[Synth] Send me an angel (.mp3) - a remix

[Oliwerko]

Hello,

For the last two weeks I've been trying to alter the popular 80s track by Real Life to my tastes (that is, using different synths and no lyrics).

Here it is:
http://oliwerko.ic.cz/Send%20me%20an%20Angel.mp3

I have to admit I'm not really satisfied with the outcome, I was hoping for something better, but well...not everything turns out to be perfect.

The arrangement is complete, but I've come into another wall here. Compression. I tried compressing the track to retain its volume without clipping, but that resulted in some strange sound effects at some places. So I used only a limiter.

The point is, I know all the compressor settings, yet I don't know what to do in practice. Like here - I have +2db peaks I want to get rid of. What am I supposed to do? Bringing the gain down results in the song being too quiet. Is brickwall limiting the way to go? I'm clueless here, I would be happy if anyone could give me a hand here.

Also, feel free to comment on whatever features of the track.

Thanks

[Rakib]

Didn't hear any distortion. But I use usually low sound and then later I can use a gainer to have it loud enough.

[uncloned]

FWIW

I use an audio editor and remove the offending peaks (if they are not too many... say 10-ish) by reducing their volume 3 to 6 db by hand and normalizing. I've been doing this for years.

[uncloned]

I don't hear distortion either.

Sounds very early scene actually.

[Sam_Zen]

If there are a few peaks I use the same method as clones, but for more I use a small shareware prog : Peak Limiter. Screenshot :
http://www.louigiverona.com/webarchive/samzen/pixz/Pklimx.png

I think limiting is better than compression, because only the strongest peaks are dampened,
while compression also rises the lower volumes, so it affects the dynamic differences.

The song : it's not my genre, but well done.

[Oliwerko]

Thanks for tips.

I was also inclining to use hard limiting rather than compression, I just didn't know if it does have any bad side-effects or not.

The distortion is not really audible, but I thought 2-2,5 dB over the limit is a bit too much.

Thanks for the software tip, Sam, I believe I've seen it somewhere already, I'll look into it!

[TheEagle]

Even if I grew up in the 80s, it's not my kind of music either. But I agree with Sam. Well done, Oliwerko.

Related to the compression/distortion problem I have to say that I never used a limiter/compressor or whatever. While composing I always keep an eye on the peak-meters. So, after rendering a finished song there is (mostly) no need for rework.

[Oliwerko]

I'm glad you like it, even if not genre-wise  

While composing I always keep an eye on the peak-meters. So, after rendering a finished song there is (mostly) no need for rework.

Funny. I always get peaks at some points, even if two relatively quiet signals interfere.

[Sam_Zen]

2 TheEagle : peak meters don't tell everything, because they work with a certain delay.

Because digital mixing consists mainly of additions of values, it can happen, that some sharp peaks are formed.
These peaks can prevent to normalize a track to max. volume.

But this limiting should be done with as less peaks as possible. If one make the threshold too low,
complete loud passages will be dampened. Then normalization results in a kind of compression.

[Oliwerko]

I remember I read something about each instrument having its own space in the spectrum. Maybe that's the problem in my case. Maybe I just have too much sound at a particular tendency. Couldn't this be the case?

What are the rules anyway? To keep the individual instruments from interfering? Is it even possible?

[psishock]

It's possible ofc if that is your intention, there are very technical ways and more simpler ones to do this too, but you can also aim for harmony where you actually want instruments to interfere. There are no "rules", it depends on your aims.

I would suggest you to forget the "loudness war" and concentrate on your composing, creative skills. Normalize the final mix, and light compress if really much needed (example because of few loose peaks).

You see, making your music loud as possible has nothing to do with its quality at all. Crappy music will sound also crappy if you crank up even the "hardware" volume to max, but very good music can be highly enjoyed even on low listening levels.

I think its advisable to start worrying about these kind of "advanced" technical aspects when you feel experienced enough with all the composing and basic stuffs.

[Oliwerko]

I'm well aware of the whole loudness war, that's why I am reluctant to use compressors in general.

I don't give a damn about +/- 2 dB in the final mix, there's always the volume knob on my speakers if I feel the music's too quiet.

I was talking more about peaks so high that when you would crank down the gain to compensate for it, the overall volume would be like 1/2 of what it should be.

I'm not really "worried", it just came to my mind as I was looking at this problem with a compressor in front of me, not sure what to do.

[psishock]

Well, the "what it should be" part is the wrong approach there. =) It's totally relative, but i understand what are you trying to archive.
Your choices are:
- Simply compressing your final mix, till most of those peaks are gone, but risking it to have distortions, unwanted sound artifacts and dynamic problems (that you will maybe or may not notice on your monitors, but other listeners could do, if have more precise speakers, ears).
- Learning and watching for thousands of technical details while you're composing, and still not able to archive the "best" sounding anyway. (i'm on about 80%+, compared where i "think" i should be, but already worrying 'bout tons of tech stuff)
- Leaving your final mix as it is, and not comparing it to modern sounds, where every track has been mastered (AND compressed as much as possible) by professional sound engineers, heavily armed with professional monitors and hardware (or software).
- Cooperate and compose together with someone who is experienced in sound engineering.

If you are on 1/2, that should be totally cool, even 1/4 from the normal(ized) volume levels are very much fine, from a non tech-savvy composer, if you ask me.

[Oliwerko]

Yeah, you have a point. Anyway, it's not really audible, so these are more issues I'm trying to understand than to really "solve".

Thanks for the insight.

[psishock]

No problemo matey. I'm not trying to discourage you, but saying that those peaks are merely the "effects", from series of stuffs that has "gone wrong" over the composing and mixing process. Solving them for good and aiming for perfect normalized sounds means, that you hafta jump into very deep rabbit hole, where you initially maybe didn't wanted to go, at all. Not "that" simple as it may look from outside.