I’ve been aware that there is such a thing as phasing problems and that they can arise when recording with more than one mic. Phasing problems are so that the more mics you use, the more aware you need to be. In my recordings, the number of mics has increased over time so it was time for a quick peek at this potential obstacle.
What’s a phasing problem?
I only know about these problems in theory, which (in naive, simple terms) is that a sound can be described as a wave with a top and a bottom. A single wave can’t have a phasing problem with itself, but lets say you get the same exact sound from two sources; the waves from each source could meet. What happens when they meet?
If the two sound waves are perfectly aligned (tops align with tops, bottoms with bottoms), then the sound would be amplified. If the sound waves are perfectly out of sync (tops align with bottoms), the waves would cancel eachother out.
Such cancellation has its uses, for example in quieting headphones which cancel out background noise from the room. They do so by listening to incoming background noise and then playing the same sound with inverted phase, thus cancelling it out. Phase cancellation has also been used to generate karaoke tracks, by “phasing out” the vocals on a recording.
For microphones, you can think of it like this: A condenser microphone has a diaphragm which moves back and forth when hit by sound. If you have two microphones, you wanna make sure that the diaphragm moves similarly for both mics. If the backwards movements of the diaphragm of one mic coincide with the forwards movements on another mic, playing back the sound of each mic together would likely cause some phasing cancellation or phasing problems. We don’t want that.
There are some basic tips on how to avoid phasing problems. The most basic ones seem to be these:
- Reduce the number of mics
- Make sure the distance from mic to source is the same for all mics
- OR – 3:1 ratio mic distance. If mic 1 has distance A from the source, you can place mic 2 at a distance of 3 x A away from mic 1 and it should be okay (unsure of how relevant this is for guitar recording, but I guess one mic pointed at the sound hole and another off the side in the neck area)
- When screening for phasing (monitoring through headphones f.ex), listen to the sound in mono (stereo alleviates phasing problems).
- If two sources are perfectly out of phase, the problem can be perfectly remedied by inverting the phase of one of the sources. Sometimes there will be a button or switch to do this on a mic and you can often do it in your DAW.
Discovering phasing problems
Phasing problems cause reductions in volume. When listening to stereo, phasing problems tend to remove sounds centered towards the middle of the stereo image. Sometimes, reverbs or delays can also cause phasing problems, for example that the sound of a guitar gets cancelled out so you mostly just hear the reverb.
But how about more subtle phasing problems that are hard to hear? Is there a way to find these?
I made an attempt at this and it seems quite simple to visually screen for phasing problems if you use a digital audio workstation (DAW). I use Reaper and after recording a single open A string on my guitar with several mics, I could check the wave files in the software and look for any out-of-syncedness that might have occurred.
I had made a new, simple mic rig with two stereo setups. At the bottom, closest to the guitar, I had a couple of Røde NT1-As set up as a stereo-pair. At the top, I had a matched pair of Røde NT5s. The rig pretty much satisfies the recommended 3:1 ratio where the distance between the NT5 pair and the NT1-A pair is about three times that of the distance between the NT1-As and the guitar.
BUT, I also threw my GM10 in there for a 5 mic total. The GM10 was rigged at relative close proximity from the strings and was my main suspect for phasing problems.
In reaper, the sound of my guitar’s A string looked like this:
When looking at the above image, see the first red line. It overlays wave tops on all the top four mics (NT1-As and NT5), but not the bottom one (GM10). As suspected, the GM10 is out of phase. Not completely, but still.
Is it audible? I don’t even know, but I wanted to try and fix the problem. I simply repositioned the GM10 back a little, maybe a couple of centimetres or so. Then I recorded an open A again and the problem seems to have improved quite a bit.
Did it matter much? Unfortunately, I don’t know, because right after this, my daughter’s full diaper demanded my attention and that was the end of my phasing experiment. However, I discovered the following:
- The NT1-A and NT5 mics are well aligned in my rig
- It was easy to screen for the problem visually using Reaper
- The phasing problem behaved predictably; I thought it would be the GM10 – and it was
To be on the safe side, I’ll probably do this in the future.
Perhaps someone out there knows how big a deal a phasing problem like what I first had with my GM10 is? Feel free to leave me a comment.