Observation: So, I'm recording these bunch of songs that I've written.....

And, they're coming out OK... The songs themselves are better than my performance, but the performances are good enough to make it listenable, I think....

So, since this is a 100% self produced effort, and I don't intend for it to go anywhere, I've been thinking that I'm going to 'master' it myself. Yea, I know - bad idea - but, It's not like I have a real deadline, and I'd like to learn as much about recording and finishing a project as I can...

Well, part of my learning experience has been to do rough mixes of the songs I've tracked so I can listen to them in the car during the week, and listen to things mix-wise that need to be changed. I'll go back on the weekend, and make some tweaks, and burn another CD....

I've normally made two versions of the same song - one mixed and normalized, and one mixed and run through the Mackie 'final mix' plug that comes with Tracktion....

I set the EQ for some real slight modifications post compressor, mostly rolling off some of the bottom end. I'll admit to not knowing much about how to tweak the compressor section, but, I just set it so that the result is sounding OK - no pumping, etc.

So, I put them on the CD one after the other. It's clear that the song that was 'mastered' with the multi-band compressor is way better sounding... Up front, loud, big sounding, etc...

So, I've got this CD of my music, and it sounds pretty good. The problem is - it hurts me to listen to!

I skip the non-compressed tracks, because I've convinced my self that the louder compressed tracks sound better.... But, after about 5 songs of the loud tracks, I gotta turn it off... I don't want to - but it's kind of involuntary! I reach for the off knob, because I've got a screamin' headache...

Well, last night I take the finished mixes, and just run them through a limiter set so I'm getting very light attenuation only about 3 or 4 places in a given song, and roll off some of the bottom pre limiter.

I made a disk of 9 tunes, and took it to the car to listen...

I listened to the whole disk, without even a hint of discomfort, and I thought it still sounded pretty darn good... It was 'loud' enough from a RMS standpoint to compete with my commercial CD's too....

So, I guess the deal is that either I'm totally inadequate when it comes to setting up the multi-band compressor (a very possible possibility), or, the thing is jacking up my music, and making it sound good only in the short term, by boosting frequencies that would eventually get into my brain and cause it to wish for the pain to stop..

I guess my plan for now is that I'm going to test my mixes out on a bunch of misc. CD players and sound systems, and for mastering, I'm just going to use the limiter to bring my levels up, and make some EQ tweaks. Screw the compression......

Not sure why I'm telling you this... :confused:

What do you think? Am I nuts, or just stupid?

Well... it seems like you've got part of a valuable lesson... over-compression -- particularly with a skewed tonal balance -- is problematic.

As you've found, what sounds impressive right off the bat -- particularly when A/B'd with a less compressed (and perhaps less loud) version -- can come back around to bite you on the backside, producing listening fatigue or even outright discomfort.

(I used to listen to a lot of home recording product in the late 90s and around the turn of hte century and it was common for me to barely be able to get through some mixes where folks had jacked the levels to the ceiling or used "stereoizer" widening effects.)

BUT... it's not the COMPRESSOR (multi-band or otherwise) that is the problem but, rather, its settings.

It's sometimes helpful to take a good look at how your files look on the computer screen. Obviously, everything is relative. But if your file looks more like a paper towel tube than something wiggly and dynamically varied, it's a good hint that you need to back off on the compression/limiting.

With regards to doing your own mastering...

If you were spending good money to put this out on stamped CDs (laying out a couple grand, presumably), then I'd say, get a pro -- or at least a semi-pro -- to do it.

But you aren't going to put this stuff out commercially, sounds like, and you don't have an "angel" backing you.

I think learning to do the mastering yourself is a great idea, given those givens.

By learning about mastering you'll also be making yourself a smarter consumer in case you ever DO send an album or other project out to get mastered. A good mastering engineer can definitely be worth the money. But there are a lot of folks out there who hung up a "mastering engineer" shingle because they didn't have "enough" studio (or engineering chops or simply don't like working around people and musicians) to build a clientelle. Some of them don't even have any better tools -- or knowledge -- than the average home recordist. They sell their services around "lowball" package prices. You've seen their ads...

One inarguable advantage of having a good ME work on your album projects [or even critical demos] is that -- if your own listening environment is compromised, doing "corrective" mastering in the same environment can compound your problems.

If your ME has a good, neutral monitoring environment, he should have a leg up on catching problems that you've "built into" your mix as you unconsciously compensate for inadequacies or problems in your own monitoring environment.

But -- again -- good mastering can be expensive. It's something you want to save for critical projects, presumably where there is a budget or a chance of recouping the substantial money you will be laying out.

And, as appealing as a professional run of a thousand stamped CDs may be, there are a WHOLE lot of folks, even bands who get out and play frequently, who are sitting on a stack of around 900 unsold CDs...

What I've learned in the process of informally "pre-mastering" some tracks to balance the levels, etc.:

Beware the automated normalize command. I slapped together a quick disk of some tunes (two solo acoustic, two band arrangements), and let Toast normalize the playlist. I knew one of the band tunes was lower in level. The acoustic stuff didn't sound too bad, but the band tune that needed a boost ended up really nasal. I think the reason for this is because of the fact that different frequencies have different inherent energy. The automatic normalize process paid no respect to this fact.

Which brings us to...

Different frequencies have different inherent energy. So using the multi-band to pump up the low end will use up your RMS headroom really fast, and make it impossible to have an adequate high end level with a pleasing contour. So to hear the high end, you might end up pushing certain frequencies that lead to ear fatigue.

I have had this exact problem as well, due to inexperience and a less than ideal monitoring environment. Listening on lots of different systems in different places is a great thing to do and eventually you will start to get an idea of how your mixes will translate to the outside world.

Unfortunatley I ruined some mixes. Save yourself a backup session of the way they were originally recorded, before you add anything to them to reference too. And try not to stray too far from this. Just focus on getting a balanced mix but retain its original natural feel. In the end, Even if it doesn't sound as polished, you will be happier with the sound and, more importantly, the performance.

Now you understand why mastering shops get paid to do what they do.

The bottom end is usually the most challenging.

FWIW- assuming that individual tracks are in reasonable control dynamics-wise then I find that no more than 2 db of master complression at under 2:1 ratio is ever a good thing. This can tighten things up without generating the problems you encountered.

Also, you want ot aim for an average RMS signal thats about 14 db below the peaks. This is generally considered to be a pleasing

dynamic structure that balances the good and evil of all this in the context of typical pop/rock music. Perhaps you have some metering that can show you this?

Good luck.

Ah... I remember the good ol' days when I thought of -14 dB RMS as the threshold of unacceptably squeezed... what a quaint notion...

Thanks for the info guys....

I've used what you've told me here to ditch the multi-band compressor/mastering plug for a 6 band para EQ, a decent compressor, and a limiter, and I've found that I can set them up fairly easily, one stage at a time.

I'm setting the compressor for about 2:1 on average, and getting about 2db on average of compression.

When I evoke the compressor, I can hear a slight tightening up of the material. It's hard to explain what I'm hearing, but it seems to work and sound 'better'.....

Then I do some subtle tweaking of the EQ (first in signal chain). Some low cut, and some mid scoop on most of the material.

I set the limiter so that I'm getting peaks at -.3db, and attenuating a few times at the most dynamic areas.

After doing this, I'm ending up with an RMS of about -12/-16RMS.

I'll listen to these for a while, and see how much better/different these sound than the ones that I just brick wall limited.

Thanks again for your help, guys.