When I render audio from Live 5, the volume of the rendered track is considerably lower than professionally produced material. I know that mastering is a complicated process, but all I want to know is how to increase the volume to a level nearing that of professionally produced music.
It sounds to my untrained ears that the stuff I render is about 2-3 db lower than professional music. Is this a result of compression or something? Does anyone know any tricks in Live or another application to increase the volume and do some diy mastering?
Simple mastering - Increasing volume of rendered file
-
hacktheplanet
- Posts: 2846
- Joined: Sat Sep 25, 2004 6:37 am
- Location: Chicago, IL
- Contact:
You need specific tools to do this : the main one being a 'limiter' which, essentially, forces the average volume of a track upwards whilst stopping the peaks from going over into the red. Can be a good thing in moderation but can completely destroy a track when done badly. The process of mastering really depends on the material - if your doing a 'DJ mix' then really all your after is gentle limiting at the outs to keep things steady, and avoid the dreaded clip, as the tracks have already been mastered. Alternatively if your doing, for example, a full 48 channel mixdown of individual instruments, drums and vocals then you'll need the tools to get those kinds of levels.the_planet wrote:When I render audio from Live 5, the volume of the rendered track is considerably lower than professionally produced material. I know that mastering is a complicated process, but all I want to know is how to increase the volume to a level nearing that of professionally produced music.
It sounds to my untrained ears that the stuff I render is about 2-3 db lower than professional music. Is this a result of compression or something? Does anyone know any tricks in Live or another application to increase the volume and do some diy mastering?
If your going down the paying route check out plug-ins such as Izotope Ozone or something like T-racks - which has pretty much everything you'll need.
In terms of freeware I'd have a look at the following:
http://www.kvraudio.com/get/1127.html - Linear Phase EQ
http://www.kvraudio.com/get/1740.html - TL Saturator/Driver
http://www.kvraudio.com/get/880.html - Endorphin m/s Comp & Limiter
http://www.kvraudio.com/get/1622.html - W1 Limiter (same as Waves)
http://www.kvraudio.com/get/637.html - Inspector Analyser and Meters
Once your happy with a mix simply use these plugs in your master track in essentially this order. EQ your overall mix if needed (which you shouldn't need to if you've mixed it properly in the first place
). Then use a subtle amount of saturation using the TL Driver. Next add a little more compression and 'density' with the excellent Endorphin and finally use the W1 to crank the gain.The last plug I mentioned is Inspector - it has various useful tools but the best is proper RMS metering. Suffice to say when your limiting material up to 0db it can be very useful to see where the 'average' volume level is - the current trends in mastering volume can vary anywhere between -15dB RMS up to a ridiculously loud -8db RMS or even above. Simple rule of thumb here is : the louder it is - the more fucked it gets.
Best bet is to take a few CD's you think sound good, and in a similar genre to your own, rip a few tracks as audio and load it up in Live. Set your track fader and the master fader to 0db default. Now load up Inspector in the Master Track and hit play - you should see two sets of meters. The larger 'peak' faders and outside them are the smaller RMS one's. That RMS figure is what you should be shooting for using the tools above.
Hope this helps
If you have something like sound forge or any wave editor it is pretty easy to simply raise you levels. Find a point in the song that has the highest peaks and use the volume plug to raise the level with out it clipping the sound. Undo your edit and then apply it to the whole song. Burn it to disk and rock it! Have Fun!
Ableton’s engineers are hard
at work developing code that will allow our software to predict the future, but we don’t
anticipate having this available until at least the next major release.
at work developing code that will allow our software to predict the future, but we don’t
anticipate having this available until at least the next major release.
-
Submersible
- Posts: 25
- Joined: Mon Aug 23, 2004 11:34 pm
That's a very elaborate procedure--basically manual normalization. You could do this in a single step with identical results just by normalizing to 0 dB.Moody wrote:If you have something like sound forge or any wave editor it is pretty easy to simply raise you levels. Find a point in the song that has the highest peaks and use the volume plug to raise the level with out it clipping the sound. Undo your edit and then apply it to the whole song. Burn it to disk and rock it! Have Fun!
The problem is that normalization is not terribly effective when it comes to raising the perceived level of the track. That takes compression and/or limiting. It's better to skip the normalization altogether and leave some headroom in the rendered mix for further processing.
-
hacktheplanet
- Posts: 2846
- Joined: Sat Sep 25, 2004 6:37 am
- Location: Chicago, IL
- Contact:
Thanks!Submersible wrote:That's a very elaborate procedure--basically manual normalization. You could do this in a single step with identical results just by normalizing to 0 dB.Moody wrote:If you have something like sound forge or any wave editor it is pretty easy to simply raise you levels. Find a point in the song that has the highest peaks and use the volume plug to raise the level with out it clipping the sound. Undo your edit and then apply it to the whole song. Burn it to disk and rock it! Have Fun!
The problem is that normalization is not terribly effective when it comes to raising the perceived level of the track. That takes compression and/or limiting. It's better to skip the normalization altogether and leave some headroom in the rendered mix for further processing.
I do understand mastering is a complex process and I would never recommend any of my comments as mastering techniques. My reply was for simply getting a little more volume out of your track but, hey I learned something.
Ableton’s engineers are hard
at work developing code that will allow our software to predict the future, but we don’t
anticipate having this available until at least the next major release.
at work developing code that will allow our software to predict the future, but we don’t
anticipate having this available until at least the next major release.
Just a quick bit....
Volume isn't so much the issue, but loudness and dynamic range.
The brain tells loudness based on the ratio of peak db to average RMS db.
The closer your RMS is to your peak the louder something sounds.
Too close and it fatigues the ears, not close enough and it sounds soft
compared to studio productions.
-Ben
Volume isn't so much the issue, but loudness and dynamic range.
The brain tells loudness based on the ratio of peak db to average RMS db.
The closer your RMS is to your peak the louder something sounds.
Too close and it fatigues the ears, not close enough and it sounds soft
compared to studio productions.
-Ben
get the free W1 (Waves L1 Clone) Limiter from www.yohng.com
play with the threshold and release timing.
play with the threshold and release timing.
more limiting less compression
If you pick the proper threshold level for a *peak* limiter, you can increase the overall track level with virtually no compression effects. This is because most mixdowns contain "sonic whitecaps", summations that cause peak spikes that last for a very short duration, fractions of a millisecond. Not only are these whitecaps short in duration, they are also infrequent, depending on your threshold level, for example 0.01% of the time. So what's happening is that you;re letting 0.01% of the track dictate the volume of the remaining 99.99% Again, where you set your limiter's threshold determines that percentage. Something like Izotope Ozone can give you a level histogram, but no tool currently lets you set a threshold and see what percentage of samples are above or below it. Anyway, by setting the limiter threshold appropriately you can get more overall energy at the expense of limiting only 0.01% (or some other small amount), the rest of 99.99% is simply gained up, no compression effect.
I have found Apple's AU Peak Limiter, which comes free with every Mac, to be effective, especially when you lower the attack and decay settings to their minimum- for peak reduction. It's also quite efficient.
One other thing to do, usually first, is roll-off the subsonic frequencies below 30 (35) Hz as this consumes much of the available energy spectrum, then do a normalize, then do the aforementioned peak limit operation. Listen to the loud sections for any distortion, if so, undo the limit operations, set your threshold to a more positive value. (i.e. from -3 to -2.5).
Again part of the magic involves deciding where to set your limiter theshold. Too much and you begin to squash significant portions of the track. Too little is no harm but at the expense of sonic detail. You'll probably want to zoom in and visually see where most of your peaks are reaching. You'll probably notice that most of the "normal peaks" are some 3db or more lower than those special peak events (whitecaps). Of course, your track-to-be-mastered should be at a higher resolution than 16-bits (24 bits). Always mix down to 24 bits unless you're hollywood, which can afford 32 bits.
the rest of mastering operation (done before the levelizing) involves gentle EQ'ing, and perhaps major surgery using multiband compressors, a hopefully rare event.
-dz
I have found Apple's AU Peak Limiter, which comes free with every Mac, to be effective, especially when you lower the attack and decay settings to their minimum- for peak reduction. It's also quite efficient.
One other thing to do, usually first, is roll-off the subsonic frequencies below 30 (35) Hz as this consumes much of the available energy spectrum, then do a normalize, then do the aforementioned peak limit operation. Listen to the loud sections for any distortion, if so, undo the limit operations, set your threshold to a more positive value. (i.e. from -3 to -2.5).
Again part of the magic involves deciding where to set your limiter theshold. Too much and you begin to squash significant portions of the track. Too little is no harm but at the expense of sonic detail. You'll probably want to zoom in and visually see where most of your peaks are reaching. You'll probably notice that most of the "normal peaks" are some 3db or more lower than those special peak events (whitecaps). Of course, your track-to-be-mastered should be at a higher resolution than 16-bits (24 bits). Always mix down to 24 bits unless you're hollywood, which can afford 32 bits.
the rest of mastering operation (done before the levelizing) involves gentle EQ'ing, and perhaps major surgery using multiband compressors, a hopefully rare event.
-dz