09-09-2014, 10:13 PM
Hey Daniel 
In the third mix, I'm hearing improvements in some places and setbacks in others. Unfortunately, ballads like this have very very complicated reverb needs, so it's very difficult to get the lush space you need without turning the tonal balance upside down... and usually the answer in those situations is the use of multiple reverbs of different types, with instruments outputting to the different reverb in different mixture levels. I'm not going to harass you for it too much at this stage in your learning because reverb is in my opinion one of the hardest things to get right... and it's the one really complicated part of mixing a ballad.
Just to give you an idea of what's involved, on my mix I have a dedicated reverb send each for the snare, acoustic guitar, piano, lead guitar and vocal tracks, as well as a dedicated delay for the vocal and a global reverb send that affects most of the instruments to some degree... though some instruments you may feel like you're hearing verb, but I've actually blended them into the pads... and the vocal delay, vocal reverb, global reverb, and lead guitar reverb have tiny automations in level, wet/dry, and length... and I was going to do a good deal more but my PC isn't powerful enough for all that reverb without a lot of trial-and-error submixing, so I chose to settle. Most mixes aren't so complicated, but a ballad lives or dies by the reverbs... this is still a brave and commendable effort for your experience level. For right now, I think you should start with turning the reverb send channel down all the way, and slowly increase the level as you make broad EQ cuts until you've got more of a feeling of blend.
Now for tonality, the vocal sounds better... I'm not sure if it's because of changes in processing on the vocal or if it's because the surrounding instruments are supporting it a little better... though I do feel that particularly the guitars and piano are a bit high in level relative to the vocal, and the bass is high in level relative to everything
. Reverb could be playing a role in that perception (because what one hears isn't always obviously related to what the meters are telling you... listen with your ears and not your eyes!), but I think you should be referencing a bit more. If your references are telling you the supporting tracks aren't loud enough relative to the vocal, and you can't seem to GET them loud enough without masking it, THAT'S when you reach for your EQs and compressors for balancing. There are no rules in mixing, but in general you'll want to give the vocal some room around 1.2 kHz in the spectrum, and use that air in the treble to help it cut 
Many of the instruments sound quite nice during certain sections, but they suddenly bloat the balance here and there, so make sure you're not doing your primary balance EQing while the instruments are soloed, because that basically never works. Get a static mix where everything decently serves the balance at all times... then automate some "color" into the supporting instruments when it's their turn to shine... at least that's what I feel works for me. I've found it's helpful to use an EQ that has a makeup gain function... so as you make your cuts, you can switch the plugin on and off and correct the level difference based on what you hear... and then if you're not sure if your EQ is helping, you can bypass the plugin for a volume matched comparison!
As for the bass and kick... my interpretation of the song may be very different than yours, but it sounds to me like you're thinking too hard about the wrong things. For me the bass and kick, although important for the sonic foundation of the mix, weren't central to the emotional message of the music, so I didn't really apply much processing to them at all... and the bass part to my ears was performed with the perfect dynamic, given it's mostly just single notes and no fills... so I didn't automate it or compress it or anything.. All I did was highpass the bass at 50 and took 5 decibels out with a narrow notch at 59. For the kick, I took out a few dB with a small peaking filter at 100 and a gentle 3 dB high shelving cut at 300. Then I focused on what was important to me: The vocals, snare, piano and ambiance. Rather than fiddle with a compressor insert on the bass to help with the transient, why not use a simple mild compressor on the entire drum buss and let the kick relax?
It's usually best to restrain most of your processing for the instruments central to the character of the music. The kick drum is going to be absolutely CRITICAL in dance, r&B, hip hop, and some rock styles, because those genres need the bass energy to drive the rhythm and make the connection. In pop music, and most genres really, you're going to spend more time on the vocals than anything else, and the snare is generally the second most important track in upbeat pop styles. Bass guitar is going to figure prominently in jazz, funk, and some rock (think RHCP), and most essentially in reggae. But there's a reason they call this style of music a piano ballad... the vocals and piano want to be the principal drivers of emotion in the main, so I'd base all of your processing decisions on what they're doing, while in most cases I'd want the rhythm section (drums, bass, rhythm guitar) mixed first. Not all engineers work this way, but for me it's easiest and best to mix to what I consider the primary character if the song... and leave uninteresting things like bass guitars playing whole notes in purely supportive roles by mixing them later on when there's less valuable mix real estate available.
Anyway, I'm not trying to discourage you and I'm certainly not telling you you're wrong... this is just the way I think about it and approach it, and I'm far from the best on the forum. You're doing good work, and the struggle I'm hearing right now in your work is one that's familiar to all of us: one our first couple tries, we get passable mixes until we learn enough rope to hang ourselves When I go back and listen to my first few mixes I think "huh, this is better than I remember." And then I look at the few mixes that followed and some of them were quite a bit worse than this! The main message I have is PRACTICE! I'd like to see you commenting on other people's mixes (and not just of this song!) because that's simply the best way to learn how to listen.... we're all friends here, and all feedback is good feedback, even if it's "wrong." Listen to someone's mix and think "man, that sounds great... what qualities do the vocals have that are helping them work so well?" or more importantly: "sweet jesus, those drums are way out of balance... what can I do to avoid that sound?" And when you make these comments (politely) on others mixes, you'll almost always learn something, even if they never reply!
Keep it up, dude. Try another mix and come back to this one with fresh ears
In the third mix, I'm hearing improvements in some places and setbacks in others. Unfortunately, ballads like this have very very complicated reverb needs, so it's very difficult to get the lush space you need without turning the tonal balance upside down... and usually the answer in those situations is the use of multiple reverbs of different types, with instruments outputting to the different reverb in different mixture levels. I'm not going to harass you for it too much at this stage in your learning because reverb is in my opinion one of the hardest things to get right... and it's the one really complicated part of mixing a ballad.
Just to give you an idea of what's involved, on my mix I have a dedicated reverb send each for the snare, acoustic guitar, piano, lead guitar and vocal tracks, as well as a dedicated delay for the vocal and a global reverb send that affects most of the instruments to some degree... though some instruments you may feel like you're hearing verb, but I've actually blended them into the pads... and the vocal delay, vocal reverb, global reverb, and lead guitar reverb have tiny automations in level, wet/dry, and length... and I was going to do a good deal more but my PC isn't powerful enough for all that reverb without a lot of trial-and-error submixing, so I chose to settle. Most mixes aren't so complicated, but a ballad lives or dies by the reverbs... this is still a brave and commendable effort for your experience level. For right now, I think you should start with turning the reverb send channel down all the way, and slowly increase the level as you make broad EQ cuts until you've got more of a feeling of blend.
Now for tonality, the vocal sounds better... I'm not sure if it's because of changes in processing on the vocal or if it's because the surrounding instruments are supporting it a little better... though I do feel that particularly the guitars and piano are a bit high in level relative to the vocal, and the bass is high in level relative to everything
. Reverb could be playing a role in that perception (because what one hears isn't always obviously related to what the meters are telling you... listen with your ears and not your eyes!), but I think you should be referencing a bit more. If your references are telling you the supporting tracks aren't loud enough relative to the vocal, and you can't seem to GET them loud enough without masking it, THAT'S when you reach for your EQs and compressors for balancing. There are no rules in mixing, but in general you'll want to give the vocal some room around 1.2 kHz in the spectrum, and use that air in the treble to help it cut Many of the instruments sound quite nice during certain sections, but they suddenly bloat the balance here and there, so make sure you're not doing your primary balance EQing while the instruments are soloed, because that basically never works. Get a static mix where everything decently serves the balance at all times... then automate some "color" into the supporting instruments when it's their turn to shine... at least that's what I feel works for me. I've found it's helpful to use an EQ that has a makeup gain function... so as you make your cuts, you can switch the plugin on and off and correct the level difference based on what you hear... and then if you're not sure if your EQ is helping, you can bypass the plugin for a volume matched comparison!
As for the bass and kick... my interpretation of the song may be very different than yours, but it sounds to me like you're thinking too hard about the wrong things. For me the bass and kick, although important for the sonic foundation of the mix, weren't central to the emotional message of the music, so I didn't really apply much processing to them at all... and the bass part to my ears was performed with the perfect dynamic, given it's mostly just single notes and no fills... so I didn't automate it or compress it or anything.. All I did was highpass the bass at 50 and took 5 decibels out with a narrow notch at 59. For the kick, I took out a few dB with a small peaking filter at 100 and a gentle 3 dB high shelving cut at 300. Then I focused on what was important to me: The vocals, snare, piano and ambiance. Rather than fiddle with a compressor insert on the bass to help with the transient, why not use a simple mild compressor on the entire drum buss and let the kick relax?
It's usually best to restrain most of your processing for the instruments central to the character of the music. The kick drum is going to be absolutely CRITICAL in dance, r&B, hip hop, and some rock styles, because those genres need the bass energy to drive the rhythm and make the connection. In pop music, and most genres really, you're going to spend more time on the vocals than anything else, and the snare is generally the second most important track in upbeat pop styles. Bass guitar is going to figure prominently in jazz, funk, and some rock (think RHCP), and most essentially in reggae. But there's a reason they call this style of music a piano ballad... the vocals and piano want to be the principal drivers of emotion in the main, so I'd base all of your processing decisions on what they're doing, while in most cases I'd want the rhythm section (drums, bass, rhythm guitar) mixed first. Not all engineers work this way, but for me it's easiest and best to mix to what I consider the primary character if the song... and leave uninteresting things like bass guitars playing whole notes in purely supportive roles by mixing them later on when there's less valuable mix real estate available.
Anyway, I'm not trying to discourage you and I'm certainly not telling you you're wrong... this is just the way I think about it and approach it, and I'm far from the best on the forum. You're doing good work, and the struggle I'm hearing right now in your work is one that's familiar to all of us: one our first couple tries, we get passable mixes until we learn enough rope to hang ourselves
When I go back and listen to my first few mixes I think "huh, this is better than I remember." And then I look at the few mixes that followed and some of them were quite a bit worse than this! The main message I have is PRACTICE! I'd like to see you commenting on other people's mixes (and not just of this song!) because that's simply the best way to learn how to listen.... we're all friends here, and all feedback is good feedback, even if it's "wrong." Listen to someone's mix and think "man, that sounds great... what qualities do the vocals have that are helping them work so well?" or more importantly: "sweet jesus, those drums are way out of balance... what can I do to avoid that sound?" And when you make these comments (politely) on others mixes, you'll almost always learn something, even if they never reply!Keep it up, dude. Try another mix and come back to this one with fresh ears
I'm grateful for comments and suggestions. Thank you for listening!