by HYPNAGOGIA » Thu May 01, 2014 11:34 pm
Ok, so guidance from the videos, here's what you need to do, in a nutshell:
1. Make sure your mix is sounding as best as possible first. Make it sounding clean and with as much dynamics as possible (by using less compression). For making it sound clean, cleaning the bass part of the frequency spectrum should help with that - use high pass / low cut filters to cut bass from anything that doesn't need it (like everything except for kick and bass). Use EQ on channels as well.
2. Disable all processing effects on the master channel, most importantly dynamic processors (compressors, limiters, expanders). If you have spatial processors on the master (like a reverb or delay) that is an integral part of your song, render two versions - with and without them, and label them properly. Alternatively, you can render a 3rd take with compressors/limiters as well and send those for reference.
3. Leave some room at the beginning and end of the song. One video suggest 2-8 bars. I'd say 4 sounds ok. In case you have long tail at the end (like a from a reverb), make sure you leave room after it's gone.
4. Set your master fader so that audio peaks around -6dB. Just in case, as a guideline, don't let it peak over -4. Don't worry about setting the fader to 0 dB, it's irrelevant. Just use it and set your song level. You can also render at lower level and then just edit volume in audio editor to keep the proper peaks. Never render peaking around 0dB - less is more.
5. Render/stream at 44.1kHz (you can use 48 as well, up to you), and 24-bit WAV file. That should be enough.
6. Listen to your mixdown to make sure everything rendered as it should. Make sure it doesn't have any noise, pops and clicks.
7. If you're doing post-render editing in an audio editor: DO NOT normalize or apply dithering.
8. Communicate with engineers: mention what you want the mastering to achieve to your song, or if you have a song you want your song to sound like. That should greatly help them with decisions about mastering methods.
Now, just to clarify something about point 4. Both videos mentions to set master fader at 0dB and lower channels instead. The guy in the first video explained that because of the floating point errors.
It is actually more work to drop each track and deal with possible volume balancing issues (basically doing the mix all over again, since Orion can't drop all faders at the same time AND keep the balance between them), than to drop just one fader. Orion can already mix channels with 64-bit accuracy (double-precision by IEEE standard), so introducing calculation errors due to floating point rounding is reduced. Personally, I've never observed (or heard) anything that would affect the mixdown by dropping the master fader.