A weird looking stereo waveform

Ok so this doesn’t look anything like the music I make inside the emailing machine:

I might not have noticed this if it weren’t for the fade in automation which produces a small bump.

What is this phenomenon called?
Why should I care?

I mix using Veils and Blinds and then send to 6 track recording using the ER-301.

Your signal has a DC offset - its average value (the “center” around which it wiggles) is not 0.

It’s fairly easy to get that with Blinds - for example if an unused channel is involved in the mix and is not exactly set to 0.

This is very easy to fix in software, you should have a DC-offset removal plug-in somewhere. Or just a high-pass filter with a low cutoff (say 20 Hz).


Thank you. That is very interesting.

This was a little mind boggling. Wouldn’t make the connection between the visual representation of a whole wave form and the frequencies it contains.

I’m also learning that Ableton’s EQ8 goes down to 10Hz in Live 10 (which I don’t own). It’s limited to 30Hz in previous versions

Someone also said this:


I’ll just find some DC-offset plugin for the time being, multiple options available! :smiley:

1 Like

Any low cut will strip off the DC component. Try it with the Ableton EQ, it should work if it is a true low cut.

This thread is really educational. I’ve learned about DC offsets and a couple methods for how to remedy that situation! It’s nice to see some positive things these days (i.e. people helping others).

1 Like

What does true low cut entail, in Hz?

People appear to be aggravated at missing out on important stuff going on at 30hz.

I’m not one of those people, I’m just curious to the interplay of dB and Hz at this point. Me coming from an expert angle, as I’m sure you’ve already noticed… :smiley:

Any low cut, even 0.0005Hz will in theory remove the DC offset entirely - think of it like the rolloff of a filter: e.g. a 12dB per octave filter will remove an additional 12dB for each additional octave (1 octave = doubling of the frequency). That is also true for high pass (= low cut) filters. For each additional octave down (= frequency/2) you’ll get the same amount of suppression.

The DC component has a frequency of exactly zero. How many octaves below e.g. 10Hz is that? Infinitely many octaves, because you can divide 10Hz by 2 infinitely without ever reaching zero. That means that any high pass filter - no matter how steep it’s rolloff is or what the cutoff frequency is - will in eventuell entirely remove the DC component.

There is a catch though: As the cutoff frequency goes lower and lower, it takes more and more time until the DC component is finally completely removed. So if you would apply a 0.001Hz high pass filter to a steady voltage, you’ll see that the filtered signal starts at that voltage and then slowly drops exponentially towards zero. It drops faster the higher you set the cutoff frequency. So in practice you’ll want to choose the high pass frequency as high as possible - just low enough to not remove any wanted frequencies from your signal.

So whatever your DAW offers as the lowest low cut frequency will do the job. If you like, you could “record” a steady voltage in your DAW and apply high pass filters with different frequencies to it and observe the results in the rendered audio.


If that sort of thing really others you, there are modules that have AC-coupled inputs such as Makenoise Rosie, which will remove static offsets without removing ultra low end.

1 Like


The recording DC idea is actually a pretty great idea.