Low output mics and in-line preamps

[rob aylestone]

We've discussed quite recently low output mics, and the Cloudlifters and similar devices that get the output up to more useful levels to let the preamps in the interfaces have an easier time.

I got a random email from Studiospares here in the UK advertising the Imperative Audio FET Pre FP1. I'd been pretty insistent about my luke warm impression of the SM7B, so figured that buying a preamp would at least be worth a try. I recorded the mic at about 2" from my lips to the foam, and on the gain setting I use for condensers and the higher output dynamics, the record level was very low. My plan was then to normalise in Cubase to -3dB, then repeat the test with the preamp. I would normalise that to the same level - so you should hear tonal and noise differences, or not.

I then had a thought. I have a couple of Coles 4104 Commentators mics - the ribbon types you often see being used in difficult locations on TV. These are lower in output than the SM7B - so I repeated the test.

I have to say I can hear a difference, and frankly, I did not believe some of the recent comments posted about the benefit a preamp can do. I was wrong. The SM7B is better with the preamp and the same difference appears on the commentators mic.

48K wav 32 bit recording done in cubase - only normalisation applied - nothing else. Comments, even "I told you so" gratefully received.


file here

 

[TalismanRich]

What I hear is a definite boost in the bass region, especially with the SM7b. The Coles seemed much closer, but there was a bit more bass and dare I say a bit more clarity to your voice. For the noise level, on both the SM7b and the Coles, I heard a definitely drop in background hiss with the preamp boost, which contradicts what I experienced in another test a while back.

To what do I attribute the difference? That's hard to say. Without having all the equipment in place at the same time, it would be impossible to speculate on why the results are different, but they are. The tonal change was VERY surprising, and might be appropriate in some settings, and not in others. It gave your voice a deeper, chesty sound. I would suspect that it might help with your double bass, although I doubt it would fix the SM58. It could be an artifact of the normalization process (was the top rolled off, which leads to apparent bass boost?). It might be better to set the record level with a tone, like 1K from a phone, just to eliminate that process. If you have the exact same recording, then normalization should be comparable. With multiple takes, it doesn't always work as well.

 

[rob aylestone]

I think I prefer the preamp in circuit, but I cannot really explain how it makes the difference it does on the SM7B - people have given all sorts of reasons for it, but I'd always thought that gain was gain, and the only other aspect could be impedance differences? I really don't know but for £50 it does seem to do what it says on the tin! Why? No idea. The ribbon mic was helped no end, and with the preamp, quite usable I think.

 

[ecc83]

I think I prefer the preamp in circuit, but I cannot really explain how it makes the difference it does on the SM7B - people have given all sorts of reasons for it, but I'd always thought that gain was gain, and the only other aspect could be impedance differences? I really don't know but for £50 it does seem to do what it says on the tin! Why? No idea. The ribbon mic was helped no end, and with the preamp, quite usable I think.

Hi Rob, I have tried to find out from Studio spares the input impedance of that pre amp. They don't know! There is a lot of agreement among studio folk that dynamic mics and especially ribbons get a sonic benefit from 'seeing' a higher impedance than the usual 1k2 to 2k5 of most AI and mixer mic pres. Many 'high end' pres have switchable input Zs of 2k and 5, even 10k (some have a low Z of 600R. WTF? No idea! ) I suspect, being a FET device that inline booster has a fairly high input Z and it is that which is 'improving' the sound.

I sent my son a FETHead pre some weeks ago and he uses it with an SM57 on classical guitar. He also uses a Makie 91c LDC and finds the boosted 57 comparable for level and noise. He quite likes the sound of the dynamic for his guitar. Into, BTW a UMC 204HD.

Dave.

 

[TalismanRich]

I'm surprised that nobody ever talks about the Cloudlifter CL-Z. This allows for variable impedance which they say lets you tune the microphone. They have a pretty wide range, from 150 to 15KOhms. It seems like a simple way to determine how the load impedance the mic sees affects the tonal balance. They even have suggested levels:

 

[Starliner]

One thing I've never really understood... if a preamp is normally used on a mic, why do you need TWO preamps in the chain (i.e.: an added pre, like a Cloudlifter). Aren't most preamps powerful enough to compensate for a low-output mic? And, how much more noise is added with the second preamp?

 

[rob aylestone]

That's the rub. Pre-amps could be designed to cover gains from modest - from the condensers, to almost extreme, like the ones talked about here. It's perfectly possible too design a linear, low noise high gain pre-amp, but they're not really that necessary for most mics. The total noise in the chain can be measured as the EIN. Using cascaded preamps could be a bad move if both had poor noise figures, but with the typical preamp noise for normal gain setting being quite good, adding a specifically designed extra preamp does work. Indeed, my own scepticism over their usefulness was wiped out this week. I can see a circumstance where adding a preamp in line could make things worse, but the design of these things optimises performance vs the potential for noise.

 

[ecc83]

It is electronically quite easy to make a very low noise mic pre amp if it has a fixed gain and especially not a lot of gain and the devices we are talking about usually have between 20 and 30dB of gain. The big challenge is making a pre amp with variable gain, not least because the gain pot itself introduces noise! It is very difficult to get a smooth gain change over much more than 50dB and indeed most AIs tend to range from +10 to 55/60dB. Mixers have an easier time of it because they can have a gain pot and another 10dB or so gain in the channel. Also mixers almost always run at higher internal voltages than AIs and so have more headroom.

The really top end mic pre amps use switched resistor sections to give gain steps, usually about 5dB intervals and then often have a second stage with 10-20dB of extra gain.

There is also a 'golden rule' to getting as low a system noise as possible...Get the level up above noise as early in the chain as possible and with the lowest noise device. After that the noise in the rest of the system has little effect.

Dave.

 

[Mr. ROUSH]

One thing I've never really understood... if a preamp is normally used on a mic, why do you need TWO preamps in the chain (i.e.: an added pre, like a Cloudlifter). Aren't most preamps powerful enough to compensate for a low-output mic? And, how much more noise is added with the second preamp?

Most preamps do have enough gain for the 7B. However, most require you to crank the gain to 75% or better, bringing up the noise floor. Therefore, adding the mic activator inline results in much less noise. It adds clean gain to the incoming signal, so you don't have to crank the preamp so high, and the noise floor stays much lower.

 

[Starliner]

Very interesting points... thanks to all for your replies!

Now it makes more sense! The impedance-matching function of the Z model always seemed to be a good idea to me. I do know sometimes it is difficult to get certain specs from manufacturers, even thought they are useful and sometimes necessary to know.

Starliner

 

[TalismanRich]

Very interesting points... thanks to all for your replies!

Now it makes more sense! The impedance-matching function of the Z model always seemed to be a good idea to me. I do know sometimes it is difficult to get certain specs from manufacturers, even thought they are useful and sometimes necessary to know.

Starliner

I was looking to see what kinds of impedances the various interfaces have. Just a few that I found:

Focusrite Scarlett 18i20 - 3k
Behringer UMC1820 - 3k
Tascam 16x08 - 2.5K
Presonus 1824c - 1.4k
Motu M4 - 2.65k
Audient ID14 - 2.8K

Other than the Presonus, it looks like 2.5-3K is pretty much the norm.

 

[rob aylestone]

Which could also explain why with my Presonus, I was unimpressed with the SM7B? Interesting!

 

[ecc83]

Which could also explain why with my Presonus, I was unimpressed with the SM7B? Interesting!

Indeed it might, but... To be really picky impedance "matching" is not the issue. In audio signal transfer we never "match" impedance. The process is 'voltage' transfer. The source impedance should be low and for microphones these days that is from about 100 Ohms to 250 Ohms and the load impedance usually at least 5 times the source and often ten or more times which is where we get 1k5 to 3k for mic pre amps. For line level inputs the normal minimum load is 10k. This voltage "matching" means almost all of the source voltage appears across the input device which means best noise performance.

The idea that dynamic mics and ribbons benefit from a load of 5k or more is a relatively recent one and present company suggests there is something to it. I have not as yet read of any one coming up with a 'mechanism' for the subjective improvement?

Dave.

 

[jimmys69]

Very interesting...

I have found with a very soft singing client i work with regularly, that with a SM7b through a Vintech X73i pre, I have to almost full on crank the pre to get decent signal into Cubase. The pre states 70db of gain that one would assume would be sufficient. I just ordered a Klark Teknik MIC BOOSTER CM-1 Dynamic Microphone Booster ($35) to see how it works. If that works well, I'll probably get the more expensive Cloudlifter unit.

I will be back. Thanks Rob!

 

[rob aylestone]

Ha! If it works, you'll probably just keep it!

 

[Starliner]

I was looking to see what kinds of impedances the various interfaces have. Just a few that I found:

Focusrite Scarlett 18i20 - 3k
Behringer UMC1820 - 3k
Tascam 16x08 - 2.5K
Presonus 1824c - 1.4k
Motu M4 - 2.65k
Audient ID14 - 2.8K

Other than the Presonus, it looks like 2.5-3K is pretty much the norm.

TalismanRich,

I might add the MOTU 896 mk3 Hybrid has an input impedance of 10k on the TRS, and 2k on the XLR's.
This info is not in the owner's manual... I got it from inquiring with MOTU in a Tech Link online.

 

[tmix]

Has anyone been brave enough to try 2 inline boosters on the same mic?
I realize it would have to be a version that passes phantom power through the unit to downstream.
Just curious really

 

[Starliner]

tmix,

Without looking at any specs, my knee-jerk reaction is I would be worried about extra noise.

Starliner

 

[ecc83]

Has anyone been brave enough to try 2 inline boosters on the same mic?
I realize it would have to be a version that passes phantom power through the unit to downstream.
Just curious really

Not a good idea IMHO. If the boosters delivered only 20dB of gain each, and most are a bit higher than that, the resulting 40dB would mean even the 'feeble' SM7b would easily put 100mV into the AI and that would be close to overloading many mic inputs.

Dave.