loud, quality tracks - am i doing it wrong

[damagedgoods]

oh yeah... and watch andivax's mixing secrets lots of good stuff in there. but he breaks the -6db concept in the video that everyone mentions here... I didn't think so b4 but that is very important.

Depends on your DAW. Most of them can go *well* above 0dB before they actually clip. Like, assuming 32 bits (so conservatively), that's about 750 dB above or below 0. There are factors that make it more complicated but in general the signal quality doesn't degrade with level per se; either you're clipping or you're not.

This is very untrue when using ableton Live. My mixes have started sounding much much better as soon as i've put all those channels down to -6db Helped loads

Hmmm, are you sure it wasn't just some kind of related psychological effect (eg lower levels on the meter -> a better overview of how loud each of your instruments is -> a more balanced mix) or that you weren't driving a plugin too hard on your master bus (like a limiter or something without appropriate gain compensation before it)?

I really struggle to believe that running channels into the red would make any difference at all as long as the master is in the green, but I'll do a null test on Live when I get home. If it proves me wrong then of course I'll eat my words!

TJ

[Opuswerk]

I'm very very sure of that. And it's not psychological at all, as it is a very audible artifact. The whole top end gets all gritty gritty. I suppose it's best to test it out with a pure sine so as to be sure.

It's also important not to have any element in the instrument chain going in the red as they can all clip there too.
For example having a simpler hitting the red, like it often does whenever one plays a chord adds distortion in the mix. Same goes if the autofilter hits the red because of a too high resonance for example. This goes for all plugins, native ones and "external" ones.
It becomes very obvious the longer the chain, as the distortion adds up at each step.
From my little knowledge what really makes a difference about all DAWs is how they react when they clip. I've heard of people pushing Reason's mixer really high in the red to obtain some kind of good distortion, to name drop, commix do that for example. But doing this with live squashes the sound in a pretty bad way imo, very harsh and very digital. However again, I need to really try and make a track in another DAW to be able to really compare that all. Maybe this weekend....

[damagedgoods]

I'm very very sure of that. And it's not psychological at all, as it is a very audible artifact. The whole top end gets all gritty gritty. I suppose it's best to test it out with a pure sine so as to be sure.

Will do!

It's also important not to have any element in the instrument chain going in the red as they can all clip there too.

To me this is more believable, as what plugins do with the incoming signal is their own business. However, I don't think I've ever personally experienced this with Live plugins; only certain 3rd party ones. As I say, I'm curious enough about this to test it out when I get home. Let me know if there are any specific reproducible steps that I can take that will guarantee some kind of adverse response to high gain.

For example having a simpler hitting the red, like it often does whenever one plays a chord adds distortion in the mix. Same goes if the autofilter hits the red because of a too high resonance for example. This goes for all plugins, native ones and "external" ones.
It becomes very obvious the longer the chain, as the distortion adds up at each step.
From my little knowledge what really makes a difference about all DAWs is how they react when they clip. I've heard of people pushing Reason's mixer really high in the red to obtain some kind of good distortion, to name drop, commix do that for example. But doing this with live squashes the sound in a pretty bad way imo, very harsh and very digital.

You'll have to forgive me but I'm always a little bit skeptical about any claims regarding "how DAWs sound at high gain", and even more so re: Reason. Here's why: DAWs do no sound processing by accident, because digital sound processing doesn't happen by accident. Even if it were conceivable that a DAW might clip the signal at some stage to +/- 1.0, that's very different to the "nice" saturation you get from an analogue mixer. If Reason really does sound good when driven hard, some developer working for them would need to have written an algorithm that says "here's my input signal, here's what the output signal looks like if you run it through a soft saturation curve". And it's unlikely that such a feature will be present without the company in question mentioning it in their marketing blurb or manual, because why would you spend time and resources developing a feature that nobody knew about?

Anyway, I will suspend judgement until I test out Ableton in a couple of hours!

[damagedgoods]

I'm not home yet so I haven't tried Ableton, but I was too curious to wait and just did a quick test with Cubase LE4 (which is all I have on my work computer):

- Two audio channels. A copy of the same audio clip in each (more or less normalized)

- Two groups. Channel 1 routed to group 1, channel 2 routed to group 2. Both groups routed to Master.

- Clip 1 has +24dB of gain. Clip 2 has -12 dB of gain.

- Each channel has channel EQ (in this case an identical high shelf at 1 kHz, +24dB, Q=12)

- Invert the phase one one clip

The observations:

With both group faders at 0dB, Track 1 solo is total cacophony, peaking at 39.2 dB. Track 2 solo, on the other hand, peaks at 3dB (because of the harsh EQ boost - apart from that it would be well-behaved because of the -12 dB clip gain).

Now, with the Group 1 fader pulled down to -36dB, Track 1 still peaks at 39.2 dB but Group 1 peaks at 3.0 dB, just like Group 2.

And - guess what - with both tracks playing simultaneously, both Group meters peak at 3.0 dB but the Master level is completely silent. Minus infinity.

Will repeat the same test on Ableton later.

[livecollective]

oh yeah... and watch andivax's mixing secrets lots of good stuff in there. but he breaks the -6db concept in the video that everyone mentions here... I didn't think so b4 but that is very important.

Depends on your DAW. Most of them can go *well* above 0dB before they actually clip. Like, assuming 32 bits (so conservatively), that's about 750 dB above or below 0. There are factors that make it more complicated but in general the signal quality doesn't degrade with level per se; either you're clipping or you're not.

You are talking about the range something can have in DB according to its Bit Depth and where they put "unity gain" on computers in correlation to distortion, what he is talking about it is a common mixing/mixdown procedure.

As for mixing a kick above -6, its just a guideline, if I have a sparse track I will push it up to -4 or -3 if i can get away with it.

[damagedgoods]

oh yeah... and watch andivax's mixing secrets lots of good stuff in there. but he breaks the -6db concept in the video that everyone mentions here... I didn't think so b4 but that is very important.

Depends on your DAW. Most of them can go *well* above 0dB before they actually clip. Like, assuming 32 bits (so conservatively), that's about 750 dB above or below 0. There are factors that make it more complicated but in general the signal quality doesn't degrade with level per se; either you're clipping or you're not.

You are talking about the range something can have in DB according to its Bit Depth and where they put "unity gain" on computers in correlation to distortion, what he is talking about it is a common mixing/mixdown procedure.

As for mixing a kick above -6, its just a guideline, if I have a sparse track I will push it up to -4 or -3 if i can get away with it.

Yup, totally agreed; I try and keep everything in the green too. But there's definitely also a common belief that DAWs will distort if driven into the red, and that's what I'm trying to disprove...

[briobox]

Could you post a sample of your work, so we can hear how it sounds? (maybe you're just too critical towards your own)

I uploaded some things here:

http://soundcloud.com/tr909

I hope soundcloud is ok? The tracks are set to be downloadable.

The last track labeled "Track ableton compressor spl attacker glue.wav" is the only one where I feel like I'm getting some results volume wise. It still sounds quite dead and narrow though. The higher drum sound has some distortion but that is purposeful - abletone tube plugin i think

The top tracks are little things that sounded great on my monitors (turned up loud) but not so good at all otherwise.

The track micro vbr is a good example as well. Far too faint. I have to crank any devise up very high to hear it well.

All comments are very welcome. Thanks again to everyone for your help and advice.

[damagedgoods]

Could you post a sample of your work, so we can hear how it sounds? (maybe you're just too critical towards your own)

I uploaded some things here:

http://soundcloud.com/tr909

I hope soundcloud is ok? The tracks are set to be downloadable.

The last track labeled "Track ableton compressor spl attacker glue.wav" is the only one where I feel like I'm getting some results volume wise. It still sounds quite dead and narrow though. The higher drum sound has some distortion but that is purposeful - abletone tub plugin i think

The top tracks are little things that sounded great on my monitors (turned up loud) but not so good at all otherwise.

The track micro vbr is a good example as well. Far too faint. I have to crank any devise up very high to hear it well.

All comments are very welcome. Thanks again to everyone for your help and advice.

All of your tracks sound like they're missing low end, but looking at the wavesforms they're obviously maxed out. I think the problem is that you're using samples with rapid and large attack transients and not much body. As an experiment, maybe try making a track using only 909 sounds and see if you can get it any louder.