- Wayne Kirkwood
- Posts: 41
- Joined: Tue Jul 21, 2015 12:43 pm
- Location: Dallas, Texas
- Contact:
RIAA and Inverse RIAA vst and Nyquist Plugins and Tools
Hey Trolls it's been awhile.
I wanted to share with you guys some work I've been doing comparing RIAA IIR Biquad plugins to analog RIAA EQ.
After finding an accurate-sounding IIR vst plugin for Audacity (and a Nyquist RIAA coefficient tool) I was able to compare DSP IIR solutions directly to analog RIAA EQ using Audio Diffmaker.
The analog EQ I used for comparison is this one: http://www.lathetrolls.com/viewtopic.php?f=9&t=5848&p=37252
I was driven by the inaccurate phase response of Cool Edit's and Audacity's FIR RIAA filters to find something better and wanted to prove to myself that IIR was equivalent to analog.
I've posted about it in detail here: http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=15&t=885
There are links to an Audio Diffmaker difference file, links to Olaf Matthes' vst plugin and some ready-to-use custom Nyquist scripts based on Wayne Stegall's coefficient generator program.
With the Nyquist scripts you can modify gain, sample rate and add a 50 kHz zero.
Olaf's vst plugin also offers Inverse RIAA.
Based on null testing the vst and Nyquist tools are essentially mathematically equal with a 60 dB correlated null depth.
Analog to DSP comparison is surprisingly close (about 20 dB correlated null depth in some tests) with the bulk of the difference in the top octave, IMHO, due to the converters used. (The test required recording both the flat, RAW, signal and the analog EQ'd signal simultaneously.)
As expected Cool Edit's FIR filter's weren't even close.
It appears to have been awhile since anyone discussed this here.
Hope you guys find this info useful.
Wayne
I wanted to share with you guys some work I've been doing comparing RIAA IIR Biquad plugins to analog RIAA EQ.
After finding an accurate-sounding IIR vst plugin for Audacity (and a Nyquist RIAA coefficient tool) I was able to compare DSP IIR solutions directly to analog RIAA EQ using Audio Diffmaker.
The analog EQ I used for comparison is this one: http://www.lathetrolls.com/viewtopic.php?f=9&t=5848&p=37252
I was driven by the inaccurate phase response of Cool Edit's and Audacity's FIR RIAA filters to find something better and wanted to prove to myself that IIR was equivalent to analog.
I've posted about it in detail here: http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=15&t=885
There are links to an Audio Diffmaker difference file, links to Olaf Matthes' vst plugin and some ready-to-use custom Nyquist scripts based on Wayne Stegall's coefficient generator program.
With the Nyquist scripts you can modify gain, sample rate and add a 50 kHz zero.
Olaf's vst plugin also offers Inverse RIAA.
Based on null testing the vst and Nyquist tools are essentially mathematically equal with a 60 dB correlated null depth.
Analog to DSP comparison is surprisingly close (about 20 dB correlated null depth in some tests) with the bulk of the difference in the top octave, IMHO, due to the converters used. (The test required recording both the flat, RAW, signal and the analog EQ'd signal simultaneously.)
As expected Cool Edit's FIR filter's weren't even close.
It appears to have been awhile since anyone discussed this here.
Hope you guys find this info useful.
Wayne
Re: RIAA and Inverse RIAA vst and Nyquist Plugins and Tools
Its just something you gotta mess with. Do it yourself.
<\\\\\\\\\\\\\\\\\\\\\[[[[[[[\/]]]]]]]\\\\\\\\\\\\\\\\\>
Recordette Sr.......Presto K-8
Recordette Sr.......Presto K-8
- Wayne Kirkwood
- Posts: 41
- Joined: Tue Jul 21, 2015 12:43 pm
- Location: Dallas, Texas
- Contact:
Re: RIAA and Inverse RIAA vst and Nyquist Plugins and Tools
I wanted to drop back in with some test results based on a different nulling approach.
Without going through all the details I figured out how to use one converter to record both the flat and analog EQ'd test signals simultaneously.
(It's called mono. Should have thought about that originally )
I used pink noise as the test signal.
One compare file had DSP-based RIAA applied (Matthes' IIR .vst) the other file EQ'd by an analog network made from 1% components.
By inverting one and summing it I produced a null file.
I then analyzed the null file using Spek.
Analog RIAA vs. IIR-RIAA Difference File Spectrum
Other than low frequencies there is no significant difference between the Analog RIAA and DSP IIR-RIAA.
At low frequencies there is some difference: The analog RIAA does not have DC gain. So what's going on in the lower octaves?
I also produced a "stereo" file. One channel is DSP, the other analog.
A simple X-Y display shows the error differences.
A persistent goniometer plot (averaged over about 45 seconds) shows spreading due to the LF phase error of the Analog EQ vs the DSP-based RIAA EQ.
Analog RIAA vs. IIR-RIAA Phase Error
With a 200 Hz HPF applied the phase error between Analog EQ and DSP-based EQ disappears:
Analog RIAA vs. IIR-RIAA Phase Error WIth 200 Hz HPF. Note the change in scale.
For all practical purposes the results are identical.
Below 200 Hz the phase error is introduced by the analog network's rolloff.
The DSP-based RIAA has response to DC.
This method appears to be a good way to verify unknown DSP-based RIAA solutions to see if they are minimum phase and accurate when compared to analog RIAA.
Without going through all the details I figured out how to use one converter to record both the flat and analog EQ'd test signals simultaneously.
(It's called mono. Should have thought about that originally )
I used pink noise as the test signal.
One compare file had DSP-based RIAA applied (Matthes' IIR .vst) the other file EQ'd by an analog network made from 1% components.
By inverting one and summing it I produced a null file.
I then analyzed the null file using Spek.
Analog RIAA vs. IIR-RIAA Difference File Spectrum
Other than low frequencies there is no significant difference between the Analog RIAA and DSP IIR-RIAA.
At low frequencies there is some difference: The analog RIAA does not have DC gain. So what's going on in the lower octaves?
I also produced a "stereo" file. One channel is DSP, the other analog.
A simple X-Y display shows the error differences.
A persistent goniometer plot (averaged over about 45 seconds) shows spreading due to the LF phase error of the Analog EQ vs the DSP-based RIAA EQ.
Analog RIAA vs. IIR-RIAA Phase Error
With a 200 Hz HPF applied the phase error between Analog EQ and DSP-based EQ disappears:
Analog RIAA vs. IIR-RIAA Phase Error WIth 200 Hz HPF. Note the change in scale.
For all practical purposes the results are identical.
Below 200 Hz the phase error is introduced by the analog network's rolloff.
The DSP-based RIAA has response to DC.
This method appears to be a good way to verify unknown DSP-based RIAA solutions to see if they are minimum phase and accurate when compared to analog RIAA.
Re: RIAA and Inverse RIAA vst and Nyquist Plugins and Tools
Thanks for another awesome thread Wayne!
Great stuff!
Great stuff!
Re: RIAA and Inverse RIAA vst and Nyquist Plugins and Tools
Hi Wayne,
Nice work. If you want to test the one I wrote some time ago, pm me with your email and I'll send it along. Mine might not be quite as good. Its IIR based, but I used a matched Z transform instead of the warped bilineal most seem to use. It made it easier to allow for adjustable turnovers. I've attached a printout of a Mathcad Analysis I did at 96Khz sample rate.
Mark
Nice work. If you want to test the one I wrote some time ago, pm me with your email and I'll send it along. Mine might not be quite as good. Its IIR based, but I used a matched Z transform instead of the warped bilineal most seem to use. It made it easier to allow for adjustable turnovers. I've attached a printout of a Mathcad Analysis I did at 96Khz sample rate.
Mark
You do not have the required permissions to view the files attached to this post.