What's in a sine wave?

  • Thread starter Thread starter ritelec
  • Start date Start date
R

ritelec

New member
Hi all.

So I'm getting some levels on drum mics.
In the past I've been working at getting the levels like -6 to -3 db.
In those takes and takes I've imported to work on (practice and learn) from other peoples sessions, there would be nice, big, strong, beautiful, spread out, wave forms in the tracks.
In reading around, I caught an article some time back about setting peaks like -6 db. I read in this forum about setting levels -18 to -12 db. It explained about the overhead you would be able to work with setting at those levels.
I have noticed with all these plugs I have (eq,comp.etc...) that when working with those strong like bull waves and the waves I get when I set to the -6 to
-3 that it does get a bit tight in that "head room" area as mentioned in the articles.
So I gave it a shot and set inputs between the -18 and -12db, and worked with the plugs. It did give that headroom to work with. I didn't get all cramped in. But those waves it produced were teeney, tiney, laughable. I was able to expand them vertically and horizontally, but they still were weak, scrawny, lame, again laughable. They didn't quite look like the waves I've been getting or like the waves on all these tutorials I've been looking at for the past two years look like.

Cut the difference? -12 to -6? Set to peak at -12 and deal with woosey visuals?
Any thoughts?

(003 rack ptle v8 and logic pro 9)


Rich
 
ritelec said:
Any thoughts?
Click to expand...

quit looking at it and listen to it. is the signal distorting in any way? if so, turn it down. is it too quiet? if so, turn it up until it clips then turn it down a little.
 
? maybe I read to much.

lol


Don't look at it! Just eat it!

lol
 
ez_willis said:
quit looking at it and listen to it. is the signal distorting in any way? if so, turn it down. is it too quiet? if so, turn it up until it clips then turn it down a little.
Click to expand...

Exactly.... listen use your ears! The meters probably aren't that accurate.
 
thats the beauty of music everyone has there own way of doing things what works for one person may not for another, also depends upon equipment, vocalist, instrument, room acoustics, etc... many many variables

in regard to recording drums, i usually put alot of emphasis on the kick, i find that a good sounding kick drum can take a not so high quality track a LONG ways, and vice versa, a crappy sounding kick can make a HQ track sound weak.

I like to record kicks raw (no signal processing) at around -8 and then once its in my tascam i transfer it to HD. I then throw a compressor on it and I often compress the bejeezus out of everything below 100hz everything above I bypass and then I usually come back with a just a tad of EQ about +6db @ 160hz with a 12db/octave slope.

makes for a nice sounding warm kick but thats what my setup entails, I couldn't even begin to add up all the sleepless nights I invested in my music playing around with the kick just to come to that.

but like im saying what works for me may not work well for you at all

snares -12 It seems as though the drummers I know aren't really consistent with there snares so i leave head room

hats -14 ditto

toms -8

cymbals -10

again thats my reference point, it takes a little finagling to get her dialed in dependent upon the drummer
 
There isnt a use for a sine wave...cept to sync video to audio as far as Im concerned.
 
Or analyzing your tone (brass player).

Those levels are probably assuming that you're recording in 24 bit with a higher dynamic range than 16 bit. I tend to favor -12, since my mics tend to have a high noise floor. But I also record in DSD which has a higher dynamic range that 24 bit. Ultimately dithered to 24 bit and 96 kHz since I don't have any super computers just yet. Are you recording to BWF, or MP3? Pretty much anything MP3 sounds teeny, and that's just how it sounds on the laptop. I shudder at the thought of ever using an iPod.

I prefer to have 24 bit and 96kHz sources minimum before editing to any degree. EQ, effects, and such. Once the edits are done, then dither that down to whatever deliverable. If I had better computer tech, I'd probably work with 192kHz, but I do occassionally like to finish things in my lifetime. If there was more software with DSD support, It'd be nice to edit in that. But I'm probably not in the market for a raid array just yet. Although I do seem to have several TB of storage capacity these days. And an 8GB stick isn't quite big enough to transfer resultants between computers and clients.
 
darrin_h2000 said:
There isnt a use for a sine wave...cept to sync video to audio as far as Im concerned.
Click to expand...

It's all sine waves these days . . . see Fourier's theorem ;)

OP, you are fine. Record with peaks at -12dBFS, do your normal mixing, in the end you'll be close to full scale (0dBFS).
 
Shadow_7 said:
Those levels are probably assuming that you're recording in 24 bit with a higher dynamic range than 16 bit. I tend to favor -12, since my mics tend to have a high noise floor.
Click to expand...

If your mics have a high noise floor, then it's less important to record hot. The goal of gain staging is to get the noise from the previous stage 10dB or so above the next stage. Given two mics with the same sensitivity but a 10dB difference in noise, you would need 10dB less gain from the preamp to achieve the same signal-to-noise ratio at the converter. That would mean your peak level would be 10dB lower.


But I also record in DSD which has a higher dynamic range that 24 bit.
Click to expand...


Yikes. Actually, DSD has less dynamic range, since it only has one bit! A crazy amount of noise-shaping pushes all that noise into inaudible regions, but the net result is still about 24dB less dynamic range in the audible band. But this isn't the place for a DSD debate, and either DSD or 24 bit PCM will exceed the amount of available analog dynamic range.
 
I'm not talking about making the noise floor on two mics match. But rather lowering the total noise floor by running the preamp levels a little hotter than you might otherwise do(-12dB instead of -18dB). Therefor lowering the total noise floor. Not always possible and a little risky (pesky drummers switching from brushes to tree logs). But I generally can't compensate with proximity since I record events from an audience perspective, not make events happen.

As far as the DSD debate, the manual for my Korg MR-1000 says that DSD has a higher dynamic range than 24 bit PCM. It even has a 3dB pad in the menu so that when you convert it to 24 bit, you don't exceed it's range. Or I could be wrong. It's why I switched from the PCM modes to the DSD modes on that unit. If it's not the case, then I could finally whipe windows off that one last machine since I wont have to use audiogate. Not to mention saving many GB of drive space. DSD runs at about 1GB per 12.5 minutes for stereo audio.
 
So size doesn't matter? (that's what she said) Use your ears. (that would be a sight).

Actually, I was trying to insert a audiosuite plug (ultra pitch on the snare) and it wasn't working. Went back and recorded snare track to another track with levels cranked. Ultra beat was able to pick it up with larger wave (but also all the noise and bleed through).
 
Shadow_7 said:
I'm not talking about making the noise floor on two mics match. But rather lowering the total noise floor by running the preamp levels a little hotter than you might otherwise do(-12dB instead of -18dB). Therefor lowering the total noise floor. Not always possible and a little risky (pesky drummers switching from brushes to tree logs). But I generally can't compensate with proximity since I record events from an audience perspective, not make events happen.
Click to expand...

The issue there is whether preamp noise, microphone noise, or ambient noise is dominant. Preamp noise should not dominate if you're using condenser microphones. For quiet condensers, ambient noise will dominate. For noisy condensers, it could be ambient or microphone noise. In either case, the setting of preamp gain (within reason) won't make a lot of difference.

If a converter has more than 110dB of dynamic range (110dB is not very good for a modern converter), it's highly unlikely that there is a difference in signal-to-noise ratio as a result of the converter's noise from a peak of -12dBFS to -18dBFS--that would correspond to dynamic range of 98dB and 92dB. Again, the analog signal is not likely to exceed either figure.

As far as the DSD debate, the manual for my Korg MR-1000 says that DSD has a higher dynamic range than 24 bit PCM. It even has a 3dB pad in the menu so that when you convert it to 24 bit, you don't exceed it's range. Or I could be wrong. It's why I switched from the PCM modes to the DSD modes on that unit.
Click to expand...

Could be true for that unit. Again, no 24 bit converter achieves anything close to theoretical dynamic range (mainly because of analog noise).

Consider that all modern converters are generating more data than either 24/96 PCB or DSD. Take, for example, this top-range chip from TI:

http://focus.ti.com/lit/ds/symlink/pcm4222.pdf

Its "native" rate is a 6-bit at 6.144MHz, that's 36.861Mb/sec. The output rate can be either 24/192 or DSD, which are 4.608Mb/sec and 5.644Mb.sec, respectively. So either output format is already a conversion from a much higher true sample rate. Note that the dynamic range is identical.

One of the reasons DSD is thought to pehaps sound better than PCM in real world applications is that less on-chip DSP is required to maintain quality in the audible range. In other words, there is a sample rate conversion going on, and a really good SRC to PCM is more taxing than the SRC to DSD. I don't know if that's true, but it would support why offline conversion from DSD to PCM could be higher quality than direct PCM out of the same chip, because offline SRC doesn't have to be realtime and could access far more power as well. Of course, you need a really good SRC to benefit.

Interestingly, that chip also supports direct output of its modulator, the huge 36.861Mb/sec could be recorded and any SRC done offline and only one, if your hard drive can take the punishment!
 
You must log in or register to reply here.
Back
Top