Hybrid Reverb: Understand the Impulse-Response Tail

What the name does, and does not, promise

An impulse response, or IR, describes how a system responds over time. Convolution applies that response to the source, so the chosen IR supplies both an initial pattern and a tail. The implemented controls here are Category, Impulse Response, Predelay, Decay and Dry/Wet. Category selects a group of responses; it is not an algorithm switch. There is no Size, Damping, EQ or algorithm blending section in this version.

That matters if you arrived looking for IR-plus-algorithm interactions. You cannot change an algorithm behind this IR or compare serial and parallel engine routing in this panel. The separate Reverb device is algorithmic and provides a different set of decisions. This lesson instead asks which qualities come from the chosen response, and which can be changed by the three available continuous controls.

Hold the response still; change its envelope

The example uses an original procedurally synthesized teaching-room response and an original eight-second synthetic pluck with rests. Neither is a recording of a real hall or performance. No uploaded recording is required. Predelay starts at 20 ms, Decay at 0.75× and Dry/Wet at 30%. Follow the space after each pluck rather than judging only the loud first instant.

Set Decay to 1.00×, then to 0.50×, keeping the response and Dry/Wet unchanged. The lower value applies a steeper falling envelope to the IR. Listen for less late energy while its underlying reflection timings remain in place. Release the Decay knob to commit the change: the device prepares a new response buffer, rather than continuously rebuilding it during the drag.

Decay is an envelope multiplier, not a seconds readout and not an IR time-stretch. A value of 0.50× does not move every reflection to half its time, nor guarantee an exactly halved measured decay in an irregular response. The algorithm shapes sample amplitudes within the existing file duration. It cannot invent missing reflections or extend the file beyond its end. Use a different response when the pattern itself is wrong.

The long setting has a cost

Try 1.20× at a reduced listening level, then return to 0.75×. Values above one lift later parts of the existing response. That can also lift low-level tail noise and make the wet signal louder. If the apparent improvement disappears when you lower Dry/Wet to a comparable level, you may prefer the shorter setting. Avoid treating the 2.00× limit as a target.

Predelay offers a separate experiment. Compare 0 and 35 ms with Decay restored to 0.75×. Only the wet path is delayed; the direct pluck stays in place. Listen for a cleaner leading edge, then for a distracting gap if you increase it further. Shorten Predelay when the room seems detached, and shorten Decay when the next note disappears under the previous tail. Lowering Dry/Wet simply reduces the surrounding response relative to the direct signal.

Know when to choose the other reverb

This embedded lesson restricts the response choices to original teaching material. It does not expose the normal studio library or require a third-party hall recording. Return to Decay 1.00× to hear the teaching response without the decay-envelope adjustment. Its deliberately generated pattern demonstrates convolution, not a guaranteed reconstruction of an actual acoustic space.

Choose this approach when an existing response has the useful pattern and you mainly need onset, tail emphasis and balance. Read the Reverb lesson when you need independent reflection levels, room-network size or frequency-dependent damping. Those are real algorithmic controls there, not hidden parameters here. Comparing the two devices is useful; pretending this panel already combines both engines is not.