Setting two db values relative to each other by percent

[chanceoperation]

Hi All

I've been reading the "Going Deep" article in SOS April 2025 and trying to implement the relative strengths of early reflections vs reverb for the three reverb settings named AMB A, AMB B, and AMB C, where the ER is given as a percent relative to reverb:

AMB A: ER 65%-75%
AMB B: ER 45%-50%
AMB C: ER 25%-36%

But the reverb I am using (Nimbus) does not have one relative dial like some reverbs do. It has two level dials, one for the early reflections and one for the reverb. These two dial are in db, with 0db being the highest value, and the dial moving into negative territory.

My math stops well short of comparing a pair of logarithmic scale values as a percentage one of the other.

So what db values (<=0) do I set the ER level and Reverb level for each of AMB A, B, and C, keeping in mind that the combined db level across all three setting will be balanced?

Andrew Culver

[Tomás Mulcahy]

Does it say what kind of dB? Or does it have a meter scale with 0dB at the top?If so it is dBFS so +3db is a doubling, -3dB is halving. So you can go from there.

[Hugh Robjohns]

Oh dear Tomás! No iced buns for you at tea time! You're thinking of power ratios. For signal levels a halving or doubling is 6dB.... even in the digital world!

Here's a list of signal level percentages and their decibel equivalents:

Percentage = decibel

5% = -26dB
10% = -20dB
15% = -16.5dB
20% = -14dB
25% = -12dB
30% = -10.5dB
35% = -9dB
40% = -8dB
45% = -7dB
50% = -6dB
55% = -5.2dB
60% = -4.4dB
65% = -3.75dB
70% = -3.1dB
75% = -2.5dB
80% = -1.9dB
85% = -1.4dB
90% = -0.9dB
95% = -0.45dB
100% =  0dB

If you want to calculate a different value the formula is:

dB = 20 x Log (percentage/100)

[chanceoperation]

So:

An ER setting of -6db vs a reverb setting of 0db means the ER has a level that is 50% of the reverb. Is that right?

What about 65%, or 30%? What is the formula to use to calculate any desired percent of relative level?

[Hugh Robjohns]

Your post overlapped mine while I did the sums and edited my post to insert the results... see above!

[Wonks]

20 x Log10 of the ratio expressed as a decimal

So for 50 percent, 20 x Log10 of 0.5

= 20 x -0.301

= -6dB

30 % = -10dB
(just taking the main dB figure)

[chanceoperation]

Wow. That's great Hugh and Wonks.

Where does the 20 come from?

[Hugh Robjohns]

Its because the decibel was originally conceived to describe power ratios, not amplitudes. The former is the square of the latter and, consequently, it is necessary to multiply the log value by 20 to generate meaningful numbers. If calculating power ratios the formula is:

dB = 10 x log (P1/Pref)

[ef37a]

Do children still do logs in school?

Dave.

[chanceoperation]

From Wikipedia article on "Decibel":

Two principal types of scaling of the decibel are in common use. When expressing a power ratio, it is defined as ten times the logarithm with base 10.[5] That is, a change in power by a factor of 10 corresponds to a 10 dB change in level. When expressing root-power quantities, a change in amplitude by a factor of 10 corresponds to a 20 dB change in level. The decibel scales differ by a factor of two, so that the related power and root-power levels change by the same value in linear systems, where power is proportional to the square of amplitude.

[alexis]

Do children still do logs in school?

Dave.

"Naturally"!

[ef37a]

Do children still do logs in school?

Dave.

"Naturally"!

"e" by gum!

Dave.

[Drew Stephenson]

We did logs in school. I revisited them at uni, and I think I've only used them once since for anything other than conversations about decibels.

[chanceoperation]

I've been mixing all afternoon with this 3-reverb method as described in the "Going Deep" article and it's really quite brilliant. It's the first pure theory mixing technique I've ever used that actually does what it says on the tin. Highly recommended.

[Martin Walker]

Yes, I set up a three-depth system last week, and got very good results, being able to place everything close, mid or distant within the same space.

Previously I just employed two reverbs - one for a small room which was mostly early reflections, and the other a hall for more distant stuff, but this new system sounds a lot more believable.

[Tomás Mulcahy]

Oh dear Tomás! No iced buns for you at tea time!

That's me told

[chanceoperation]

Now that I am the expert , let me draw your attention to two things mentioned in the article that might be overlooked but are essential.
First, use exactly the same reverb algo on all three sends (change only the pre-delay, filter cutoff, and ER / reverb relative levels).
Second, do reduce the widths of the middle and far auxes after the reverb outputs (the farthest is narrowest).
These are part of the secret sauce I believe.