More Thoughts on Robo Mastering

It seems whenever the topic of auto mastering comes up, a fire gets lit. It's easy to forget that real people on all sides are working with a common goal in mind: to help people make better recordings with less effort and less out of pocket cost.

And I don't really have a beef with Auto Mastering programs. I just have rarely been happy with the results I've gotten from them. However, I have gotten results that I've kept and released, and AAMS in particular was one of the programs I used to get results on my most recent Album "The End of Silence". I also used EZ Mix 2 on one of the Masters and in the mixes for certain things that I wanted to hear, which my others plugins couldn't do.

The end result was that I used a combination of things, not just AAMS and EZ mix to get the results I wanted out of the final Masters. Most important of all, I used my ears and a monitoring system that I trust and know.

To this end, I want to point out something - getting truly great results out of your mixes takes at least these five things:

1) Recordings that are well done
2) Mixes that sound great and are well balanced
3) Mastering that fills in the gaps and tweaks the balance of the EQ curve.
4) A monitoring system you absolutely trust
5) Objective 'ears'

This is much harder than it looks, so I encourage you to learn as much as you can from sources that are reputable. I love recording forums, but many times the advice I've gotten on them has led me astray.

For The End of Silence, I had mixes I was happy with overall, but there were also problems due to some bad sampling software that introduced a ton of aliasing into the Mix. Because I couldn't afford to just go out and get new sampling software, I had to fix the aliasing in the mix using EQ. This was no small task. I also did some early takes where I had done a sloppy job at recording the instruments. I re-tracked and re-mixed lot of the material this past year to get things up to a standard I was happy with. I would say that I'm overall pleased with the results, but I also think I can do better. That's perhaps a perfectionist tenancy I have, but it also propels me into more learning.

I found that I was also able to use some Mid/side processing techniques to fix more of the aliasing during mastering. The final masters cleaned up nicely and are 100% better than the original mixes. Many of those tracks were run through AAMS, but it also took me making decisions about what I liked and didn't like each time I ran a track through the program. And those decisions did not come until I had listened meticulously to each master on several playback systems that I trust. Some of the masters lacked punch, others lacked clarity while others had some combination of both.

So what I ended up doing was subduing the loudness maximizing from AAMS, then I ran my final mixes through the program until I found something I was happy with. I then imported the 32-bit masters from AAMS into REAPER, and completed some final EQ adjustments, compression, and limiting to get what I chose as the final versions.

On one song, "Our Father", the results I got from AAMS were pretty well exactly what I was looking for. I only took out a bit of harshness in the mid and sides, and did the final loudness maximizing. It was done very quickly.

On another song, "Disappear", I ran it through LANDR and wasn't at all happy with the results. Then I ran it through AAMS and wasn't happy with the results. Then I ran it through Mastering Box and generally liked what I heard, except there was an insane amount of bass distortion and the mids were too scooped. So, what I decided to do with that one was loudness match each with the original and make manual adjustments until I had something I liked better than the other automated masters.

This begs the question: Is auto mastering worthwhile? I think the answer is yes and no. It is not an either/or situation.

I think it's important to test the automated waters and see if the results are better than what you are able to do on your own. If you can afford it (sometimes we will do it for free), get a mastering engineer to master a song and then compare the pro or semi-pro master with the automated masters. Whichever one you like best, go with it! Because this is really about your music and what will make you the most comfortable releasing your art into the public sphere.

How Not to Screw up a Mix

I learn a lot through trial and error. But that also means a get frustrated often. Another way of putting it is that I learn what not to do by doing it, sometimes over and over until my brain finally catches on that "doing x,yz will not give you the results you want"!

Here are a couple of the biggest lessons I've had to learn:

1. recording everything hot will give you a "cooked, smeared, un-detailed" sound, 
2. gain stage your plugins. Clipping one plugin can sound bad. Multiply that by a channel with 5 plugins and compressors, and you'll have a real problem! The compressor (not to mention compression and limiting during mastering) will make the clipped sounds nice and loud. 
3. Follow an established protocol for your work flow.  I have created my own, and you should too. Maybe I'll post mine up if there's any interest. 
4. Regarding #3, don't just have a protocol or work flow that you don't ever look at. Study it, know it.
5. Don't bite off more than you can chew. Sometimes you don't realize you've done this until it's too late ... and that's OK. But do your work in chunks and take regular breaks. this will help your brain, your ears, your eyes and your relationships.
6. Follow through on your promises/release dates etc. I had been working on an album for 8 years. That's insane. I set a release date for August 2016. I stuck to it and I'm glad I did! now I have 6 more songs percolating. 
7. Always reference your mix to a *good* pro mix in the *same style* as your music.
8. Always loudness match the reference tracks to your track. This means turning down the reference do it has the same RMS loudness as your song. This is incredibly important.
9. Always reference each mix on a variety of playback systems,  but especially in a vehicle. Vehicle sound systems have an uncanny way of exposing mix issues. I recently listened to a remix/remaster of Pearl Jam's first album. It sounded fine enough on headphones, but the whole album totally fell apart in the vehicle. I much prefer the original mixes, despite all the reverb (it was the 90s). The remixes are tinny, way overcooked and terribly honky.
10. Use an AB plugin like TB Pro Audio's AB_LM (freee) or Ian Shepard's Perception plugin ($149) to loudness match before and after your FX chain.

Level Matching and "Robo Mastering"

I ran a poll recently on a couple of home recording forums where I asked people to identify the best sounding master out of 5 possible masters.

I used the following to "remaster" an old song called "Future"
AAMS
LANDR
Masteringbox
A Professional Mastering Engineer
A Master that I produced myself

If you want to do this blind test yourself, check out my web page:
http://www.levityproject.com/home-mastering/mastering/audio-mastering-shootout

Since was a blind test, no one knew which one was which. This made things interesting ... and the results surprised me (sort of). One of the auto services actually got more love than the rest, but the reason why might be more predictable than you would think. I'll explain in a later post.

To complete the test, you will need to download Reaper, which can be found at http://reaper.fm/download.php

There are two Reaper Files in the test. One is not level matched, and one contains masters that I have level matched so they are all the same perceived volume. Try out both and see which master you like best in each.

Once you made your choices, check the results page to see which master you chose! Leave a comment to let me know which you preferred.

Leave a comment here or on the facebook page about which master you chose. I'll explain why you may have chosen it in a future post, but I believe the results will be fairly consistent (as they were when I ran the tests in some audiophile forums).

Photo Source: Wikmedia

Microphone Hack - Make it Sound Like a Neumann!

This is one of my favorite hacks of all time. I have some decent condenser mics, but what I've really wanted was a Neumann u87. It's almost $4000US, and I can't afford to spend that kind of money on a microphone. 

Well, I found and impulse model that completely negated my desire for that mic! I attached the model so you can use it, too!  You will need either SIR or Reaverb to load the impulse model into your DAW.

I like Reaverb myself. Here's a video on how to use it!

Here's what to do

1. Record yourself using a large diaphram condenser or sm58. Stand 15cm/5in away from the grill. To tame sibilance, rotate the face of mic to the left or right slightly. 
2. Load up the Neumann U87 impulse on your take using sir or reaverb
3. Be amazed! The sound is awesome. It's not perfect, but it's really good!
4. Hack #2: if you are using an sm58, load up the impulse with -sm58 in the title. This negates the sound of the 58 and adds the character of the u87. Fantastic!!

Click here to download the Neumann u87 impulses