A different mic position to train ToneDexter . . thoughts?

SongwriterFan · #1 06-17-2021, 08:41 AM

Has anybody thought about this, or better yet, tried it?

I got to thinking about something . . mainly, I want the output of the ToneDexter to sound like the guitar "as I hear it" when I'm playing it. I have no idea how it sounds to others in front of the guitar. When I buy an acoustic guitar, I buy it based on how it sounds to ME, when I'm playing it.

So, why not train the ToneDexter by position the mic around the position of my head/ears?

Does that make sense? Or is there some reason this wouldn't work well?

I'm thinking of pointing the microphone down, just above my head. I guess I could choose closer to the height of my ears, but then I'd have to choose between my left side and right side.

gfirob · #2 06-17-2021, 09:38 AM

This makes no sense at all, to me. A guitar is designed to project sound outward, towards the audience and away from the player. My experience is that a guitar is almost always disappointing from the player's perspective and you can only really hear what it sounds like when somebody else plays it.

tadol · #3 06-17-2021, 10:36 AM

I’d say it wouldn’t hurt anything to try it - so go for it!

The one piece I’d add is that when you hear your guitar, you’re hearing more than just the sound coming straight from the guitar, so I’d try it also with an omnidirectional mic, so you also pick up the way the room affects the sound. I have no idea if it’ll work or not. But I’d love to hear your report on it!

James May · #4 06-17-2021, 11:25 AM

This idea does have merit. I have done this myself and have gotten excellent results on certain guitars, but not all of them.

I found that an omni mic will work better in this position. There is a thread where I talk about recording multiple mics at the same time. In that thread there are some pictures showing the mic positions and one of them is indeed similar to where my ears are when playing.

Gordon Currie · #5 06-17-2021, 01:19 PM

I've gotten nice wavemaps by positioning a microphone over my right shoulder pointing down towards the 14th fret, on two guitars with sound ports.

I still prefer the wavemaps I get from the classic 'listeners position' for live performance.

If you aren't performing, then this could be useful. Give it a try!

SongwriterFan · #6 06-17-2021, 04:19 PM

Quote:

Originally Posted by James May

This idea does have merit. I have done this myself and have gotten excellent results on certain guitars, but not all of them.

I found that an omni mic will work better in this position.

Interesting. The only omnidirectional mics I own are "calibration mics" that came with other products. But they are both powered by USB-C and don't have XLR outputs. Here's the link to one of them.

https://www.minidsp.com/products/aco...urement/umik-1

But it comes with a mini-USB connector, so to get it to XLR, I had to buy these items:

https://www.amazon.com/gp/product/B07TLBTXXJ

https://www.amazon.com/gp/product/B08L8L5DLH/

https://www.amazon.com/gp/product/B01J2OSKFQ/

Sheesh! Hopefully that will all work.

James May · #7 06-17-2021, 06:10 PM

Using all those adapters might work. However the latency of the USB to XLR maybe long enough to be an issue. I don't know.
Keep in mind that for the price of all those adapters you can almost buy yourself another calibration mic that is old school analog.

SongwriterFan · #8 06-17-2021, 06:17 PM

Quote:

Originally Posted by James May

However the latency of the USB to XLR maybe long enough to be an issue. I don't know.

Hmm, I didn't think of that. I guess I assumed there must be at least some attempt in these devices (ToneDexter, VoicePrint, etc) to try to figure out any latency.

Quote:

Keep in mind that for the price of all those adapters you can almost buy yourself another calibration mic that is old school analog.

Do you have a couple in mind? I wouldn't mind doing that.

I also notice that both of these mics I own come with "calibration files" that are used by the software (one is for a DSP on my computer speakers, the other is for a surround sound system). That makes me wonder just how far off "flat" they are. That is, how much improvement is seen by using the calibration file vs not using them at all.

shufflebeat · #9 06-17-2021, 06:47 PM

I did my last set of wavemaps in a hurry (after a firmware update) using a Karma Mics Silver Bullet, the super-inexpensive mini-omni. If you can find one or more of these grab them with both hands.

The wavemaps were some of the best I've achieved, in significant part I'm sure thanks to the firmware improvements but whatever the reason those maps have replaced the previous ones which I spent quite a time over.

James May · #10 06-17-2021, 08:25 PM

Quote:

Originally Posted by SongwriterFan

Hmm, I didn't think of that. I guess I assumed there must be at least some attempt in these devices (ToneDexter, VoicePrint, etc) to try to figure out any latency.

Do you have a couple in mind? I wouldn't mind doing that.

I also notice that both of these mics I own come with "calibration files" that are used by the software (one is for a DSP on my computer speakers, the other is for a surround sound system). That makes me wonder just how far off "flat" they are. That is, how much improvement is seen by using the calibration file vs not using them at all.

Tone Dexter doesn't currently know what delay your mic has. It will train nonetheless, but whatever delay it has will show up in the wavemap. If it is 3ms, then not a problem. If 10ms, then it will be annoying.

Dayton Audio EMM6 is a good choice.

These mics are usually flat enough already that the cal files would not really be necessary.

SongwriterFan · #11 06-17-2021, 08:31 PM

Quote:

Originally Posted by James May

Dayton Audio EMM6 is a good choice. These mics are usually flat enough already that the cal files would not really be necessary.

Thanks. That's actually one that I had put into my Amazon cart after cancelling the order for all the other stuff. Just bought it!

Quote:

Tone Dexter can't know what delay your mic has.

Interesting. You've probably already looked into something like this and it doesn't work. But just in case, here's something to consider.

First, I'm NOT a geophysicist, but I do work with them, so I'm familiar with some of their techniques in processing/etc. One is called auto-correlation / cross-correlation. Maybe it has some application here that could be applied to determine the delay?

https://wiki.seg.org/wiki/Crosscorre...utocorrelation

tadol · #12 06-17-2021, 10:11 PM

Honestly, it’s one of the finest features of the Tonedexter - it’s incredibly easy to use, you can experiment with all kinds of options both with guitars and pickups and mics, and if you don’t like one arrangement, you can try another very easily. The biggest complaint I can come up with would be that James needs to write a nice little app to keep track of wavemap info so you can easily refer back to what you have saved!

Andy Howell · #13 06-18-2021, 03:04 AM

I'e experiemented with training using a series of different mic positions and honestly haven't found any real advantages over a mic pointed at the 14/15 fret neck/body join.

shufflebeat · #14 06-18-2021, 07:18 AM

Quote:

Originally Posted by Andy Howell

I... haven't found any real advantages over a mic pointed at the 14/15 fret neck/body join.

My other "go to" position is behind the bridge at a greater distance, much more fullness in the mids and sometimes nice for strummy chords.

It's a good way to have two characters for one instrument or sometimes if the neck joint position doesn't suit the instrument, usually smaller bodies.

James May · #15 06-18-2021, 10:10 AM

Quote:

Originally Posted by SongwriterFan

Thanks. ...

Interesting. You've probably already looked into something like this and it doesn't work. But just in case, here's something to consider.

First, I'm NOT a geophysicist, but I do work with them, so I'm familiar with some of their techniques in processing/etc. One is called auto-correlation / cross-correlation. Maybe it has some application here that could be applied to determine the delay?

https://wiki.seg.org/wiki/Crosscorre...utocorrelation

You are correct. I misspoke. We certainly could calculate the latency and subtract that delay. However we choose not to do that for a number of other reasons that are too complicated to go into. We had not anticipated a use case for digital microphones with latency.

I've edited my response and changed "can't know" to "doesn't currently know".