+6 db increase vs 10x gain...

Hi guys,
I'm trying to figure this out.  I have a signal coming out of a mixing board effect send.  I am wanting to feed that in to my modular synth and then back out of the synth into the board.  So Modular synth levels are 20v p-p (max).  Wikipedia tells me that +4 dbu gear is line level of 3.472 p-p.

I am plugging the effect send into a module that says the signal is "increased by 6dB".

What does a 6db increase equal to in terms of p-p?  Since 6db is twice as loud, do I just double it to 6.944p-p?

6dB is 4X the voltage (twice the power)

No is not, it's 4X the power! (twice the voltage)

so, 18.888Vpp if you started with a signal that's 3.472Vpp

Note that wikipedia isn't always right, dBu is nominal level. 4dBu is 1.227 V RMS. If your signal is a sine wave you have 3.47Vpp, if your signal is a square wave you have 2.45Vpp

JS

PS: I think my numbers are right, I just messed with the words being a hard day's night...

Mbira said:

What does a 6db increase equal to in terms of p-p?  Since 6db is twice as loud, do I just double it to 6.944p-p?

Click to expand...

6 dB is double the voltage, or a gain of two.

Math:

AvdB = 20 log (Vout/Vin)

or

Vout = 10 ^ (AvdB / 20)

joaquins said:

6dB is 4X the voltage (twice the power)

Click to expand...

Nope, 6 dB is double the voltage. (When talking power, a 3 dB increase is twice the power.)

Andy Peters said:

6 dB is double the voltage, or a gain of two.

Math:

AvdB = 20 log (Vout/Vin)

or

Vout = 10 ^ (AvdB / 20)

Click to expand...

This hurts my head...so does that mean I can just double 3.472 p-p to 6.944?

That would make sense because what is happening is that I am not getting enough gain...

> +4 dbu gear is line level of 3.472 p-p.

Except "+4dB" is a fictional average from VU meter days.

Speech/music peaks will be 10db-18dB higher. 10dB often enough to notice on AM radio. 18dB over may not happen every day. 16dB is a frequent figure for all but very critical work.

16dB is 6.3 of voltage.

3.472 times 6.3 is 22V p-p.

So a "+4dB" speech/music signal will overload a 20V synth's range, maybe a few times a day.

Considering "why?" you might want to pass audio through a synth, I'd say rare clips are not a big deal. And you will know if they are screwing-up the sound. I would have no fear doubling the signal level; more if that isn't ugly yet.

However we used to run TEAC -10dBV levels through the ARP without added gain. Most synth circuits are not level-fussy.

Good job guys...  ;D

yes +6dB is 2x voltage,,,  Since power equation has voltage squared  +6 dBW (watts) is 4x power.

+4 dBu = 1.23v   

While perhaps too much information dB is purely ratios (in fact power ratios but too convenient to not use for voltage ratios too).  dBu is ratio relative to 0dBu or ratio relative to 0.7746V, the same voltage as 0 dBm (1mW) into 600 ohm load,  an archaic legacy audio termination/reference.  +4dBu is 4dB more than 0dBu so 1.58 x 0.7746 = 1.227V

JR

Sorry, brain fart... I'll edit now

JS

Mbira said:

So Modular synth levels are 20v p-p (max).  Wikipedia tells me that +4 dbu gear is line level of 3.472 p-p.

Click to expand...

There's your trouble right there. You're comparing a maximum voltage (synth) to a nominal level (+4 dbu audio) that has headroom.  That's like specifying a ceiling that's either 8 feet high or "comfortable for average height humans", ie also about 8 feet.  Both of these standards really relate back to power supply rails, which most often are +-15vdc in both types of gear.

But what might really be the issue is that synth oscillator outputs tend to be jammed right up against that ceiling, versus musical inputs that average far lower. So you might want to heavily compress the signals going in and then add gain.  Running a typical mixer output up near red on the meters would probably work just fine.

> if your signal is a square wave you have 2.45Vpp

"Audio" uses "sine" almost exclusively.

While RMS of a square has great meaning when powering heaters and incandescent lamps from some non-standard power sources (especially low-price UPSes), we don't do that for audio level measurements.
_____________________________

> power supply rails, which most often are +-15vdc in both types of gear.

+1.

Thanks for all the replies.  I can only say what is actually happening here for us.  I have the QU-16 A&H mixer sending a balanced mix out and the meters are going to 0 (not above) because I don't want to clip and I want to have enough headroom.

Sometimes I send an instrument like marimba and sometimes I send vocals through this "effects send".  This is currently going into an input module that claims to "boost the signal by 6 db".  This boosted signal is now my sound source inside the modular and is feeding effects and filters, etc, and is then going back out the module that "attenuates the signal by 6db" to come back in to our mixer.

The 6 db boost is not enough.  It is too quiet.  I need more gain.  It is like 1/4 the volume of other things in the modular like the VCOs. 

This is a eurorack setup and has +-12v rails.

Push the meters until the sound comes back bent.

Also, turn-down the VCOs. They default to Absolute MAX level. No acoustic source can be this big without clipping.

You probably do not want the -6dB return loss.

You're probably losing 6dB because of the balanced->unbalanced thing.

The Qu has pretty good headroom.  Don't worry about pushing the meters over 0dB.  OK, you might have some audible clipping if you go all the way into the red, but yellow is totally fine.  Modern live sound gear is very forgiving.

Mbira said:

Thanks for all the replies.  I can only say what is actually happening here for us.  I have the QU-16 A&H mixer sending a balanced mix out and the meters are going to 0 (not above) because I don't want to clip and I want to have enough headroom.

Click to expand...

I haven't been paying close attention to your journey.

Apparently the qu-16 is a digital mixer and the meter is calibrated for 0dB (0VU?)= +4 dBu (1.23V).  They spec a max output of +22dBu so a very comfortable 18dB of headroom above the +4 dBu.

The meter is a "fast" peak meter (their description). So setting it so the peaks only nudge 0VU, means you are running a rather cool signal... Old school VU meters were average responding so when pushing 0VU average, the peaks could easily be 10 dB hotter (or more, much more). 

IIRC you are messing with some kind of percussion instrument so the crest factor of peaks to average could be larger than typical sound sources.

Sometimes I send an instrument like marimba and sometimes I send vocals through this "effects send".  This is currently going into an input module that claims to "boost the signal by 6 db". 

Click to expand...

It is common for active balanced outputs to bump gain +6dB (from 2x the output voltage swing.. 1x on each output leg).

This boosted signal is now my sound source inside the modular and is feeding effects and filters, etc, and is then going back out the module that "attenuates the signal by 6db" to come back in to our mixer.

Click to expand...

OK modular means "synth", and this synth drops signal -6dB.. but I thought you just said input module boosts signal +6dB? Is the input module the "synth"?  I am getting a little confused.

The 6 db boost is not enough.  It is too quiet.  I need more gain.  It is like 1/4 the volume of other things in the modular like the VCOs. 

This is a eurorack setup and has +-12v rails.

Click to expand...

I googled eurorack and didn't find any specs.

Since I have long since lost track, I will offer some general advice,, if it isn't loud enough turn it up louder.

The QU meter "peak" indication is calibrated for -3dBFS (so 3dB before digital clipping). Just tickling the -3dB peak lights on occasional musical transients is not clipping.  You don't want to see steady peak lights but occasionally flashing is OK. 

If you are creaming it, it will probably be audible.

JR

I don't think anyone has mentioned unity gain. The easiest way to figure out what is going on is to put up a tone at the beginning of the signal path at whatever reference level you are using and check for level differences along the way. Once you have unity sorted out you can go back and figure out where the clip point of each device is. As long as each device passes signal at unity gain headroom will be an apples to apple comparison.

> mentioned unity gain.

Not a "unity gain" case. His problem is the marimbas are not as loud as the synth's VCOs. He could have perfect +4dBu in and out on the marimbas but they still can't match the +19dBu of the VCOs.

> If you are creaming it, it will probably be audible.

If it isn't audibly dented, it isn't "creamed" yet.

PRR said:

Not a "unity gain" case. His problem is the marimbas are not as loud as the synth's VCOs. He could have perfect +4dBu in and out on the marimbas but they still can't match the +19dBu of the VCOs.

Click to expand...

He said  he was using the synth module as a sound processor for the source from the mixer. I saw no mention of VCO's. I assumed since the sound source was from the mixer the module was more like a VCF, ring modulator,  phaser, etc.

JohnRoberts said:

...I will offer some general advice,, if it isn't loud enough turn it up louder.

Click to expand...

8)

> I saw no mention of VCO's.

Just so everybody is on a same page:

Mbira said:

instrument like marimba ...vocals ... 6 db boost is not enough.  ...It is like 1/4 the volume of other things in the modular like the VCOs.

Click to expand...

(Emphasis added)

PRR said:

> I saw no mention of VCO's.

Just so everybody is on a same page:
(Emphasis added)

Click to expand...

Ahh, then I'd say the solution is probably a compressor on the send. That will help with any level and headroom problems.. Marimba and vocals can have large peak to average ratios.