Experimentation in Reverberation
To remain consistent in our experiments, a Rode NT2 was used in all experiments. An omni-directional polar pattern was used to capture the characteristics of the space we were recording except for our dry source recording, which used a cardioid polar pattern.
Euphonium Reverb
Euph Reverb Showreel by firemanfloaf
Experiment 1- Studio 160cm from source.
Experiment 1 Experiment 2
From this distance we can paint a pretty accurate picture of what the space is. The studio is custom built to absorb as much sound as possible, resulting in a short pre-delay time and minimal early reflections. Dense reflections are barely audible.
Experiment 2 Studio 296cm.
At a significantly further distance, the essence of room is captured more successfully. However, we soon notice that the timbre of the Euphonium is jeopardised to a great degree. It shows us that when distance miking, we need to find the right balance between articulation and resonance.
Experiment 3- Studio dry 0.2m.
Experiment 3 - Dry
The reason for this was to capture the euphonium without any sonic reflections so we could then apply it to our various reverb strips.
Experiment 4 dry with Logic's PlatinumVerb.
This old-school style of reverb application utilises no impulse responses whatsoever and is in fact, merely a glorified synthesizer. Nonetheless, when added to our dry Euphonium track, it is clear that a substantial amount of adjustment is needed. Even with our bus assigned to post-fader, the reverberation effect appears continuous, eradicating the realism of a particular room, as reverbs main purpose should be to create an image of a real space. Despite our best efforts, even with subtle application, PlatinumVerb comes across too synthetic and forced, creating a metallic sound in some areas.
Experiment 5. Dry with Cafe Space algorithm.
Because we are using a dry sound that has been bussed to a reverb channel, we get the clarity and detail of the instrument. With the reverb applied in a separate channel, we get the happy medium we were missing from experiments 1 and 2.
Experiment 6- actual cafe space. 3.0m.
Experiment 6
The visualisation of the space is more obvious than our previous experiment. However, the room is too reverberant for the instrument with regards to resonance, something that we have no control over. The slightest of tweaking would have sufficed in this instance.
These results were nearly identical when replicated in the recital room and foyer hallway. To my ears, it is better to record a solo instrument dry, and then apply the reverb subtly in a different channel. This allows for automated tweaking during mixing, something that cannot be done in a real space. There are of course, disadvantages for this application of reverb. As our dry source is close miked, we pick up all sorts of unwanted sound. For example, on a Euphhonium we pick up the sounds as the fingers press down on the valves, the player breathing etc. On guitar we pick up fret noises and scuffeles from the player. This would not happen if we recorded at distance. Also, it is easier to hide mistakes in the actual space, whereas with a dry sound bussed , the detail is more clear and not masked with the real acoustics.
Perhaps another way to record would be to close mic in the recital room with an ambience mic at around 6m. This would however cause phase issues so in essence, the best method is to bus. It's easier, controllable and above all, clearer.
Similar tests on vocals
Vocals Reverb Showreel by firemanfloaf
Similar test on nylon guitar
Guitar Reverb Showreel by firemanfloaf
Tweaking the parameters
Experimenting with reverb
Vocals Reverb Showreel by firemanfloaf
To prepare the reverb effect, an auxiliary channel was created with the Logic’s space designer plug-in. The original source was then bussed to that auxiliary strip. For demonstration purposes I have selected a bright hall preset, not because of the name, but because I have used it in the past and found it appeals to many of my recordings. The reverberation toggle was set to full (0.0db) and the dry toggle was set to mute. This was in order to isolate the reverberation and so we would not hear an increase in track volume, thus keeping the reverb effect separate from the track. Had we just applied the space designer setting to the original track, we would have just heard a straight manipulation of the track , which in turn would have produced some disastrous results. So, in theory it is better to bus to an auxiliary.
After much experimentation, it was clear only a small amount of gain was needed on the bus toggle (c20%). The next thing to be done was to experiment with the reverb length. However, it was not an important focus in this case as we are only applying it to a single track . Had this been a complete mix with multiple instruments, more consideration would have to be given to make sure the reverb length did not jeopardise the mix. Still, after sweeping through various reverb lengths, 1.436s seemed to suit this particular track as it provided the track with warmth and sounded natural without effecting the timbre of the original track.
On the auxiliary strip, EQ was applied to reduce some of the higher frequencies as with most sonic reflections, the higher frequencies die out, leaving the resonance in the lower frequencies. This enhanced the richness of the room and slightly adjusted the characteristics of the space.
The spectrum analyser within the EQ was used to monitor the reverb, paying close attention to any prominent frequency- response peaks as they would make the whole effect too obvious.
Convolution and creating impulse responses
In order to compare our tracks recorded in a real environment with their identical synthesised reverbs , impulse responses were taken from the Foyle Arts, cafe space, foyer hallway and recital room.
A singular Genelec HT208 monitor was placed in each environment, with a Singular Rode NT2 (omni-directional) placed approximately 3.0m away in order to pick up the room's characteristics along with our generated sound. A sine sweep was then played through the monitor sweeping from around 10Hz - 20kHz. These impulse responses were then deconvolved using Apple's Impulse Response Utility. These impulse responses have been applied to our 3 test instruments (Voice,Euphonium and Acoustic Nylon Guitar).
A summary of each space's characteristics
The Recital Room has a high ceiling and wide walls, resulting in a longer pre-delay time. This space was the most stentorian of the 3. Although there were many small obstructions in the room (e.g. chairs, piano), the identity of the hall is prominent, ideal if we wanted to create that specific environment.
Cafe Space
The laminate wood flooring of the space, coupled with PVC walls produced the earliest reflections. The weaker reflected sound waves combine with the pre-delay onset to produce an ugly 'glass bowl' effect.
Foyer Hallway.
This was the most intriguing out of the 3 spaces. Not only is it a long corridor, it also has steps to the side which lead to additional square space but confusingly has a high ceiling. This sonic confusion is evident in the impulse response. Upon listening to the auxiliary reverb on solo, it is evident that the early reflections combine with, albeit very subtle, late reflections from this additional space. Although this prominently dense space is intriguing to the ear, with regards to our test instruments, it would not fit into any of our exhibits.
