Professional Analog Mastering.
Professional Analog Mastering.
226K+ subscribers
So, In the last video we released, titled ‘Clean Mastering,’ I made the claim that plugin latency can cause pre-ringing distortion.
Now this was a claim that a lot of people who watched the video didn’t like - I understand why it doesn’t seem like this would be true - that is that plugin latency can cause pre-ringing distortion, but I wanted to quickly show you some evidence that what I said was correct.
So what I’ve done is take a Drac file - which is a high amplitude, single sample file.
Then, I’ve run it through various plugins - some don’t introduce latency, and some do.
What I’ve found is that almost all of the plugins that do introduce latency also introduce pre-ringing distortion - with the exception of 1 plugin.
Watch the video to learn more >
The first plugin I tested was this Weiss MM-1 maximizer - it introduces about 90ms of latency and amplifies quieter parts of the signal.
If we observe the Drac file that has been run through the plugin and zoom in, we’ll notice that pre-ringing distortion is occurring.
The reason this distortion occurs is because artifacts that are generated after the signal are then compensated for by our DAW’s automatic delay compensation. When our DAW moves the signal back to its original starting point, these artifacts are centered on the start of the original signal.
This is how we can have a signal occurring before the original signal even begins.
Watch the video to learn more >
Moving on to the FabFilter Pro-L2, set to its modern algorithm, with TP limiting enabled and 8x oversampling, we’ll notice that, again, we observe pre-ringing distortion.
This is due to the plugin’s roughly 60ms of latency, in combination with a high-pass filter introduced by oversampling.
What’s interesting here is that we’ll notice how the latency of a plugin actually corresponds to the severity of the distortion.
Watch the video to learn more >
Next, we have the Pro Q 3 Equalizer set to its maximization latency setting - yet if we observe the waveform, we’ll notice no pre-ringing.
This is because I made no changes to the frequency response.
As soon as I do, we’ll observe pre-ringing distortion. Now note that with this second EQ test, I used a 3dB bell filter at 600Hz. The fact we can observe pre-ringing suggests that we don’t need a severe filter to cause this type of distortion. All we need is for the plugin to introduce latency and, alter the frequency response.
Watch the video to learn more >
The 1 plugin that introduced latency and did not cause pre-ringing was Newfangled’s Saturate plugin. It introduced about 10ms of latency, but even when I was engaging the clipper, I observed no changes to the Drac file other than it’s amplitude.
This would suggest that the design of the plugin, maybe some behind-the-scenes processing, can mediate this distortion - so I’m going to test this plugin some more later to try and figure out what that might be.
Watch the video to learn more >
The FabFilter Pro-C2 normally does not introduce latency - but if we enable lookahead, it now has 20ms of latency. When we observe the Drac waveform, we’ll notice asymmetrical pre-ringing - which, to be honest, I’m not sure what is causing the distortion to react this way - maybe lookahead is altering the artifact.
Watch the video to learn more >
When it comes to Saturn 2 - it only introduces 0.2ms of latency unless a linear phase is enabled. On the right side, we’ll see the expected result of moderate saturation. To the left of the original signal, again, notice that pre-ringing is occurring, but its severity, or lack thereof, is directly tied to the latency of the plugin.
Reverb can do this as well - this is actually how I first came across this distortion when I was making some impulse responses. Notice that a fair amount of signal is present before the original Drac sample. This suggests that temporal effects can introduce some form of pre-ringing distortion as long as they introduce latency as well.
All of this suggests we need 2 things to cause some form of pre-ringing distortion.
First, the plugin needs to alter the frequency response, either with filters, harmonics, or some other additive or subtractive form of processing.
Second, the plugin needs to introduce latency which is then compensated for by the DAW.
Watch the video to learn more >
One last thing - in the previous video, I did give some wrong info that I want to correct here. I said that increasing the sampling rate would reduce the effect of pre-ringing by lowering the delay time of each plugin. Now, this is true when it comes to our monitoring latency, but it looks like Plugins scale their latency to the sampling rate.
For example, if I were to increase the sampling rate of my session by 4 times, a plugin would just multiply the number of samples by which it delays the signal by 4, resulting in the same delay time and the same amount of pre-ringing distortion. That said, increasing the sampling rate will still reduce aliasing distortion, so it’s a good idea to do if you want a clean-sounding master.
Sometimes with these videos, I’m covering things that rarely, if ever, get talked about - so I have to do my own research and testing, which means I can’t always link where I found certain info. But I hope this video helps clarify what I was talking about in the ‘Clean Mastering’ video, and if you have any additional questions, please let me know in the comments.