Fun with vinyl measurements

[somebodyelse]

Did you specifically choose an LP that you know is from the same master as the CD? @Thomas_A

And that has nothing below 40Hz on either?

 

[Thomas_A]

Did you specifically choose an LP that you know is from the same master as the CD? @Thomas_A

I do not have many copies both from CD and LP but I am pretty sure that these copies are identical in terms of spectral content. Looking at the second track sweep of CA-TRS1007 it will match very well with this result.

 

[melloncolliecat]

Yeah, from what I've heard they used a digital master for the original LP, so makes sense.

 

[Thomas_A]

And that has nothing below 40Hz on either?

Thay look basically the same below 40 Hz. I normally use a susbsonic filter for the vinyl.

 

[somebodyelse]

Thay look basically the same below 40 Hz. I normally use a susbsonic filter for the vinyl.

They do. The question was whether you picked the recording because of that, or whether it was just coincidence - there are plenty of CDs that have content below 40Hz.

 

[Thomas_A]

They do. The question was whether you picked the recording because of that, or whether it was just coincidence - there are plenty of CDs that have content below 40Hz.

Well I do not have many CDs and vinyl that I have the same of. I usually not buy stuff (if I do buy) that I already have. I don't see the point...

I have Eric Bibb 45 RPM "Spirit and the Blues" and CD version. But it will probably no contain much below 40 Hz.

Below 40 Hz it looks like this anyhow. There is a peak at 17 Hz both in the CD and vinyl version. If that tells anything - it should be on the master as well.

 

[Thomas_A]

You can user trace arithmetic and show the difference , easier to se the difference then

 

[Balle Clorin]

Intermodulation distortion two sytems

ATOC9MLii in SME V and SHure V-15 IV +Jico SAS-B on denon 51F

Test record LB 201 VEB DDR. Flat recording

 

[Balle Clorin]

This is a AI generated python script for presenting FFT and IMD-INTERMODULATION DISTORTION- using a test record LB202.
Test track has 315 Hz and 3150hz in 4:1 ration. no RIAA eq. Use at own risk, AI can help you modify it to other test frequecies, like 60Hz and 7000hz
Rename .txt to .py


DO NOT ask me to explain anything, it is all AI genereted on my request. I do not have test files to verify the results..use at own risk

This result is in the same order as a simple visual/manual sideband evaluation gives, but lower the REW calculation reports. I guess it can work to se relative changes ,but not absolute correct values

The first file is based in amplitude modulation effects , the second is based on spectrum FFT more like REW does

but the script dos this-using average funtamentals ,funny. I thought it should be relative to F2 the 3150hz

The IMD‐percentages are always computed from the linear amplitude ratios, not directly from the dBFS values. Here’s the recipe:

1. We measure four linear amplitudes from the FFT:
   • A₁ = amp at f₁ (Hz)
   • A₂ = amp at f₂ (Hz)
   • A_diff = amp at |f₂–f₁| (Hz)
   • A_sum = amp at f₁+f₂
2. We pick a reference A_ref for dBFS (in our script that’s max(A₁,A₂)), and compute dBFS:
   dBFS_x = 20·log₁₀(A_x / A_ref)
3. We form the average fundamental amplitude:
   fund_avg = (A₁ + A₂) / 2
4. The 2nd-order IMD percentages are then
   IMD_diff% = 100·(A_diff / fund_avg)
   IMD_sum% = 100·(A_sum / fund_avg)

Because Aₓ/fund_avg is a pure linear ratio, the printed “0.995%” comes from:

• f₂ @ –17.7 dBFS ⇒ A₂ = 10^(–17.7/20) ≃ 0.1337
• f₁ is at 0 dBFS ⇒ A₁ = 1.0
• • ⇒ fund_avg = (1.0 + 0.1337)/2 = 0.56685
  • diff @ –45.0 dBFS ⇒ A_diff = 10^(–45.0/20) ≃ 0.00562
    ⇒ IMD_diff% = 100·(0.00562 / 0.56685) ≃ 0.992% → ≈ 0.995%
    Likewise for the sum sideband:
  • sum @ –36.4 dBFS ⇒ A_sum = 10^(–36.4/20) ≃ 0.0151
    ⇒ IMD_sum% = 100·(0.0151 / 0.56685) ≃ 2.66% → ≈ 2.684%
    In short, dBFS just tells you how big each Aₓ is relative to the loudest fundamental. The IMD% is then the linear ratio of the sideband amplitude to the average fundamental amplitude, times 100%.

doing it relative to F2 increases % of course
── Left Channel ────relative to f2 3150hz─────────────────────────────────────
f1 @ 316.57 Hz : 0.0 dBFS
f2 @ 3140.34 Hz : -17.7 dBFS
diff @ 2827.59 Hz : -45.0 dBFS IMD_diff% = 4.299%
sum @ 3456.74 Hz : -36.4 dBFS IMD_sum% = 11.593%
Avg IMD% = 7.946%

── Right Channel ────relative to f2 3150hz─────────────────────────────────────
f1 @ 316.57 Hz : 0.0 dBFS
f2 @ 3140.34 Hz : -17.8 dBFS
diff @ 2827.83 Hz : -37.6 dBFS IMD_diff% = 10.200%
sum @ 3456.84 Hz : -31.8 dBFS IMD_sum% = 19.782%
Avg IMD% = 14.991%


f1 @ 316.57 Hz : 0.0 dBFS
f2 @ 3140.34 Hz : -17.7 dBFS
diff @ 2827.59 Hz : -45.0 dBFS IMD_diff% = 0.995%
sum @ 3456.74 Hz : -36.4 dBFS IMD_sum% = 2.684%
Avg IMD% = 1.840%

── Right Channel ────relative to average level of F1 and F2─────────
f1 @ 316.57 Hz : 0.0 dBFS
f2 @ 3140.34 Hz : -17.8 dBFS
diff @ 2827.83 Hz : -37.6 dBFS IMD_diff% = 2.338%
sum @ 3456.84 Hz : -31.8 dBFS IMD_sum% = 4.534%
Avg IMD% = 3.436%

 

[morillon]

The hungaroton ? ( 300/3000)

 

[Thomas_A]

Wonder if there is a more appropriate signal for IMD since most LPs are cut mono below 100-150 Hz.

 

[morillon]

Wonder if there is a more appropriate signal for IMD since most LPs are cut mono below 100-150 Hz.

How would the mono aspect of low frequencies pose a problem in the observation and interpretation of 2-tone imd ?
;-)

(https://www.audiocheck.net/audiofrequencysignalgenerator_dual.php)

 

[Thomas_A]

How would the mono aspect of low frequencies pose a problem in the observation and interpretation of 2-tone imd ?
;-)

(https://www.audiocheck.net/audiofrequencysignalgenerator_dual.php)

I was thinking that IMD tests are made from vertical modulation tracks. A 60 Hz vertical modulation does not occur in music tracks since it is mono-coupled.

 

[morillon]

I was thinking that IMD tests are made from vertical modulation tracks. A 60 Hz vertical modulation does not occur in music tracks since it is mono-coupled.

At observe the ultimate have one with 60hz/4k


( at6607 more interresting 1.8/2.2khz)

 

[Thomas_A]

With respect to the 470pF/36k and 240pF/36k, I went back to my original setting 240pF/47k. Even if it is a lift in the outer tracks of the CA-TRS1007, I find it sound a bit better. Could be imagination, but the overall sound over the whole LP less congested with the 240 pF/47k

 

[Thomas_A]

At observe the ultimate have one with 60hz/4k


( at6607 more interresting 1.8/2.2khz)

That is lateral, so it will show lateral IMD. It is not at all the same gross level as vertical. Which is good of course.

 

[morillon]

Who use str112 ?
(Or the so very special ttr103?)


(In old time we can found a 16k/16.300hz too, 300hz... ;-) logic...like the 1.8/2.2k, 400hz, from at-6607)

 

[Balle Clorin]

Wonder if there is a more appropriate signal for IMD since most LPs are cut mono below 100-150 Hz.

my LB202 is 315 & 3150hz . But I would presume that test record do not use elliptical filters(mono in bass)

 

[JP]

With respect to the 470pF/36k and 240pF/36k, I went back to my original setting 240pF/47k. Even if it is a lift in the outer tracks of the CA-TRS1007, I find it sound a bit better. Could be imagination, but the overall sound over the whole LP less congested with the 240 pF/47k

View attachment 480480
View attachment 480481

What profile? An MR on TRS-1005 shows less than half a dB difference (IIRC) between inner and outer.

 

[Thomas_A]

What profile? An MR on TRS-1005 shows less than half a dB difference (IIRC) between inner and outer.

This is MR with the CA-TRS1007 tracks. Don’t know why the ”big” difference. But checked also with the REM record, CD vs LP. Difference 1-2 dB in highs.