Input noise spec?

  • Thread starter Thread starter ofajen
  • Start date Start date
O

ofajen

Daddy-O Daddy-O Baby
When preamp makers specify the Equivalent Input Noise (EIN) for their preamps and just give number in dB, say, -130 dB, is that assumed to be in dBu or dBV? I'm assuming dBu. Thanks for the help.

Cheers,

Otto
 
Never mind about the units... it appears that dBu is presumed. The bigger questions remaining are whether their is a presumption about the type of filtering used, if any, and making sure that gain level, bandwidth and source impedance are specified. Can anyone shed light on the question of what filtering should or should not be used? My sense is that their shouldn't be any filtering, but I haven't thought through how the filtering choice will affect the spec.

Cheers,

Otto
 
ofajen said:
Never mind about the units... it appears that dBu is presumed. The bigger questions remaining are whether their is a presumption about the type of filtering used, if any, and making sure that gain level, bandwidth and source impedance are specified. Can anyone shed light on the question of what filtering should or should not be used? My sense is that their shouldn't be any filtering, but I haven't thought through how the filtering choice will affect the spec.

Cheers,

Otto
Click to expand...

The standard for source impedance is 150 ohms.

My personal feeling is that the standard should be dBV, as dBu is an anachronism (as an aside, the standard pro level should be set to 0dBV, only 2dB away from where it is). Mics are generally spec'ed in dBV (or mV, which I find annoying, and customers who can't convert to dBV would find useless), so using dBu for preamps just unnecessarily adds another 2dB to the equation.

But anyway, whatever reference they use for level, they should use the same for all other specs.

I also feel A-weighting is appropriate, for two reasons: first, many many mics are spec'ed with A-weighted noise (the exceptions tend to be the high-end European stuff, Schoeps, DPA, etc), so it makes noise computations straightforward. Second, that is appropriate for both mics and preamps as the ear is not sensitive to very high or low frequency noise.

A-weighting will drop an unweighted noise reading by as much as 6 or 8dB. So for a manufacturer of prosumer gear, it would be suicide to use an unweighted spec, since almost none of their competitors do.
 
I nominate mshillarious to make an unambiguous, coherent, standard so we dont have to constantly compare apples to oranges and multiply translation errors

Seriously!

This is a MAJOR pisser for tech support, to constantly deal with trying to figure how different systems relate
 
Generally it's dBu, 150 ohms, measured with a wideband voltmeter with a filter bandwidth of 20 to 20kHz. The gain of the preamp should also be mentioned, as the EIN can also be affected by the amplifiers output noise, in fact that noise can be the dominant source in low gain applications.

A-weighting is not typically used, even though it looks magically better.
 
Thanks, Dan, you were next on my list to bother. :) As I'm sure you know, the standard is IEC 60268-1, but it is, to quote David J, "malleable." The spec allows:
1) bandlimited 22-22k flat,
2) A-weighting with an rms meter,
3) CCITT weighting with a quasi-peak filter, or
4) 1/3 octave weighting with an rms meter.

You can pick any or all. That's just great. :rolleyes:

I'd be curious to see at least the first two applied to every unit and see how they vary, depending on the spectrum of each unit's noise.

Cheers,

Otto
 
Yeah, standards.

Don't get me started. Or internationally recognized BS. ISO-9000 what??

But, back to noise. Almost everything is pretty quiet these days. There is some difference in the spectrum of the noise, and pretty minor differences in the levels. The specs are practically meaningless in determining how they sound different. It's more a matter of hearing the noise yourself, against the signal you have, and then saying yep, that's fine.

My NV is not the quietest preamp, my older MP-2MH was pretty quiet, within a couple of dB of theoretical, but by measurement, they're only about 3dB apart. In use, the noise difference almost never comes into play.

My published specs on both units are all at 40dB gain, 150 ohms, 20-20K, no weighting.
The measured differences between them and any other manufacturers units will be within fractions of hairs. The paper numbers are really pretty useless, there are too many variables.

The noise of a preamp that is well designed is heavily influenced by it's source, hence the 150 ohm source in most published specs.

In reality, microphones have output impedances that vary all over the map, at least the ones where preamp input noise matters. If you are using condenser mics, the preamp noise simply is not a part of the equation, because the mic is already a preamp, and it's noise will swamp out what the outboard preamp contributes.

With dynamics or ribbons, the output impedance can be in the very low ohms, like an SM-7 or an RCA, or it can be high, in the 300 ohm range like a Royer.

Those output impedances can make the preamp have several dB difference in output noise
just by the virtue of them in the cable instead of something else.

Add the variables in the output level between mics and try to figure the "real" impact of the overall system noise to the recording by reading spec's, well, good luck.

At the extemes it really takes careful pairing of mics and preamps for the best performance, and there aren't specs published that will predict it with certainty.

I'd be happy to publish the spec's according to a standard, I'm not hiding anything, it's the physics of the situation that dictates the performance, not me cheating, but as you've seen, there isn't one everyone agrees to, for various reasons.
 
Dan,

I didn't know you tread here? I had to think for a minute - I know that name hmm.

Then it dawned on me with all the great info you provided:D

Good to see you here.

T
 
Well here's the thing about various standards: for preamps, it shouldn't matter too much which standard you pick, so long as the two you are comparing are using the same standard. That's because preamps tend to have a similar noise profile--whitish noise--even at different noise levels. So we can probably roughly define the difference between A- and unweighted readings as 3dB (the difference between say 10kHz and 20kHz bandwidth).

For mics that's not true, they can vary from pinkish to whitish at different frequencies. So you could argue that it's mics that should be unweighted, but that would skew things on a practical level too. I don't have an anechoic chamber, and I can't even get low-frequency ambient noise quiet enough for a real noise reading below 100Hz. But you know what? Thanks to A-weighting, I don't really need to, and the reality is that most studios wouldn't be able to do better than me, so acoustic noise will swamp whatever self-noise exists at those levels.

The trouble of course arises with ribbon mics, so I tend to think noise ratings should be at high gain settings. With lower-end stuff, I think you need to be careful. I think Dan might exist on a higher plane than some of us :o and there are a few popular recommendations for budget preamps that really do have high noise ratings. You should be less concerned with gain than noise in a ribbon mic to preamp to digital converter setup, given that 40dB of gain is plenty to get a ribbon above your converter noise (I should hope!).

Is there a magic number? I would think anything below -120dB, whether u, V, and/or A or other is probably good enough.
 
You must log in or register to reply here.
Back
Top