An impulse response measurement can offer insights into a specific question: “If an impulse is sent into the input, what will be the output?” Because of this, these measurements are commonly used to analyze characteristics from audio electronics or loudspeakers. For more information on what an impulse response is, read TüN® 4 & MAX™: What is an Impulse Response?.
Even though electricity travels incredibly fast, electrical signals take time to pass through devices. For example, if an audio signal is passed through an amplifier, the time it takes for the amplifier to amplify the audio signal is not instant. There are several different components inside of an amplifier that the audio signal must pass through before it reaches the output stage. That time is referred to as ‘latency’. Latency can be measured and quantified by measuring the impulse response of a device. All electronics will have some amount of latency. This is normal and should be expected.
In some cases, when attempting to measure a device, the impulse response may drift across the graph. This can sometimes happen when the latency of the device is unstable and changing constantly during the measurement. This can happen for several reasons and in many cases, it is unavoidable. The user should confirm that the measure and reference signals are properly connected before proceeding.
In cases where the drift is originating from the device, it is best to attempt other measurement techniques that can be used to acquire meaningful data, such as viewing an RTA and summing channels to see how they interact, or using an open loop measurement to acquire useful phase data to understand exactly what type of filters may be present. In some cases, one of the channels of the device can be used as a reference signal and a useful impulse response can be generated that way.