As others have already stated, when recording at 24 bits there is really no reason you should be clipping, other than you are recording too hot. 24 bit gives you plenty of headroom. So, before recording, rehearse your take to make sure your peaks do not go over the neighborhood of - 10 to -18 db on your input meter.

I record through a hardware UA 1176 after my preamp, with the limiter set so my input meters do not go beyond -12 db. If you do not want to use a limiter going in, then just make sure you keep your input levels well under 0db. 

When recording in 16 bit it was necessary to record fairly hot. But, this is not necessary in 24 bit.

so, I'm trying, people. smiles. I will record a  group for hours. someone starts using a vocal Mic for a violin or sax. these things happen. I'm not clipping because I'm intentionally pushing head room. sometimes mics get moved. sometimes musicians take you bysurprise. it happens to all of us, I'm sure. and once a new song is begun, I hesitate to redial the level before I know I must. I much prefer to make level  changes between songs.

Back to the OP

If it is not too bad you can zoom in and cut the clipped bits out of the track, then either delete the hole made by the edit or fill it in with a small piece of the adjacent clip. Some prefer to use dedicated editors for this function. Dedicated editors have automated tools to assist in the process.

gswitz


so, I'm trying, people. smiles. I will record a  group for hours. someone starts using a vocal Mic for a violin or sax. these things happen. I'm not clipping because I'm intentionally pushing head room. sometimes mics get moved. sometimes musicians take you bysurprise. it happens to all of us, I'm sure. and once a new song is begun, I hesitate to redial the level before I know I must. I much prefer to make level  changes between songs.

I hear what you are saying. It can be a challenge when recording a group of musicians. When I had my commercial studio, often while setting levels the band kind of held back while playing. Then, when the record button was pushed they played more aggressively. It just takes time to dial all the levels each time you begin a new song, especially if new instruments are introduced, or new mics added. Some groups/bands are easier than others.  


May I ask what you are using as a "front end" (preamps, compressors, limiters) for each mic?

gswitz


so, I'm trying, people. smiles. I will record a  group for hours. someone starts using a vocal Mic for a violin or sax. these things happen. I'm not clipping because I'm intentionally pushing head room. sometimes mics get moved. sometimes musicians take you bysurprise. it happens to all of us, I'm sure. and once a new song is begun, I hesitate to redial the level before I know I must. I much prefer to make level  changes between songs.

Are you recording in 24 bit? There is really no downside to starting with 30dB or more of headroom when using 24 bit. So if you're clipping in 24 bit, you almost certainly are pushing the headroom. 

The rule when recording using 24 bits is:
Don't even try to "optimize" things - the optimal recording level is anywhere:
1. You aren't clipping.
2. The quantization error level is below the level of analog and environmental noise. In most real world recording situations, 24 bits will allow you 60-80dB of headroom (or more). 

So the point is - it's not even worth worrying about #2.

I'm using a RME Quad Mic Pre. Two of those channels go straight to the Fast Track Ultra using 2 ring quarter inch cables. Two go to my DBX 162 SL and from there into the Fast Track Ultra.

The quad mic pre has a green light which lights up when you have a useful line level, but for some instruments it almost never lights when you are recording below clipping (like a banjo which has lots of loud spikes and a relatively low line level otherwise).

The DBX has a 'stop level' setting that uses 'peak stop plus' to limit. This works very well and I set this at 0 DB - the lowest available setting. After the compression steps and before the peak stop, there is a gain nob, which I usually try to leave around 0. I've just looked at the compressor and the thresholds are around -20 and the compression ratio is 1.5:1 (less than 2:1).

Usually people monitor through headphones direct through the Fast Track and I record with all FX bypassed so the computer runs as quietly as possible (almost 0 processing used in the recording).

So, based on the comments here, I'm still recording too hot if I'm safe down to an -80 DB peak. That's really quiet. I just tried to zoom to -80DB on the recordings from the other night, and Sonar doesn't even allow you to zoom to that precision. -57 is the closest I can get (96 zoom factor). At that level of zoom, room silence is above -80 DB.

Most of my levels were below -12 DB. Right now, for some reason, I can't change 3 of the tracks off of showing percentage rather than DB as the scale for the wave form. Not sure why.

And the other night, there was one song, where I joined in. I moved one of the vocal mics in front of me and played and sang into the vocal mic. I was happy and played hard (no one was watching the levels) and there were a couple of clips.

If I record everything quieter, I guess gain staging could come into play... so... like, I would run from the Pre Amp into the compressor with a reasonably strong signal, and then I would roll down the gain on the compressor to -12 or more. Then, I would get a max input around -24 with the loudest and limited moments being at -12. I guess that's what I'm being told is best.

Yes, I'm recording at 24 bit 88.2.

Yes, I've recorded at 24 bit since I stopped using my akai dps12 around 2003 or so.

Yes, I've made a lot of recordings with very low levels, but I have also found that in some cases the levels were too low (even using 24 bit). it is possible if you are me to mess this up. I have found that using the pre amp I can usually get more level safely (especially if I have 2 gain nobs giving me gain staging control so that I can fine tune the level pretty well).

Anyway, thanks to everyone for the guidance. I'll back off the levels some more. I think one of the people who turned me up a little again was the guitarists guide to sonar which pointed out that tracks recorded at 88.2 might come out of the Amp Simulators sounding a little better. I tried and agreed.

Here are some steps towards getting great consistent recording levels:

1     Add a VU display system or device to your set-up and calibrate. This can be a plug-in. A real VU is a little nicer. It will show the rms level component of the signal.

2     Choose a ref level. A good place to start is K system metering and that ref level can be -12, -14 or -20 dB FS. Calibrate so a tone at the ref level shows 0 dB VU on the meter.

3     Recording levels do not have to be 'anywhere' they can be very somewhere! Most instruments will easily allow you to adjust for a 0 dB VU level during record level set-up. The built in headroom takes care of peaks. Very transient percussive instruments are best set using the normal peak metering available.

4     16 or 24 bit does not matter only that 24 is better as the digital noise floor lowers so the -20 dB ref level can be easily used. For 16 Bit I tend to do a lot of work at the -14 dB ref level.

Most DAW's do not show VU or rms components very well. We only had VU before with analog and used it very very well. Now it has been removed. Dumb. It should be put back and used. You will never have problems with all the right levels being on your tracks prior to mixing. You will never have to add gain here or there because everything is already correct. When you work this way you are keeping all your rms levels consistent and letting peaks end up where they do. You will never clip a plug-in again going in and out. The VU meter can tell you more about your music than peak metering ever will. It is a powerful tool.

Buses levels also add nicely and average around the ref level as does the master buss. Mastering is a dream from a well set-up K system using VU metering options. K system also involves a consistent sound pressure level in your studio.

This system I have described is the best and only way to get your recording levels all correct and make for the easiest adjustment as well during recording. Even if you get a level from an artist during testing and they get louder in the recording, the headroom that is built in to the -20 dB ref level specially will easily take care of most situations like that.

Jeff Evans


Here are some steps towards getting great consistent recording levels: 

Good advice. Step one is to meter the levels at the track input!

Step one is to meter the levels at the track input! .......This is important and for a virtual VU meter it is the first plug-in that is line in your effects bin. You need the input monitoring enabled in order to see it.

And looking at the OP's avatar, voice and guitar and two great contenders for VU input level setting. The VU meter works very well with vocals. The VU meter I think was designed originally for monitoring voice levels.

If you connect your input sources to a mixer first then a real VU can be situated on the main stereo outs. You do have to calibrate the level being sent from the mixer to the DAW so that it matches the VU you are reading. With digital mixers this is taken care of automatically.

It is difficult to go wrong setting levels. Just get the performers to do their thing as close as possible to the energy that is going to be used in the actual recording and set the input gain so the VU is just reading 0 dB VU on the loudest parts of the signal. Your done. Most instruments work this way. Very fast sounds or short transients may not make the VU move enough or at all so your peak meters come into their own. It is really then you should be reading them but the rest of the time the VU is king. 

You get so used to working with them that you forget about peak metering and don't look at it hardly at all. But you don't see any red clip lights either. 

Thanks!!