Tracking Vocals Without Headphones: Audio Examples
The audio files available on this page accompany the article in SOS October 2015 about recording vocals in the control room, in front of the speakers and without headphones. As this technique inevitably results in a certain amount of bleed of the monitor mix into the vocal mic, we looked at various techniques that might help reduce this. As described in the article, we set up two microphones so they could capture takes at the same time. This is to illustrate the difference between using a dynamic or capacitor mic for the different techniques. Judging the quality of the vocal performances over the takes is obviously not a scientific process, as there were quite a few takes and thus plenty of variables, but you should be able to get a good idea of how the recording quality and speaker bleed were affected. All the vocal takes were recorded without compression, or any EQ, and recorded via Focusrite ISA pre–amps and Lynx Aurora A–D converters.
This is the instrumental track that formed the monitor mix to which all the vocal takes were performed.
As a reference, we recorded a vocal take with the speakers off using headphones, using a Shure SM7 dynamic microphone. As well as illustrating the qualities of the traditional spill–free vocal recording, it gives an idea of the space in which the vocals were recorded. Not completely dry, but with a controlled enough ambience, that should work well come mix time.
The same take but with a Neumann U87 capacitor microphone. Notice the different tonal character of the two mics, with the capacitor mic having a slightly brighter sound with a more pronounced ‘top end’. The SM7 has a touch more low–end.
We initially tried a few different playback levels within the room. This was the quieter option, and we could talk in the room fairly comfortably at this level. The spill is lower in level, of course, but the singer, Griff, found the monitor mix too quiet for him to ‘get into’ his performance.
This vocal was recorded with the speaker playback at the lower of the two levels, this time via the U87. The bleed is slightly more noticeable with this mic.
This is the louder playback option that we used for the rest of the session, captured by the SM7. You’d have had to talk fairly loudly to be heard at this playback level. There is more bleed than the quieter option, of course, but Griff felt this was the point at which he could sing comfortably and get into his performance without there being any foldback of his vocal part in the speakers.
The louder playback option captured via the U87 mic. Notice the different level of bleed than with the SM7.
Ensuring the playback level and mic placement was exactly the same as with the vocal recording, we re–recorded the instrumental track back though the mics in the room — this could then be polarity inverted to cancel the spill on the vocal takes very effectively.
The results of the same technique but with the U87. Notice how, although the level of the low end and mid-range is reduced, the top end has not been so successfully cancelled out.
The technique of using EQ on the playback to eliminate the spill from the top end of the frequency spectrum. An EQ with a low–pass filter set at 8.5kHz was used on the master bus. Although this compromised the playback recording in the room slightly, it meant the top–end ‘air’ section of the vocal was now spill–free.
The results of the same EQ technique but with the U87 capacitor mic. Preserving that top–end part of the vocal should be even more beneficial with this microphone. Notice how the snare is the most prominent element remaining in the ‘bleed’ from the speakers now.
We used the EQ technique combined with the earlier cancellation technique. This seemed to provide a good reduction in spill from the critical areas of the vocal take. Notice how the snare is still fairly audible, however.
The results of the same technique, this time using the U87 recording.
In an attempt to counter the snare spill, we used another EQ setting, combining the low–pass filter with a 3–4dB scoop at around 5kHz. Notice how the snare is noticeably less prominent in the spill.
The same technique with the U87.
The second EQ setting, combined with the technique of inverting the polarity of the re–recorded instrumental. The effect of the spill should be fairly negligible now and the snare is no longer prominent.
The same technique with the U87.
Using the U87 in figure–of–eight mode, the mic was positioned so that the side nulls of the polar pattern were pointing at the speakers. Notice how the tone of the vocal is different with this pattern. The spill is at a fairly manageable level however, even without any EQ or cancellation techniques.
The results of using the cancellation technique with the figure–of–eight setup. There is a small but noticeable reduction in spill.
We tried the same technique with two singers positioned either side of the mic. As the artists we were working with like to sing together sometimes we tried the technique in this setting. The spill seems manageable and they really enjoyed singing together like this.
To look at the effect of having some of the singers’ vocal being fed through the speakers, with the instrumental playback, we set the U87 to a cardioid pattern, with the singers looking straight at the speakers. With the mic signal folded back through the speakers, we raised it to the sort of level that a singer might want when asking to hear more of their vocal in their cue mix. (Obviously, we were careful to avoid feedback!)
The effects of the cancellation technique with this recording. At the louder playback level, the cancellation effect is quite pronounced, but notice how the vocal itself is now being affected, not just the instrumental.
We recorded a final take with Griff using a handheld SM58 and encouraged him to move about a bit. The level of bleed is very low, due to the close proximity of Griff’s mouth to the mic, but the quality of the vocal take captured is relatively poor. This might be improved by using a higher–quality handheld condenser mic.