Relation between Change in dB & Change in percieved loud

Anything about MUSIC but doesn't fit into the forums above.
Post Reply New Topic
RELATED
PRODUCTS

Post

I am trying to trace the Relation between Change in dB and Change in percieved loudnes across the entire dynamic range and want to find out if it is linear or like some would say logarithmic ...

What i mean is if I push tne fader up of a particular track from -20 to -10 and if i perceive a doubling of loudness , should I also perceive a doubling of loudness for a 10 db increase anywhere in the dynamic range like -50 to -60 or even -80 to -70.....

ok ok, i know it is a lot of theory, but its a practical scenario that led me thinking about this.

while mixing a song i have set proper levels for each track such that the balance is perfect. Now i notice that my final LR is only peaking at say -20 db (hypothetical) and the mix really isnt all that loud, although it is perfectly balanced. So i want to get the overall level up by pushing up the levels of all my individual tracks in the mix.

now in my mix some tracks have high levels like -12 dB or -14 dB, while some tracks have low levels like -50dB or -42 dB... Now heres the dilemma

Will i be bringing about a linear increase in overall level, by pushing up the level of individual tracks by say 8 dB , Or is it something like tracks with higher levels require less dB level increase as compared to tracks with lower levels....

So in my example, where i have a mix with tracks of varied levels, can i achieve a linear increase in overall level ( loudness ) by simply pushing up all individual tracks with the same level. Or my mix will lose its balance when i do that....

Hope i have explained myself properly... lookin for some explainations.. :shock:

Post

Check out the Fletcher Munson equal loudness curves. It's a far from simple relationship between amplitude and perceived loudness. Worse than that, you can't dictate to your listeners how loud they will listen to your track, so almost inevitably the listener will not hear the exact same colouration of sound.
Music with dinner is an insult both to the cook and the violinist.

Post

..couldn't you just puch up the master fader, that should increase the overall sound, w/o affecting the balance, tho' as griels correctly asserts that the F-M curves are a consideration.

Probably doesn't help, but I thought I'd step in!
Image

Post

If I understand correctly, for a given signal, applying 6dB of attenuation, halves the volume. (So the "10dB" in the original post is mistaken.)

I think the original question can be phrased:

Does summing a stereo signal that peaks at -6dB with one that also peaks at -6dB result in a mixed signal that peaks at 0dB?

And the answer must be "probably not". It doesn't take into account where the peaks are: the two parts might not even be playing at the same time.

What I'm not sure about is if you sum two continuous signals - say a 440Hz tone and an 880Hz tone - both with a -6dB level, so you get 0dB then? Does the difference in pitch make any difference?

Anyway, I'd recommend just normalising the overall mix, especially if it's from a purely digital source.

Post

pljones wrote:If I understand correctly, for a given signal, applying 6dB of attenuation, halves the volume. (So the "10dB" in the original post is mistaken.)
hi jones.. that was just an assumption. i am aware that 6 db increase double loudness... i was tryin to convey a certain level change bringin about a certain change in loudness.. thats all... all figures stated to be considered hypothetically..

yes i do have certain obvious solutions like normalising , compression etc. But i am not considering these for this particluar situation.
duncanparsons wrote:..couldn't you just puch up the master fader, that should increase the overall sound, w/o affecting the balance, tho' as griels correctly asserts that the F-M curves are a consideration.

Probably doesn't help, but I thought I'd step in!
thanks duncan for checkin in :wink: .. yes yours is a perfect solution... and it will work perfectly to the extent that i have headroom on my LR which is normally around 6 to 12 db in most software.

i'm okay with it, and i do that at times... but i'm sure hardcore engineers will agree that pushing up LR , and keeping down the channel levels is like murdering the signal to noise ratio... Besides it can also happen that i may not have enuf headroom.. so then now how do i deal wih it..

The thing i'm trying to crack here is , if i push up all the channel faders say by a certain level, will my LR fader go up by a proportionate level. will the relation between the loud, moderately loud and soft instruments in the mix remain the same, or to rephrase will my balance remain unaffected.

well after all my reading, i think it wont. pls correct me if i am wrong. Since certain instruments are at a lower level and certain are at a higher level. so i will have to fine tune my balancing after pushin up the levels of all channel.

Besides this is what happens to me practically. after pushin up all channel levels, i am never satisfied with the balance and always make small changes// But i was just wondering if theoretically every channels' level must increase proportionately, irrespective of whether it was at a higher or lower level in the mix.
:shock:

Post

hmmm..Master fader is a multiplier, not an adder - that is, taking is up 6dB will stretch what was taking up -6dB, will now go all the way up to 0dB, and being a multiplier, all the relative proportions will be maintained. It is true that any background noise will become more 'there', but unless there is something quite radical, it will probably not be percieved..

:)

DSP
Image

Post

The background noise will increase wheather you use the master fader or the individual track faders. But it will still be in the same relative volume to the sigal.

You can't change the signal to noise ratio by moving a fader on an already recorded sound. You can only effect signal to noise at the time of recording.

Post Reply

Return to “Everything Else (Music related)”