The above video is a quick demonstration of how your signal-to-noise ratio impacts the level of noise your analog modeled plugins will deliver. You may have seen this video floating around Facebook, which is where it was exclusively hosted until now. I thought it was high time to add it to the blog to make it easier to find.
Sending a lower signal into them will put the signal closer to the noise floor. When that signal is brought up to normal mix levels, the noise floor of the plugin rises with it.
Sending a hotter signal into them will put the signal higher above the noise floor. When that signal is brought up to normal mix levels, the noise floor is very very quiet.
A lot of folks have said gain staging in the box doesn’t matter with floating point systems. I agree that it doesn’t, as long as you UNDERSTAND it. To read more on why gain staging in the box isn’t that big of a deal, read “Gain Staging: What To Know And Why You Shouldn’t Stress Out About It“.
This is one reason people shouldn’t adhere to rules like “normalize all tracks to -18”. First of all, people rarely every tell you what -18 is supposed to mean. -18 peak is a lot quieter than -18 RMS. In the analog days, 0vu was typically an rms measurement. In order to keep all those noise producing analog components from creating noise buildups, people followed standards that placed the audio in a sweet spot that gave a high signal to noise ratio while maintaining the integrity of the original signal (i.e. without adding unwanted non-linear saturation or distortion…the key word is “unwanted”, because all of them added some of that. How much you’d want was really up to you, as long as you didn’t exceed the signal strength your hardware could take and blow a circuit, which was risky behavior on high price gear used back then.)
But digital measurements are different than analog measurements. -18dbfs can equal 0Vu, if you want it to. So can -12dbfs, or -20dbfs. Different developers will use different 0vu points because 0vu in a linear system is a completely arbitrary measurement point.
Meanwhile, if you’re feeding a signal into a plugin that has a built in noise floor, you need to understand this feature and how to minimize it. I don’t look at meters when feeding signal into these plugins, but I will push signal into it until I find a place that has no noticeable noise floor while giving me the sound I want. If I want a clean sound, I stop before it starts distorting. If I want an overdriven sound, it’s 1’s and 0’s and not a physical circuit so I don’t care how hot it’s going in as long as it sounds good. In fact, in the above video, I like the sound of the guitar as it’s saturated. But I also like the sound just before it hit saturation, which is warmer and more exciting than at any other clean gain point in the demonstration.
But the key to being able to ignore any guideline is to understand the guideline and why it’s there. If you don’t understand signal-to-noise ratios, and you’re just shooting for -18dbfs max peak on your tracks, when you run that audio into a long chain of analog modeled plugins that each have a noise floor you’ll experience a buildup of noise. This is why so many wish companies would include a noise-off button, and some companies do. But when the plugin has no noise floor, the plugin no longer reacts and responds the same way the hardware version of the plugin does. So should developers make their modeled plugins sound and behave like hardware and hope that you understand signal-to-noise ratios? Or should they take the noise out and have people complain that they can’t get “that sound” out of the plugin they poured their blood sweat and tears into?
I say the onus isn’t on them. It’s on us. We should help people understand how these issues were addressed in analog gear, and how these issues can be addressed in digital recreations of analog gear. If your plugin has a noise floor build up, you’re likely running too quietly into the plugin. Grab a trim plugin, push more into it, and you’ll have a better signal-to-noise ratio to work with.
Products from Amazon.com
- Price: $569.00
- Price: $239.00Was: $299.00
- Price: $995.00
- Price: $995.00
- Price: $649.00
Products from Amazon.com
- Price: $199.00
- Price: $99.00Was: $139.00
- Price: $299.00
- Price: $171.02
- Price: $148.60