Skip to main content

Reducing noisefloor before going digital (technical)

Hi I have a dynamic omni mic and a preamp. The mic is rated 500 ohm impedance and the preamp input is 10k ohm. I read on a website it is ideal for a mic/pre to have a ratio of 1/10 impedances but my setup is 1/20. Im just trying to figure out ways to lower the noise floor before recording as my current noise floor is workable (-51dB) but still a pain when tracking and compressing.

If I increase the impedance the pre sees to 1k ohm is it possible to lower the noise? I tried a test on my preamp, by recording it's input into my wave editor, but I didn't plug a mic into the pre, I only completed its circuit with a variable resistor. I noticed that at 10k ohms the noise floor was about -48db, but turning the resistor all the way open to 0 ohm the noise completely disappeared, no signal at all. What can I do to lower the noise without having to use the digital FFT noise reductions?


camsr Wed, 12/13/2006 - 22:00
I got a tip from my electronics instructor today, he said connect the hot side of the pre input to the common ground and then adjust the pres volume and watch it for internal noise. I guess this will work for my unbalanced input. He also explained to me the value of a balanced connection when running super long cables.

Kev Wed, 12/13/2006 - 22:29
also try 150 - 200 ohms between 2 and 3
and let the input float against ground

fully floating differential input

look back at the John Hardy specs and think about the methods of testing

cables don't have to be that long for it to be worth while
you could be in a noisy environment

balance lines and Common Mode Rejection is a good thing for small signals

often people do like to use unbalanced were distance is short
but the signal levels are much greater than that of Mic signals
Mastering Suites are often single ended
and HiFi is single ended ... even the most expensive

camsr Wed, 12/20/2006 - 16:57
Im getting a pair of headphones for xmas, Im looking forward to either the Grado SR 60 or maybe the Beyerdynamic DT 231. Fits my price range and according to they both have a desireable response in the bass.

About the Shures, are they balanced? Also are they condensers or dynamics?

Kev Thu, 12/07/2006 - 12:00
get a real mic and a real mic-pre ?

what are you using ?

no matter where the noise floor is, there is still the issue of signal to noise ratio and it may be that your gain structure is all wrong or the combination you have for what you are recording could be better

If I increase the impedance the pre sees to 1k ohm is it possible to lower the noise?
and the way you have described will also lower the signal level

camsr Thu, 12/14/2006 - 01:55
Kev wrote: also try 150 - 200 ohms between 2 and 3
and let the input float against ground

What do you mean "between 2 and 3"? I dont know where I would connect to common ground from the tape deck cause the case is non-conductive, and the plug for the deck itself is only a two pronger. And I think the outlet it's plugged into is the only one in the room thats on a different circuit. Did he mean connect the input to the ground of an outlet?

Kev Sat, 12/09/2006 - 01:02
camsr wrote: .. these imaginary preamps and mics really suck
yeah funny
and so do toy mics and toy mic-pres
that some manufacturers try to pass of as pieces of real sound equipment

... Frankly I didnt understand a word ...
perhaps you don't

read the specs
and the terms of reference

-51dB (see above) is a relative term but you haven't specified ... relative to what ?

E.I.N., 20kHz bandwidth, unweighted RMS, 600 ohm load 150 Ohm source -128.7 dBu
dBu ... the u specifies the reference
dBm and dbV
and many more ways to use the term dB(ref)


Frequency Response, ref: 1kHz 20Hz -0.05 typ, -0.1 max dB
here they specify the reference ... ( ref 1kHz)

The John Hardy M990 Mic Preamp is a real Microphone Preamp.

let's move on

... but turning the resistor all the way open to 0 ohm the noise completely, no signal at all.
so you shorted the input
no signal and no noise to speak of
why doesn't that surprise me ?

I tried a test on my preamp, by recording it's input into my wave editor, but I didn't plug a mic into the pre, I only completed its circuit with a variable resistor.
so you recorded the input to the pre-amp ?
you recorded the output of the pre-amp with no source except the variable resistor
almost a worthwhile test

lets refer to the John Hardy specs
E.I.N., 20kHz bandwidth, unweighted RMS, 600 ohm load 150 Ohm source -128.7 dBu
IF you had of set the variable resistor to 150 ohms you could have compared your result with the John Hardy

you would need to know what load your computer audio interface presents to the Mic-pre output
and that should be relatively high for it to be dBu
the note at the bottom of the page
Unless otherwise specified : 25°C ambient, 150 ohm source, 600 ohm load.

confused yet ?
this may help

which microphone and which microphone pre-amp are you using ?
... then we can talk specifically to your problem

camsr Sat, 12/09/2006 - 13:29
Okay, the -51dB was actually dBFS, and the impedance of the soundcard input is 20k. Im pretty sure this soundcard is -10dBV, and the preamp is +4dBu so that is already a problem. I've tried correcting it with a T-pad but I got directions from some bozo on the internet and he gave me a schematic for a -6dB VU t-pad, when I actually need a -12dB VU. ( So I still need to fix that.

The mic is from radio shack (commence laughing), a generic dynamic omni with a 500ohm impedance. The preamp is built into my JVC cassette deck (TD-W318) and doesn't use phantom power. All lines are unbalanced.

Kev Sat, 12/09/2006 - 14:25
was actually dBFS,
... impedance of the soundcard input is 20k.
... pretty sure this soundcard is -10dBV,
... preamp is +4dBu
so that is already a problem.
very good info
and does look like you have enough knowledge to move forward
no offense meant

so the JVC cassette deck (TD-W318) is balanced ? ... no unbalanced
All lines are unbalanced
I've tried a quick google but not having much success

even so
the cassette deck will have an output control and you can manage some of your level into the sound card with that.

:shock: :roll:

where to start ?
you are making it hard on yourself
first try to optimise the connection from cassette deck to sound card

perhaps with a good cassette and strong level music loaded, ( test tone cassette perhaps ? )
try to get it set so that with the output knob in a normal position the levels into the sound card and the audio software are good and without clipping

it could require the T pad but a simple 20 k pot and think resistive divider may get the result you want

for general pad info start here
the unbalanced r1 with r2 shunt ( if the cassette deck is RCA output)

either way 6dB here or there is not going to solve your problem

more later


The mic is from radio shack (commence laughing), a generic dynamic omni with a 500ohm impedance.
Houston we do have a problem

I just can't know the construction of that Dynamic Mic or the mic input circuit of the JVC and what it is optimised for.
Generally speaking it probably isn't far from what it should be and to get the noise floor from -50dBFS down to -80dBFS+ is not going to happen without some major changes ... given there isn't a basic fault somewhere

a good healthy sound source into the mic and with the normal knob positions on the cassette deck it should meter OK and therefore the levels into the audio software should be close

it not

don't tell me you are trying to get dialog from 20 feet away ...

optimise the Cassette Deck to Audio Software first ... then start work on the Mic to Cassette Deck

Kev Sat, 12/16/2006 - 13:42
try the mic directly to the sound card

without being there and trying every trick I know
working with these low spec pieces of gear ... it is very hard to know which or where your weakest link is

gut feeling tells me ... " get a real mic "
that is where life starts and a Cardioid Dynamic like an SM58 or 57

better still
lets start again
what is the end goal of the whole exercise
is there any budget to complete it ?

camsr Sun, 12/17/2006 - 00:25
I think its a combination of both the elements, not to mention how much this preamp colors the sound. Esses get blasted through!

Budget huh? Well, I got like 20 bucks? It IS christmas time.

I did do the mic direct to the soundcard. The input was very low as you could imagine so I boosted it in software and the noise floor, at an audible "lead quality" volume, was still -51dB. And thats before compressin. So I DO believe it is the mic. I just don't have alot to replace it so Im trying my options. The next mic I buy will have a balanced line and Im gonna get or build a preamp for it. One thats designed for its impedance rating.

Kev Sun, 12/17/2006 - 12:46
yes the lack of linearity and poor frquency response can lead to those esses and scratchy sounds

the EL Cheapo Mic can also do this but I don't expect the frequency response to be that great so it may be a dulling effect

even at line levels with a real mic and pre-amp the SoundBlaster card is going to degrade the sounds you record

without knowing what and how you want to record
... all I can recomend is SM58 or 57
perhaps a Tandy PZM
and do my balanced and phantom mod

building a Mic to Line ... yes I have them too
search for the Green-pre or KDMP ... JLM Audio for a more expensive Baby Animal

look out I'll get flammed here
a secondhand M-Box or M-audio two channel interface ... obviously I'm a PT user

A simple M-Box with PT-LE and Reason Adapted and an SM58 and a modd'd PZM
and I can do things that impress most beginners
oh ? ... we would need headphones and speakers or some sort

camsr Mon, 12/11/2006 - 12:11
Kev thank you so much for your help. I've read that pad link and found out I need a new one. The one I got now is a 600 ohm and my cassette deck outputs RCA at 1000 ohm (this was tested with the deck shut off using an ohm meter, not absolutely sure if its right). Appearantly I have a 1/20 impedance on both the mic-to-pre and pre-to-soundcard, not sure how to interpret this but I guess it's pretty good?

Kev Mon, 12/11/2006 - 13:38
you can't measure input and output impedances this way

a DMM (digital multi-meter) is a DC device (mostly)

the impedance we are interested is over the audio range and often done at 1kHz

Have you actually tried going from TapeDeck +4dBu unbalanced to SoundCard -10dBV unbalanced ??

that will put the TapeDeck 14dB stronger than required but the output control of the TapeDeck may be enough to get a better result than what you have now.

even so
a simple 10 or 20 K ohm pot
with the signal from the TapeDeck across the outer terminals and the wiper feeding the SoundCard.

this will allow you to set for nominal levels on the TapeDeck and then trim to suit the SoundCard.

then ... without moving the pot position ... measure the relative resistances of each end of the pot and then choose some static resistors to make lead.

Personally I believe you should make a passive interface box for the SoundCard to allow for all sorts of sources ... I'd even include a transformer option

camsr Mon, 12/11/2006 - 15:07
The deck doesn't have an output control, if it did I wouldn't be having this problem


lol yeah I know the impedance is resistance to AC and not DC but those values seem resonable. And yes I have gone directly from the tape deck to the soundcard and it clips badly. Adjusting any software settings is pointless because the DAC is getting too many volts.

I will build another T-pad and let you know how it goes. Thanks for the help