Do I need a DI box?

  • Thread starter Thread starter gbeam
  • Start date Start date
It's a passive DI, and the fact I'm not hearing a difference is why I'm asking ;)

I think his point is that if you are getting results you are happy with, there's no problem - at least on the surface.

I do understand all about the unease we feel when we get good results from something most people say won't work.
 
I think his point is that if you are getting results you are happy with, there's no problem - at least on the surface.

I do understand all about the unease we feel when we get good results from something most people say won't work.

Well, back to my original posting, I was wondering whether or not a DI box was necessary as to me, it looked like the input on my soundcard was capable of handling a hi-z output (guitar).

Now I'm trying to understand/find out *what* a guitars actual output impedance is. I understand you need a DI box when the output impedance is high and the imput impedance is low, and I understand why (the loss of highs, the "tone sucking", etc), so now it's whether or not the input on this card is considered low/high impedance...
 
You don't want to match impedances, you want to bridge them. Thus the recommendation to have 10x input impedance as the output impedance of your guitar.

Matching impedances maximizes power transfer, but you don't particularly care about powering the line input. You want to transfer voltage, and to minimize the voltage drop, ideally you want a much higher input impedance.

If you take a simple case of a non-reactive source, then if you match impedances you will lose half your signal to impedance of the source. But guitars aren't non-reactive sources, so the signal loss is not uniform. A guitar pickup is an inductive source, so it will form a basic low-pass filter with the input impedance of the card. The lower the input impedance, the lower the corner frequency of the filter. You want a high input impedance to get that corner way high out of the audio band.

Why haven't you heard a difference using a DI? Two possibilities I can think of offhand: first, you may be using an instrument with active pickups. In that case, the output impedance is much lower than 10K ohms, so you have no problem.

Second, it's possible the DI you are using also has an inappropriately low input impedance, so the signal degradation occurs whether you run through it or not. This may be the case with inexpensive cable-connector type impedance converters.
 
Why haven't you heard a difference using a DI? Two possibilities I can think of offhand: first, you may be using an instrument with active pickups. In that case, the output impedance is much lower than 10K ohms, so you have no problem.

Second, it's possible the DI you are using also has an inappropriately low input impedance, so the signal degradation occurs whether you run through it or not. This may be the case with inexpensive cable-connector type impedance converters.

Thanks for all the info - I wasn't aware that you'd *want* 10x the input impedance in order to preserve the sound. I assumed having them be equal would be enough, but now I understand why I'd need to use a DI.

Pickups aren't active, so that isn't the case. DI is a Yorkville YDI-1P. Can't find any specs on it, but it's a very very common DI where I'm from.

I've been running through the DI for ages, and when I said I heard no difference between going direct and with a DI, this was from memory. What I failed to put into the equation was that in the past, I used to route everything through a mixer, so it wasn't really "direct".

I'm going to actually record some tests tonight both direct and through a DI and perhaps I'll recant my earlier statements :D
 
Last edited:
I read that a guitar had an impedance around 10 kilo ohms, which is what the card appears to be as well. I guess that's incorrect?


yes, but you're not looking to match impedences. you'd want something much higher impedence than zour source. here's a handy mnemonic:



low to high will fly
high to low won't go.
 
Back
Top