Group recording session

Vocal/piano duet recording session

#

Recording Session Information

#

The music was performed by soprano singer Jessica Baldry and pianist Toby Baker. They performed Me Llaman la primorosa from Rossini’s El Barbero de Sevilla, and Faure’s Les Roses d’Ispahan. The recording session took place in the Elgar Concert Hall and ran for 40:44. It was recorded on a Zoom F8, using an LCR mic setup for the piano, a stage A/B pair for reverb, and an A/B pair and spot mic for the vocals.

Ultimately, I decided to not use the vocal A/B pair, as I thought the stereo pair didn’t fit as well within the frequency spectrum of the piano as the spot mic did. The stage A/B pair did an excellent job capturing the natural hall reverb, which is the primary reverb used in this recording session. The recording was edited in REAPER.

Microphones

#

The microphones used were as follows:

The AKG C414 XLS II capsules are quite hard to align due to the shape of the protective mesh, and its footprint can be quite large, but it makes up for these shortcomings by being very versatile because of its alterable pickup pattern, allowing it to be used in many different setups.

Edit Locations

#

Below is a table for each piece detailing the edit locations for the final comp.

Annotated scores

#

Below are copies of the scores for each piece, annotated with the respective edit points.

Issues that couldn’t be resolved

#

Despite our best attempts there were unfortunately some timing issues that couldn’t be resolved in the recording session. Below are tables for each piece detailing these.

Further Editing Details

#

I used two REAPER sessions to edit the recording, first a MIX session for mixing the recordings and adjusting the levels and EQs of individual mics. I then imported this as a sub-project into an EDIT session where the various takes where comped together to fix any performance errors.

Mix Session

#

This is the mixer of the MIX session. All the individual tracks are included in track folders to allow easier editing of EQs for each microphone pair. Note the VOCAL AB wasn’t used in the end, so it’s muted. Each track is hard panned to its microphone respective direction. Most of the levels are left at their default values, with a slight reduction in the STAGE AB volume, and the CENTER channel of the PIANO LCR, to allow space for the VOCAL SPOT in the centre of the mix.

 An EQ with a high pass filter was applied to the **STAGE AB** folder at about **250 Hz** to reduce the muddiness of the reverb. The top end is also cut slightly to reduce harshness when stacked with the other mic pairs.  Again, a small high pass filter is applied to the **PIANO LCR** and **VOCAL SPOT** to reduce the muddiness of the piano, in both the piano microphones, and the audio bleed in the vocal microphone.  The **VOCAL SPOT** was placed slightly too close to the singer, and as such picked up quite a lot of mouth noise and had some harsh plosives and breaths. To reduce this, I briefly reduce the volume of the vocal spot during rests using volume automation as pictured. Below is a table of the edits in the **MIX** session where I reduced the vocal spot’s volume.

A small amount of compression was also added to the VOCAL SPOT to reduce the level of the louder sections, especially in Me Llaman la primorosa. I used ReaComp compressor and found a threshold of -17.6dB and a ratio of 5.59:1 worked quite well, without over compressing the vocals.

Beginning at 15.16, 33.22 and 38.49 in the MIX session I also periodically lowered the volume of the PIANO LCR slightly, to better balance it with the voice.

Edit Session

#

Importing the MIX session into the EDIT session as a sub-project renders the MIX session down to a single track, which allows much easier editing and keeping track of takes. I make sure to set a unique colour for each different take, so that it was easy to tell where edits are made and which takes I used.

Both pieces were edited on their own tracks, with their own REVERB buses to allow both pieces to have different reverb settings.

The splices for the comp happen right before the beginning of a note, with very small crossfades to mask the edit point. A big challenge with this edit was working around large vocal phrases, and piano pedal changes that carry over the end of a phrase.

The volume levels were left at their default values, because all the mixing had been done in the MIX session. The levels of the REVERB buses are adjusted respectively to mix the artificial reverb with the natural hall reverb.

For Me Llaman la primorosa I used two different EQs, before and after the artificial reverb, to adjust the sound of the reverb track. I boosted the lows and cut the highs of the reverb both before and after. I decided to use a little bit of artificial reverb to compliment the already exellent hall reverb.

I decided to use a smaller amount of artificial reverb due to the speed of the piece, in order to keep the clarity of the music. I used an impulse response based reverb, with a decay time of 2.5s, and applied -18dB of gain inside of ReaVerb.

For Les Roses d’Ispahan I wanted to go for a more ethereal sound, so I used a long 3.9s impulse response, again with -18dB of gain, but the level of the reverb is higher in the final mix. I used an EQ to cut some of the high end before the reverb, so as to keep the clarity of the vocal material in the mix, without too much reverb on it.

Rendering

#

Finally, I rendered the piece to a single WAV file, using Time selection bounds, 48khz Sample rate and 24bit PCM bit depth.

Reply by Email