Craig, thanks for sharing the video.  I wish everyone would watch it at least once to understand your main point about music interaction. 

This brought up some question and ideas in my mind. 
 
Craig had some specifics in his video, that he needed a cut around 200hz, add in some high end from around 1.26Khz and a bit more around 1.46K to the end and put in a cut around 2.7khz. How did he know all this? Well he was probably looking at the graphs in channel view and using his knowledge and experience.
 
What if you don't have the ears, experience, and vast knowledge of Craig? Or what if you're getting a bit older and maybe don't hear as well as you used to and aren't as skilled at looking at the graphs? Maybe Landr is a good way to go. 
 
Then there's the dreaded har-bal! I have used har-bal for a while and wanted to see what it could do. So I took the first song off Craig's video and pulled in the unedited version to see what I could find. Here's har-bal's graph of it.

edit: for some reason I can't get the inline images to show up, so here's a link to the 1st graph
https://drive.google.com/...view?usp=sharing 
 
As you can see from har-bal's analysis. There is indeed a big peak around 200 hz, it looks like it might be centered a bit before 200hz. It also clearly shows a low point that starts around 800hz and extends to about 2khz but with a peak in there around 1.2khz and again around 2.4khz with the high freq tailing off from around 4khz to the end. 
 
In my view, it clearly points out that Craig's analysis was exactly right, but also gives more detail as to exactly where things need work. 
 
It took me about 2 minutes in har-bal to modify it to look like this on the graph.
 

edit: for some reason I can't get the inline images to show up, so here's a link to the 2nd graph
https://drive.google.com/...VneGc/view?usp=sharing
 
Also, If you also notice in the upper right hand corner of the first pic. The "processed loudness" is at -20.67db. In the 2nd pic I corrected this to -14.14db. This was quickly and easily done in har-bal's histogram view raising the level to just before compressing.
 
This was all just an experiment on my part to compare Craig's findings in the video with what I could discover using har-bal. Is it going to replace a Craig or someone like a Bob Ludwig? NO! Is it an easy way to quickly make big improvements to a song and possibly understand what needs work? YES!
 
Here's a link to the file processed by har-bal if you want to compare to the ones on Craig's video. You'll notice there isn't the added reverb from Craig's job as that's not something har-bal does.
 
https://drive.google.com/file/d/0B3mQDhLlz1HnZlM4TWtFdXU1b0E/view?usp=sharing
 
gabo

Yes, Craig. Thank you. This is the type of educational material I was referring to. Of course you are using a third party maximizer in this but the gist is exactly what I was talking about.
 
You yourself of course are quite busy so I wouldn't expect you to take anything like this all on yourself but if these types of volumes could occasionally be created by various pros using the tools most of us will have on hand (or easy access to) then Sonar immediately becomes a much more useful tool (on top of it's already massive usefulness).
 
Even if the Templates and Chains aren't created and included in the Sonar releases at least folks can create their own based on what they see on screen.
 
Totally moving target but that's why seeing how multiple experienced people would go about such things on different types of material using the tools us schleps have around can help us make our own plans and get comfortable with it all.
 
Maybe a monthly blog/eZine entry where someone "masters" a tune in Sonar and explains what they are doing just as you have here would be a boon to... well us schleps just trying to put out decent demos.
 
Cheers and thanks again.

Wow. Stunning improvement, particularly the 1st demo.  And thanks for the details of what you did. It's not many master chef's that share their cooking secrets as you do. 'much appreciated.
 
Rick
 

What struck me was that both of these songs are well crafted and the performances are good, BUT... they both are deficient in the mix department.  Obviously this is subjective and others are free to disagree, but a little more care at that stage could have really improved the final product. 
 
You did us all a service with this post, because it shows that mastering (small m) can improve the final product, but it can't go back and really fix things like lost dynamics and less than optimal EQ and level decisions at the mix stage.
 
thanks for posting this!

So much good stuff out there about mastering.  It's really helps to know what you're doing (I've been doing my own mastering for 4-5 years now and consider myself a certifiable hack).  Cough Cough.
 
1st things first - it all starts with a good mix.  A bad mix can be healed only partially.  Conversely, a good mix need not be thrashed in the Mastering Process. 
 
It's quite the education process when you encounter musicians who want stuff mastered (for free, .....always!) and when they give you their mix they are unable to tell you what mixdown processing was performed, and more importantly (overall), why it sounds gawk awful.  I am helping a classical guitarist with his mixes....recorded all over the place, no continuity of sound, etc.  Musicians!!! (ha)
 
Many hats of to Anderton for his efforts here.  Learned a lot from you Craig!
 
BTW: I'm an Ozone 7 / Insight user....remarkable product with good presets to then help you dial in, if need be.
 

gbowling
This brought up some question and ideas in my mind. 

Gabo, that post is fascinating and your contribution is duly noted and much appreciated.
 
For what it's worth, when I was doing the actual mastering all the choices regarding frequencies to boost or cut was done by ear, not by looking at graphs. After all these years I can "get in the ballpark" for where the EQ should be, but also have a technique that helps greatly in identifying those frequencies very precisely. But, do note that just because there's a bump or dip doesn't necessarily mean there's a problem. There are some situations where that is natural and desirable in a piece of music.

Anderton
For what it's worth, when I was doing the actual mastering all the choices regarding frequencies to boost or cut was done by ear, not by looking at graphs. After all these years I can "get in the ballpark" for where the EQ should be, but also have a technique that helps greatly in identifying those frequencies very precisely. But, do note that just because there's a bump or dip doesn't necessarily mean there's a problem. There are some situations where that is natural and desirable in a piece of music.

That's pretty impressive! I can usually hear a ballpark of what frequencies need work, but usually not close enough to start tweaking EQ settings without some visual feedback. My ears hear "something in the low end" needs some work and from experience I would generally guess that it might be in that "mud range" around 200hz. Then I might tweak around on the EQ till I find what I like.
 
Would be interesting to hear more about your technique for identifying the frequencies.
 
I'll also agree that a bump or a dip doesn't necessarily mean it needs "fixed."
 
But the analysis part of har-bal does help me in determining things that need work and generally does a better job of it than any other tool I have. It can also be a great "learning tool" especially for people who haven't got as much experience. As I said, it's not a replacement for a Craig Anderton or a Bob Ludwig but it can help you learn things faster than trial and error.
 
Cheers
gabo

Out of interest, has anyone put these through LandR?
(Internet is slow and unreliable here, so I can't try it)

The Before and After mastering results were an amazing transformation, especially on the second entry. Just curious if Craig prefers certain genres over others. I would love to hear one of his masters on a smooth modern jazz tune.