Quote History Quoted:
I would really hesitate on this meter. I looked at the instruction manual. There appears to be 3 detectors: slow, fast, and impulse. While it says it will record peaks, there does not seem to be an actual peak detector. At best, the impulse detector has a response/rise time of 35 MILLIseconds, while for peaks, you need a rise/response time less than 20 MICROseconds. You need to query the manufacturer about the peak rise/response time.
I think it is going to read around 30 dB low. See if you can get a demo with firearms. At 1 meter left of the muzzle with A or C weighting, an M4 carbine should be 164-165 dB, a Ruger 10/22 around 141, a Beretta M9 around 160-162, etc. Unweighted, look at about 5 dB higher.
You need a true peak detector if measuring firearms. A meter without a peak detector may be perfect for accurately measuring commoner sound levels, but not for really short duration peaks. One of the less expensive alternatives that is closer to being right is the Larson-Davis LxT1-QPR with a rise/response time of under 30 microseconds. With A or C weighting, it is close enough at 1-2 dB low (worse unweighted, but weighting is what the industry uses and MIL-STD 1474D permits).
Just my opinion, based on a number of mistakes I made many years ago.
View Quote
Peak is no time weighting.
Therefore rise time is 0second.
Time weighting is made by LRC circuit.
Peak does not pass these circuit.
Peak detect wave form directly.
For clarification
For instance sample rate 48 kHz [20.8µs]
20.8µs is sampling frequency before root-mean-square.
Sound level meter takes in wave form every 20.8µs.
And Leq, Lmax, Lman, Lpeak are operated before they pass root-mean-square circuit.
Block diagram for a sound level meter
Time weighting and root mean square detection circuit