K Scale Calibration in Ableton

[junqueira]

I'm trying to calibrate my studio in Ableton using the K Scale. I downloaded Bob Katz -20db pink noise file. When I load it into a track and play, it peaks on the master fader around -10/-11 db. To play a mastered file at about the same level, I have to turn it down to around -14db.

My question is why does it not play back in Ableton at -20db? What's the point of calibrating to -20db it's going to peak at -10db? When I record vocals, am I shooting for -20db or -10db?? So confused...thanks!

[Stromkraft]

I'm trying to calibrate my studio in Ableton using the K Scale. I downloaded Bob Katz -20db pink noise file. When I load it into a track and play, it peaks on the master fader around -10/-11 db. To play a mastered file at about the same level, I have to turn it down to around -14db.

My question is why does it not play back in Ableton at -20db? What's the point of calibrating to -20db it's going to peak at -10db? When I record vocals, am I shooting for -20db or -10db?? So confused...thanks!

The -20dB levels are RMS levels, not peak levels. Look at the graph of the Live meters:



See that clear green bar? That's the RMS level. The faint darker green is the dynamic range up to the peak levels, which is the value you see at the top of the Live meters.

Here's the great plug-in Sonalksis FreeG that also have numerical readouts of RMS levels illustrating this set of values.

[junqueira]

I'm trying to calibrate my studio in Ableton using the K Scale. I downloaded Bob Katz -20db pink noise file. When I load it into a track and play, it peaks on the master fader around -10/-11 db. To play a mastered file at about the same level, I have to turn it down to around -14db.

My question is why does it not play back in Ableton at -20db? What's the point of calibrating to -20db it's going to peak at -10db? When I record vocals, am I shooting for -20db or -10db?? So confused...thanks!

The -20dB levels are RMS levels, not peak levels. Look at the graph of the Live meters:



See that clear green bar? That's the RMS level. The faint darker green is the dynamic range up to the peak levels, which is the value you see at the top of the Live meters.

Here's the great plug-in Sonalksis FreeG that also have numerical readouts of RMS levels illustrating this set of values.

Thanks for that explanation! So when recording vocals, what's a good level to aim for? Should I be focusing on RMS rather than peak?

[Stromkraft]

I'm trying to calibrate my studio in Ableton using the K Scale. I downloaded Bob Katz -20db pink noise file. When I load it into a track and play, it peaks on the master fader around -10/-11 db. To play a mastered file at about the same level, I have to turn it down to around -14db.

My question is why does it not play back in Ableton at -20db? What's the point of calibrating to -20db it's going to peak at -10db? When I record vocals, am I shooting for -20db or -10db?? So confused...thanks!

The -20dB levels are RMS levels, not peak levels. Look at the graph of the Live meters:



See that clear green bar? That's the RMS level. The faint darker green is the dynamic range up to the peak levels, which is the value you see at the top of the Live meters.

Here's the great plug-in Sonalksis FreeG that also have numerical readouts of RMS levels illustrating this set of values.

Thanks for that explanation! So when recording vocals, what's a good level to aim for? Should I be focusing on RMS rather than peak?

In my opinion, yes. What does the K meter methodology suggest? Personally I've selected -18dBFS RMS as my starting point when tracking. In which I'd include Live instruments and plug-in instruments as well as any external audio sources.

-20dBFS RMS for vocals makes sense to me as the frequency content is a factor for how loud a specific RMS level feels. Higher frequency material will appear somewhat louder in the mix.

While I very seldom need to worry about peak levels this depends on the singer's typical dynamic range and the microphones used as well as whether you're using a compressor on the input signal or not (before the track).

I don't find Live particularly helpful for this way of working, but it's very doable.

[2pauluzz2]

Not to push anyone to buy additional plugins, but just throwing this in here in case anyone reading this uses Fabfilter products: their limiter has very useful k-scale metering.

[Tarekith]

Most of the Voxengo stuff does too.

[2pauluzz2]

Oooh I just found a very nice one: http://forum.cockos.com/showthread.php?t=176243 (https://github.com/Rolias/KMeter-OSX).

Hats off Rolias!

[fishmonkey]

i like this one:

http://www.meterplugs.com/kmeter

[junqueira]

In my opinion, yes. What does the K meter methodology suggest? Personally I've selected -18dBFS RMS as my starting point when tracking. In which I'd include Live instruments and plug-in instruments as well as any external audio sources.

-20dBFS RMS for vocals makes sense to me as the frequency content is a factor for how loud a specific RMS level feels. Higher frequency material will appear somewhat louder in the mix.

While I very seldom need to worry about peak levels this depends on the singer's typical dynamic range and the microphones used as well as whether you're using a compressor on the input signal or not (before the track).

I don't find Live particularly helpful for this way of working, but it's very doable.

I know you stated that you don't generally worry about peak levels, but do you have a peak level limit that you won't go past in the mixing phase? I guess if you're going for a 'loud' master, you want to try to get your peak levels pretty close to your RMS levels

[Tarekith]

The usual recommendation is to leave roughly 6dB of headroom on the master meter of your DAW, that's peak not RMS. Just to make sure you're not clipping.

I have no idea how you would get a peak value close to RMS values unless you're talking about sine wavs.

[2pauluzz2]

I think he means trying to get the distance between peaks (transients) and the average level of a track close together, which is also what I use as a rough guideline.

I'm not sure how to interpret your comment Tarekith; did you interpret Junqueira's comment literally, as in, get the actual values close together?

[Tarekith]

I'm still trying to intrepret it actually

[2pauluzz2]

Excuse the low res pic:



He's referring to the 'crest factor' / getting the transients that will be pushing into your limiter, to be close to the average level of the track.

[Stromkraft]

In my opinion, yes. What does the K meter methodology suggest? Personally I've selected -18dBFS RMS as my starting point when tracking. In which I'd include Live instruments and plug-in instruments as well as any external audio sources.

-20dBFS RMS for vocals makes sense to me as the frequency content is a factor for how loud a specific RMS level feels. Higher frequency material will appear somewhat louder in the mix.

While I very seldom need to worry about peak levels this depends on the singer's typical dynamic range and the microphones used as well as whether you're using a compressor on the input signal or not (before the track).

I don't find Live particularly helpful for this way of working, but it's very doable.

I know you stated that you don't generally worry about peak levels, but do you have a peak level limit that you won't go past in the mixing phase? I guess if you're going for a 'loud' master, you want to try to get your peak levels pretty close to your RMS levels

Yeah, I usually mix into a Limiter that is set to -0.3dB and a gain increase for things I expect to pseudo-master myself so that's my roof. For things I are about to give to a proper master engineer it's typically -6dB or even less if I remove the Limiter.

I don't worry that much about crest factor in the final mix, but depending on material it's 7-12dB in most final mixes. I do look at crest factor on individual tracks and use techniques like parallel compression.

Typically a crest factor less than 6dB sounds like a mush. Thankfully I more often than not have already realized I have gone wrong before that. Of course, it always depends on the material and the role they play in the track.