Is a DI box necessary for recording guitar direct into my audio interface?

[AmyM65]

I've recently begun delving into the world of amp sims and I sometimes see online claims that using a DI box improves tone? Is this a legitimate claim or is it just snake oil? If it helps, I'm recording using an Audiobox USB 96 interface, which I believe has high impedance inputs. Thanks!

[Sam Spoons]

As you say you are beginning delving into amp sims then, as your audio interface has HiZ inputs, I wouldn't worry for now. If you are recording (I assume) electric guitars they need the HiZ inputs and a DI is another way of providing one but your AI has them built in. There are other benefits on a live stage but probably nothing that helps in your setup for now.

I am keeping things simple here but if you want to know specifics or more detail just ask.

https://www.soundonsound.com/techniques/guitars-di

And welcome to the forum

[Wonks]

As Sam says, use the Hi-Z setting and you should be fine. On other interfaces with just a choice between mic or line inputs, an active DI box can provide a high impedance input for your guitar. Ideally a guitar or bass with passive pickups needs to plug into something with around 1 Megohm input impedance to avoid 'loading' the pickups and dropping output and significant levels of treble.

A standard line input has a typical input impedemce of betwern 10k and 47k ohms. OK for guitars or basses with active pickups or preamps, but not passive pickups.

So in that instance an active (not passive) DI box that can proxide an input impedance of at least 500k ohms, fed into a mic input, will give the correct guitar/bass tone. You can also use an FX pedal with a buffer in bypass mode running into a line input to provide the same function (true bypass pedals won't work here).

[AmyM65]

Right, I'm not entirely sure the resistance on my interface is high enough then. Nowhere does the manufacturer describe its inputs as Hi-Z, the manual says it has an input resistance of 0.5 Megaohms. So as you say, ideally I'm aiming for 1Megaohm.

[Wonks]

Yes, it could be a bit higher, but a lot of Boss pedals (and other makes too) have a 500 k ohm input impedance and nobody complains. It's high enough. It's an order of magnitude above a normal line input impedance, and that's what's important. You'd have to listen really hard to notice any difference, if any, between 500k and 1M.

[Sam Spoons]

[ef37a]

Yes, it could be a bit higher, but a lot of Boss pedals (and other makes too) have a 500 k ohm input impedance and nobody complains. It's high enough. It's an order of magnitude above a normal line input impedance, and that's what's important. You'd have to listen really hard to notice any difference, if any, between 500k and 1M.

Agreed. Much depends upon the pickups and whether you use the volume pot, leave it at max and the input load has minimal effect but a doubling or halving of the guitar cable length would probably have more effect than 500k to 1 meg and that not a lot.

Dave.

[Hugh Robjohns]

500k is fine. To be honest, 250k is probably enough in many cases.

The reason for the notional 1M expectation is that the traditional DI box sits in parallel with the input to the guitar amp and the guitar feeds both units at the same time. In that scenario, it is vital that the DI box doesn't reduce the overall combined input impedance (to avoid tonal and loading effects on the guitar), and so its input impedance must be much higher than that of the amp's own input. 1M Ohms was therefore a convenient and safe number to aim for.

But if the guitar is only feeding the DI box (or an interface input) and nothing else, there is no parallel loading to worry about, and a lower input Z would be perfectly acceptable.

H

[Wonks]

I feel it's more that guitar amp inputs traditionally have been designed with a nominal 1M ohm input impedance and passive guitar pickups were designed to feed into that impedance. Leo Fender, Jim Marshall and Dick Denny certainly weren't thinking about DI boxes in parallel when they designed their first amps (or rather applied the various layouts from the RCA valve circuit design book).

But certainly if taking a DI and also needing into an amp, the input impedance of the DI also needs to be high.

Edit: My Two Rock amp has a 5-position pickup loading selector switch, which lowers the input impedance from the full 1Meg as the 'pickup loading' is increased. Even the first notch down from full has a slight affect on the sound, which gets more marked as the impedance lowers. Now what I currently don't know is by how much the impedance is reduced on that first notch down.

[ef37a]

As with much in guitar industry electronics, things are done more for fashion or adpuff reasons (or just lazy copying)
For active DI boxes at least the input resistance could be 10 meg or even higher and then the loading of a second 1M input negligable. You can even arrange for the slave jack to switch out a resistor and thus keep the guitar loading pretty well constant.

Passive DI's of cousre can rarely be more than 250k with the common 1-2k mic preamp input but then the loading plus the often cheap transformer adds a "charm" that some players like!

Dave.

[Hugh Robjohns]

I feel it's more that guitar amp inputs traditionally have been designed with a nominal 1M ohm input impedance and passive guitar pickups were designed to feed into that impedance.

I'm not convinced there was much in the way of 'intelligent design' when it came to guitar electrics! I think most of it is the way it is simply because that was the way it was at the time... It happened that way and more or less worked in an acceptable fashion, rather than being designed and planned with good electrical engineering concepts in place.

Yes, a lot of amps have a 1M Ohm terminating input impedance, but I've also seen a fair few schematics for amps with notably lower input impedances too.

The thing is, though, that if you plug a guitar into an active DI box with a 1M impedance, and then connect the DI box's (direct) link output to a guitar amp with a 1M impedance, the total load on the guitar is halved to 500K.

If the 1M load impedance was so critically important, you'd have thought active DI boxes would need to present a proper bridging load of 10M or more (which is perfectly doable)... but they don't. (A number of active DI boxes provide a buffered link output to feed the amp, specifically to avoid the double-loading issue of conventional boxes).

I guess there are just so many variables of pickup design, guitar volume and tone controls, cable capacitance, amp input impedance and so on and so forth that it's really hard to pin down what is actually working 'correctly' in the engineering sense.

[Sam Spoons]

I'm not convinced there was much in the way of 'intelligent design' when it came to guitar electrics!

Or guitar players.......

[garrettendi]

Or guitar players.......

As a fellow guitarist I can also vouch that we're not very bright...

[Wonks]

Or guitar players.......

As a fellow guitarist I can also vouch that we're not very bright...

Have you considered drumming instead?

[Sam Spoons]

Yup, I'm useless at it.......

[The Bunk]

Yup, I'm useless at it.......

You'll be perfect then.

[Sam Spoons]

[Watchmaker]

ouch

[garrettendi]

Yup, I'm useless at it.......

You'll be perfect then.

[Eddy Deegan]

[Dave.P]

Just to say randomly came across this thread yesterday and it tempted me to try a DI box on my Strat before going into my interface - still experimenting, but wish I had done this before!

Thank you for making me spend more money

[Rich Hanson]

Thank you for making me spend more money

Welcome to the world of G.A.S.