: the mark of an MM cartridge loaded by too much capacitance – its coil inductance and the capacitance of the cable, the arm wiring and the stage's input resonate, here near 11 kHz, and above it the treble falls 12 dB or more an octave. A shorter or low-capacitance cable moves the peak up; a stylus that loses the treble falls more gently, from no one point.

![Crosstalk view: L→R and R→L along the one-channel sweeps, from 800 Hz to 3.5 kHz R→L about −28 to −31 dB and L→R −31 to −36 dB; above 5 kHz both rise, L→R to −20 dB at 7.5 kHz and −5 dB at 17 kHz, R→L to −9 dB](figures/record_crosstalk.jpg)

The crosstalk is −28 dB or better up to 4 kHz, then grows with the same resonance. L→R is lower than R→L below 3.5 kHz and higher above 6 kHz: a stylus leaning a little, but the two curves cross, so turning the headshell would trade one band for the other.

![Azimuth view after the alignment: four plays of the Ortofon tracks 5–8 with nothing turned, at −1.3, −1.5, −1.5 and −1.4 dB (L→R about −31.7, R→L −30.3 dB), all inside the green ±2 dB band: Azimuth OK](figures/record_azimuth.jpg)

After the alignment four plays of tracks 5–8, with nothing turned between them, read −1.3 to −1.5 dB: R leaks a little more into L than L into R, inside the ±2 dB band, and the same from play to play within a tenth or two of a decibel. The azimuth is right; a finer turn of the headshell – or, where it does not turn, a thin shim under one side of the cartridge – would bring it nearer to zero, each step checked with another play of tracks 5–8. The tracking force and the anti-skating were set first: a stylus sitting unevenly in the groove moves this reading too.

![Tracking view: THD of L and R over the tracking tracks 50 to 100 µm, both 0.3 to 0.5 % from 50 to 90 µm; at 100 µm L 0.6 % and R 2.3 %, under the 3 % line](figures/record_tracking.jpg)

Both channels track the record's highest level, 100 µm, under 3 % THD; the right one rises first (2.3 % against 0.6 %). Before the alignment the arm had not been balanced to zero when the counterweight was set; after it, at the same 3 g, both channels rise together to 1.5 % (below). The left channel's 0.6 % is no better result: the stylus reads both groove walls at once, so tracking is judged by the worse channel and by how alike the two are. A low figure on one channel beside a high one on the other means the stylus does not sit evenly in the groove: from an unbalanced arm the dial does not show the real force. The same pattern, stronger, appears with the force set to 2.5 g (below: 0.4 % against 2.6 %), so before the alignment the real force was most likely below 3 g. After the alignment the worse channel improves from 2.3 % to 1.5 %, and the two channels rise together.

![Waveform view: two periods of the 2.7 ms square wave, L and R lying over each other, the short pulse downwards, clean rounded edges without ringing and slightly sloping tops](figures/record_waveform.jpg)

L and R lie over each other: the cartridge is wired right. The short pulse points down – the tube stage turns the polarity over. No ringing after the edges; the sloping tops are the stage's subsonic filter.

### Tracking force in the measurement

The same cartridge with two settings of the counterweight, its recommended force being 3 g. At 2.5 g on the dial (anti-skating left at 3) the stylus no longer sits evenly in the groove; set to 3 g after balancing the arm to zero (it floats level, the dial turned to 0, then the weight to 3 g), with anti-skating 3, both channels read alike again.

![Response view, Ortofon sweeps, two settings: at 2.5 g the right channel falls from 6 kHz, an octave before the left one; at 3 g with the arm balanced both stay within +1 to +2.2 dB from 2 kHz to 11 kHz and fall together](figures/record_force_response.jpg)

![Crosstalk view, two settings: at 2.5 g L→R 3 to 9 dB worse than R→L up to 5 kHz, the two closing above 8 kHz; at 3 g within 2 dB of each other up to 2 kHz, L→R then 3 to 6 dB above R→L](figures/record_force_crosstalk.jpg)

![Tracking view, two settings: at 2.5 g the right channel jumps to 2.6 % THD at 100 µm while the left stays at 0.4 %; at 3 g both rise together from 0.5 % to 1.5 %](figures/record_force_tracking.jpg)

| | 2.5 g on the dial | 3 g, arm balanced |
| --- | --- | --- |
| Treble, R against L | R falls from 6 kHz, L from 12 kHz | both from 11–12 kHz |
| Crosstalk at 1 kHz (the pair of tracks played last), L→R / R→L | −28.1 / −32.9 dB (difference +4.8 dB) | −30.0 / −30.7 dB (+0.7 dB) |
| THD at 100 µm, L / R | 0.4 / 2.6 % | 1.5 / 1.5 % |

What gave it away was one channel behaving unlike the other in three views at once – treble, crosstalk and tracking: a stylus sitting unevenly in the groove, which points to the tracking force and the anti-skating rather than to the azimuth. The azimuth view alone would have suggested turning the headshell, which would have hidden the cause. The dial on the counterweight reads true only from a balanced arm.

### The same turntable with the Analogue Productions record

Both sides played through in one run, all 16 tracks recognised: speed −0.07 %, wow and flutter 0.072 % weighted RMS (A10), channel balance +0.2 dB, IMD 4.1–4.5 % (A9), crosstalk at 1 kHz −41 and −35 dB (A2, A3).

![Response view of the Analogue Productions bass sweep A7, 1 kHz down to 20 Hz in steps of a sixth of an octave: L and R within −1.3 and +0.8 dB of flat, falling only below 25 Hz](figures/record_ap_bass.jpg)

The bass sweep (A7) lies within −0.6 and +0.8 dB from 1 kHz down to 25 Hz and at −1.3 dB at 20 Hz, through the stage's RIAA and its rumble filter: every step of the record is a point.

![Resonance view: the vertical sweep B5 relative to 1 kHz, flat from 1 kHz down to 50 Hz, rising to a peak of +6.2 dB (L) and +7.2 dB (R) near 25–28 Hz, then falling steeply to −14 dB (L) and −8 dB (R) at 16.5 Hz; all 16 tracks heard](figures/record_ap_resonance.jpg)

The vertical sweep (B5) shows the arm and cartridge resonance at **25–28 Hz, +6 to +7 dB** (24.8 and 27.9 Hz in two plays: the steps lie a sixth of an octave apart) – well above the 8–12 Hz a cartridge and arm should be matched to. A stiff (low-compliance) cartridge on a heavy arm puts it there; the stage's rumble filter cuts below 20 Hz, so the true peak may lie a little lower. Such a peak lifts the deepest bass and makes the arm sensitive to footfall; a lighter headshell or a more compliant cartridge brings it down.

![Tracking view of the anti-skating ramp B1: three steps of 315 Hz at 0, +5.7 and +11.3 dB, THD of L and R together from 0.5 % to 1.6 %, far under the 3 % line](figures/record_ap_skating.jpg)

The anti-skating ramp (B1) comes in three steps about 6 dB apart; L and R stay together and under 1.7 % THD to the top – the anti-skating is balanced. The silent groove (B6) gives rumble −57.0 and −54.4 dB A-weighted against the record's 1 kHz reference tone (A1, 7 cm/s), above the −60 dB the task marks as OK: the record's own surface noise is in it too, so this is the turntable and the record together.

## The square wave

Ortofon track 15 is a square wave whose short pulse takes 3 of every 10 parts of the period. The Waveform view draws one period of each channel, averaged over the whole track, from a quarter of a period before the rising edge.

- **L and R in phase** – the traces lie over each other: the cartridge is wired right. A mirror image means one channel's wires are swapped.
- **Short pulse up or down** – the absolute polarity of the whole chain, the phono preamplifier included. Many preamplifiers, tube ones above all, turn it over, and the sleeve does not say which way the record is cut: AudioLab tells it and does not judge it.
- **The edges** are only to look at: ringing after an edge is the cartridge's treble resonance, a rounded edge missing treble, a sloping top the bass rolling off (the preamplifier's subsonic filter).

## Limits

- **Blocks of about 0.35 s** (16384 samples at 48 kHz, 32768 at 96 kHz), whatever the FFT length in Settings: a longer block would hear too much of a sweep at once. Even so a block hears about a sixth of an octave of a 10 s sweep, and a sharp change – the Analogue Productions record's level error above 16 kHz, a narrow resonance – reads smoother than it is.
- **Lateral and vertical** cuts look alike by level: tones, sweeps and noise are told apart by the phase of L and R (in phase lateral, out of phase vertical) – click the track first if in doubt.
- **Stepped sweeps:** the sweeps of the Analogue Productions record (A6, A7 and B5) step a sixth of an octave about every 1.3 s instead of gliding. AudioLab follows them as sweeps and reads each step where the tone holds, so a step is a point of the response, not a tone of its own. Hum beside a low step (near 50 Hz) pulls it a little and does not end the sweep. The anti-skating ramp (B1) is cut in three steps about 6 dB apart; they are read as the ramp, its THD per step.
- **33⅓ rpm only** for now; at 45 rpm the tones lie 35 % higher and are not recognised.
- **The silent groove** (Analogue Productions B6) is the loudest quiet stretch of at least 4 s heard after B5, the track before it: quiet on side A is the arm being moved, the lead-in grooves between the tracks are shorter, and with the arm lifted – turning the record over – only the phono stage's own noise is left, quieter than a groove. Played without B5 before it, B6 is not found by itself: click B6 before it plays to be sure.
- **Not read:** VTA (Analogue Productions A9 is cut laterally and gives no VTA reading); the spectrum of the pink noise (look at it with the Spectrum task); the flutter spectrum (Speed & flutter).
- A test record wears with every play, the high levels and the treble most: check doubtful results against a second copy or a known turntable.

## Trying it without a turntable

Choose the interface **AudioLab Simulator** and the simulated device **Turntable + phono preamp (test record)**; the task switches to it by itself. The simulator plays the chosen record shortened, through a cartridge with an azimuth error (L → R −27 dB, R → L −24 dB at 1 kHz), a right channel 0.4 dB low, mistracking that starts in the right channel first, an arm resonance at 10.5 Hz, a treble resonance at 18 kHz that makes the square wave's edges ring, and a platter with wow at 0.556 Hz.

## The groove, interactively

How one stylus reads two channels from the two walls of a single groove – and what the azimuth, the tracking force and anti-skating, the cartridge load, the arm resonance and the speed do to a measurement – can be tried in the browser: [Stereo groove simulator](groove-simulator.html). It draws the groove and the stylus moving in it; the setup examples are illustrations of cause and effect, not targets for a particular cartridge.
