LANDR is an auto mastering service by AI.
I would write my own strange way of using it.
Firstly, LANDR is an innovative service which offers you to mastering your tracks by AI and deep learning system.
Although I registered 2 years ago (Mar.2015), I have left it until recently. Because I didn’t have a good impression at that time after using a few times.
But the time has passed since then,
Now many musicians seem to affirm it. Recently I also use it frequently.
Although of course, some type of music is cut out for but others are not, but sometimes I really surprised at conclusion especially in dance music.
Apparently, the systems are learned and updated repeatedly, then it became outstanding now. It means this will be more better from now on, doesn’t it?
So, about my personal using way, I’m not joining the monthly paid plan.
For free, they can master 4tracks per a month with low resolution mp3. I’m using it.
Moreover, I don’t use it for mastering actually.
I upload my track to LANDR when my mixing is 80% complete.
A short time later, I can get mastered track. And I would download loudest version.
-To be honest, I want just little bit more loudness. (In the stock music market, many authors make their tracks quite loudly.)-
I use this mastered track as a reference point of mixing form here.
In other words, seems like an engineer is sitting next to me and we are discussing about the mixing of this track.
For instance, after long keeping work, I even cannot sometimes understand which is ahead kick drum or bass. I’m cutting too much mid low of the guitar then it clutter, and so on. The thing is I’m sometimes losing my objectivity.
But when I listen to the LANDR mastered tracks, I would notice my mistake.
Like “Oh, I had put the kicks too ahead!”
I hear LANDR’s saying “If it’s me, I do like this..”, and then I start my mixing form there again.
It’s like the third person’s perspective.
After that while I keep my mixing, I put the LANDR track as a quality reference point. I aim to do my mix and mastering better than LANDR track in the result.
Especially in beat track like EDM, to make better than LANDR is really difficult. (Sometimes I cannot.)
Thus, I also think LANDR as my mastering teacher.
Maybe this type of technology would improve more and more, then near the future, AI mastering might become a normal thing…
But advantage of the human is that they will get used to the awesome something and will make new balance, idea or combination. Then the AI will learn that soon after. Then human…
If that, the important essence and demand of human musician or artist might change gradually.
Back to LANDR, I only did low resolution mp3 like I wrote before. I think I should try high resolution mp3, wav and HD wav.
I want to know how they differ. Is it matter of data type or mastering algorithm? When I try that I write here.