Help me with my guitar sound

[Ghilou]

Hi,

Every guitar track I record sound "muffled". I practice by playing covers and sometime record them. I can't get a normal sound as far as meds and trebble are concerned. I did a lot of fooling around, and tested a bit of everything: different sound card (including one guitar multi effect) different guitars, different cable etc. I even tested my signal chain for frequency response.

What I can say so far is :

- My standard signal chain (Input -> JVM 410 preamplifier -> Serial Loop Send -> Soundcard line input) has no obvious problem. I generated a sine sweep at the input and recorded the output : the frequency response of the JVM preamp is normal.

- With the above point checked, I plugged my guitars directly into the soundcard. The guitars sounds muffled, and plotting the spectrogram does indeed show the aforementioned lack of trebble and medium. I lack experience here, that might be normal. But it's still consistent with the rest of my test, that is, the very begining of the signal chain would lack the mids and high. But note that EVERY GUITAR and EVERY CABLE I tested have the same problem...

- When hearing what comes out of my speakers (a 1960 cab into the JVM head) I dont feel it lacks any trebble or mids... It sounds ok to my ear (well it sound AWESOME to my ear but hey, everyone likes different amps). Indeed, when I record with a mic - my webcam mic, I have nothing better - It sounds less muffled. It still shows a significant drop in meds and trebble but it sound more like a mic proximity issue - although I stuck the mic as close to the HP as I could.

Ok here is a litle clip to show what I perceive as a problem. What you will hear is:

- Dry guitar signal, directly from the soundcard input
- "clean channel" preamp signal, w/o speaker simulation
- "distortion channel (OD2-Red)" preamp signal, w/o speaker simulation
- "distortion channel (OD2-Red)" preamp signal, with speaker simulation

I played each part separatelly, apart from the two last. I didn't adjust the levels I trust you to set the volume as needed q:

http://nowwhat.free.fr/guitar-recording2711.mp3


The last part is what I allways end up getting: A guitar track that sounds ok but lacks the overral presence a normal guitar track should have.

What am I missing?

 

[Firaxis]

Have you tried using impulses?

Maybe turn the gain down a little on the amp to get rid of that fizz.

 

[vespiz]

It still shows a significant drop in meds and trebble but it sound more like a mic proximity issue - although I stuck the mic as close to the HP as I could.

Don't know if this is the issue, but..

From Wikipedia, the free encyclopedia

The proximity effect in audio is an increase in bass or low frequency response when a sound source is close to a microphone.

Tried recording the guitar twice? WITH bassguitar?

 

[Ghilou]

Have you tried using impulses?

Maybe turn the gain down a little on the amp to get rid of that fizz.

Yeah I've been using impulses from time to time but they do not add frequencies I seem to be fundamentally lacking.

Don't know if this is the issue, but..

Tried recording the guitar twice? WITH bassguitar?

Yes, been there done that...

Results with these techniques were underwhelming at best.

What I'd like to know is if, from the sample I posted, you can perceive that lack of presence. I mean, if I suck at mixing I can live with that, but I have the feeling I can't dial a presence that wasn't there to begin with.

 

[Trevoire520]

- My standard signal chain (Input -> JVM 410 preamplifier -> Serial Loop Send -> Soundcard line input) has no obvious problem. I generated a sine sweep at the input and recorded the output : the frequency response of the JVM preamp is normal.

^There's your problem. A line input is fine for the output of your amp head as that is a line level output. But your guitar needs a hi-z instrument input. Plugging a guitar into a line level input will result in a major loss of treble.

 

[Ghilou]

^There's your problem. A line input is fine for the output of your amp head as that is a line level output. But your guitar needs a hi-z instrument input. Plugging a guitar into a line level input will result in a major loss of treble.

I suspect something of the kind, but I can't wrap my head around all this.

Guitar in line-in is not good due to impedance mismatch. that much I can understand (although I used an active EMG set for this test, which are low impedance as far as i can tell)

But my standard configuration which is "fine" from your own word, still lack trebble and overall presence. And for the record, in this configuration I can sweep a sine through the chain and I get good results (standard meds and trebble).

Digging on the net, I realised that the JVM loop as a output impedence of 1k ohm, wich is pretty high when considering the standard input impedence of a consumer grade soundcard (around 10k). But it's still within good limits.

Besides, I had the exact same problem with my Zoom G9 in the loop, acting as an USB soundcard. And the zoom has a mega ohm (guitar) input.

I do troubleshooting for a living and I just hate when I can't find the problem q:

 

[Trevoire520]

How old are your strings?

 

[Ghilou]

not very old. Beside I get the same results with all of my 3 guitar, so I don't think it's guitar-related.

All are mahagony solidbodies, a little on the "dark" side then, but I doubt the effect would be so big.

Here is a screenshot of various clips I recorded. All of my guitars and some cable tests are featured there (notice the one curve with really severe drop in trebble at 7k? It's my Stagg low quality cable).

All of this clips have the exact same setup. That is guitar into jvm input, jvm send into Zoom G9 input acting as a sound card (you see how i get the same muffled sound as with a standard consumer grade soundcard). NO speaker convolution/simulation NO equa, no effect whatsoever. Just the plain preamp sound.

http://nowwhat.free.fr/Screenshot.png

What do you think?

 

[Firaxis]

It doesn't look that bad (well apart from the stagg cable) but then again I didn't think the JVM clip was that bad either, maybe a little more treble and presence or less mid etc. on the amp? Are the lines with the top end roll off running through a cab sim? That is roughly the shape I would expect if they were.
edit: The flatter lines look like a preamp signal should as well

It certainly sounds better than some of the tones I get when I've been playing guitar for too long to make an educated tone decision.

What plectrum are you using? That can certainly give you less attack. Perhaps you could try guitarhack impulses, there is a lot of top end in those.

 

[Ghilou]

It doesn't look that bad (well apart from the stagg cable) but then again I didn't think the JVM clip was that bad either, maybe a little more treble and presence or less mid etc. on the amp? Are the lines with the top end roll off running through a cab sim? That is roughly the shape I would expect if they were.

Yeah I was told that allready. But no, there is no cab sim here.

edit: The flatter lines look like a preamp signal should as well

Does it? I mean, why not, I can live with the fact that I need to learn how to mix better. What I would like to avoid is trying to boost a fundamentally flawed signal.

It certainly sounds better than some of the tones I get when I've been playing guitar for too long to make an educated tone decision.

What plectrum are you using? That can certainly give you less attack. Perhaps you could try guitarhack impulses, there is a lot of top end in those.

I've used the guitarhack impulses from time to time but I allways lack the presence. So I tried using a parametric EQ to bump the mids a bit and it does add some air to the sound, but I'm not really good at using that kind of tools.

 

[Headof75]

Listening to your clip of the raw signal it does sound a little dark to me - I would expect to hear some top end/string detail/rattle on there and I can't. Your signal also sounds as though it overloads/clips a couple of times, which is never a good thing. My advice is to download a few of the good di's that people have posted on here - any from Ola and the recent Mago and Lasse thread were all well recorded with no clipping or anomalies. Pull them into your DAW and compare their tone, appearance, output etc. That should give you some clue as to where your problem lies and should show you what to shoot for.

 

[Ghilou]

My advice is to download a few of the good di's that people have posted on here - any from Ola and the recent Mago and Lasse thread were all well recorded with no clipping or anomalies. Pull them into your DAW and compare their tone, appearance, output etc. That should give you some clue as to where your problem lies and should show you what to shoot for.

That of course was an excellent idea. I wanted to try something like that but never took the time.

So I downloaded Olaf's last year's christmas gift (thanks for the great work!). I don't know if these are the best DI out there but I figured that if it is good enough for him, it sure has to be good enough for me.

Pulled the right guitar track into my DAW (Ardour) and used a spectral analysis. First surprise here:

My signal is bright green. If anything, I have more presence (and less low end, but that wasn't a concern) than him. This could easilly be an artifact of my clipping or my signal running hotter, but the fact is that there is no outstanding mediums and trebble difference. A good DI guitar signal sounds a bit dark.

Then I reamped his track through my JVM. All knob at 12 o'clock and I tried to dial a decent amount of gain (I suspect the DI coming out of my soundcard is a bit compressed/weak compared to an actual guitar signal so it sounded a bit brittle). I took the premp signal and, again, compared it to my recorded tone :

Can't be much more spot on. I mean the "meat" of the guitar tone looks different but apart from that it's exactly the same thing.

So if the above preamp signal is dark (which is not even a sure thing, given the various opinions on this thread), then it's dark because of my preamp and/or pream->cable->soundcard chain. Both of them I have tested with a sine sweep signal showing no sign of trebble or medium loss.

So I guess that my guitar tracks are going to sound dark by nature, and I'd better try to deal with it.

 

[Headof75]

Glad you tried that even if all it does is give you some peace of mind.
I think I probably agree most with Trevoire's input on this thread - make sure to put new strings on for serious tracking and also look hard at your input chain at the moment. What you're doing isn't ideal no matter how you look at it. You'd be better either just recording the output of your JVM with a traditional cab/mic approach or getting something with a dedicated hi z/instrument input and recording clean signal directly to disk, then reamping or using a modeller. Depends on your budget, location and neighbours

 

[Ghilou]

Well I'm not sure mic recording is an option in my flat. I could probably get away with "loud TV" volume for short period of time, but it would probably be a bit weak to drive a standard cheap dynamic mic. And I'm not sure a condenser mic would be a good investment for that purpose q: Not to mention that if I were to record the amp I'd try to get some feedback in my sound (I just love feedback) which would mean much higher volume than what I can get here. Hell, I can't get proper feedback even at rehearsal level (unless I dial way too much gain, which is really easy with the JVM).

I was under the impression that getting the JVM preamp signal wasn't all that different from using studio preamps and/or digital simulations, with an added CAB simulation. Reamping and DI are a great idea but for now it's a bit above my head. But I will definetelly be buying a real sound card with a real line in, sounds like a must have.

 

[Headof75]

Check out this thread before you buy anything - there's plenty of options out there and it's worth doing some research before you buy -http://www.ultimatemetal.com/forum/equipment/513601-getting-audiointerface-faq.html
If you can afford it I'd recommend what I've got (of course!) RME Babyface.
Look here.