Getting a good (guitar) sound across multiple listening devices. Advice?

[RyanH]

This issue definitely applies to recording and mixing in general, but for me it's focused on guitars.

I haven't finished a song in literally years, despite working on music 10-15 hours per week, and with like 20 almost-finished songs in the queue. The reason: trying to get the f&$*ing guitar to sound right!

The problem mainly started once I began using synths - it's so hard to get guitars to fit perfectly with synths. But my main issue right now is that I'll get the guitar to blend really well in, say, my decent-quality headphones, and then it sounds bad on the monitor speakers. So I adjust the settings to sound great on the speakers and then it sounds bad in the headphones. The problem gets worse with each pair of speakers/headphones/earbuds I add.

Has anyone had similar frustrations and settled on a method or workflow that got them to a satisfying conclusion? I feel like I will die before I finish anything else and someone will have to sort through folders full of file names like "120 BPM Synth Pop vii," having no idea what .qtr opens with.

[SpotlightKid]

Firstly, there is no one "right" guitar sound (or sound of any other instrument). It all amounts to what is appropriate for the song, its genre, and the mix and the feel you're going for.

For example, if you're doing anything like rock, punk or metal, all rhythm guitar parts are often doubled, tripled or quadrupled. But it wouldn't be appropriate for jazz or funk, where the rhythm guitar is part of the rhythm section and not the main harmony foundation and needs to be tight and have a focused sound. Lead lines and solos need to be treated totally differently than rhythm guitar, often treating them more like vocals.

Also the frequency range the guitar covers depends on the genre, whether you want a modern sound or emulate the 70s/80s/90s/etc. sound, and its function in the song.

So it is not very efficient to throw random tips at you. Instead, I suggest you explain what styles you're working in and what kind of guitar parts you want to improve and also what you've been doing so far when recording, arranging and mixing.

[Impostor]

But my main issue right now is that I'll get the guitar to blend really well in, say, my decent-quality headphones, and then it sounds bad on the monitor speakers. So I adjust the settings to sound great on the speakers and then it sounds bad in the headphones.

Has anyone had similar frustrations

Yes

and settled on a method or workflow that got them to a satisfying conclusion?

No. Unless ignoring the problem counts as a solution. I rate this problem's intractability a notch below the hard problem of consciousness.

[erlkönig]

Well there is no general advice on getting a decent guitar sound (especially without knowing the material you are working on). But there are some usual suspects:
A good sound starts with a good arrangement and a good performance. If you listen to your song in mono, do guitars, keys and the rest covering each other? It's very hard to separate e.g. a legato chord from a Hammond and legato chords from a distorted guitar.
If you had a bad day on recording your guitar, no technical thing will make it better.

Oftenly, guitarists tend to overuse reverb/delay/chorus. Most times: less is more. As a rule of thumb: delaytimes should somehow relate to the tempo of the song.

Another issue: different reverbs (sizes/predelay/damping) on keys, guitars, vocals and so on.

Do you listen to your guitar solo? That should just be something to find some very special issues, like with a microscope. Your guitar must fit to the rest, so it should be mixed with the rest.

Is that problem occuring just with this special song? I mean: are you familiar with your monitorsystem?

It's very difficult to give an advice without having heard it...these were some issues that came to my mind firstly, so it's more like a lottery, without knowing your song. Could you upload a piece?

[glowrak guy]

...focused on guitars...and settled on a method or workflow that got them to a satisfying conclusion?

A high quality car stereo is a really good way to test your sounds. Go to a BMW/Mercedes car lot, if needs be! Or buy a balanced system for your chariot.

If like me, you can't ID artists and their typical gear during casual listening, you'll want ampsims with accurately named and implemented presets, so you can use and modify them, and build a frame of reference for discussing and comparing your music with other musicians. BlueCat's Axiom ampsim has a 'Legends' sub-menu that is helpful. (the demo fades out and in periodically, so can be useful before purchase. The recent Guitarix Neural Amp support also opens a floodgate of real-world gear models with recognizable names, like Diezel, Plexi, Bogner, yada yada, some really great sound foundations among the many captures. Great sounds can survive the mix.

Get used to recording guitar with percussion, matching guitar tone/style to the 'room' and type of percussion, so the guitar is boss. Then add in a FEW other parts, one by one, none of which dominate the most important individual part. You want the collective parts to blend well. So start by limiting the number of parts. If for example, you listen to a live rock band from the '70's, There will often be just drums, bass, a rythm part on guitar, organ, or keys, some melodic part on guitar or keys, and vocals if part of the song.

Harrison Mixbus is a bargain for mastering, loading sour songs will run them through a proven console system, that is not difficult fathom. IK Multimedia's Lurssen Mastering software is another fine tool, but like Axiom, is in wine/reaper/yabridge territory. As is IK's MixBox, currently on sale for $30 at https://audioplugin.deals/ It has many 100's of pro presets to parse, in a drag & drop gui, with wet/dry on each module, so simple to craft modify great effect chains as desired.

[tseaver]

@RyanH

But my main issue right now is that I'll get the guitar to blend really well in, say, my decent-quality headphones, and then it sounds bad on the monitor speakers.

Just a half guess -- I'm not great on getting EG tones I like recorded either -- but how do your "reference" songs sound in your monitor speakers? My studio room is only partially treated (mostly bass trapping, with some first reflection handling) and some of my favorite tunes definitely sound "off" on the monitors, compared to the cans.

@glowrak guy's suggestion of listening somewhere like a good car audio system should help you get perspective: given that your vehicle has a reasonable system, you likely listen to stuff you like there, and know what it should sound like.

[RyanH]

Firstly, there is no one "right" guitar sound (or sound of any other instrument). It all amounts to what is appropriate for the song, its genre, and the mix and the feel you're going for.

For example, if you're doing anything like rock, punk or metal, all rhythm guitar parts are often doubled, tripled or quadrupled. But it wouldn't be appropriate for jazz or funk, where the rhythm guitar is part of the rhythm section and not the main harmony foundation and needs to be tight and have a focused sound. Lead lines and solos need to be treated totally differently than rhythm guitar, often treating them more like vocals.

Also the frequency range the guitar covers depends on the genre, whether you want a modern sound or emulate the 70s/80s/90s/etc. sound, and its function in the song.

So it is not very efficient to throw random tips at you. Instead, I suggest you explain what styles you're working in and what kind of guitar parts you want to improve and also what you've been doing so far when recording, arranging and mixing.

My issue isn't so much about finding the right sound as it is about things sounding great on one listening device but crappy on another. However, thanks for the reminder about doubling. I've done that in the past but not for this song I was struggling with when I wrote my post.

[RyanH]

No. Unless ignoring the problem counts as a solution. I rate this problem's intractability a notch below the hard problem of consciousness.

Haha, I find your reply very comforting despite its hopelessness. Sometimes it just helps to know that the problem is not yourself. My very few musician friends don't record music, so I only get this kind of feedback on LinuxMusicians.

[RyanH]

If like me, you can't ID artists and their typical gear during casual listening, you'll want ampsims with accurately named and implemented presets, so you can use and modify them, and build a frame of reference for discussing and comparing your music with other musicians. BlueCat's Axiom ampsim has a 'Legends' sub-menu that is helpful. (the demo fades out and in periodically, so can be useful before purchase. The recent Guitarix Neural Amp support also opens a floodgate of real-world gear models with recognizable names, like Diezel, Plexi, Bogner, yada yada, some really great sound foundations among the many captures. Great sounds can survive the mix.

I'm using an old Vox Tonelab multi-pedal that has a tube in it. The presets are pretty bad, IMHO, but I've been creating good-sounding custom ones. That's where I'm encountering problems. A preset will match the rest of the song's tonality on the speakers, but then not the headphones, or vice versa. The differences are not subtle, either. I considered switching to Guitarix, but I'm using the Tonelab's expression pedal for wah and sometimes turning the delay on and off, so can't really work it with Guitarix. However, I might pick up a standalone wah, in which case I might make the switch.

Get used to recording guitar with percussion, matching guitar tone/style to the 'room' and type of percussion, so the guitar is boss. Then add in a FEW other parts, one by one, none of which dominate the most important individual part. You want the collective parts to blend well. So start by limiting the number of parts. If for example, you listen to a live rock band from the '70's, There will often be just drums, bass, a rythm part on guitar, organ, or keys, some melodic part on guitar or keys, and vocals if part of the song.

For this this song, and some of my other recent ones, the synths, bass, and beats are the drivers and the guitars are there to accent them. Before, when I was just using drums, bass, and guitar, it was so much easier to match sounds. I think maybe if I switch to Guitarix things will be a bit easier. At the very least, I'll be working with one console instead of two.

Thanks for the tips!

[RyanH]

Just a half guess -- I'm not great on getting EG tones I like recorded either -- but how do your "reference" songs sound in your monitor speakers? My studio room is only partially treated (mostly bass trapping, with some first reflection handling) and some of my favorite tunes definitely sound "off" on the monitors, compared to the cans.

@glowrak guy's suggestion of listening somewhere like a good car audio system should help you get perspective: given that your vehicle has a reasonable system, you likely listen to stuff you like there, and know what it should sound like.

I have to admit that I haven't been using reference songs yet. I planned to wait until the mixing stage for that. The speakers generally sound great regardless of what I put through them. I tested a reference song on the headphones, which sounded very good, but also very different from the speakers (and all my earbuds, which also sound different from each other). I guess it's like I'm having a hard time finding an anchor... too many variables.

As for a car, I don't have one. I do have access to one, though. I tested another song that way when I was at the mixing stage, but went back to square one when I realized the guitar sounded wonky.

[RyanH]

Thanks for all the input! It feels great just to be able to air the issue and hear your perspectives. Talking to myself was of limited value (although not entirely without value).

[Linuxmusician01]

I don't know if you use an physical amp and a microphone to record your (electric?) guitar but getting the sound as good on the recording as it sounds live is a known problem/frustration of producers. See this video by Spectre Sound Studios on it, may not be your style of music but it might help:

https://youtu.be/9l6YKjZtzTs?si=Gw9qmvFkmMJl3eU6&t=500

Spoiler alert: the trick appeared to be a mix of a Shure 57 and a Sennheiser 421 microphones recording the guitar sound produced by the Celestion T-75 speaker (known from Marchall amps). This producer of small starting out Metal bands also argues in many vids that the amplifier, guitar pick up element, etc. do not affect the sound that much: it's what actually producees the sound itself in the end that does: the speaker.

I never played (electric) guitar, let alone recorded it. But the guy from Spectre Sound Studio's makes sense to me. More so since he did a blind test of "clean" electric guitars through the same setup. Everybody got wrong which was the Gibson and which the Fender.

[SpotlightKid]

IMHO getting a good amped electric guitar sound recorded is a solved-problem, regardless of whether you're using free or proprietary software, since the advent of impulse response (IR) convolvers, guitar speaker cabinet IRs and neural amp modellers. Even with the previous generation of amp modellers, based on circuit modelling, like guitarix etc. you can get a great guitar sound when using a good cabinet IR.

The rest is arrangement and mixing, i.e. interlocking of parts, layering, FX, EQ-ing and spatialisation (delay, reverb). And the principle techniques used in these stages aren't really much different than those employed for any other instrument, IMO.

[RyanH]

I don't know if you use an physical amp and a microphone to record your (electric?) guitar but getting the sound as good on the recording as it sounds live is a known problem/frustration of producers. See this video by Spectre Sound Studios on it, may not be your style of music but it might help:

https://youtu.be/9l6YKjZtzTs?si=Gw9qmvFkmMJl3eU6&t=500

Spoiler alert: the trick appeared to be a mix of a Shure 57 and a Sennheiser 421 microphones recording the guitar sound produced by the Celestion T-75 speaker (known from Marchall amps). This producer of small starting out Metal bands also argues in many vids that the amplifier, guitar pick up element, etc. do not affect the sound that much: it's what actually producees the sound itself in the end that does: the speaker.

I never played (electric) guitar, let alone recorded it. But the guy from Spectre Sound Studio's makes sense to me. More so since he did a blind test of "clean" electric guitars through the same setup. Everybody got wrong which was the Gibson and which the Fender.

Hey thanks for the link, I'll check it out. I realized after my last post that I kinda buried the lede when it comes to my gear. The Vox Tonelab is an amp and cab modeler (I called it a multi-effects pedal, which it is, but the main feature is the modeling). I'm using it direct, but I did think that many of my problems could possibly be solved by putting it through an amp and changing mic positions.

[RyanH]

IMHO getting a good amped electric guitar sound recorded is a solved-problem, regardless of whether you're using free or proprietary software, since the advent of impulse response (IR) convolvers, guitar speaker cabinet IRs and neural amp modellers. Even with the previous generation of amp modellers, based on circuit modelling, like guitarix etc. you can get a great guitar sound when using a good cabinet IR.

The rest is arrangement and mixing, i.e. interlocking of parts, layering, FX, EQ-ing and spatialisation (delay, reverb). And the principle techniques used in these stages aren't really much different than those employed for any other instrument, IMO.

I have an obsession - probably not a smart one - with trying to record things more or less how I would play them live - in this case, using the Tonelab rather than something like Guitarix. Also the Tonelab has an actual tube in it, and it has a much better noise reducer than the Guitarix plugins I've used (although that was a couple of years ago). And I want to use the Tonelab's before and after-amp pedals, although the only crucial one is the wah.

I might get a standalone wah, stop worrying about my recordings being 100% reproducible live, and try Guitarix again.