6

I recently bought electric drums and I tried recording them through a MIDI interface. I'm not sure what the gain levels actually do, but from a quick internet search, it's related to the level of signals that go through the interface.

How do I know the gain levels I should set when playing/recording? Should I set it to the highest possible where there is minimal noise? Does it differ by instrument?

What factors should be considered to determine the correct gain level when recording instruments?

  • Are you only sending midi data out from the drums? Which make/model of electric drums are you using? – Dave Oct 31 '14 at 17:05
  • @Dave, Roland TD-11KV. I'm not recording it through a speaker then a mic. I connect my audio interface to my drum kit's control piece, and then the audio interface to the PC. So I assume it's MIDI? Please correct me if I'm wrong. – Zaenille Oct 31 '14 at 17:08
  • 2
    I believe that you are sending audio (not MIDI) signals out from your drumset into the (audio) input of the interface. I've never heard of a gain adjustment on midi input/output. – Dave Oct 31 '14 at 17:13
7

Some people suggest to just record everything as hot as possible, as close to the max as possible, but that might not be the best procedure. Some possible issues: you risk ruining the take by clipping the signal. Some devices induce more noise when working at their limits You have to predict the dynamics of the signal in advance or process it with limiters or compressors, just to name a few.

In digital recording one important factor is the bit depth.

In the average home studio, recording at 24 bits, you want your signal to average -18dBFS. Peaks can be around -10dBFS. At 32 bits the that range is lower, at 16 it is higher. This leaves more than enough room for your signal to have a large dynamic range, while staying far enough from noise.

From SOS article How much headroom should I leave with 24-bit recording?

The basic idea is to treat ‑18dBFS as the equivalent of the 0VU mark on an analogue system's meter, and that's where the average signal level should hover most of the time. Peaks can be way over that, of course, typically kicking up to around ‑10dBFS or so. Drums, being largely transient peaks, will be kicking up there regularly.

If the material you are recording is well controlled and predictable in terms of its peak levels — like hardware synths tend to be, for example — you could legitimately reduce the headroom safety margin if you really want to. But in practice there is little point.

The only advantage to recording with less headroom is to maximize the recording system's signal‑noise ratio, but there's no point if the source's signal‑noise ratio is significantly worse than the recording system's, and it will tend to be that way with most analogue synth signals, or any acoustic instrument recorded with a mic in a normal acoustic space. The analogue electronic noise floor or the acoustic ambience will completely swamp the digital recording system's noise floor anyway.

Recording 'hot', therefore, won't improve the actual noise performance at all, and will just make it harder to mix against other tracks recorded with a more reasonable amount of headroom. One issue that comes up a lot is the confusion between commercially released media (CD, MP3, for example), which have no headroom margin at all (they peak to 0dBFS), and the requirement for a headroom margin when tracking and mixing.

From article Proper Audio Recording Levels

Stop recording so hot. Instead of trying to get your tracks to peak at -2dBFS, have them peak between -20 and -12dBFS and your recordings will almost undoubtedly sound better. Mixing will be easier. EQ will be more effective. Compression will be smoother, more manageable and predictable. You're in the age of 24-bit digital recording - Relax and enjoy the headroom. Even if your only concern is the volume of the finished product (which would be a shame, but it happens), recordings made with a good amount of headrom are almost undoubtedly better suited to handle the "abuse" of excessive dynamics control. QUIETER recordings have more potential to be LOUD later. It's because they're usually better sounding recordings in the first place.

If microphones are involved, other factor is the ambient noise. If you are not recording in an acoustically isolated room, chances are that your room has plenty of ambient noise. The hotter the signal enters the mic, the larger the signal to noise ratio will be. So, to reduce the noise, the singer can sing louder, the amp can be turned up louder, the drummer can play louder, you can get the mics closer to the source. Just be careful, mics can be damaged if the sound is too loud. Check your mic's max SPL and make sure you are not near that limit. Note that your recording levels in the DAW should still be around -18dBFS at 24 bits.

Other factor is the noise of the devices that you are using. Some devices (pedals, amps, acoustic instruments, whatever) behave differently at different amps. Some will induce some noise and/or distortion at high amps. If you don't want that noise or distortion, the correct gain level here is before the noise and distortion kick in.

In short

If this is too technical and cryptic for you, just watch your input meter when recording and make sure it peaks at around -10dBFS and it stays around -18dBFS and you will be fine. You can even go a little hotter if you want. Recording at max levels is risky and not necessary at 24 bits (make sure you are recording at 24 bits at least).

-18dBFS

-18dBFS

  • Yeah, that was my problem when I tried to record a day or two ago. Even a weak strike to the drums cause the input meter to max out. – Zaenille Nov 1 '14 at 0:50
0

0dbfs does not equal 0vu. Yes you can push your signal right up up to peak at 0dbfs and the DIGITAL LEVEL will not go clip; however, the A/D converter has analog circuits in it, namely a preamp which does clip it's ANALOG SIGNAL before the signal gets converted into digital long before you hit 0dbfs. Professionals don't record right up to 0dbfs the way most amatuer, semi-pros, pro-wannabes (amateurs who think and pretend to be professionals. One of the ways they do this is by charging clients full big studio rates for an inferior amateur product and then call themselves PROS.) do. They record at an AVERAGE (RMS) level of -18dbfs (the 24 bit version of analog's 0vu.) And peak at -6dfs. In the days of 16 bit you would average at -12dbfs and not peak beyond -3dbfs. Of course Pro converters back in the late 80's actually put out 16 bits (or no less than 15bits). Back in those days there was no loudness war and so no need to normalize. And this is the main reason why compact disks from the 80's are so low in volume. They didn't compress and they didn't jack the digital volume to peak at 0dbfs. This is why a remastered CD will always be louder than the original version even if there is no compression or additive EQ - normalization. So don't always assume that louder means compression.

Back to your question sir. 24 bit digital has no "sweet spot" but the analog section of your 24 bit A/D/A Converter does. That "sweet spot" where your converter sounds its best (warm) is an average level of -18dbfs (in the U.K. it's -20dbfs) and a peak of -6dbfs. This is the level your converter is designed to operate at. And you will find it will sound it's best in this range. If you find you are going over your -6dbfs peak too often (once or twice is no big deal. Remember, it's the signal before it hits the converter; and that's analog.) than lower your recording level accordingly. Analog does not have a maximum like digital. In analog recording it was: sounds bad...sounds better...really good...Perfect.(0vu).....Not as good (+3vu).......Not good...Awful (+5vu)....Horrible. Turn that blasted recoding level down (+7vu). The analog section of your A/D Converter is the same.

Remember the ADAT. Ever wonder why they only went up to zero? Since -3dbfs was the 16 bit peak, 0dbfs on the scale was more then sufficient. Need further proof about these levels? Look at the Professional portable wave recorders from Tascam. Not the ones for amateur or hobbyists. At the digital scale you will see they have the 5 most important numbers on it: -18, -12, -6, -3, 0. The rest is just dot..dot.....and more dots. The recorder can record at both 16 and 24 bits. 24 bit: RMS -18dbfs, peak -6dbfs. 16 bit: RMS - 12dbfs, peak -3dbfs. And of course 0dbfs - last chance to stop the madness.

Don't worry about excess noise. Even at 18 bits the signal to noise ratio is 108 db, which is much quieter then your mike preamp or analog preamp interface. The digital noise floor is always lower than the analog noise floor in any A/D/A Converter. And think about it....The average DR rating of a track these days is 7. Unless you plan to do something like Pink Floyd's, "Dark Side Of The Moon" with just the sound of a single heartbeat....I wouldn't worry about noise.

Sorry to be so long winded but I wanted to explain properly why recording in digital is not this simple "just don't go over 0DBFS" If this was the case every DAW mixing software would just have clip indicators on them. Now you know what all the numbers our for. I believe there is either a plug-in or some option in most DAWs to switch the metering to VU.

Hoped I have helped. Good luck!

-2

The term for what you're asking about is "gain staging", this is an important part of audio engineering, and there can be many nuances.

For digital recording, you generally want to have the maximum amplitude of the recorded signal to be as close the the maximum allowed. This maximizes the signal to noise ratio, and thus offers cleaner sound. If you're mixing multiple tracks, you may have to pull the faders back to avoid clipping in the summed signal. So this is where you want to get to.

Now, depending on your setup you may have several places where you can affect the gain; the first of these is probably on the drums themselves, the next is at the digital interface, and finally there may be internal software faders. You need to manage these to achieve a high (close to clipping) final signal level without clipping at any of the intermediate stages.

  • By maximum amplitude, you mean the gain? Or the volume? I'm sorry, as I'm quite new to recording so I'm not yet familiar with the terms. :) – Zaenille Oct 31 '14 at 17:12
  • By amplitude, I mean the level of the signal (which can be affected by gain(s), and is closely related to volume). – Dave Oct 31 '14 at 17:16
  • 1
    This sounds like dangerous advice, since finding a "close to clipping" level without clipping is hard. This way of working increases the risk of digital clipping, which is way worse than background noise. You have to have pretty low levels to get problems with that. – Meaningful Username Nov 1 '14 at 13:20

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.