next »

[Daniel Pellicer]

Hello there!

I played drums in a band a few years ago but I actually haven't done a lot of rock with virtual instruments yet. The thing is Im not sure how to do it.

This first try I did is a 20 second introduction for a fan-made anime. I use two layers of guitars for the rhythm part, one guitar with a second channel with delay for the main guitar, this 4 panned left and right, left and right. Then bass and drums in the middle. The last instrument is a synthesizer and its also in the middle.

After the mix I compressed and leveled it, I didn't cut any frequencies.

My questions are:

- How many guitar layers should I use?
- Should I pan everything or only guitars, leaving bass and drums in the middle?
- Should I compress as much as possible or is there a limit?
- Should I cut lower frequencies in bass and drums, maybe in everything?
- Should I use reverb for all guitars or some with and some without?
- How can I give more punch the drums?

https://soundcloud.com/danielpellicer/anime-intro-20-secs/s-uBngm

Thanks for the answers

[Díoltas]

First off I think guitar vstis are just horrible. They will never sound good. To date I havn't heard a single one that sounds close to a real guitar. That said, if you're happy with the fake guitar sound and it's the sound you're looking for (I'm sure it has its place) then that's fine by me. Mixing a virtual guitar might be different to mixing a real one since it's a lot more rigid than a natural sounding instrument but here's some things I do for mixing my own real guitars.

For rhythm guitar and a big guitar sound double tracking is pretty much a requirement (record 2 unique takes of the same rhythm guitar and pan one hard left and one hard right). copying and pasting a vsti and doing this will not work and possibly even recording 2 takes of a vsti won't work depending how good or bad the vsti is with handling unique articulations etc. Lead guitar can also be double tracked if you're feeling adventurous but with not so hard panning. Single tracking lead is fine too and fine tuning some reverb and delay to match the feel of the track. EQ your guitar so it doesn't overlap with bass guitar frequencies.

Bass is mixed in the middle with little to no reverb/delay.

Drums should be panned according to their position in the drum kit. it's usually acceptable to pan snare centre along with kick which is always centre.

EQ and compression are a whole other ball game and there's plenty of guides out there if you google them. Compression is an art in itself.

[Audiosprite]

Hey Daniel,

I don't think anything here needs any more compression, or the drums any more punch. What would do it for this recording I think is to simply lower the guitars all a good amount, maybe softly attenuate the 1-4kHz area to give the synth some room (you can probably lower the synth a db or so if you do this too, giving the mix more balance).

Rock is about electric guitars for sure, but if you listen to some of your favorite rock albums you might find that rhythm guitars aren't actually that loud in the mix. Reference those mixes, see what you get.

[Daniel Pellicer]

Hey Daniel,

I don't think anything here needs any more compression, or the drums any more punch. What would do it for this recording I think is to simply lower the guitars all a good amount, maybe softly attenuate the 1-4kHz area to give the synth some room (you can probably lower the synth a db or so if you do this too, giving the mix more balance).

Rock is about electric guitars for sure, but if you listen to some of your favorite rock albums you might find that rhythm guitars aren't actually that loud in the mix. Reference those mixes, see what you get.

Cool! Thank you! I will try that.

First off I think guitar vstis are just horrible. They will never sound good.

I dont agree with you. I think a few colleagues here do great things with VST.
for example:
https://soundcloud.com/zack-parrish/was-sagt-der-fuchs

Anyways thanks for the tips

[Kyle Preston]

Another point, less specific to your question but equally important, is if you can get ahold of the instrument you're trying to emulate, whatever it is and just play around with it.

You'll realize very quickly what you're probably neglecting in the mix. Yes, you have to process guitars, bass and drums properly but you should also make sure the midi data has the 'feel' of a guitar, bass, or drums, etc. For me, this lesson came about while messing around with string vstis; pushing a note on a keyboard is different from a real live cello in SO many ways. If you can break down and explore all of these differences, you can probably create a pretty solid replica.  

Also, panning the instruments is a matter of taste, but I like to pan things the way they would be if I were standing in front of the band.  Vox center, bass slightly to the right, guitars to the leftish, high hat to the left, crash and floor tom to the right, and kick is dead center.

...vstis are just horrible. They will never sound good.

While I'm a guitar player and don't like using guitar vstis, this is an ignorant statement that contributes very little to the discussion. I've used them in a pinch and no one caught me doing it.....

[Díoltas]

I think vstis are just horrible. They will never sound good.

You misquoted me and left out the important part of what I said. "I think". i.e. My personal opinion. Everyone's entitled to one.

In retrospect, however, after listening to that mix of Zack Parrish's I will admit that "horrible" was perhaps too harsh a word to use. The newer sample based libraries have come on leaps and bounds and can sound quite good when used correctly. Older guitar vstis were quite horrible though and this was more what I was thinking of when I posted that as the OP sounds like he's using an older vsti, or perhaps not using it to its full potential. Please understand that I was only trying to help the guy and to point out the limitations of the instrument so as to better understand how to make the most of it within a mix.

[Kyle Preston]

Dude,

1) Not criticizing you for having an opinion; your opinion is the thing being criticized.

2) Not a misquote, notice the ... Not to mention this is all on the same page, anyone can read what you initially said

3) The statement "vstis are just horrible" is not an empirical statement; it already implies your own opinion

4) This is a misquote:
 


Apologies for shifting the focus away from the original topic; something strange happens to me when someone is wrong on the internet.

[ZackParrish]

- How many guitar layers should I use?
Depends on the type of  track you are writing. But if you are doing rock/metal or whatever then two will suffice, panned between 50-70% to the right and left respectively to give better spacing to the mix.

- Should I pan everything or only guitars, leaving bass and drums in the middle?
Bass should be kept in the center, if not slightly to the right or left, or you can use stereo modelling to widen it's sound and push it away from the center a bit. Drums are typically panned according to their relative position to the real layout of a drumset. So your hihat would be on the left, crashes on the right and left but not too much. Snare would be to the left but not as far as the hihat, the kick would be dead center, etc. Most VSTs of drumsets are already panned so you wouldn't normally need to pan anything manually.

- Should I compress as much as possible or is there a limit?
Compression should always be used lightly and only in cases where it enhances your mix. If you mix things appropriately you won't need compression at all. In cases where you record something like vocals, or a live guitar playing, it can come in handy for leveling the dynamics of the recording... but outside of that it's really not necessary.

- Should I cut lower frequencies in bass and drums, maybe in everything?
...Why?  What relevance is there to cutting the lower frequency? Unless you are correcting an already bass heavy track from poor balancing in the instrumentation, or a badly mixed VST/Sample, then there is not point for this.

- Should I use reverb for all guitars or some with and some without?
Sparingly. For lead lines(like a solo) reverb or delay is usually okay but for the rhythm parts you really don't want to slap on a lot of reverb because it can have adverse effects on the mix. A small bit of reverb with a short decay on it is useful for boosting the width of the guitars so they sound fuller. The bass I would advise not using any at all.

- How can I give more punch the drums?
Mid range frequencies are your punch factor in EQ, but if you want to boost the drums then mix it on the individual part of the drums you want boosted. It's usually not a good idea to EQ the entire drumset when certain parts of it are poorly mixed. Boosting treble will make your cymbals ear bleedingly hissy, boosting the bass too much can cause your kick to clip the track. If you wanted to brighten the snare you would want to do just the snare and not the cymbals/kick too. If you want the snare to have more of an audible punch you'd boost the mids/high mids but be weary of doing it too much. Raising the mids can quickly make your mix unbearable to listen to.


Of course, these are all my opinions seeing as I have no formal training in any of this stuff. I'm almost completely self taught so feel free to shrug off everything I said.

Also, if you want some good guitar VSTs check out Ample Sounds' guitars, or Impact Soundworks. I use Impact Soundsworks Shreddage X and Shreddage IIX for electric and Ample Sounds' Acoustic guitars for well... acoustic stuff. They have nice electrics too but I had already purchased the shreddage libraries before finding them. IS also has a nice bass guitar, and I believe Orange Tree Samples and Ample Sounds both have some nice bass libraries as well. Just make sure whatever(if ever) guitar libraries you buy that they do not have prerendered effects on them. Those are the WORST as you lose a lot of control with how they sound.

Okay... I'm done.

[Daniel Pellicer]

Thanks Zack! There is a lot of useful info there . I will try to apply all those tips.

By the way, im actually using shreddage2 its only I still learning about it, so the sequencing is pretty bad. Lets see how good can i make it sound

[Tony Cordes]

If I can add my two cents, I've found in mixing drums (well, mixing anything, but I first started applying this to drums) that each kit's components have a different "sweet spot" on the frequency spectrum, depending on the kit, size of the drums, how they're tuned, etc.  Even the size of the drumsticks will change what frequency bands are most prominent.
Something I'll do, using a parametric EQ, is to select a narrow frequency band and raise the level on it by something ridiculous like 10db, then shift it up and down the frequency spectrum until I find the particular sound in, for instance, the snare's spectrum, that I'd like to hear more of.  Then I'll set the level of that EQ parameter back to zero and attenuate it only a bit (maybe 1 or 2db or so) in an instrument that I want to remain prominent in the mix.  So, for example, if this particular snare's "meat" is around 300Hz, I'll leave that alone in the snare's EQ and lower that just a bit in the guitars or some other instrument I find to be contentious with the snare.
Guitars in rock should probably have a high pass filter on them.  In almost all songs that I mix, that happens, because their low end fights with the bass and the kick, and just ends up putting a bunch of mud in the mix.
Even so, I try to use as little EQ as possible.  If simply adjusting the faders gets everything to sit in the mix just right (it can happen, even if rarely!), then I stay away from those EQs.  One thing I try not to do is start adjusting the EQ before getting everything to sit as well in the mix as it can without.
In my experience, too, effects like compression/expansion should come after EQ.  They just end up confusing things when the EQ isn't already balanced.  That's not always the case; my bass, for example, gets a specific compression on it every time, because my style is rather dynamic and makes it difficult to hear in the mix while I'm tracking.  On the whole, though, I've found that my mixes turn out a lot better when I save dynamics processing for after spectral processing.
Also, I think you'll save yourself a lot of headache if you save the spatial effects (reverb, delay, etc.) for last.  I've found few better ways to make a mix totally unintelligible than to layer on the big reverbs first thing.
Ultimately, it has to sound good, so remember not to rely too heavily on formulae- even those that worked well for you in another mix, because every mix is different.  The golden rule of recording is that there really are no rules.  Less is more, but sometimes more is more, too.  Just approach each mix with an open mind and use your ears before you use your memory.

[Daniel Pellicer]

Less is more, but sometimes more is more, too. 

XD. I liked that one