hey yall i have several questions...

i make some mixes that sound 'ok' to me, but i wanna step my game up a lil', look at things a little bit surgically or sumthin'

i read this post on gearslutz about keeping the fader @ 0 and using a gain/trim plugin to 'gainstage', they also recommended this free plugin to do the job that goes by the name of sonalksis freeg. what are your thoughts on this?
it has to do with the resolution of the faders when they're really low in volume or somethin'. are some of you working in this way? i would really like to hear some of your experiences *yes*

also if i understand correct, a compressor is used to lower the dynamics, thus taming the peaks, thus lowering the volume but if i just 'gain stage' correctly there is much less need for the use of a compressor right? most of my friends compress about everything...(a lot)? why?

bye!

 

also i'm talking like a pro here, but can someone tell me a little bit more about gainstaging in general also, i'm just starting to delve into this topic you know

 

Gainstaging is (in classic studio practice) leveling your inputs to a defined level when they enter the console. Consoles always have trimmer pots as the first thing in the signal chain. In digital, the practice has often been to run everything as hot as possible, even using floating point file formats to try and ignore overs. This method never works well and pro's don't ever do this. They still gainstage to a defined level. Well what level? Big question, there are different standard in the US, Europe and the rest of the world. A typical Euro level would be the EBU (European Broadcasting Union) standard, which is roughly 0dbRMS = -18dbFS. Another standard is Bob Katz's K-metering systems K20, K14 and K10 which is 0dbRMS=-20dbFS and so on.

The basic idea is track your channels to disk using around 21 out of the 24 bits available for the bulk of the audio. Not using 32 float format is also a good idea as you can't use more than 24 bits resolution in a 32 bit float format and you at least 25% better performance on your drives. The idea is ensure that there is sufficient overhead above the signal RMS to avoid peaking problems in any part of your signal chain. Many pro plugs assume you will have signals in the RMS = -24dbFS to - 18dbFS range and distort if hit much over this.

So, how to do it. Samplitude/Sequoia has input trimming pots just like a console, that are applied just before the signal hits the plugin tree. Most DAWs don't, so you need a tool like Sonalksis FreeG (never used it) or Sonimus Satson that lets you do a level trimming ONLY. Another good choice is to have really good metering as the next thing in your plugin chain. Waves Durroughs is good, but anything that gives you both RMS metering and peak levels. You're aiming to get the RMS level peaking at around your chosen gainstage level, say -18dbFS, whilst at the same time (if using EBU standards) ensuring no peak goes over -6dbFS. If you're tracking a soft synth you can usually level down using the master volume control on the synth. If tracking a live source, you usually need to do this on your mic pre output side. The nearer to the source that you can level down, the better and keep the console (equivalent) trim pots for fine and last-recourse adjustment. Keeping the peaks under -6dbFS control may prove problematic, in which case you need a ceiling (non-boosting) limiter on the input side of the VST chain. Waves L1 is the ideal solution for this, there are many, many other choices, as you are limiting to a ceiling not boosting gain.

You apply the same rules to the master buss. You mixdown to a 21bit print and send that (or use it yourself) as the copy to mastering. Then, and only then, do you slam the pedal to the metal, the four to floor, and loudness the bejesus out of it.

Conversely you can run as hot you like, using all of the 24 bits because it must sound better right? Well it doesn't. Only tomfools and amateurs do this.

 

thanks for clearing this out for me a bit! i think you've explained to me very well, but i do have one burning question, i don't understand the 21/24 bit thing. could you elaborate specifically on this some more? your effort to share some knowledge are appreciated bro!

 

bit depth rate: http://en.wikipedia.org/wiki/Audio_bit_depth

 

There's 24 bits in most DAW user's audio format - so the temptation is to try and use all of them, by running hot, to get the best resolution you possibly can. The basic idea behind gainstaging is that you give up the top three bits (or so) in exchange for more overhead, less overs and intersample overs, and give uour self mixing room. Thats what 'printing on 21 bits' is about, using a lower amplitude to get better, controlled sound. Given that your going to end up on 16bit CD or mp3, the lost bits don't make a bit of difference whn you convert down, but make all the differnce when you're mixing. Use a lower defined mix level, but turn your monitoring up.

 

oh i get it now, this is about that loudness war going on, "if you want it louder, just turn up you monitoring". thanks for clearing that out for me once again, i was thinking already in the wrong directions, but it's all ok now, i understand. thanks!

 

by the way...

i also heard that plugins have some kind of 'sweet spot' too, like they have for e.g. the best resolution if input sound @ (...)db. or something like that..
now does anyone know between which db these plugins are functioning optimal?

 

Sweet spots are when refering to EQ (and timbre/character/presence of sound). Just stay under 0 dBFS going into a plugin.

If you are running hot pre-master you need to turn things down (channels, subgroups, returns, etc) or reduce the gain/level (with a level/gain plugin) on the first master insert slot. That way you can keep the master at unity gain and the following master insert slots will have appropriate gain going into them.