Yesterday, a colleague of mine dropped by to get a few pointers on bass mixing. As many of you know, bass mixing is one of the most difficult aspects of mix engineering. Not only is it very contingent on the source tracks and the style of arrangement, but it also requires a significant degree of work in post processing to achieve what many would call a 'professional' low end.
As we went through the process of getting the bass from its raw DI state, to a place that was about 80-90% toward our desired end product, I realized that this would make a great tutorial to share with everybody.
The tracks in question are not ideal - hence, they create a good real-world representation of what many of you will run into on a day to day basis.
What he brought over were two stems. Stereo drums and the raw bass DI.
The performance on the DI was less than ideal. Inconsistent, lacking in power and drive. The tone itself was also less than ideal. Stuffy, resonant and indistinct. The goal was to take the DI track and do all the things we needed to do in order to lock it into the drums.
Here are the before and after clips:
Bass Processing Before & After
And here is the philosophy behind it:
Bass DI
Often when I receive a bass DI track like this, which is lacking in clarity and has obvious tonal issues I will do a light amount of processing to it before reamping. Normally I would run it through an equalizer and remove the crux of the junk between 160 to 180Hz and also a wide lower midrange scoop between 250Hz to 500Hz, depending on the tone of the bass. I may also compress it in order to even the dynamics of the playing before we go any further.
In this case, in order to save time I did not do any of this. The bass DI was reamped in its raw state.
Reamping
I ran the bass DI through a Sansamp PSA-1.1 unit. This is a versatile distortion processor that I've used on bass many times.
In this case we found that heavy amounts of distortion did not fit with the music, so it was dialed simply to emphasize the 'clang' portion of the bass tone. Slight overdrive.
You will all be using your own distortion processors, so there are no specifics to give at this point. Simply dial the tone to best match the scope of the project.
Make sure to print a click impulse or a stick hit through the unit so that you can phase align it with the raw DI track later.
Referencing
From this point on, you may wish to import a raw bass track which represents to you what a good bass tone is. The rock band multi tracks can be invaluable here. Otherwise, simply pick a song of your choosing and skip to a section where the drum and bass are playing alone. This will help to give you an idea of how to interlock the two.
Splitting
Now we have two bass tracks, along with our reference bass tones loaded into the session. You can see the layout as such:
https://dl.dropbox.com/u/285689/Forum/Tutorials/Bass Processing Tutorial/Bass-0.jpg
The role of the Bass Group tracks will become evident a little later.
For the moment what we want to do is go back to our Bass DI track. Because this DI track is so far from ideal sounding, I decided to load an instance of Ampeg SVX onto it, in order to aid in shaping the tone through broad strokes toward our desired end result.
After doing so, and tweaking the sim a bit, I still found the balance lacking. One of the biggest tonal issues with this DI is the tremendous ring and build-up between 160 to 180Hz. So I instantiated an equalizer after the sim and removed a large amount of content at 174Hz. I also set a lowpass at 3kHz in order to stop the top-end sounding too brittle.
You can see the result of this here:
https://dl.dropbox.com/u/285689/Forum/Tutorials/Bass Processing Tutorial/Bass-1.jpg
Now that we've sculpted our DI track a little more, we can go to our grit track, which in this case is the PSA-1.
There isn't much detail I can go into here, as it will be highly dependent on your own distortion processor. All I can suggest is that you high-pass this track quite heavily, and let the Bass DI track handle the low-end. My high pass in this case was around 500Hz. You may also want to low pass this track somewhere between 3 to 4kHz with a 6dB or 12dB/oct filter in order to soften the high end. This is especially necessary if you're dealing with heavily distorted tracks.
Now you simply bus these two tracks together into what we've called Bass Group 1.
This will create your fundamental tone, and what you will be spending the rest of your time processing.
Processing
You can see screenshots of the final processing chains here:
https://dl.dropbox.com/u/285689/Forum/Tutorials/Bass Processing Tutorial/Bass-2.jpg
https://dl.dropbox.com/u/285689/Forum/Tutorials/Bass Processing Tutorial/Bass-3.jpg
Level Riding
This part of the process is not necessary, but using a level-riding tool like Waves' Bass Rider, or manually riding the level of each note can be very handy before taking the track to the next step.
The idea is to even out the dynamics as much as possible - which is crucial in an uneven performance such as this - and lessen the load on the compressors before we move on. This way our bass track is allowed to breathe a lot more freely in the end.
Saturation
Because this is a pop-rock/punk type arrangement, I decided that the use of console and tape saturation would be a good idea. In very dense metal arrangements where clarity is key, you sometimes want to avoid doing this as it will increase the sense of density of the bass guitar, but also 'smear' it to a degree - lessening your ability to get the utmost tightness and note clarity from the source. In this case however the density and smear is a desirable quality.
Here we have used one instance of Slate's VCC set to the Brit 4k mode, and an instance of VTM set to use FG9 tape, 30 ips, 2" 16 track. Both saturation plugins are hovering around 0dB on their respective VU meters, but there's no reason you can't push them harder.
Limiting
The idea behind having a limiter right after the saturation plug-ins is in order to control any errant transient peaks before the track goes into our main compressor/s. We want to lighten the load in increments. In this case, as the performance was so dynamic, this was a natural choice.
Compression
In this case we have two compressors doing the brunt of our work.
The 1176 D is an actual hardware compressor which I dialed for 7 to 10dB of gain reduction. Attack was roughly around noon, and Release was near 3 o'clock.
At this point I wanted the compressor to still retain transient content, in order to keep the bass sounding punchy. The release was set to be a bit slower in order to keep the performance sounding more level, and prevent any errant and undesired transient peaks from destroying the 'bed' of the performance.
The 1176 was a good choice in this regard, as it's a very dense sounding compressor and suits simple rock material well.
The second 1176 came in plug-in form, and was set to effectively function as a lite-limiter. This helped control the transient spikes, and leveled out the performance even more. Gain reduction here was between 1 to 2dB, in order to not stifle and muffle the performance too much. The key here is to aim for control, rather than obliteration of your bass.
Equalization
Here is where things get very source-dependent.
In our case the most natural move to make was a broad cut at 300Hz. This removed the crux of the useless muddy content around the lower midrange.
The second move was to further subjugate the ringy, resonant upper-bass content at 174Hz.
Now we removed the errant low-end rumble below 60Hz. You would be surprised how little sub-bass content you actually need in order to have a perceptibly 'large' bass tone. The trick is usually to cut enough upper bass and low-mid junk away so that your mid-bass frequencies sound very tight and clear. Your bass drum generally helps complete this illusion by slotting nicely between the 50 to 70Hz area.
Now we set a high shelf to remove any high end that's creeped up during this process. As this is a largely subtractive equalization methodology, you may find yourself needing to adjust this shelf as you continue throughout your processing.
Cuts were made at 56Hz, 112Hz and 209Hz. They were simply done to even out the tone of this particular bass. After our prior processing, the bass rang out too prominently at 112Hz, so it was attenuated. The 56Hz cut helped reduce clashing with the kick (which itself was quite peaky at around 50Hz). The 209Hz cut would help let the snare bottom end through, which quite commonly peaks between 190 to 210Hz.
Limiting Stage 2
The limiter here is effectively the last bit of transient control. It's dialed to even out the peaks and keep the performance consistent. It's essentially an extension to our second 1176 compressor, offering the final bit of control. I generally insert these after all the EQ processing to the bass, in order to level out any slight resonances or peaks that our EQ processing might have created.
Equalization Stage 2
I decided that at this point the bass still did not sound 'ready' enough to showcase, so I created another group (as we ran out of insert slots on the last), and instantiated yet another equalizer.
The idea here was to tweak up a few issues left within the tone. A 250Hz cut in order to further subdue the lower midrange junk. A 150Hz cut in order to remove more mid-bass ringing. A 120Hz low-shelf in order to increase the density of the bass. Cuts at 71Hz and 86Hz in order to tighten up the low-end, and allow the kick to come through more naturally.
You will generally find yourself balancing with these fine-tweaking equalization stages throughout the mix, until you finally find the overall balance which works for the project.
Generally these equalization stages would happen before the final limiter, but in this example I got lazy and inserted the EQ after it.
Spectrum Analysis
This can be invaluable for low-end processing. I commonly keep an instance of Curve EQ at the very end of my bass chain so that I can see what's actually happening in terms of frequency content.
In this case what the analyzer tells me is that the bass tone might still have a bit too much prominence at around 45Hz. That perhaps I've removed a little too much content at around 70Hz, and 300 to 500Hz. Perhaps there is also some left-over weirdness happening at 200Hz.
But this is all song and mix dependent. At this stage you have to look, ask yourself what's happening and LISTEN. If you hear problems, let the analyzer help you find them. But don't get too obsessed with a 'pretty' looking spectrum. Sometimes flattening it all out simply leads you to more problems.
Gain Staging
One crucial aspect of heavy processing like this is to gain-match after every new processor. Make sure to compensate for lost level by using the make-up gain on your EQs and compressors. Always practice good gain staging.
Linear Phase Equalization
I can't stress this enough. In my opinion the best way to achieve a tight, solid low-end is to use linear phase EQ for the majority of your processing. It will simply sound cleaner, leaner and also retain more transient punch.
Use Headphones
Unless you're mixing in a perfect room, with perfect monitoring you will have trouble getting the 'ideal' low-end. Don't be afraid to put on some cans and do a lot of your low-end balancing through them. They will generally showcase peaks and nodes quite clearly, without the interference of natural room modes to create drama and need for interpretation.
Final Words
Bear in mind that this is a real-world example of what a majority of engineers might run into on a day to day basis.
The DI tone and performance is far from ideal, and as such so will the final tone always be. We can do a lot in order to clean up, but there is no replacement for a good bassist playing through a good instrument, captured through good gear.
The end result we achieved after this particular chain of processing is perhaps 80 to 90% of the way toward what I would want a final bass tone to sound like. It still needs work, and that part is your job. To find your own idea of what sounds 'right'. This tutorial was mostly to equip you with the tools and methodology to go about getting there.
Thanks for reading!
If you want a full-fledged tutorial on bass mixing, you can find it (amongst many others) in my e-book, the Systematic Mixing Guide, which can be purchased from here:
As we went through the process of getting the bass from its raw DI state, to a place that was about 80-90% toward our desired end product, I realized that this would make a great tutorial to share with everybody.
The tracks in question are not ideal - hence, they create a good real-world representation of what many of you will run into on a day to day basis.
What he brought over were two stems. Stereo drums and the raw bass DI.
The performance on the DI was less than ideal. Inconsistent, lacking in power and drive. The tone itself was also less than ideal. Stuffy, resonant and indistinct. The goal was to take the DI track and do all the things we needed to do in order to lock it into the drums.
Here are the before and after clips:
Bass Processing Before & After
And here is the philosophy behind it:
Bass DI
Often when I receive a bass DI track like this, which is lacking in clarity and has obvious tonal issues I will do a light amount of processing to it before reamping. Normally I would run it through an equalizer and remove the crux of the junk between 160 to 180Hz and also a wide lower midrange scoop between 250Hz to 500Hz, depending on the tone of the bass. I may also compress it in order to even the dynamics of the playing before we go any further.
In this case, in order to save time I did not do any of this. The bass DI was reamped in its raw state.
Reamping
I ran the bass DI through a Sansamp PSA-1.1 unit. This is a versatile distortion processor that I've used on bass many times.
In this case we found that heavy amounts of distortion did not fit with the music, so it was dialed simply to emphasize the 'clang' portion of the bass tone. Slight overdrive.
You will all be using your own distortion processors, so there are no specifics to give at this point. Simply dial the tone to best match the scope of the project.
Make sure to print a click impulse or a stick hit through the unit so that you can phase align it with the raw DI track later.
Referencing
From this point on, you may wish to import a raw bass track which represents to you what a good bass tone is. The rock band multi tracks can be invaluable here. Otherwise, simply pick a song of your choosing and skip to a section where the drum and bass are playing alone. This will help to give you an idea of how to interlock the two.
Splitting
Now we have two bass tracks, along with our reference bass tones loaded into the session. You can see the layout as such:
https://dl.dropbox.com/u/285689/Forum/Tutorials/Bass Processing Tutorial/Bass-0.jpg
The role of the Bass Group tracks will become evident a little later.
For the moment what we want to do is go back to our Bass DI track. Because this DI track is so far from ideal sounding, I decided to load an instance of Ampeg SVX onto it, in order to aid in shaping the tone through broad strokes toward our desired end result.
After doing so, and tweaking the sim a bit, I still found the balance lacking. One of the biggest tonal issues with this DI is the tremendous ring and build-up between 160 to 180Hz. So I instantiated an equalizer after the sim and removed a large amount of content at 174Hz. I also set a lowpass at 3kHz in order to stop the top-end sounding too brittle.
You can see the result of this here:
https://dl.dropbox.com/u/285689/Forum/Tutorials/Bass Processing Tutorial/Bass-1.jpg
Now that we've sculpted our DI track a little more, we can go to our grit track, which in this case is the PSA-1.
There isn't much detail I can go into here, as it will be highly dependent on your own distortion processor. All I can suggest is that you high-pass this track quite heavily, and let the Bass DI track handle the low-end. My high pass in this case was around 500Hz. You may also want to low pass this track somewhere between 3 to 4kHz with a 6dB or 12dB/oct filter in order to soften the high end. This is especially necessary if you're dealing with heavily distorted tracks.
Now you simply bus these two tracks together into what we've called Bass Group 1.
This will create your fundamental tone, and what you will be spending the rest of your time processing.
Processing
You can see screenshots of the final processing chains here:
https://dl.dropbox.com/u/285689/Forum/Tutorials/Bass Processing Tutorial/Bass-2.jpg
https://dl.dropbox.com/u/285689/Forum/Tutorials/Bass Processing Tutorial/Bass-3.jpg
Level Riding
This part of the process is not necessary, but using a level-riding tool like Waves' Bass Rider, or manually riding the level of each note can be very handy before taking the track to the next step.
The idea is to even out the dynamics as much as possible - which is crucial in an uneven performance such as this - and lessen the load on the compressors before we move on. This way our bass track is allowed to breathe a lot more freely in the end.
Saturation
Because this is a pop-rock/punk type arrangement, I decided that the use of console and tape saturation would be a good idea. In very dense metal arrangements where clarity is key, you sometimes want to avoid doing this as it will increase the sense of density of the bass guitar, but also 'smear' it to a degree - lessening your ability to get the utmost tightness and note clarity from the source. In this case however the density and smear is a desirable quality.
Here we have used one instance of Slate's VCC set to the Brit 4k mode, and an instance of VTM set to use FG9 tape, 30 ips, 2" 16 track. Both saturation plugins are hovering around 0dB on their respective VU meters, but there's no reason you can't push them harder.
Limiting
The idea behind having a limiter right after the saturation plug-ins is in order to control any errant transient peaks before the track goes into our main compressor/s. We want to lighten the load in increments. In this case, as the performance was so dynamic, this was a natural choice.
Compression
In this case we have two compressors doing the brunt of our work.
The 1176 D is an actual hardware compressor which I dialed for 7 to 10dB of gain reduction. Attack was roughly around noon, and Release was near 3 o'clock.
At this point I wanted the compressor to still retain transient content, in order to keep the bass sounding punchy. The release was set to be a bit slower in order to keep the performance sounding more level, and prevent any errant and undesired transient peaks from destroying the 'bed' of the performance.
The 1176 was a good choice in this regard, as it's a very dense sounding compressor and suits simple rock material well.
The second 1176 came in plug-in form, and was set to effectively function as a lite-limiter. This helped control the transient spikes, and leveled out the performance even more. Gain reduction here was between 1 to 2dB, in order to not stifle and muffle the performance too much. The key here is to aim for control, rather than obliteration of your bass.
Equalization
Here is where things get very source-dependent.
In our case the most natural move to make was a broad cut at 300Hz. This removed the crux of the useless muddy content around the lower midrange.
The second move was to further subjugate the ringy, resonant upper-bass content at 174Hz.
Now we removed the errant low-end rumble below 60Hz. You would be surprised how little sub-bass content you actually need in order to have a perceptibly 'large' bass tone. The trick is usually to cut enough upper bass and low-mid junk away so that your mid-bass frequencies sound very tight and clear. Your bass drum generally helps complete this illusion by slotting nicely between the 50 to 70Hz area.
Now we set a high shelf to remove any high end that's creeped up during this process. As this is a largely subtractive equalization methodology, you may find yourself needing to adjust this shelf as you continue throughout your processing.
Cuts were made at 56Hz, 112Hz and 209Hz. They were simply done to even out the tone of this particular bass. After our prior processing, the bass rang out too prominently at 112Hz, so it was attenuated. The 56Hz cut helped reduce clashing with the kick (which itself was quite peaky at around 50Hz). The 209Hz cut would help let the snare bottom end through, which quite commonly peaks between 190 to 210Hz.
Limiting Stage 2
The limiter here is effectively the last bit of transient control. It's dialed to even out the peaks and keep the performance consistent. It's essentially an extension to our second 1176 compressor, offering the final bit of control. I generally insert these after all the EQ processing to the bass, in order to level out any slight resonances or peaks that our EQ processing might have created.
Equalization Stage 2
I decided that at this point the bass still did not sound 'ready' enough to showcase, so I created another group (as we ran out of insert slots on the last), and instantiated yet another equalizer.
The idea here was to tweak up a few issues left within the tone. A 250Hz cut in order to further subdue the lower midrange junk. A 150Hz cut in order to remove more mid-bass ringing. A 120Hz low-shelf in order to increase the density of the bass. Cuts at 71Hz and 86Hz in order to tighten up the low-end, and allow the kick to come through more naturally.
You will generally find yourself balancing with these fine-tweaking equalization stages throughout the mix, until you finally find the overall balance which works for the project.
Generally these equalization stages would happen before the final limiter, but in this example I got lazy and inserted the EQ after it.
Spectrum Analysis
This can be invaluable for low-end processing. I commonly keep an instance of Curve EQ at the very end of my bass chain so that I can see what's actually happening in terms of frequency content.
In this case what the analyzer tells me is that the bass tone might still have a bit too much prominence at around 45Hz. That perhaps I've removed a little too much content at around 70Hz, and 300 to 500Hz. Perhaps there is also some left-over weirdness happening at 200Hz.
But this is all song and mix dependent. At this stage you have to look, ask yourself what's happening and LISTEN. If you hear problems, let the analyzer help you find them. But don't get too obsessed with a 'pretty' looking spectrum. Sometimes flattening it all out simply leads you to more problems.
Gain Staging
One crucial aspect of heavy processing like this is to gain-match after every new processor. Make sure to compensate for lost level by using the make-up gain on your EQs and compressors. Always practice good gain staging.
Linear Phase Equalization
I can't stress this enough. In my opinion the best way to achieve a tight, solid low-end is to use linear phase EQ for the majority of your processing. It will simply sound cleaner, leaner and also retain more transient punch.
Use Headphones
Unless you're mixing in a perfect room, with perfect monitoring you will have trouble getting the 'ideal' low-end. Don't be afraid to put on some cans and do a lot of your low-end balancing through them. They will generally showcase peaks and nodes quite clearly, without the interference of natural room modes to create drama and need for interpretation.
Final Words
Bear in mind that this is a real-world example of what a majority of engineers might run into on a day to day basis.
The DI tone and performance is far from ideal, and as such so will the final tone always be. We can do a lot in order to clean up, but there is no replacement for a good bassist playing through a good instrument, captured through good gear.
The end result we achieved after this particular chain of processing is perhaps 80 to 90% of the way toward what I would want a final bass tone to sound like. It still needs work, and that part is your job. To find your own idea of what sounds 'right'. This tutorial was mostly to equip you with the tools and methodology to go about getting there.
Thanks for reading!
If you want a full-fledged tutorial on bass mixing, you can find it (amongst many others) in my e-book, the Systematic Mixing Guide, which can be purchased from here:
