Community member Joerg Witzsch has got in touch about a recent reverb related project he recently worked on, which was somewhat out of the ordinary and wanting to share it with the rest of the community. Over to you Joerg…

I recently recorded a live-performance of our "New York Voices" - like SATB-quartet "The Gershwinites" - just piano and four singers. For the Gershwin-classic "Embraceable You", we have a wonderful arrangement, that needs to be sung in a very "floating, dreamy" kind of way - as for the singing, voices need to blend perfectly, dynamics need not to be extreme and a major focus for us was, how to use our singing techniques to make the sound of our voices support the different parts of the composition.

When it came to mixing the recording, something major was missing for me - that dreamy quality we had in the concert hall. But just adding reverb (or different combinations of reverb plug-ins) did not do the trick. It just created too much of a noticeable room on the forte sung parts and it was missing the dreamyness in the softer parts.

So I remembered, how we approached our desired "live sound" in that very song: one critical part was using different singing techniques to alter the vocals' sounds to "imitate" dynamics, rather than singing a specific passage super-forte - we used ‘sound’ associated with dynamics more than ‘real dynamics’ themselves. Of course, we still had pianissimo and forte passages in the arrangement, but we did not fully expand to those dynamics.

What I ended up doing in my MIX, was the following…

1. Sending the vocals' stereo mix to an additional subgroup.

2. Reducing the dynamics of this very subgroup by using…

   1. Waves' Vocal Rider to get an averaged out, but still musical sounding "loudness", followed by…

   2. Two stages of compressors with different settings.

3. That way, I ended up with a vocal mix, that sounded dynamically pretty flat, uninspired, uninteresting, just flat out "undynamic".

4. The only thing, this subgroup-mix still had kept, was the different tonality of our voices I have spoken about above.

5. I then fed this sub-group's mix into a reverb and mixed the output with the main stereo vocal track.

6. I adjusted the output-level of the reverb to sound good with the pianissimo/mezzopiano-passages of the arrangement.

That trick provided, what had been missing when using "just reverb on the full dynamic main stereo vocal mix"…

1. When the singing was soft, it created just the needed amount of reverb and room to support that "floating, dreamy" kind of quality I wanted for the song.

2. In the forte-passages, it did NOT create those super-noticeable, long reverb-tails, that make the reverb stand out so obviously: since the reverb was fed with a majorly dynamic-reduced signal and adjusted to fit the pianissimo/mezzopiano passages, the forte-passages were not producing "more" reverb, but instead, sounded a bit "dryer" (the main stereo vocal mix was more prominent).

3. I fine-tuned this reverb-effect by adjusting the compressor's settings to reflect, how much reverb I wanted to allow in the louder passages.

4. So in the end, the reverb was supporting the softer passages of the song, but did not stand out in the louder passages.

This image above shows the various compressors, I have been working with.

Note the EQIII, which for me is important to filter out the "boom" of the performance venue.

Also note DBX 160: the compression settings are taken from two voices, tenor and mezzo, and show, that I had to cope with the dynamics of voices individually (they also "ring" differently in the reverb section, so compression settings at that stage becomes important, imho, before voices are fed to the reverb plugins)

Scheps Parallel Particles initially was just used as an experiment and turned out to be of help to give the voices just the right "sound" going into the reverb subgroup. This affects the "color" of the underlying "compressed" reverb sound.

Vocal Rider was directly used on the reverb subgroups, before the reverb plugins

The PuigChild 670 was sitting on the master-bus of the voices (just a little "glue")

The image above shows the reverb plug-ins I have been using. Initially, this seemed a bit of an overprocessing to me, but I was just experimenting and hit home on the reverb subgroup with this combination.

It is noteworthy, that this combination is not about creating a normal, nice room / reverb on the voices' ‘direct’ signal, but they are targeted to "smooth out" things on the reverb "compressed" subgroup. Each reverb plug-in sits on an individual subgroup, so I can individually control output levels and are not "chained", so they do not feed off each other.

More About Joerg Witzsch

He was born in 1962, is a german radio host for national radio SWR1 and he is very active on stage both as a performer in and director of musicals. For many years, he has also reviewed gear for German HiFi-magazine "stereoplay". His passion for music naturally expanded into some studio-work, recording and mixing.