Mic Pre Amp Gain - How Much?

Steinberg Lounge
#1

I use a Tascam US-1800 for my mic pre-amp with a Sennheiser e-840 mic for the vocal.

Got a simple question. Is it best to have the mic gain turned up as far as possible on the US-1800 before clipping to get the best sound from the mic? Or can the mic trim be turned down lower and then make up any gain on the VST compressor built in cubase?

Is there a big difference in quality of sound captured between making the mic hot from the mic pre amps vs making it hot within cubase mixing? If so can someone explain to me in simple terms why?

Thanks in advance for the cheerful and great help.

Gary

#2

Generally, you want to go as loud as possible without clipping as long as the signal is still analog. That means the noise you are recording will be at its lowest relatively speaking.

#3

With a simple setup like that Mic>Preamp/Converter>Computer the gain setting should be determined by the input metering on Cubase!

Basically just make sure it doesn’t clip :stuck_out_tongue:

#4

Two things really

Always try to leave enough of headroom. You don’t need a hot signal if you are recording 24bit or higher files. Hot signals are a legacy of recording in 16bit. Peaks at -6 with an average around -18 or -20 will be fine. This is basic gain staging that everyone should learn and should give you a balance between the noise floor and over hot signals.

Try to find the sweet spot of the preamps. Again this is more than likely not too hot with prosumer gear but will also depend on the mic. You can always raise the gain inside cubase.

The problem with daw meters is that what is considered a ‘normal’ level looks comparatively low on a digital peak meter. Try some kind of vu or rms meter to check your levels. PSP’s free vintage meter calibrated to -18 is quite good.

If in doubt err on the low side. You will find that your recordings will generally be better. Also mixes come together quicker IMHO. You will find that if you use vst instruments you will have to lower the output a lot to match working at a lower level.

There are great threads on this topic over at ‘The Womb’ and ‘Gearslutz’

Hope that is of help

Cheers

#5

Oh PLEASE! Stop this ancient myth! There is NO reason to record at low levels on modern DAWs. Gain staging? Who cares? You have zillion dBs of headroom in modern floating-point DAWs. Just keep signal below 0dBfs all the time in

  1. A/D
  2. D/A
  3. Digital I/O (SPDIF/ADAT/TDIF/etc…)
  4. Exporting into fixed-point file.

Of course, if the audio interface is not capable of giving clean signal up to 0dBfs, this does not apply … but then it’s time to buy a better interface (even any SoundBlaster can do it).

Oh PLEASE! The Ultimate Source of Misinformation!

#6

Ha, I knew someone would jump down my throat when I mentioned Gearslutz. No problem. :slight_smile: As you correctly point out there is plenty of misinformation over there. I don’t post there but do have a read from time to time. One needs to be able to filter out the crap and the fan boys. I am always willing and open to trying things out and learning for myself though.

Regarding the original question.

Got a simple question. Is it best to have the mic gain turned up as far as possible on the US-1800 before clipping to get the best sound from the mic? Or can the mic trim be turned down lower and then make up any gain on the VST compressor built in cubase?

Is there a big difference in quality of sound captured between making the mic hot from the mic pre amps vs making it hot within cubase mixing? If so can someone explain to me in simple terms why?

I believe I gave the correct technical answer. As always it’s best to use your own ears.

All electronic equipment is built to a budget and savings have to be made somewhere in lower prices gear. All I can say is that not going into my interface preamps too hot results in a better sound. I did used to record hotter. Lower cost gear in particular (such as the interface in question) often functions nicest within a smallish range.

When using vst instruments it makes no difference as far as I can see but it does when recording real instruments. I believe the preamp is not inside the daw but is in the analog world. Keeping the peeks low ensures you don’t clip/distort or stress the input stages, especially when recording yourself when it’s difficult to perform and watch meters. As you point out there is plenty of headroom inside the daw.

This is my experience anyway. Your experience might be different.

Regards
Dave

PS. All rules are made to be broken. :slight_smile:

#7

Aloha,
and +1 to both.
{’-’}

#8

Excuse me, but your technical answer was:

Which is just repeating the old mantra from 16/24 fixed-point DAW period, where you had to leave headroom to your recording, because there was no headroom on DAW’s signal path. So it’s your approach, which is legacy, but unfortunately living on as The Whole Truth And Nothing But The Truth.

Of course if your preamp+A/D combination is cr*p and doesn’t provide clean signal up to 0dBfs, you have to back down. But there’s even cheap (I think they are cheap, because the components would cost approx $3) TASCAM preamps which are just fine driving A/D conversion up to 0dBfs (I have 24 of those inside my DM4800).

#9

It’s actually not a myth, it’s what happens when you gain stage in analog correctly on the way in. Signals with little or no transient info will land / peak in the daw much lower and signals with transients peak much higher obviously.

The problem is that too many DAW users don’t understand the relationships between peak metering, RMS metering and converter calibrations. Converters are calibrated from -15 to -20 with the latter being the pro level since most professional level analog pres have much more usable headroom. Many cheaper prosumer pres / soundcards can’t really be driven as hard, which is probably why their converters (on the all-in-one preamp audio card devices) are calibrated a little higher.

The problem isn’t the scale or the tracking level. You can track pretty low and be just fine. The problem is cheap gear with poor metering, some DAW’s without RMS metering on the inputs (or anywhere else), and people who don’t fully understand what they’re looking at… so they think “more is better” and crank up the level.

#10

I’m somewhere in the middle on this one. Dave is correct about gain-staging and all, but Jarno is also correct that in today’s floating point world the level of the signal showing up inside the DAW is mostly irrelevant (see below). However, on the outboard analog side, and going into the A/D, it’s still crucial to watch your levels – if you push your analog too hard you’ll get a poor signal regardless of what the digital level is.


There’s also something to be said about maintaining comfortable levels within the DAW because it gives you better fader flexiblity… and, although Jarno is largely right that the DAW output is irrelevant IF you have decent D/A, you have to watch it on some of your lower end pro-sumer stuff so you don’t distort their analog section.

Also – and I invite Jarno’s commentary about the following, because it strikes me as yet another modern audio urban myth – I’ve read a lot of people claiming that plug-ins have a “sweet spot” in terms of what level you should feed them. I’ve never had such problems, however; whenever I’ve had a plug-in that’s audibly clipping it’s a limiter on the output and I assume this is because a DAW’s output is 24 bit fixed

I say use your ears – when a mic pre or other analog input is being pushed too hard you’ll certainly hear it. Same with DAC


I say this as a person who has ignored the clip light on the output of my DAW for years

:laughing:

#11

Sure. But those are two different processes, tracking & mixing. I often will use clip gain to put my fader in a potentially more comfortable place for a track as relates to the fader range not being linear.

One thing (imo, mmv, not arguing at all) really has nothing to do with the other. If when mixing you prefer for a track to operate in a certain fader range for some tracks, there’s pre-fader gain/trim all over the place… on the clip itself and on just about every plugin.

I suggest that “newbies” avoid worrying about that kinda minutia when tracking… and just turn their speakers and headphone amps up a little. :mrgreen:

#12

-10 to -15 average level is what I record at. Less if clipping, but not all clipping sounds bad :wink:

#13

Of course you’ll have to watch your levels. But if your integrated A/D/A interface/preamp does not handle signal levels up to what represents 0dBfs, it’s poorly designed, no matter if it’s cheap or expensive. As I said, even SoundBlasters can do it. And I have no reason not to believe US-1800 can’t. While TACSAM’s quality control, support and drivers may be horrible, they really can build excellent cheap preamps.

This is absolute cr*p, unless your plug-in is written to behave like that on purpose. For example Steinberg’s Magneto has a “sweet spot” by design, but StudioEQ doesn’t. I’ve tested it myself: fed two copies of same signal (one with reversed polarity) to StudioEQ, one boosted to peak at +60dBfs, another at -60dBfs, matched levels in groups and summed these. Result: total silence

Indeed! And don’t believe, when someone says: you have to record at -XXdBfs.

#14

As I said earlier.

#15

The converters are fixed point. I agree that you can not clip inside a 32bit floating point modern daw. A/D converters and interfaces do contain analog components beside the digital converter. I may be misunderstanding you but are you saying that the Soundblaster converters are as good as something like an Apogee Rosetta? At £269 for 16 inputs I would class the Tascam as prosumer rather than high end. I am not doubting that the Tascam is a good unit at all. Definitely a very capable interface though, as is my 1820m. This level of modern gear is outstanding for the cost but, from my experience, most audio circuits sound best when working at comfortable levels, unless it’s a guitar amp. :slight_smile: :sunglasses: Inside the daw itself is a different ball game.

Does anybody know what level Tascam calibrate their converters to, as a matter of interest?

You can count me as another yet to be persuaded that all plugins only work properly at a low peak level input. Some do and some don’t to my ears.

I mentioned mixing using an average level meter is because that gives a more realistic idea of the loudness of a signal. I find it easier for my workflow to start off by setting the trims so that all tracks read similarly on a vu type meter and then work from there with the faders.

A lot of this going off the original topic.

#16

There is, as far as I can see, no data on calibration levels for the us-1800.

The pertaining document only states that the overload light comes on at -2dBfs to give you (advanced warning of clipping) my words. So even the signal lights on the device are calibrated to the digital scale. As the unit does not have proper metering.

Thus the only reliable metering is in the DAW itself, whether peak of if you choose, RMS. No doubt the THD will rise close to 0dBfs but as no information is given for that, it would be up to ones ears to say if it was acceptable or not.

#17

Yes they are. And because of that, you should not hit (above) 0dBfs.

Yes they do. And analog stage of properly designed A/D converter should be operating (relatively) linear manner up to 0dBfs.

Yes you are. They definitely are not. They have inferior S/N, distortion, jitter, etc figures. But even with them there’s no need to record at low levels. Take a look at http://www.youtube.com/watch?v=BYTlN6wjcvQ (starting at 41:15). (uncompressed audio files available at http://www.ethanwiner.com/aes/)

So do I. But being a pro-sumer doesn’t imply: you’ll have to record below -6dBfs.

No. And it really doesn’t matter. If you’ll have straight chain: Microphone -> Integrated preamp+A/D+interface, you don’t need to know, at which voltage levels the signals are there inside the box.

Some do and some don’t. Well-coded floating-point plugin behaves exatly the same at any levels, if not coded to behave other way on purpose.

Nothing wrong with that. If it suits your workflow, you should do it like that. But your way is not The Whole Truth And Nothing But The Truth. It’s your way. My way is just play everything out from the tracks, no matter what level they are and adjust levels with faders. And use master fader to set the final level. Only if inserting something analog on the path, I may use digital trims in my console to prevent D/A from clipping. But my way isn’t either The Whole Truth. We all have our own working methods.

#18

Just took a look at US-1800’s manual.
Line in: -10dBV or +4dBu => -16dBfs
Guitar in: -6dBV => -16dBfs
Figures for Line in are same as with my TASCAM DM-4800. Maybe this is where TASCAM calibrates their converters, then. Hard to say from statistically unsignificant sample.

#19

1 watt at 1 metre. And you’re dealing with the rules governing the end-product output.
Speakers from sublime to cruddy (high-end tech jargon :mrgreen: ). You make your mind up where you’re aiming for.
Car stereo? Hifi enthusiast? Radio? MP3 headphone? Club? (mono?) or all of them. Each will require a different approach.

#20

88.3dBSPL on my monitors (if I’m not mistaken with my math: 113dBSPL@300W => 88.3dBSPL@1W).