How to Mix Instrument Frequencies for Best Sound
Mixing is both an art and science. Why? It is an art because there are no limitations in being creative. It is a science because there are methods to be followed.
Supposing you have completed the recording and panning process. It is the proper time to start mixing the frequencies of the instruments. A song which is not mixed properly can result to poor quality sound recording.
I did myself some educational recording and mixing of my own song “At the highway” (credits: www.musicforlicense.net),
I played and recorded all the instruments (guitar, bass and drums) in multi-track and the vocals are performed by Jeanine Maningo (author: How to become a singer at becomeasinger.blogspot.com).
There is no bass guitar involved. Listen to the sound clip below which the frequencies are still not mixed for clarity. (Although panning and recording process are done)
This mix does not include the bass guitar mixing as it is being discussed in: tips in mixing bass guitar like a pro. Check this post!
A quick critic to this mix is as follows:
1. Guitars dominating the mix.
2. Muddy guitar sound.
3. Kick sounds so weak looks like punching a pail.
4. Vocal lower frequency range in conflict with lower guitar frequencies and kick.
Though this is a demonstration of music production at its simplest form, it has been illustrated, that muddiness of the mix can be corrected at the early part of mixing process and should not be a part of mastering process.
Below is the corrected mix with proper mixing settings applied:
What can you say? Clear is it? Yes I admit it is. The kick and the rest of the instruments are not fighting with each other, so guitar sound is so clear and not conflicting with either the vocals. Overall, the sound is not muddy. The secret in doing this, is very expensive in recording schools. But I will reveal below:
Rule #1. Each music instruments has it’s own center frequencies and range. Use a parametric equalizer
Rule #2. Cut and boost conservatively depending on the resulting sound.Q setting of a parametric equalizer is important.
Here we introduce “Q” and parametric equalizer, what are those things?, Q is a measurement of how narrow or wide the frequency adjustment on a parametric equalizer. Parametric equalizer is a mixing tool that will enable you to manipulate frequencies of instruments and balance it in the mix, just like what a paint brush will do to a painter.
To simply understand Q:
a.) a Q of 1.0 could be considered as medium wide.
b.) a Q of less than 1.0 is considered to be wide frequency adjustment.
c.) a Q of 1.4 is average adjustment.
d.) a Q of greater than 2.0 is a narrow adjustment.
Also there is what we call as “shelving”. It can be classified as low pass filter or high pass filter. A low pass filter will preserve low frequencies and cut frequencies higher than the cut off.A high pass filter will preserve higher frequencies but cut frequencies lower the cutoff frequency.
A center frequency is a frequency where maximum amplitude occurs. In layman term, it is a frequency most noticeable by the ear of that instrument. Let say voice frequencies. It is around 300Hz to 3000Hz. It is a pretty wide range, but that is the bandwidth of a telephone line. It is why the telephone lines are optimized for voice transmission.
The center frequencies of that bandwidth are around 1650Hz~2000Hz. It is where the voice frequency is strongest. And during a mixing process, that certain spot in frequency range is reserved for the voice to avoid conflict with other instruments.
Also take note the amplitude adjustment in parametric equalizer are measured in dB (decibels).
Below are the list of common instruments I used to mix “At the highway” and the EQ adjustment :
Short Snare
Freq1: 100Hz, Boost 1dB, Q 1.0
Freq2: 2500Hz, Cut 3dB, Q 1.4
Freq3: 8000Hz, Boost 2dB, Q 1.0
Purpose- 100 Hz serves to fatten the snare sound; cutting at 2500 Hz will minimize conflict with vocal frequencies. Boosting at 8000Hz will add some crisp (audible snare strings)
Kick
Freq1: 75Hz,Boost 6dB,Q 1.0
Freq3: 400Hz, Cut 6dB, Q 1.0
Freq4: 4000Hz, Boost 1dB, Q 0.6
Purpose- Boosting at 75Hz creates super-punch for the kick., Cutting at 400Hz can prevent conflict guitar lower bass and improves kick sound by removing the cardboard like sound. Boosting at 4000Hz can make the kick sound to be heard on small speakers such as headphones.
Guitar
Freq1: 160Hz, high pass filter, Cut 6dB
Freq2: 3000Hz, Cut 9dB, Q 1.0 (Cutting this prevent direct mud sound with vocals which occupies the same frequency range)
2. Right-Click then Copy
3. Paste the HTML code into your webpage
February 12th, 2009 at 2:05 pm
Hi,
I’ve read sources that say you should consider re-recording a part if you need to boost any frequency more than 5db. You are suggesting boosts a good bit higher than this. Is there any truth to what I’ve read?
April 12th, 2009 at 5:23 pm
Hi DZ,
Thanks for commenting the post. Anyway it is true provided it will create distortion above 5dB. I have met some amateurs recording well below the standard and that results to weak sound. Boosting it any further (during EQ) does not guarantee positive results and that in case, will be re-recorded.
I will not recommend boosting above 6dB if it causes some distortion in the resulting mix or some clippings. If you are in doubt try turning the volume up and see if distortion occurs.
August 20th, 2009 at 2:17 am
Hello, I have problems mixing synths,I do understand frequencies to some extent when it comes to synths or when i create them.My question is how can i separate the instruments apart to my production.I have friends who feel i play everything on the same frequency,but yet i play them in different parts or octaves of the keyboard?
August 21st, 2009 at 5:33 pm
Hi Jeff,
You need to separate them in tracks in order to mix them properly. Do not mix them together then do frequency adjustment, instead, separate those tracks first using your multi-track editor then do frequency adjustment.
Give them frequency assignment, for example I have 3 synths,
a. Bass synth —> from the lowest to 300Hz
b. Mid range synth —> 300Hz to 5000Hz
c. Treble synth —->5000Hz to above
So when I mix them, for 1st track Bass synth, I do a low pass filter at 300 Hz, so it means it pass all frequencies below 300 Hz and attenuate above (which I do not need because it is a bass)…I do the same for mid range and treble.
The result is a cohesive mix.
September 18th, 2009 at 9:06 am
Hi Emerson
I have problems knowin wen to add delay to my mix. Just finished recordin instrumental 4 my song(5 tracks: kick, snare, bass, panned tamborine, melody and a slightly panned synth sound). As far as am concerned it sounds pretty tight already. However because most tracks nowadays need effect I think I might have to add at som point. Wuz wondering if u can suggest to me a way I can use delay without loosing the feel of the track and without overusing it. Basically scared of slowin down the tempo?( Note. This song has a pop feel and the bass drives the song.)
another question. Am considering using a flanger effect on my vocals. backup or lead I haven’t decided yet however wuz wonderingif u can suggest the best way to do this without causing phase cancellation, I.e do I just put it on the track or do I use an aux. Any help is appreciated.
September 20th, 2009 at 6:48 am
Hi John,
If you need to add delay to the mix. It is a good practice to add some space and ambiance to your stereo mix particulary melodic instruments except bass or those panned in the center (though with some creativity you can apply delay to vocals in a doubling technique) , so I suggest this method ( i use this technique in my mix):
a. In your mix, say you are mixing tamborine… put one track to the left and one track to the right.
b. After that, put a small delay to either of the track (either left or right). A small delay could be 10 ms or 5ms or even 15. Do not overdo as it will ruin the sound. Only put one delay on one track, minor stereo differences creates some spacey sound improving the ambiance.
c. Experiment some settings until you have notice that it sounds full and spacey.
Do this with all melodic instruments. In my experience, I prefer to add delays on melodic tracks and those that are not panned in the center. It is not always required to add delays, if you decided that it push or improve the song, then use it. Thanks for reading and commenting my post.
Emerson R. Maningo
March 10th, 2010 at 6:02 am
Hi Emerson,
I have a 5 piece band (guitar, bass, keys, drums, Sax along/w vocals) that play live music. However, i am having a hard time hearing myself when i sing with the band as it seems that my vocals get drowned out in the music no matter how much gain i have. the other two guys who simply sing back up come through crystal clear and always louder than me even though my volume on the mixer is louder. is this a problem of frequency? if so, can you suggest a good frequency for voice that can cut through the mentioned instruments or let me know if 1650Hz~2000Hz is ideal no matter the voice type.
Thanks
Oren
March 16th, 2010 at 6:47 pm
Hi Oren,
The live band mixer guy is not doing well the job. If you play live and using a mixing console, here is what I suggest to make your vocals clear no matter how loud those other instruments:
First: You need to cut 3000Hz using the mixing console for these instruments (guitar, keys, sax), the amount of dB depends on your ear, but I suggest -3dB to -6dB cut. By cutting frequencies on those instruments, you are making a hole in the overall frequency spectrum at 3000Hz, this room is for the vocals to properly sit in the mix.
Second: Using your mixing console, boost your vocals at 3000Hz only, 2dB is recommended, and I think 3dB is too hot. Depends on your ear.
If your mixing console comes with parametric equalizers effects, then set the Q to 1.4 for all frequency adjustments. 3000Hz is the only frequency your mixer needs to tweak to get the job done.
March 20th, 2010 at 6:00 am
Thanks Man…will try that and see how it plays out.
much respect,
Oren
April 14th, 2010 at 4:22 am
Hello, I thought this was an extremely useful page. Currently a friend is mixing the recording he did for my band, even though it sounds alright, the band agrees precisely on what you say here, seperating the frequencies so it has a less muddier sound. However, unfortunately we can not seem to agree with our friend and among us, on how to make the instruments and voice sound “in different planes”. I was wondering if you would be willing to listen to a part of the track and give us some advice or tips on how to acheive what we want. Thanks.
April 19th, 2010 at 9:53 pm
Hi Yao,
Sure, please post the YouTube URL where the raw recording (still not mixed) can be found. I will listen to it and write a detailed constructive comments on the mix and post it in this website.
May 15th, 2010 at 7:05 am
Hi Emerson,
first of all, thank you for making this great site! Anyway, I’m an amateur trying to get a nice sounding mix for a one clean guitar (roland jazz chorus clean) + one vocal duo. I thought that since the guitar is the only accompanying instrument, I should cover all bases with it, including low end, mids and highs, leaving some window for the vocals only. But when I listen back to it through the home stereo, it resonates it and has kind of a boxy sound. I figure I’m doing something wrong, but is hi-passing the low end radically the only solution, or how could I get a smooth, “wide range” sound with all the frequencies needed? (I’m using Reaper.)
Thanks in advance!
May 18th, 2010 at 10:09 pm
Hi Vihar,
OK what you are doing is a bit tricky. If I will to EQ your mix, I would do the following:
Vocals:
-6dB cut (high pass filter) at 200Hz
+2dB boost Q=1.0 at 3000Hz
+2dB boost Q=1.0 at 15,000Hz
Guitar tracks:
-6dB cut Q=1.4 at 3000Hz
-3dB cut Q=1.4 at 800Hz
-2dB cut Q=1.4 at 200Hz
I think the problem is a “mud” between the bass frequencies of the vocals and guitars. Both tracks do have bass frequencies. The solution is to cut the vocals bass frequencies using a high pass filter, leaving only the guitar. Since vocal structure is still strong at the following vocal frequencies, I cut the guitar at 200Hz, 800Hz and 3000Hz to avoid masking problem.
Also consider panning the tracks, I place the vocals at the center (setting =0) and guitars at -50, +50 (left and right) in stereo). Panning can also avoid mud. Try that, and see if it solves your mixing problem.
May 20th, 2010 at 4:07 am
Thank you very much for your reply! Yes I kept trying around those frequencies, and if I removed enough the bad resonance went away but so did the fullness. Finally it turned out that the cabinet impulse responses I’ve been using for the DI recorded guitar were too muddy, even when playing the guitar track by itself, without the vocals. I’ve found new, flatter sounding ones so it’s much better now. The panning idea works great! I also put two slightly different sounding impulse responses of the same guitar cabinet on the left and right side, so I have a nice subtle stereo depth thing going on. Thanks again for your pro advices, and keep it going!
June 19th, 2010 at 5:06 am
Can you please explain what is a “phase cancellation” problem ?
I think I have this problem in a song. When I check my mix in mono, my guitar tracks (left and right) disappear. Is this phase cancellation ?
If so, how can I fix that?
Thanks
June 22nd, 2010 at 6:20 pm
Hello Ivan,
I just written a detailed a blog post pertaining to Phase Cancellation in Audio Mixing & Mastering: How to prevent this.
Thanks for dropping by.