DIs sound awful - why?

[olihunt]

http://dl.dropbox.com/u/3059614/comparison.mp3
Okay, the riff plays 6 times. The first time is a DI I got off of someone else (nerveend.com), the second time is me playing that riff with my 7 string. The third time is me playing it with my 7 string after lowering the pickups, which seems to sound better. The next 3 times are the first 3 times repeated but just the DI, instead of through an amp. They all go through the exact same ampsims with the exact same settings. Why does my guitar sound so rubbish? It is an RG7321 with a D-Activator pickup in the bridge, going into a Focusrite Saffire (the original white faced one). I've taken it to a guitar tech but he couldn't hear much of a problem, and thought it was to do with intonation. He's doing a full setup as I needed one anyway, but surely there is something else going on here?

[soundguruman]

The DI has no preamp, like a guitar amp with no gain, and no tone circuit, and no harmonics or distortion.
The transformer used in most DIs is a piece of crap with terrible frequency response. Even though the manufacturer tells you it's good, it's not.
In fact you could buy a better transformer at radio shack for $1.99
Theoretically, a good transformer (which is well over $100) would produce a better result.
BUT the best sound comes out of the amp and speaker, and it is best picked up with a microphone.

Over the years, there have been a lot of DIs, and they all sound like du du.
Also there have been many attempts to create a "line out" on the amp, for connection to the mixer, PA, and these all sound like crap too.

BUT the DI DOES have a good use, like interfacing a keyboard to a PA system, or changing unbalanced connections to balanced. OR even sometimes you can DI the bass guitar, and get away with it.

The lack of a good transformer is the main problem. Decades ago, there used to be better matching transformers on everything. These usually plugged into an octal sockets. These had much better frequency response. Perhaps the good transformers became too expensive to manufacture.

[olihunt]

I don't follow? I'm talking about the recording I get when I record my guitar through a recording interface. My 6 string (mahogany with EMGs) sounds fine through the ampsim, so does the DI i got off of someone else, it's just this 7 string guitar that sounds bad. I don't think my interface is the problem, as it works fine with my 6 string. When I say DI I'm talking about the recording, not a DI box.

[soundguruman]

There's 3 main things:
The pickups, the wood that the guitar is made out of, and the actual internal wiring of the guitar.
If the 6 string sounds OK, then there must be a difference in one of the above. Check all that stuff. What is the configuration of the 7 string?
What kind of pick ups does the 6 string have?

[dai h.]

hi, maybe if you could explain the exact signal chain used on the 3 segments of your recording, it would help.

[imaradiostar]

Does the 7 string have active or passive pickups?

They just sound different, which means the processed sound will be different.

This all begs the question- why not use a real amp? Isn't that what most of us are here for?

jamie

[olihunt]

Okay, the first DI is someone elses 7 string DI, not sure what pickup they use but it sounds good. They are all going through the same ampsim, so even if you don't like ampsims, it's not the fault of the ampsim. My pickup is a D-activator, and on my 6 string which sounds good through my interface with the same ampsim settings has EMGs, so I'm not expecting them to sound the same. This is what someone said about my 7 string's DI if it is helpful: '7stringDI.wav is distorted in a bad way, some sort of high-pass filtering (this can be due to many reasons) on top of pickup (or maybe interface preamp) internal clipping (specific shape of strong peaks). Also asymmetric clipping.' That doesn't sound good to me. The reason I don't use a real amp for recording is I live in a studio flat with many people very nearby, can't be too noisy, and I don't have any mics. Also modellers tend to sound better on 7 strings due to valve amps not handling the low notes so well. Signal chain is: http://dl.dropbox.com/u/3059614/righthandside.jpg and http://dl.dropbox.com/u/3059614/lefthandside.jpg and here is it in a mix with the other guy's DIs: http://dl.dropbox.com/u/3059614/Squid.mp3 (original here: nerveend.com, sorry I have to post this, that is his conditions, I'm not spamming ) I don't think this is relevant because my DIs clearly sound much worse. I've taken my guitar to get a full setup and I should have it back in 5 hours. Hopefully that will solve any problems with pickup height. Not sure if he is looking at the electronic but I hope so. The wood is basswood. I have heard really good recordings with this exact guitar, even with the stock pickups which are supposedly crap, through amp sims. Also my 6 string is the mahogany 6 string version of my 7 string, same model, so should be comparable quality. Pretty sure wood isn't the problem. There's a small chance it could be the pickup, but it's supposedly good. Unless the pickup is a dud I doubt that is the problem. The only option left is the setup or the electronics. Getting it fully set up, so if it is still bad afterwards, it must be electronics. Will report back when I get it back.

[J M Fahey]

Dear olihunt.
PLEASE donīt feel insulted by what Iīm saying, imagine you go to the Doctor and he says someting:
you are getting the answer to what you asked, isnīt it?

1) the first three segments sound quite close to each other.
Why?: because distortion clips everything to about the same, and looks like you got the post processing quite the same.
No problem there .
Please donīt tell me sample #2 "sounds awful" because itīs not so. And #3 is *almost* the same.

2) Now to samples 4 (nerveendīs original), 5 (yours pickups high), and 6(yours picups low).
Is that so? If not, please correct me.
To begin with: let me be honest: I donīt like that guitar sound. None of them.
Anyway, I can work on that, have been working many years now helping musicians get sounds they had in their heads, but couldnīt achieve live or at the recording studio:

3) Neverendīs (#4) sounds precise , balanced and defined.
What does it mean?:
precise: he fingers the notes just in time, pushing the strings precisely against the fretboard, getting the notes he wants at once.
balanced: lows/mids/highs do not overwhelm each other .... they are "balanced"
defined: you can easily "separate" the different notes in your mind.

4) Now to sample #5 (yours):
sorry to say so but it has problems:

a) it has way too much bass mid, "ugly" bass mids. It does not add "body" at all, which maybe was what you expected, but turn mids and highmids (the "typical guitar frequencies) muddy.
If you could use a graphic EQ or even better a Parametric, set it to about -4/-6dB with low Q and sweep from, say, 250Hz to 300/350Hz I think you would clean your signal a lot.
b) sorry but your playing (at least in this sample) is sloppy, lacks precision (as defined above).
You seem to finger the notes just *a little* too late, and instead of getting the exact note at once, you seem to "reach" it as if you were applying a little vibrato or as if you pressed the string halfway, picked and then pressed a little more.
This causes an effect, coupled to the excessive low mids, as if there was some kind of modulation effect applied.
It muddies the sound.

Sample #6 is about the same.

[olihunt]

No offense taken, I learnt the riff very quickly for this comparison so it is sloppy. I'll try out that EQ tip when I get my guitar back, thanks a lot.

[David Schwab]

Does the audio interface have a high impedance input for passive guitars? Many don't. Even the ones that do, like my M-Audio Firewire Solo, don't sound good with guitar.

The active guitar with EMGs has a low impedance output that can drive interfaces well. The guitar with the D-activator is passive. So it may be loading the pickup enough to alter the tone so that you don't like it.

I use a Roland VM-3100Pro mixer which has a high impedance input for guitar. It also has amp sims built in. But it works well when using software amp sims. If I plug right into the M-Audio interface, it doesn't sound as good through amp sims with my passive guitar.

[J M Fahey]

Interesting.
Maybe thatīs the cause of the muddy low mids hump I hear.
Agree, plugging the guitar into a high impedance (1M) buffer may change the sound a lot, for the better.

[g1]

"The first time is a DI I got off of someone else".
I don't know what you mean by DI here but it is causing a lot of confusion. A DI is a direct box. It is generally used to run a guitar or bass into a PA mixer. Please refer to your recordings as recordings, or sound files. They are not DI's.
As David mentioned above, it may be your problem is because you need a direct box, or DI, for your passive pickups.

[dai h.]

What David is talking about is what came to mind, so (if that is the problem) a mic pre or channel strip type thing etc. with a high impedance input (rated 500k ohms or higher) would probably help. If there isn't, then might try a direct box (countryman, or whatever you can get your hands on) to help match the impedance (not sure what happens exactly when this is wrong--less signal, more distortion?--but from experience this seems to usually sound poor).

Also, (I don't mean to be another "grammar police", but) it IS confusing the way the term "DI" is being used. It does seem apparent that a few of us thought "DI" was referring to a piece of kit (a "DI box") instead of a "DI'ed (or direct-injected) signal", so for clarity's sake (and to help others help you) it would help to try to be less confusing.

[Steve Conner]

They are recordings made by plugging the guitar directly into an analog-to-digital converter, with a bunch of modelling software where the amp, speaker cab, microphone, mixing desk and tape machine ought to be. Calling them DIs seems quite logical in that context.

A DI box is also a gadget for plugging your guitar directly into a PA system for live performance, which is where the confusion comes from.

Anyway, I had a listen to the clips on my good system and here are my findings. I'm ignoring any playing issues and concentrating on the sound.

The "good" guitar has a somewhat tinny, midrange-focused sound. The "bad" one has a huge bottom end, it sounds more like a bass. JM calls it a low-mid hump, but to my ears it goes down into the bass too.

This bassy tone might sound good played clean, but it'll be scrappy and ill-defined once it's been distorted. And indeed that's what it sounds like when distorted.

I don't know if this bass boost is because the guitar itself is designed to sound bassy, or because of the pickups, or the circuitry in the Saffire's instrument input. The impedance mismatch issues that Dave mentioned might be a possibility: a too-low input impedance could be killing the upper mid boost that hot pickups are supposed to provide, making the bass seem bigger in comparison.

Anyway, you can fix it by putting an EQ at the start of the amp modelling chain and cutting out bass until your guitar sounds like the "good" one.

If you have one of those Boss "Chromatic Tuner" pedals (and I've never met a guitarist who didn't ) you can also try that between your guitar and the Saffire: they have a buffer built in that will raise the impedance.

[J M Fahey]

Unfortunately, each human activity seems to have its own slang, which may be confusing or meaningless somewhere else.
We are all (well, most of us) "hardware" oriented, such as actually building a Champ if we need the Champ sound and so on.
And for us DI is a piece of hardware used to connect a high impedance instrument to a low impedance mic preamp.
But in the Guitar Software world, a DI is the output of an instrument, raw, unprocessed (as if obtained using a DI box, hence the name)
Nothing against it, I only demand that the definition is stated somewhere, preferrably at the beginning of the text, to avoid confusions as early as possible.

As an extra free addition to the dictionary: an "impulse" is what the softare guys use to simulate the sound of a certain real, "physical" speaker cabinet.
I guess the name comes from sending an impulse through the actual cab (presumably through a clean SS amp, for repeatability), picking it with a microphone, and subjecting said received audio pulse to some FFT analysis whose results can be used by approppriate simulation software.
Only guessing, but donīt think Iīm too far from the truth.