Microphone Frequency Response

[pfsmith0]

I measured the frequency response of the mic available in the HD500 and attached a PDF of the results here. I also have all the data in an excel file here. I used the Class A-15 adjusted for flat frequency response and the 1x12 Blackface 'Lux cab. You can clearly see the extended low frequency response of the ribbon mics and the hi-mid enhancement of the 57. I'll be using these results to measure what the can DEP parameters are doing but thought I'd share my data here in case anyone wants it.

 

-------------------------

Sorry, the PDF file didn't show up so I attached it again.

Mic Frequency Response.pdf

[wilson69978]

Hello pfsmithO:

 

Wow, that looks so technical. I have 3 questions my friend,

 

1. what equipment or programs etc did you use to measure all this. ( just very curious how it is done, not questioning the results)

 

2. What do the x and y axis actually measure, one looks like frequency, what is the other?

 

3. How does one use this data to his or her advantage in patch creation, or in recording , mixing etc?

[MartinDorr]

Great, really appreciate this work.

 

One question: Have you ever measured whether the D/A conversion for the HD's analog outputs have any significant impact to the signals when compared to SPDIF?

 

In regard to the need to use a really low input signal to keep the HD amps and models in a linear operation mode I found that if I reduce Channel Volume so that peak signal levels stay below -12dBFS, all HD models I ever measured (all preamps and somme EQs and compresser/boost models) stay in the linear operation range. If peak signals get into the -9dBFS and higher range the models start to compress, soft clip, and finally distort.

For peak model output signal levels starting significantly above -12dBFS the increasing of gain on Channel volume (or Master Volume, mixer, and/or volume pedal) will change tone rather than just signal level. Not sure you can confirm or have observed this effect (I believe I saw reduced level increases in your EQ curves in the higher signal level ranges).

Assuming this observation is correct I think it can be used to raise the input signal to a level that more closely reflects what a guitar would provide (and changing to channel volume so that the amp stays in linear operation range where HD signal level adjustment dials do not change tone).

 

Martin

[pfsmith0]

Gosh, I never answered these! Sorry...

 

I used an Audio Precision APx-525 to measure this. It's an excellent very high-end piece of audio test equipment. We have about 7 of these at work, each about $25k. I plugged the generator output into the HD500 (Guitar, aux, mic depending on what I wanted to measure) and routed the SPDIF back into the APx (unless I was measuring the output levels at the various connections).

 

I did measure the clip levels of the various inputs and outputs here but it is hard to tell if the clip levels are due to the converters or the supporting analog circuitry. It appears a -30dBV guitar input gives a -39.3dBFS SPDIF level. This means the ADC will clip at +9.3dBV. But looking at the outputs, they all clip at 8.3Vpp, which is also 9.3dBV, So the clip levels seem to be due to the ADCs, not the DACs.

 

I did not do any measurements on tonal changes as the input gets close to fullscale.

 

X axis = frequency. Treble is toward the right. Bass is toward the left.

Y axis = magnitude (aka amplitude aka loudness). The higher the graph the louder that frequency is.

 

The main use for these particular curves is to give you a broad feel for what the various mics do. For example, the 57 will be somewhat bright because it has higher levels on the right side of the graphs. The 409 will give you great bass since it has a higher response toward the left.