Why Not a Kemper/AxeFx/Helix?

  I'll admit; I've thought about one of the above not for the sound, but for the (ironically) limitations of having everything in a box.  That's all.  

 "Chip, that doesn't make sense, those things having hundreds of parameters!"

 Yes, but - not *near infinite* parameters.  And they do have limitations: a lot of choices, BUT what you're given in the Box.

 Outside of the box: I've had ... let's say 100 amps.  100 pedals.  Some of which, had I been able to keep a handful would now be worth more than my mortgage.  The Boxes give you a semblance of that, yes. 

 But limited resemblances.  They can perfectly mimic a sound under certain conditions.  If you put a certain analog pedal in front of them - then the model doesn't work right. They can't alter your pickups, or your guitar.  Nor can they perfectly replicate every possible room sound, and every possible mic position. 

 In the real  world - in my house - I have a near infinite amount of combinations not found in those Boxes.  

 I have 2 different SM57 microphones .  A bunch of other oddball mics not found in the modellers: an Apex U47 tube mic clone.  A matched pair of MBHO small diaphragm condenser mics.  An MXL U-87 clone.  A Shure SM-81.  MXL R144 ribbon mic.  Some others.

 A Grace preamp.  A Presonus MP20 preamp modded with Jensen transformers and Burr Brown op amps. 

 A ton of pedals with non-exact copies in the above mentioned Boxes.  And amps and speakers.

 And an oddball place to record them in.

 I have spent an inordinate amount of time "testing" combinations.  Ultimately I won't go to using a Box because even at worse, at least the sound I get sounds "real"; as opposed to the struggle for "real" with the boxes.  The downside being with the infinite choices the above yields, more than in one of the Boxes.

 The Boxes feature mic placements in inches.  I'm here to tell the reader, if you didn't know, moving a mic 1 degree, a quarter inch in any direction, changes the sound "a lot".  You also have more choices than "dust cap, dust cap edge, cone, cone edge".  It goes on forever.

 So for me, one of those boxes might actually be a good idea, because they're actually limiting.  Most people find them good enough, or great.  I should get on with it...

 But that's not what instigated this blog.   I previously wrote about a.i. being used in a VST by 2020; well, that's definitely going to happen.  

 Now I'm going to go further.

 By 2021, a.i./machine learning will make recording and playing guitar a process unrecognizable by our perspective today.  2 years from now.

 The 2021 Modeling Box won't just be about models of whatever piece of gear.  It will be about nailing models of existing recordings.  It almost won't matter what guitar you put in front of it.  With 10 ms latency you'll have exactly whatever sound you want. 

 Same goes for recording.  All channels - bass guitar, drums, vocals - will yield perfect renditions of any recording.  So much so the a.i./ML plugin will substitute "something" no matter how far off the source is.  By 2022 it will be trivial to make a recording sound like anything with practically no technical knowledge.

 By 2022 this will be in all amps.  Which will lead to something "new" but old, and scary I'll explain in my next post.