Logic levels spike on attack - compressors don't seem to work

[akpasta]

Hello,

I'm noticing that whenever I record a track, any instrument, at whatever volume-- even low volume-- when the signal first goes from silence to attack, the volume is way higher and sometimes makes a clipping noise, even if it's well below the threshold. In order to get clean tracks I have to give some signal during the click-in and time it right so the first hit isn't super loud.

I'm using an Apogee Element 24, I even use the soft-limit and it doesn't make much of a difference. I've also tried using the limiter in the compressor tool in Logic and the Limiter plug-in, but they don't seem to function as they should, they don't have an effect on the attack. On the compressor I've even moved the 'attack' all the way to 0ms and the limiter to -10 or however low it goes. It doesn't seem to have much effect.

My attack levels spike at like -12 and then the normal signal is like -24 to -30 or so. So I'm well below the clipping threshold, but the attack is spiking in a weird way and I can't figure out what's going on.

[vvv]

I've no idear what's going on with your system, but the only time I've ever seen that is when I have a compressor's attack set too slow ...

[akpasta]

I've no idear what's going on with your system, but the only time I've ever seen that is when I have a compressor's attack set too slow ...

Ya, I've got the attack set to 0ms, and the threshold at -10 so it should be on all the time.

[A.David.MacKinnon]

Seems like the problem is on the input. If that's the case a plug in comp/limiter won't do much because the clip is already there.
It almost sounds like the issues you can get with a cold solder joint. No signal until something loud jolts the connection enough to keep it connected.
I'd try a different recording interface to see if that's the issue. Then I'd try different recording softwere to see if the problem is a Logic issue. After that I'd work my way through everything in the chain until I found the problem.

[digitaldrummer]

and if you haven't already, you may want to log a ticket with Apogee (or however they work..). Maybe they have ideas.

[akpasta]

So this is potentially a little embarrassing (although I have a decent enough excuse), when I examined things more closely and carefully I believe I've traced the issue down to being caused by using an MXR Dynacomp between the guitar and the interface.

I also use a lot of electric 12 string, so compression is a must. I always run a dynacomp into an amp or direct. It's virtually imperceptible to the ear, or with analog recording, but evidently the "attack" on the dynacomp is not quite fast enough for digital recording and the signal doesn't compress fast enough when you strike a note from total silence. I never tried testing back and forth with the pedal on or off because my ear could never detect a difference!

I suppose my next question is, is there anything I can employ as a workaround so that I can keep using the dynacomp, which I like very much except for this one issue, or should I explore compressor pedals that have more control over the attack parameter?

Thanks!

[A.David.MacKinnon]

You might be better off doing the compression in two stages. Dial back the dynacomp a bit and do the rest of the compression in Logic with something that has more controls.
Or use a tube amp. Tubes and a speaker will help round out spike-y signals.

[Nick Sevilla]

Now that you found part of the issue, there is a solution.

You need to learn all about gain staging. Do it now.

If you limit (that is the correct term) your incoming signal, you will end up messing it up, you won't like the sound, since
you are not doing this before you began recording your guitar signal.

The solution is learning the dynamic range of the guitar sound that you like.

Clearly, you were sending a signal that was too loud to begin with to the Apogee. You may want to insert a gain pedal or a console channel or anything that can enable you to just turn the entire signal DOWN before it hits the Apogee converter. So that is maintains the sound you like while allowing the correct level to get to the converter, and not make it distort.

Not another compressor, and not a limiter. No.

[akpasta]

Now that you found part of the issue, there is a solution.

You need to learn all about gain staging. Do it now.

If you limit (that is the correct term) your incoming signal, you will end up messing it up, you won't like the sound, since
you are not doing this before you began recording your guitar signal.

The solution is learning the dynamic range of the guitar sound that you like.

Clearly, you were sending a signal that was too loud to begin with to the Apogee. You may want to insert a gain pedal or a console channel or anything that can enable you to just turn the entire signal DOWN before it hits the Apogee converter. So that is maintains the sound you like while allowing the correct level to get to the converter, and not make it distort.

Not another compressor, and not a limiter. No.

Thanks for the response. I've read about gain staging, and gain unity. So I think I get it. Ultimately it's a matter of keeping levels down at a safe level at all stages of the process, and that's easy. I suppose the area I'm falling down is that I'm not paying attention to gain staging at any point before I hit my DAW interface, and the signal before my compressor is too far above what my compressor is pushing out.

I also see the point you're making that there's no work-around in the DAW to fix a signal that's too hot somewhere before the DAW.

I suppose one solution would be to ensure the signal going to my compressor pedal is about the same level as the signal going out of the compressor, so it does not have to limit the signal so much. When mic'ing an amp, that would mean making sure the sound coming from my amp was roughly the same level with the pedal on or off (I'll try this and see how it goes). However, I notice this is also an issue recording "direct," ie when my guitar goes straight to the compressor, and then straight to the DAW. How would you manage gain-staging in this regard? I can't very well turn my guitar volume knob down and it seems the signal from my pickup shouldn't be producing gain that is above what the pedal's attack can handle. So this is one area where I'm still confused.

Perhaps it's something bout the MXR Dynacomp where it is also amplifying the signal prior to compressing it? I dunno. For the record, I was not suggesting an additional compressor, but a different one, that had more control over the settings, such as attack.

[floid]

https://www.electrosmash.com/mxr-dyna-comp-analysis

Scroll down to the section on the envelope detector block and the 150k resistor.

[Nick Sevilla]

Thanks for the response. I've read about gain staging, and gain unity. So I think I get it. Ultimately it's a matter of keeping levels down at a safe level at all stages of the process, and that's easy. I suppose the area I'm falling down is that I'm not paying attention to gain staging at any point before I hit my DAW interface, and the signal before my compressor is too far above what my compressor is pushing out.

I also see the point you're making that there's no work-around in the DAW to fix a signal that's too hot somewhere before the DAW.

I suppose one solution would be to ensure the signal going to my compressor pedal is about the same level as the signal going out of the compressor, so it does not have to limit the signal so much. When mic'ing an amp, that would mean making sure the sound coming from my amp was roughly the same level with the pedal on or off (I'll try this and see how it goes). However, I notice this is also an issue recording "direct," ie when my guitar goes straight to the compressor, and then straight to the DAW. How would you manage gain-staging in this regard? I can't very well turn my guitar volume knob down and it seems the signal from my pickup shouldn't be producing gain that is above what the pedal's attack can handle. So this is one area where I'm still confused.

Perhaps it's something bout the MXR Dynacomp where it is also amplifying the signal prior to compressing it? I dunno. For the record, I was not suggesting an additional compressor, but a different one, that had more control over the settings, such as attack.

Sure.

I would keep the Dynacomp, and the rest of the signal chain, IF it is the SOUND YOU WANT.

The issue then becomes one of bringing down the final level so that it never ever distorts the Apogee converter. I usually accomplish this by having the final gain stage be a console channel with a fader, so I can easily control the levels to the converter inputs.

This is not a matter of anything other than a final volume control at the end of your signal (and presumably the tone you like) before the converter.

Cheers

PS yes, learning gain staging throughout your signal chain is only a good thing. At the very least, you learn which devices are working, and which are not, to your liking.

[joninc]

I suppose my next question is, is there anything I can employ as a workaround so that I can keep using the dynacomp, which I like very much except for this one issue, or should I explore compressor pedals that have more control over the attack parameter?

the pedal has an output knob - turn it down