Hey guys, I'm producing a beat and I really like the overall sound of it with an exception for this guitar piece (I'll upload a small clip below). If you notice there is a huge spike in volume with the initial transient with the wave form confirming it visually along with my db meters. I've stuck a pro c 2 on the mixer track with the fastest possible attack time and tried different ratios but I can't get rid of that spike.

The levels of the piece are fine except for that first transient. Am I missing something in my approach? A transient designer? Another compressor? Bare in mind I'm quite a noob compared to some (maybe most) of the people on here so go easy on me if I've neglected doing some thing obvious.

Theres a small chunk of the guitar I've recorded below (Starts about 2 seconds in). You can download it aswell if you feel the need to analyse it in your own Daw. I've also included a screenshot of the transient which is bugging me.

Audio Clip:
https://instaud.io/3Phb


Visual of the transient spike:

 

Zoom in and edit the volume of that transient.

 

Didn't really think of that. Perhaps automation would work too, same principle really.

Don't suppose you'd be able to tell me why it's happening in the first place?

 

Looks like it was recorded (or bounced) with a compressor set wrong. Slow attack and very slow sustain.

Just zoom in and edit the volume. It's easier/quicker than automation. Cut, cut and pull down the clip volume (I guess Fruityloops works like all other DAWs).

 

This often happens when you render a file and use plugins with a slow attack time, like for instance tape emus.

I had no prob catching this spike with Reaper standard limiter and 5ms look ahead.

You can also make a small fade at the beginning of the rendered file.

If this spike is caused by tape emu or kind an automation will only work for the last aka main fader.

 

Apologies for my ignorance guys. The file I uploaded was from Edison just to give you a visual of the wave form. However the actual 'guitar' part is in midi. The attack time on pro c 2 is at it's fastest. I also tried to apply some lookahead an have also got the same results. Increasing the ratio just brings it all down but the initial transient spike is still there. Adjusting the threshold has no effect also in reguards to this one note being a pest...I'm more confused than anything as the levels after that spike are perfect.

 

Are we not supposed to limit it in order to preserve dynamic range? Correct me if I'm wrong.

 

I tried everything suggested to me but the limiter was the best solution for this. I'm completely confused about the compressor not taking that initial transient down. As I said above I tried all kinds of attack times, lookaheads and ratios but there would still be a huge spike even after setting the ratio to 100:1 and bringing the makeup gain up. Never experienced this problem before (and I hope I never do again). If anyone has any thoughts as to why this maybe happening I'd love to know. I understand compression is tricky at times but I used all logical approaches and it still wouldn't give. I even tried other compressors and tried using more than one. Still the same result...Given myself a headache (probably some of you guys too).

 

If you limit a track you normally also limit the so called dynamic range. The less you compress and limit the bigger this range is.

 

I have had this type of problem working with Nebula compression before. Sometimes, for reasons unknown it will miss certain transients. Some guys even say that's correct and that the original hardware it simulated would do the same, and so it's normal to check how things were compressed and then clip or limit the stragglers, depending on the amount of edginess you're going for.

But why PRO-C (which is quality) is missing the peak even with lookahead on, baffles me too.

 

Things are always relative to what you want to get in the end. Is it reducing the dynamic range? Yup. But in this case you're wanting to get rid of the peak, it means nothing for the song and is just annoyingly sitting there taking up space and making mixing the track more difficult than it should be, since your channel will be constantly popping into reds because of that untreated one annoying peak.

So the idea in mixing is to try and preserve *musical* dynamic range that feels good to the ear, whilst completely obliterating annoying dynamic range that you feel the song doesn't need. The absence of dynamic range when things are too compressed sounds like flatness, bland boringness, old TV program type quality. The presence of a good amount of range sounds like awesome, lots of balls, power and movement. When there's too much dynamic range things sound sparsed out, pointy, clicky, weak.

It's kinda the same thing as advice that says don't boost by more than 3-4 db with an EQ. Makes a lot of sense, it normally helps preserve the naturalness of the sound. But the other day I was playing with boosting a bass instrument by 12 (!!) dB at 30hz, with a Pultec. I was doing it with no guilt, because it made the thing sound exactly how I wanted it to, even though that's kinda an insane move to pull if you go by the book. Coming back to how it's all relative to what you want to get in the end from your adjustment.

 

I experienced this phenomenon with different compressors. That's probably why there are compressors and limiters. They work indeed differently. Even limiters can have different choosable algos, like those from FF and DMG Audio.
At the end of the day what you change are just displayed values. I once saw several measurements of dialed in values and the real effect. Was pretty insightful.

 

i had the same problem with bouncing out kicks i just render 1 bar and cut 1 beat (not the first) 2.3.4 is fine
the 1st beat sounds slightly diffrent