1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Most noise in tracking chain mic-preamp-converter?

Discussion in 'Studio Techniques' started by Steven Rowat, Oct 27, 2009.

  1. Steven Rowat

    Steven Rowat New Member

    (and please forgive the cross-post to Logic Pro Help)

    I'd like to reduce background noise produced in my tracking chain, but I'm not sure which elements of the chain are a best bet for replacement.

    The chain is a simple three stages:

    1. --> dynamic mic (Sennheiser 441 -- older version, from the early 1990's)
    2. --> preamp (Mackie 1202 mixer OR MOTU 828 mk2 interface)
    3. --> A/D converter (MOTU 828mk 2)

    I've done fairly careful comparison testing on two or three occasions, but I haven't been able to pinpoint any of these stages as being a larger source of noise, because I have no good way to separate them. All I've learned for sure is that my noise level is more or less identical with both preamp routes. But where the noise is generated I am unsure.

    So, my questions of the group are:
    1. Which of the mic -> preamp -> converter is likely to be the larger source of bg noise? (Both in general and in my specific setup).

    2. Would a Rode NT1-a, (with a self-noise of 5dB-(A)) be likely to be quieter than the Sennheiser? (I have read that dynamic mics are generally far quieter than condensers, so perhaps I will not gain anything by changing mics; I'm unsure.)

    3. Has anyone found the MOTU 828mk 2's noise contribution to be noticeably larger (say 4 db or more) than another interface?

    I'm not looking for a huge difference -- 6db would be nice and worth changing equipment for. 3 or 4 db maybe.


    Steven Rowat
  3. Markdvc

    Markdvc Administrator Staff Member

    It depends to an extent on the source material, but I would start off trying a different microphone. The Sennheiser 421 is very well established and widely used, but more often than not on fairly loud sound sources. A condenser such as the Rode will give you a stronger output signal compared with most any dynamic mic, meaning, the microphones' self noise is less of an issue, compounded by the fact that you will then need less preamp gain - which would make the disadvantages of a noisy Microphone all the more apparent. To cite some extremes, if you are recording an AC 30, Trumpet or Drumset this shouldn't be an issue, but for a Harp, or micing the room for church music it certainly will be. This will compound itself when mulittracking, and will be made even worse by using the same microphone repeatedly, summing the same noise over several tracks.

    One of my favourites is the Sennheiser MKH series - I have used their MKH 40 succesfully on most everything except kick drum. Recording a guitar amp with it, I may end up with the preamp gain all the way down, and the microphones pad switched on so as not to produce too hot a signal and overload the AD stage - the Mic's output is that high.

    After that you may want to look at preamps, but if the 1202 pres are anywhere in the ballpark of the 1604, this probably won't make as much of a difference compared with the choice of microphone.

    HTH, kind regards

  4. Steven Rowat

    Steven Rowat New Member

    Thanks Mark; two points about this. My mic (see my post) is a 441, not a 421. I also owned a 421 at one point; it had a more sculpted frequency curve and apparently was used for live performance more. The 441 is a significantly higher price and has a flatter response, and I've found it useful for many years for recording vocals (and other things). My hunch is it has less bg noise than the 421, although I can't do a comparison now.

    This leads into your next point, which I'd like to try to clarify because I don't understand it fully: how the mic self-noise is less of an issue if less preamp is required. Is this because the noise of the preamp?

    Suppose the Senn 441 and the Rode both have a self-noise of 5db A (I have to guess at the Senn, but let's suppose this).

    To me, this means that whenever we amplify the recorded signal, no matter what level, there's 5db of mic noise in both; even if the Rode uses a higher power level while it's recording. So relative to the recorded signal, the original mic-noise stays the same for both.

    Then we amplify using the preamp, by, say, 60db for the Senn and 40db for the Rode (which, just to grab figures out of the air, might be 20db louder due to its having active power).

    My way of understanding your statement is that this 20db difference in the preamp (60db boost vs 40db boost) will add appreciably more preamp noise to the Senn signal than to the Rode? Is that what you mean?

    Steven Rowat
  5. Markdvc

    Markdvc Administrator Staff Member

    Sorry Steven, I confused myself more that anything else, mixed up two Sennheiser classics :redface:

    Turning up the preamp amplifies the signal entering it. If a higher propertion of that signal is noise, it is getting amplified too. A microphone with a high output level and low self noise brings a double benefit - it requires less preamp, and becuase of that, what noise it does create is not being raised as much anyway. As a rule of thumb, a decent condenser will have lower noise than even a good dynamic such as the 441, but in addition, the condenser will as a rule have a higher output signal.

    Just to illustrate this:

    I connected a Sennheiser 441 ( I also have one) as well as a Sennheiser MKH 40 into the first two Mic channels of my Metric Halo ULN 8. I used Logic's Test Oscillator to produce some pink noise, set the two mics up at the working position in my control room, woith their capsules seperated from each other by about 3 inches and the mics about3-4 ft from the monitor speakers and turned the pink noise up, then adjusted the trim of each Mic preamp to get decent signal levels. Below is a screenshot of the two channels in MIO console. Note that the output levels are similar, but note the trim svalues. The 441 is channel 1, and is set at 68, the MKH 40 channel 2 (as illustrated by the p48 being switched on) and is only at 38.5. In fact, the channel 2 signal is quite a bit higher than channel 1, so I could probably have lowered the preamp gain somewhat:


    As you can see, in order to generate the same output signal level, the 441 requires the pre to be set waay higher than is the case with the MKH 40.

    I found some specs on both microphones.

    The 441

    The MKH 40

    Compare the sensitivity values. Unfortunately, there are no equivalent noise levels shown for the 441.

    Does that clear things up a little? Sorry for the initial confusion!

    kind regards

  6. Steven Rowat

    Steven Rowat New Member

    No problem, whenever I try to remember the 441's number I usually say '421' also and then feel unsure and have to go look at it.

    Thank you for going to this trouble. But for me the plot thickens. I see that there are two kinds of noise here in your test -- , ambient background noise, which is I believe where your added 'pink noise' falls, and mic self-noise, which is what I initially was thinking about and which is what would happen in a perfectly silent room. My understanding is that mic self-noise is actually thermal-generated noise from the mic electronics, and as such is inherently worse in condenser mics (since dynamics don't have active electronics).

    However, it occurs to me now that these two noise types constitute a new, fourth confounder in my '3 stages' I defined in the original post. So that the stages should have been:

    1. ambient noise
    2. mic (adding self noise)
    3. preamp (adding preamp noise)
    4. converter (adding converter noise)

    So it's probably too complex to try to predict and I should just order the d*** Rode and do my own tests. ;-) However, seeing as you've gone to all this trouble, the least I can do is tell you where I agree (and disagree) with you.

    So, in your statement:

    Yes; but my presumption to test was that the two mics have the same self-noise. To me that means that the same % of the signal is noise; db being a relative scale; so, say, for a 100db signal, 5db will be self-noise in both cases.

    Then, turning up the preamp will not change this proportion -- unless the preamp itself adds its own noise at a higher proportion for higher gains.

    Now, I'm unsure whether the addition of ambient noise changes this relation: but it doesn't seem that it should.

    And as far as I can see, your test didn't tell me the result I'd need to know to decide that -- although it may be there in the data.

    What I mean is: yes, the trims are different. But what does that tell me about noise, realistically? In other words:
    : Given an identical drum hit, or spoken word, or whatever, with no pink noise added, once the trims were adjusted so the levels were the same, what would the ambient+self noise of the two mics be? Is one higher than the other, or are they the same?

    I think that's the test result that will answer the question pragmatically for me. In theory. :) .

  7. Markdvc

    Markdvc Administrator Staff Member

    Just to clarify, I merely chose pink noise as a convenient, constant signal spread over all the audible frequencies, primarily to simplify setting up the preamp levels. I could also have used a sine wave. I didn't want to use music or any other changing signal for the sake of simplicity. Just because it is called pink noise doesn't mean that it was anything comparable to ambient background noise, I was wearing ear protection :)

    OK, let's try to clarify this some more.

    What is important here is Equivalent Noise Level. THis is a measure of the audibility of the microphones' noise level, and is dependent not only on the noise level of the microphone taken in isolation, but also takes the sensitivity of the mic into account. If the microphone is very sensitive (usually desirable), then the S/N ratio between the source signal we are recording through the mic, and the noise of the microphone itself, is high. This is the case with the MKH 40.

    Typical values for Equivalent Noise Level are 15 - 30 dB A weighted for Dynamic mics, and 12-24 db A weighted for condensers. In other words, the typical condenser will usually have a better S/N ratio, or if you prefer, be more sensitive than the typical dynamic

    Is the same for any microphone, we can safely ignore this.

    Explained above

    Glad you mentioned this. Turning up the preamp will also raise it's noise level. The less you have to turn it up, the cleaner the signal it passes on in terms of noise added by the preamp itself. So, the higher the signal level entering the Preamp from the microphone, the less you have to turn it up, and as shown in my test, a high quality condenser can supply a signal that is considerably hotter then even a high quality dynamic such as the 441.

    as with 1., equal for both microphone types.

    Not really IMO.

    Yes, by all means satisfy yourself!

    Hopefully the above will have clarified that this is not the case, or, to be more precise, the sensitivity is relevant here.

    Exactly. Hope it is clear now, if not, I better let someone else have a go ;)

    kind regards

  8. Steven Rowat

    Steven Rowat New Member

    Alas, yes and no. :) . After unearthing and re-reading this page:


    in particular the section on 'Noise', I find I'm even more confused; trying to compare the 'sensitivity', what I find in the table 3 on that page is that as sensitivity goes up (assuming larger numbers in mv/Pa means 'more sensitive'), equivalent input noise goes up also. And that dynamic mics still come off as less noisy (via the step process at the end of the page for determining EIN).

    I apologize for starting this thread, because I don't believe I have the theoretical background necessary to follow all the conversions of units through numerically and keep clear what's happening.

    I think I'll go finish recording my song now. :).

  9. Peter Ostry

    Peter Ostry Administrator Staff Member

    Well, if you read Rane notes you are yourself responsible for any brain damage they might cause ;)

    This is true if you measure the mics alone. But they have also a weak output signal compared to condensers and you need to turn the gain far up. For quiet sources like a nylonstring guitar your preamp gain may be between 50 and 60 dB and most preamps start to produce audible noise at these levels.

    Even on stage you see often condenser mics, sometimes the mic of a singer is the only dynamic one. And sometimes the singer has a condenser and the sax player a dynamic mic. The engineers have their reasons for that, the question is not only condenser or dynamic or more or less noise. The challenge is rather to choose the right mic for the particular circumstances.

    Technical specifications tell something but not all. Noise or not, a dynamic mic at a kick drum is fine but a dynamic mic 4 feet away from a concert harp is poor. A first class dynamic mic on stage may sound impressive for your voice but may fail for recording where a mid-class condenser or ribbon can sound just right.

    If everything is loud, some noise is barely audible (often not at all). But if you go romantic or the sounds are isolated, the SNR (signal to noise ratio) is important. And if your preamp is hissing from too much gain there is little chance for romantic emotions. Therefore there is little chance to do that with a dynamic mic.

    Not complex if we talk about the practical usage. A good universal condenser mic on a good preamp can record about anything in reasonable quality. You cannot expect this from a dynamic mic.

    Røde is known for smooth and very quiet mics (I don't know their K2 which is another breed). You mentioned the NT1-A, this is a widely used good microphone.
  10. Steven Rowat

    Steven Rowat New Member

    Hi Peter,
    And thank you for sharing your experience. So, based on your description, my guess is that a low-self-noise condenser like the NT1-A might be better than a dynamic like the 441, both for bg noise and/or for ability to register fine differences in sound, in the following situations that I might use it for:
    a) finger-picked acoustic guitar, at about 12 inches.
    b) a distant second mike to pick up room ambience.
    c) vocals that will be compressed substantially in mix.
    d) field recording of relatively distant nature sounds.
    Does that seem reasonable?

    Best regards,

    Steven Rowat

Share This Page