Taming an uneven kick drum

analysers are part of the story in a different way

- they are showing a spectral view of the signal which is different to what the meters on each channel in your daw are showing (they show an instantaneous level if a Peaking meter or an average level if an RMS meter)

- the signal has been decomposed into discrete bands of frequency energy (as many as 512 bands which is about 0.235 of a semitone compared to the usual 31 bands of the 1/3 octave (4 semitones per band) graphic equaliser over the range 20-20000Hz) - i.e. the signal is so decomposed that, perhaps, it shows too much information rather than too little

any spectral view of our signal will show high bass levels and high high frequency levels and comparatively lower levels between 400Hz and 4kHz, if we have mixed well - simply because we have mixed to the response of the human ear - if everything is the same level then we have made some error in our mix, meaning we need to start again
 
Thanks. I do understand how an analyzer works. Also my previous question wasn't about mixing, or why the frequency shape is the way it is in a correct mix. I don't understand why, in order to calculate how much headroom I have left before 0db, I must add up all the sounds in different frequency ranges in order to get an accurate picture, when the analyzer accurately represents how all tracks are not equally contributing to the problem in equal increments?

Also, if the picture has been wrong all this time and we should measure loudness differently e.g. your post about 0dbrms, then I guess I need to recalibrate the meter on the analyzer? But have no idea how to measure the rms and spl levels, I guess I'll look in the settings...

Cheers
 
Last edited:
on your spl levels: do you have a sound level meter? that is the only way to know for sure what your current mix setup is generating; I have two, one slightly more complex than the other and both measure my setup at 83dbspl @ 1m for my nominal 0dbrms of -12dbfs

even the cheap ones from the Chinese manufacturers/sellers on ebay are good enough for the job of checking your home studio levels (both of mine fall into this category)

it's actually not that the analyser is giving a false picture but because it is showing an un-weighted view of the signal it seems to show there is plenty of headroom for some things when there isn't...

i.e. we know that we need bass energy be about 500 times as loud as mids and highs, which is roughly 27db higher than the other parts of the spectrum, we also know that the spectrum above 4kHz starts to require the same kind of energy (+10db - +20db above 1kHz reference point) to be heard as being equally as loud. If we have done our mixing right we can see this on the analyser.

- what happens next though is that when we sum all channels the summing process is not concerned about what the frequency range is only what the signal energy is on each channel, so you cannot rely on the analyser to tell you you have the summing mix correct only that you have your frequency balance correct

ideally, the two should meld into a perfect unison of expression, but we have all heard the stories of those who try to use high numbers and then end up clipping the whazoo out their mixes as a result of trying to push the loudness envelope

starting with much lower average channel levels (i.e. a nominal 0dbRMS value of -12dbfs or -14dbfs or -20dbfs) gives your mix a chance to actually achieve it's full potential

However, we know from watching some youtube vids that there are "producers" out there who think you need to have the entire spectrum flat across the board or, worse, the infamous scooping out the low mids to make the mix less muddy (inverse philosophy of make everything the same levels - in both cases it is because they do not understand the scientific basis for what they say supports their position or they misapply the science to what they are doing

i.e. very few folks here seem to understand that sound levels are logarithmic, that is, that a db value is an exponent (a power) of 10, not a linear value.

E.G. a shift upwards of 10 db is the same as saying increase the level linearly by a factor of 10[sup]10/10[/sup] which is 10[sup]1[/sup] or simply 10.

The division by 10 in the exponent is important as the deci-bel is 1/10th of a bel, so to do any conversion to a linear scale we need to bring teh value back to bels first.

Shifting upwards by 20db is the same as saying a linear increase of 10[sup]2[/sup] or 100 fold
 
Last edited:
Don't know how you make the time to help people so thoroughly, Bandcoach... Our Youtube searches are often dictated by misled ambitions, and your input really helps.

Could you please clarify: if we were to start with a (not so) arbitrary -14db (with the kick being -14 and other elements proportionately lower to that) and so have that much to play with, how many db of our own processing can we do before mastering?

What of the group processing e.g. plugins that will raise the levels of, say, all the combined pad sounds? And the one applied to individual sounds? Is it a matter of creative or rather technical choice to only keep subtle plugins and remove the one used for creative effect? Or is there a formula that, for example, says "keep individual compression/eq/warming, but remove the ones on group buses before sending off for mastering"? In short: should my -14db already include individual + group effects (but not master ones), or do those come on top after that mark? If yes, by what margin should the levels increase from a technical standpoint, if such a concept exists at all... I don't insist on getting answers to each question, just trying to illustrate the lack of knowledge best to avoid asking again ;)

Thank you and a very happy new year from from Moscow to Sydney.
 
thre is no right or wrong answer to your questions other than that you want to leave some headroom, even -0.1dbfs as your peak level leaves headroom

the only way to get an answer that is compatible with your goals is to ask your mastering engineer before finalising your mix

imho, any processing that improves your mix should be left intact as mastering is as much about preparing for your distribution medium as it is about sonic adjustments - i.e. the sonic adjustments are project wide settings to bring everything into the same acoustic/sonic signature

as for finding time to help others; the best way to learn something is to try and teach it: if you can teach it well then you truly understand it: i.e. it is my way of maintaining and improving my knowledge base: I learn new things along the way almost everyday (in some cases they are not so much new as new understandings are forged, which is important as well)

happy new year back to Moscow from Sydney
 
Last edited:
Everything's been made abundantly clear, thank you! I've felt what you said to be true, but your reassurance was invaluable.
 
You must log in or register to reply here.
Back
Top