Missing frequency on Shure sm7b?

[TalismanRich]

I set up my Tascam to read the preamp with gain turned up full, and recorded the noise. Normalizing took 58dB of boost. Here was the frequency response of the noise in Reaper.

I also checked the noise is a spectrograph. This was with 12dB gain boost (at 58dB it was just a mass of green).


I don't see any notches or irregularity in either.

Finally, I took an acoustic guitar recording made with a Sennheiser e935 and looked at that.
The band just over half way up is from 26k to 30kHz. I did the recording at 88k/24bit. There was nothing at 6K to speak of.

 

[jimmys69]

Not sure what all this means...

Pink noise recorded from my monitors into Cubase via Steinberg UR824 interface. SM7b direct and through 'cloudlifter' Klark CM-1. Same with SM57 with and without cloudlifter.

All normalized to 00db.

Not sure if I got everything right, but here it is.

 

[rob aylestone]

It’s in the audio drop down. A simple line graph of frequency vs strength. This is quite interesting. Jimmy also has these darker bands that start quite sharply. We don’t hear this, but does this really mean at a certain frequency the response suddenly drops? The frequency response curve published does not do this, so is it the interface or some anomaly of the DA process responding sharply to a gentle change in curve? I’ve got a mixed project on today in both of my studios, so I’ll grab the pink noise and record it in my sound studio on a few mics, using the presonus interface I use here, then repeat the test in the video studio when I’m there this afternoon.

Im not sure if I’m chasing something irrelevant, or important to know. I have always played down the role of the audio interface, as I just want low noise but maybe these bands are actually room resonance? A mic in front of a speaker making the roasting noise I have always taken as so loud and close, room effects would be minimal, but maybe not?

I don’t think we’re going anywhere with this but why would one mic design produce repeatable but distinctly odd results?

 

[ecc83]

Can someone send a 16bit* .wav of the SM7b please? Listening to noise would be best but jusr 'air' noise would do. Then I can stuff that into RMA and get a clean graph. And I mean a .wav FILE I can download, not something I need to record K.I.S.Sirs

Idea: The 7b has an LCR filter on board. I wonder if there is an unsuspected resonance happening?

*16 bit because RMA cannot handle 24bits. I can convert in Samplitude Pro x 6 but would rather not process the signal further? Bad science.

Dave.

 

[rob aylestone]

here is pink noise from studio monitors 48K, 16bit on the audio studio presonus, and the video studio tascam (which also has smaller 5@ monitors.

I'm now more certain the 'issue' is with the Tascam interface - as repeating the same test, using the same noise file on the two interfaces with the SM7B produces very different effects, and one result is simply bizarre. I'll post that shortly.

This forum doesn't like wavs - so the two files are here.

The image below shows the same mic at my two locations - the left side is the SM7B in the audio studio with Presonus Firepod (which I really like) and the right side shows the same mic in the video studio - which has a Tascam 1641 with RCF 5" monitors - which I use for video editing. The microphone was in front of the speakers, and recorded the pink noise and sweep tone from the same files on the server. The graph is from Cubase, installed on both systems. The presonus Cubase spectrum analyser trace is top left, the tascam vesion top right.

The spectrum analysis tool in audition shows a nice distribution on the presonus, and the horizontal bands on the tascam at certain frequencies.

The bizarre result at the bottom is a simple 20-20K sweep, as recorded by the SM7B, but while the presonus one looks (and sounded) normal, it was obvious when I played the same file on the tascam something was very wrong. The tone went up, but then another appeared, going down, and then repeated - the mic captured what came out of the speakers. The single frequency became multiples that interacted with each other. I have never heard this in my life before.
I've attached it as an mp3 (you can hear the computer chime in the background so please ignore this.)
My thoughts are that the interface is a problem - either by design or fault, but on normal audio tracks, it still sounds great.
View attachment sm7b tascam 20-20k sweep_01.mp3

 

[ecc83]

Here are the spectra for the two interfaces. Assuming the mic was not moved during the tests the Tascam plot has more hum and so I would guess a poorer common mode rejection ratio? Otherwise I can see little problem with them.

Dave

 

[rob aylestone]

Cheers Dave - the Tascam does have that little notch between 6 and 7K, as the OP noticed. Very strange. Any suggestions/comments on the really odd response to the sweep tone? Never heard that before?