Wave form height in time line

How to make it taller?

Is there some setting that I can change to make the sound wave form taller in the time line? When I play the tracks, I hear them just fine and when I am recording, the signal level in the VU meters looks good also. However, the wave form in the time line is barely more than a tiny squiggle. At low volumes, it’s basically a straight line. If I look at the track with Sound Forge, it looks normal.

A point of clarification: I’m talking about the height of the wave form relative to the height of the track on the time line. In other words, the ratio between the height of the wave form and the height of the track. I now I can increase the height of the track and, thus, the height of the wave form. What I want is to make the wave form taller while keeping the track height constant.

normally that would imply your recorded volume is too low. you could try going into the preferences and looking around. not sure how to change that though.

Short answer: “Normalize”, or else record hotter.

Long answer:

If the wave files look anemic, then you’re probably not getting very high recording levels. However, if you have a very good analog gear chain and very good 24-bit soundcard, this isn’t a problem. (Also, if they SOUND good, then they certainly can’t be too bad!) If you want, you can normalize to make the waves nice a beefy looking, making many editing tasks easier. This won’t improve the sound quality – in fact, it will reduce it but by an amount that’s too small to worry about. For 16-bit tracks, normalization also won’t hurt, but if you need to normalize, you probably should have recorded hotter.

What I’m trying to say is that the waveform size is proportional to the average peak levels. So, if they’re skinny, then it’s because you recorded at a relatively low level. If your gear chain is very good, you have lots of elbow room. If it’s not, you don’t.

Let’s get quantitative. The biggest waveform is 0dBFS (FS stands for Full Scale, in case you didn’t know). -6dB is half that. -12dB is a quarter of full scale, -18 dB is one eighth, and so on. That is, cut the size in half for ever 6 dB (roughly – the actual value is closer to 6.02, not that it matters).

So, the best way to get big waveforms is to record hotter. Of course, you always want to leave a bit of “headroom” or safety margin in case you get a great performance with a loud peak, and you don’t want to tolerate any clipping. How big a margin you need depends on what you’re recording – which you probably know. For instruments/players that are unpredictable in dynamics, you need to leave more of a margin. Most of us shoot for a peak level of -6dB or better, and hope for -3dB or better. If you peak higher than -1dB, it’s a good idea to scan your track (using the “Normalize” scan button) to see what the real peak is; if it’s 0.00, you probably clipped and you should probably redo it. During recording, n-Track doesn’t catch every clip, it just does the best it can.

However, sometimes we just can’t get that high. When I record my acoustic guitar with an SM57 through my MOTU 828, even though I mike very closely, the peak levels are around -15dB. Well, the SM57 has a very low self-noise floor (as do most dynamic mikes), and the S/N ratio of the MOTU is very good, so I don’t worry too much about the low level, and I’ll often normalize to help see the wave data better. I’m using 24 bits, of course. Interestingly, when I use my TASCAM mixer’s mike preamps and line inputs I don’t have this problem, but I think the MOTU preamps sound a little better.

Many purists will say never to normalize because it adds quantization noise. They’re right, but you can calculate the amount of noise it adds, and it’s not enough to worry about. (It’s the same as the noise added by putting the fader anywhere but unity gain. Double this noise and it’s still nothing to worry about in most cases.)

Bottom line: There’s nothing wrong with normalization. However, there might be something wrong with NEEDING to normalize. Whether there is or not depends on the quality of your whole input signal chain, including how well you’ve managed the input gain structure. (That is, whether you’re able to run each piece of gear at optimum signal levels.)

Right g69!

How you control your soundcard’s record level depends on what soundcard you have. If you’re using a built-in soundcard, use n-Track’s “View -> Soundcard’s Mixer -> Recording controls”. This just brings up a Windows control panel – one that’s normally kinda hard to find, though!

Great input! Thanks.

I had a chance to experiment today and, sure enough, if I get the peak level to around -3dB, I get a nice fat wave form. Thanks again for the input; it was very helpful.

BUT, I find that tracks recorded at nominal levels at 24 bits (very dynamic drums for example) are still hard to see the way I’d like to see them ocassionally when editing close in. I’d REALLY like a vertical zoom like the kind wave editors have. Dynamic tracks properly normalized still might not show up in the timeline well enough sometimes when zoomed in at a quiet spot.

It’s always something… :D

Right – this is an aspect of Flavio’s algorithm for showing wavefiles that I don’t like. If you load the same wave files into CoolEdit96, those percussion parts show up just fine.

n-Track does something like “sampling” of peaks rather than always showing the peak value for the interval represented by a given column. There’s an advantage to this, it tends to show average peak levels which is often nice to see. However, I’d like a true peak display, where it always repshows the peak value for all samples represented by that pixel column.