twilightsong wrote:One thing that will definitely improve a vocal track is to try and isolate the mic from room reflections when recording.
As described in Michael Paul Stavrou's 'Mixing with your mind", we used our mics with figure-8 patterns so successfully that we could almost treat each track as a separate recording for stereo placement purposes, so good was the isolation.
We built our studio with 200mm acoustic fibre walls so that there is little reflection, but the condenser mics can pick up the 'atmosphere' around each instrument. We can then place them in a stereo field, and add REVerence (what a nice warm reverb) to build the room in which the whole 'performance' takes place.
Michael's book, being one of the most practical and empowering for aspiring recording engineers, also highlighted that vocalists rarely sing out the front of their mouths, but instead bias towards one side, which is determined by using his technique of moving a mic around the vocalist's head and listening for the position with the best timbre. The bias tends to be greater the older one gets. Other benefits of having the mic off axis is that it tends to avoid plosives which DO come out the front, and allows the face to be more clearly seen, which is good for YouTube videos. Look at the vocal mic placement in our video for an example.
He also describes:
a) a simple technique for finding optimum speaker placement to avoid room problems, with just your ears.
b) a fantastic compression-reverb trick that enhances vocals the louder they get. This is used on my wife's voice in the video, applied to a EMT140 plate reverb only 30 degrees to either side of centre.
The book is great, and available from http://www.mixingwithyourmind.com
. 'Stav' as he is known, has worked with all the greats in London's top studios, but is now back in Australia and based in Sydney. I once asked him some questions by email, in response to which he arranged to meet with me to 'set me straight', but not before almost making me sign an NDA!
A couple of tricks I have for handling problems in the recorded vocals are:
These have the curious property of making the whole stereo sound field sound like it is swaying in the wind. The waveform looks like it has a very low frequency at the start, slowly petering out. The simple fix I have is to just drop, but not silence, the front part of the wave form to reduce the first few 10ms gradually increasing to normal level at the end of the low freq pattern. I created an Envelope preset which I apply to a selection covering the two points above. You may have experiment and undo until you get the right sound, capturing the power of the song sound-wave, but avoiding the leading sledge-hammer.
It can be annoying, but I find it is only so if it increases to become very noticeable. If it is made to look more like a mesa (flat-top mountain) of about a third of its peak, it seems very natural, and is not overtly noticeable. Again, an Envelope preset with just the right amount of dip in the middle, applied to a selection that starts and ends about a third the way up. It seems a whole lot better to treat the few that occur this way, than apply a dynamics plugin across everything, which is sure to have some unintended artifacts!
c) Lip smacks and mouth noises:
RX, especially RX3, has its Spectral Repair module that makes short work of these. Just use the Replace tab, open to maximum pre/post width, select just outside the thin vertical line that these look like, and Process. Voila!
Also makes short work of guitar string touches, which are wider than the mouth noises.
Generally, I apply these tasks permanently to a copy of the file -- first two in the Cubase sample editor online (and Apply Edits...) and the third in RX offline -- than audition it in the mix instead of the other, and if OK, remain using it instead . I still have the original if I need it.
Hope these help.