Hi, I’d like to share my latest turntable measurement procedure that I’ve been working on. It’s still under development, so I’ll focus more on describing the procedure than on the results or overall quality of the tested turntable.
Any comments or suggestions are more than welcome!
Tested turntable: Audio-Technica AT-LP140XP
Vinyl record: “Turntable Test Vinyl Record – Laboratory Edition,” provided by Unitra LINK
Signal analyzer: RME Fireface UC and Audio Precision APx500 FLEX
In terms of repeatability and measurement quality, I’m aware that a lot depends on the physical setup of the cartridge and stylus. Even the test vinyl is intended more for setup than for precise measurements. On the other hand, most turntables are delivered partially assembled and only require attaching the headshell and setting the tracking force and anti-skating according to the manual. This provides fairly repeatable conditions and allows evaluation of a specific model at a satisfactory level.
All measurements were conducted without a preamplifier; the turntable was connected directly to the RME interface and grounded to it. The RIAA curve was applied only for frequency response measurements (this will be noted where relevant).
As you can see, I should adjust the cartridge.
FFT Length: 96000, FFT Window: AP-Equiripple, Averages: 10
FFT Length: 96000, FFT Window: AP-Equiripple, Averages: 10
Track 1: 20 Hz – 20 kHz, log sweep, −20 dB
Track 2: 20 Hz – 15 kHz, log sweep, −14 dB
Track 3: 20 Hz – 2 kHz, log sweep, 0 dB
Each signal covers a different frequency range and level, which is related to the physical limitations of vinyl records caused by RIAA equalization. Track 3 allows for a good signal-to-noise ratio but has a limited frequency range. Track 1 covers the full hearing range but the signal is reduced by 20 dB, making it very low, especially without a preamp. Track 2 is a sweet spot, taking into account the limitations of the phono cartridge in reproducing high frequencies.
Below are measurements including both raw and equalized (according to the RIAA curve) frequency responses.
Track 1
Track 2
Track 3
Side D includes two additional tracks for frequency response analysis. The first is a logarithmic sweep (20 Hz–15 kHz, −14 dB) with a slow rise, lasting over 100 seconds. It allows for more detailed frequency characteristics and local nuances. The second track has the same parameters but uses a linear frequency sweep.
Below you can find the graph. As you can see, I captured this signal using a simple signal recorder function, so the measurement is in the time domain. To match frequency with amplitude, I added a second Y-axis on the right side, represented by a dotted curve.
Here are the results:
Track 1, 33 1/3 RPM – outer edge of the record
Track 6, 33 1/3 RPM – inner part of the record
Track 2, 45 RPM
Please let me know what you think about the procedure. Especially, what would you change, remove, or add.
Thanks!
Any comments or suggestions are more than welcome!
Tested turntable: Audio-Technica AT-LP140XP
Vinyl record: “Turntable Test Vinyl Record – Laboratory Edition,” provided by Unitra LINK
Signal analyzer: RME Fireface UC and Audio Precision APx500 FLEX
In terms of repeatability and measurement quality, I’m aware that a lot depends on the physical setup of the cartridge and stylus. Even the test vinyl is intended more for setup than for precise measurements. On the other hand, most turntables are delivered partially assembled and only require attaching the headshell and setting the tracking force and anti-skating according to the manual. This provides fairly repeatable conditions and allows evaluation of a specific model at a satisfactory level.
All measurements were conducted without a preamplifier; the turntable was connected directly to the RME interface and grounded to it. The RIAA curve was applied only for frequency response measurements (this will be noted where relevant).
1. RMS and FFT
Let’s start with the basic RMS value and FFT response for a 1 kHz signal. To capture these parameters, I used side C, track 3, which contains a reference 1 kHz, 0 dB, mono in-phase signal.As you can see, I should adjust the cartridge.
2. No-signal measurement
Side B of the vinyl is dedicated to verify anti-skating, but it also includes a silent track that allows measurement of self-noise. Here are the results.FFT Length: 96000, FFT Window: AP-Equiripple, Averages: 10
3. Frequency response
Side A includes three test signals with a pilot tone that allow measurement of the cartridge’s frequency response:Track 1: 20 Hz – 20 kHz, log sweep, −20 dB
Track 2: 20 Hz – 15 kHz, log sweep, −14 dB
Track 3: 20 Hz – 2 kHz, log sweep, 0 dB
Each signal covers a different frequency range and level, which is related to the physical limitations of vinyl records caused by RIAA equalization. Track 3 allows for a good signal-to-noise ratio but has a limited frequency range. Track 1 covers the full hearing range but the signal is reduced by 20 dB, making it very low, especially without a preamp. Track 2 is a sweet spot, taking into account the limitations of the phono cartridge in reproducing high frequencies.
Below are measurements including both raw and equalized (according to the RIAA curve) frequency responses.
Track 1
Track 2
Track 3
Side D includes two additional tracks for frequency response analysis. The first is a logarithmic sweep (20 Hz–15 kHz, −14 dB) with a slow rise, lasting over 100 seconds. It allows for more detailed frequency characteristics and local nuances. The second track has the same parameters but uses a linear frequency sweep.
Below you can find the graph. As you can see, I captured this signal using a simple signal recorder function, so the measurement is in the time domain. To match frequency with amplitude, I added a second Y-axis on the right side, represented by a dotted curve.
4. Wow and Flutter
Side C includes three additional tracks dedicated to wow & flutter measurements. Tracks 1 and 6 both use a 3150 Hz signal and allow calculation of W&F at the beginning and end of the record for 33 1/3 RPM. Track 2 also uses 3150 Hz but for 45 RPM speed. Audio Precision provides a utility for calculating W&F based on AES 6-2008 (equivalent to IEC 60386).Here are the results:
Track 1, 33 1/3 RPM – outer edge of the record
Track 6, 33 1/3 RPM – inner part of the record
Track 2, 45 RPM
5. Crosstalk
Side C also includes two tracks for crosstalk measurement. Both contain a 1 kHz, 0 dB signal in one channel. My results aren’t very good. I need to adjust the turntable. Ideally, I should get a more similar result for both channels.Summary
As you can see, I need to improve my turntable setup. I’ll try to recalibrate it and post updated results afterward.Please let me know what you think about the procedure. Especially, what would you change, remove, or add.
Thanks!
