Thought I'd post my question on here since it seems this community has really knowledgeable people in it. I been doing mixes for quite some time now. When I'm in the mixing process getting my levels and eq and compression everything going, it actually sounds beautiful. It sounds flawless. Until I get to my mix buss and its time to turn it up with a limiter. When I do this my mix gets destroyed. Not only does it get destroyed, but when I compare it to other commercial releases, there songs are WAYYY louder, even when I've squashed the hell out of my track already. Am I missing something here? How are these people getting there mixes SO loud, but still clean sounding? I've been looking for this answer for years and I can't figure out how they are doing this.
I use compression on the mix bus and everything. I even tried using multiple compressors to take the load off of just using one. I tried that with using multiple limiters as well. Nothing works and I just can't seem to compete when it comes to getting my mixes loud and coherent.
Please someone, anyone
Yes you are missing critical things.
The bad news is that this issue spans across many dimensions of engineering, the good news is that you can still fix the issue entirely by addressing the issues in all of those dimensions.
I'll break it down for you.
The DAW system is limited by the combined sample rate, latency, CPU, internal fx processing bit depth, oversampling rate. The software fx bus is limited by 64-bit precision on newer systems and 32-bit precision on older systems. On the true 64-bit systems/plugins the output is dithered down to 32-bit precision. By applying fxs in the hardware domain you raise the precision of these processes, hence you have higher levels of resonance in the signal left after fx processing.
Here are some calculations for you:
A) Pro Tools HDX running at 192 kHz with a hardware based limiter (infinite bit depth, infinite oversampling, infinite = 10000 as illustration) without CPU being the bottleneck
Input to output roundtrip latency: 67 samples -> 0.3 ms
->
1 / (0.3) * 1 * 10000 = 33000
B) Pro Tools HDX running at 192 kHz with an AAX 64 bit based software limiter at 8 times oversampling without CPU being the bottleneck
Input to output roundtrip latency: 67 samples -> 0.3 ms
->
1 / (0.3) * 1 * 48 * 8 = 1300
C) Presonus AudioBox 1818VSL running at 44.1 kHz with a 32 bit software limiter without oversampling, with the CPU putting a 10% bottleneck
Input to output roundtrip latency: 265 samples -> 6 ms
->
1 / 6 * 0.9 * 32 * 1 = 5
So we are up to this point talking a performance diff of 33000 (A) vs 5 (C).
Furthermore, the software summing process is limited by the 64-bit precision and the summed signal must then be dithered down to 32-bit precision and then finally dithered down to 16-bit precision. By summing in the hardware domain you raise the precision of this process, hence you have higher levels of resonance in the signal left after summing. Let's see what the impact is:
A) 33000
Instead of summing in the software domain, we decide to sum in the hardware domain. Summing precision: infinite. (infinite = 10000)
->
33000 * 10000 = 330000000
B) 1300
We decide to sum in the software domain. Summing precision: 64-bit, dithered down to 32-bit, dithered down to 16-bit.
1300 * 32 = 42000
C) 5
We decide to sum in the software domain. Summing precision: 64-bit, dithered down to 32-bit, dithered down to 16-bit.
5 * 32 = 160
At this point the performance diff between setup A and setup C is: 200 000 000 %
But this is not all. Then you have additional impacts as well, the "signal capacity of the limiter/audio interface and the input gain - threshold - output gain balance" quality factor, the factor for the headroom requirement for the target playback format, the precision of the engineering moves due to the monitoring performance and so on. That 200 000 000 % diff figure will eventually become so big that of course you will notice a huge difference in the "presence-clearity-beauty" factor between a commercial mix and a home recording mix.
Please note though that this basically just answers the presence question, there are of course much more dimensions at play here before the listeners go "I love this music", we are now discussing mostly vertical dimensions, but there are horizontal dimensions as well. A horizontal dimension is for instance the production quality - it spans across the vertical dimensions, although I usually approach it as if it was a vertical dimension, simply because the human mind with its current DNA setup is not designed for true multi-dimensional thinking.
The answer is not to think that hardware is all you need, or that presence is all you need. Hardware is only a part of it, so is presence, but those two are pretty critical when it comes to modern pop music. The emotion is in the overall charge density that the listener locks into.