Hey guys,
I've wanted to do something like this for a little while. Originally in video form, but that may have to wait. It's basically a supportive tutorial, using the methods outlined in this guide: http://www.ultimatemetal.com/forum/...-series-1-poking-holes-high-gain-guitars.html
I've been mixing an exceptionally difficult project over the last little while. The guitars on this one have been giving me trouble due to a few factors:
- Budget DI recording chain.
- Very bright guitar/pickups into very bright/fizzy amp
- High track counts, which means lots of potential clashing
This is why I realized it's a perfect example to use for this little tutorial. It illustrates how you can go about 'cleaning up' a raw tone into something a bit more mix friendly, while at the same time illustrating the limits of what you can do in post processing. As much as you can tidy up a tone, you cannot change its inherent character or harmonic structure, so you still end up with many of the characteristics endemic to the raws. In this case it's an abrasive, peaky high-end, and flat, 2-dimensional midrange, which can be subdued to a degree, but never entirely eliminated.
As a note, please realize that the mix of all the instruments outside the guitars here is from the band's pre-production demo. This isn't a reflection of the final mix of the CD - we're only using this backing track for perspective on how rhythms can be mixed to fit into an arrangement.
Raw guitars with pre-production backing track: http://dl.dropbox.com/u/285689/Foru...ssing Tutorial/Judgment Rough - Cobra Raw.mp3
Processed guitars with pre-production backing track: http://dl.dropbox.com/u/285689/Foru...Tutorial/Judgment Rough - Cobra Processed.mp3
You can see a picture of the processing chain here: http://dl.dropbox.com/u/285689/Foru...orial/Ermz---Guitar-Processing-Screen-Cap.jpg
Please refer to the picture as you read through the explanation of why certain things were done.
Filtering
Firstly, there is something not shown in the picture. That is the high-pass filtering, which was done at an earlier stage. The guitars are quad-tracked, broken up into 2 pairs. I grouped those pairs individually, prior to the overall guitar bus. Those earlier groups are where I put high-passes of 18dB/oct at 113Hz.
The reason I staged the high-passes earlier was so that there was no undue low-end causing the compressor to act and 'pump' the guitars.
A lowpass at ~7.5kHz appears later in the chain, on the main EQ.
Saturation
The first element I hit the guitars with is saturation - two stages of it. The first stage is VCC Neve saturation, and that's to give the guitars a subtle depth, width, and to round off the transients, helping to glue the guitars into the mix. I originally used the 4k saturation, but it created tubbyness around the lower-midrange and otherwise just didn't lend itself well to the tone of these guitars. The Neve saturation created extra depth, while also subtly rolling off some abrasive highs.
The second element is Nebula CLC Input Line saturation. This is essentially a frequency shaping stage, which further prepares the guitars for what's to come. The overall effect moves the tone further into the midrange - helping roll off more abrasive high-end before the equalizing has even begun.
Broadband Compression
This is something I don't always employ, but I've come to realize its benefits when dealing with less than ideal circumstances.
The Waves SSL Bus Compressor is one of the only software compressors I've found to work well on distorted guitars. The irony here is that my hardware GSSL does not work anywhere near as well as the Waves version for this purpose.
The overall effect of setting a slow attack time and automatic release on this compressor is to tighten up the overall sound. It helps focus the transients and give a sense of punchier, cleaner playing. It also unifies the tone of the guitars and lets them sit within the mix more easily. If the compressor had a stereo 'unlink' function, it would make for a great day-to-day guitar bus compressor.
Equalization
This is where things get interesting.
When I first listen to the raws, I try to zero-in on the most abrasive, or detracting element from the tone. In this case that was clearly the huge fizz spike at 6.2kHz. After this was attenuated, I focused on the 'flat' midrange character, centered around 600Hz.
After these were worked out, it became a matter of needing to clean up the higher midrange. I started with the wide scoop at 2.5kHz to create more space within the tone, and further attenuate the overall 'junk' midrange content. When this was achieved, it became obvious that some harshness presented itself around 4kHz, which I then notched out.
At this point the low-end started to become an issue. There was a muddy build-up between 120 to 250Hz, which I attenuated somewhat with a cut at 200Hz, to cut away the content that would clash with the bass, but leave some of the 'fuller' lows around 100hz intact.
This tone has an inherent cloudyness centered at 370Hz, which I targeted after all the other major issues were sorted. This is a very dangerous area, as too much subtraction here can completely hollow out a tone - leaving it powerless. The lower midrange is a difficult region to balance. 1dB can be the difference between a mix that sounds tight and balanced, or tubby, or thin.
When I realized that this EQ alone, despite how much I wanted it to, didn't get the guitar tone all the way there, I had to instantiate another EQ. The overwhelming factor at this point was a midrange 'flatness', which I targeted with a wide scoop at 900Hz. This cleaned up the mids, but let some of the more abrasive content across the high midrange come through again.
Realizing that broad sweeps alone wouldn't solve it, I started to take a notched approach again, taking away from nodes centered at 3.4kHz (right in the middle of the critical vocal range - naughty guitars!), and 5kHz, which had some fairly abrasive fizz left over from what the cut at 6.2kHz originally didn't get.
After all this was balanced, the low-end started creeping back up, so I set up a low-shelf at 230Hz to allow me to attenuate low-end conveniently as it rose up with each new stage of midrange processing.
What's important to realize here is that I was always going back and forth with the balancing. I didn't immediately start with a 10dB cut at 6.2kHz. It would've been closer to 5dB at the start, but as I worked out the midrange and certain frequencies started becoming more prominent than others, these cuts need adjustment. Each new stage of processing usually necessitates more adjustment in prior stages.
All subtractive
Please note that the EQ processing here is 100%, completely subtractive, exactly as outlined in the original guide. Not a single frequency was boosted at any point. This approach centers solely around removing frequency junk, or 'poking holes' if you will.
It's not about desiring to alter the tone itself, but rather trying to reveal all the desirable aspects of the tone, hidden beneath all the frequency junk.
Remember that no matter how much you process, you will not change the inherent character of the tone. Your best option is always to track well, and then smooth out the rough edges using your processing.
In this circumstance that is remarkably difficult to achieve... impossible in fact. The tone itself will never be 'perfect', and that's why it's good to show you where the limitations lie.
Further processing
The tone you hear is only about 80% complete.
It still requires a multi-band compressor across the mid-bass - that's something which only makes sense to tweak with the final mix rolling along. Without the final kick and bass sounds for reference, there isn't much point tweaking guitar lows.
It also has some funny abrasive stuff happening across the upper mids. It has to be accepted, however, that some element of this will always live on in the tone. It is incapable of sounding 'soft' without removing all riff clarity in the process.
Finally, I'll likely try an L1 at the end of the chain, chiefly to see whether it will help glue the guitars into the mix further and gain some free headroom, or whether it will cause the punch of the tone to fall apart. It's a simple 'try it and see' situation. If it works, it stays, if it doesn't, it goes.
-----
Anyway, thanks for bearing with me. I hope that despite the less than ideal starting circumstances this guide serves as a useful reference to some of you.
I've wanted to do something like this for a little while. Originally in video form, but that may have to wait. It's basically a supportive tutorial, using the methods outlined in this guide: http://www.ultimatemetal.com/forum/...-series-1-poking-holes-high-gain-guitars.html
I've been mixing an exceptionally difficult project over the last little while. The guitars on this one have been giving me trouble due to a few factors:
- Budget DI recording chain.
- Very bright guitar/pickups into very bright/fizzy amp
- High track counts, which means lots of potential clashing
This is why I realized it's a perfect example to use for this little tutorial. It illustrates how you can go about 'cleaning up' a raw tone into something a bit more mix friendly, while at the same time illustrating the limits of what you can do in post processing. As much as you can tidy up a tone, you cannot change its inherent character or harmonic structure, so you still end up with many of the characteristics endemic to the raws. In this case it's an abrasive, peaky high-end, and flat, 2-dimensional midrange, which can be subdued to a degree, but never entirely eliminated.
As a note, please realize that the mix of all the instruments outside the guitars here is from the band's pre-production demo. This isn't a reflection of the final mix of the CD - we're only using this backing track for perspective on how rhythms can be mixed to fit into an arrangement.
Raw guitars with pre-production backing track: http://dl.dropbox.com/u/285689/Foru...ssing Tutorial/Judgment Rough - Cobra Raw.mp3
Processed guitars with pre-production backing track: http://dl.dropbox.com/u/285689/Foru...Tutorial/Judgment Rough - Cobra Processed.mp3
You can see a picture of the processing chain here: http://dl.dropbox.com/u/285689/Foru...orial/Ermz---Guitar-Processing-Screen-Cap.jpg
Please refer to the picture as you read through the explanation of why certain things were done.
Filtering
Firstly, there is something not shown in the picture. That is the high-pass filtering, which was done at an earlier stage. The guitars are quad-tracked, broken up into 2 pairs. I grouped those pairs individually, prior to the overall guitar bus. Those earlier groups are where I put high-passes of 18dB/oct at 113Hz.
The reason I staged the high-passes earlier was so that there was no undue low-end causing the compressor to act and 'pump' the guitars.
A lowpass at ~7.5kHz appears later in the chain, on the main EQ.
Saturation
The first element I hit the guitars with is saturation - two stages of it. The first stage is VCC Neve saturation, and that's to give the guitars a subtle depth, width, and to round off the transients, helping to glue the guitars into the mix. I originally used the 4k saturation, but it created tubbyness around the lower-midrange and otherwise just didn't lend itself well to the tone of these guitars. The Neve saturation created extra depth, while also subtly rolling off some abrasive highs.
The second element is Nebula CLC Input Line saturation. This is essentially a frequency shaping stage, which further prepares the guitars for what's to come. The overall effect moves the tone further into the midrange - helping roll off more abrasive high-end before the equalizing has even begun.
Broadband Compression
This is something I don't always employ, but I've come to realize its benefits when dealing with less than ideal circumstances.
The Waves SSL Bus Compressor is one of the only software compressors I've found to work well on distorted guitars. The irony here is that my hardware GSSL does not work anywhere near as well as the Waves version for this purpose.
The overall effect of setting a slow attack time and automatic release on this compressor is to tighten up the overall sound. It helps focus the transients and give a sense of punchier, cleaner playing. It also unifies the tone of the guitars and lets them sit within the mix more easily. If the compressor had a stereo 'unlink' function, it would make for a great day-to-day guitar bus compressor.
Equalization
This is where things get interesting.
When I first listen to the raws, I try to zero-in on the most abrasive, or detracting element from the tone. In this case that was clearly the huge fizz spike at 6.2kHz. After this was attenuated, I focused on the 'flat' midrange character, centered around 600Hz.
After these were worked out, it became a matter of needing to clean up the higher midrange. I started with the wide scoop at 2.5kHz to create more space within the tone, and further attenuate the overall 'junk' midrange content. When this was achieved, it became obvious that some harshness presented itself around 4kHz, which I then notched out.
At this point the low-end started to become an issue. There was a muddy build-up between 120 to 250Hz, which I attenuated somewhat with a cut at 200Hz, to cut away the content that would clash with the bass, but leave some of the 'fuller' lows around 100hz intact.
This tone has an inherent cloudyness centered at 370Hz, which I targeted after all the other major issues were sorted. This is a very dangerous area, as too much subtraction here can completely hollow out a tone - leaving it powerless. The lower midrange is a difficult region to balance. 1dB can be the difference between a mix that sounds tight and balanced, or tubby, or thin.
When I realized that this EQ alone, despite how much I wanted it to, didn't get the guitar tone all the way there, I had to instantiate another EQ. The overwhelming factor at this point was a midrange 'flatness', which I targeted with a wide scoop at 900Hz. This cleaned up the mids, but let some of the more abrasive content across the high midrange come through again.
Realizing that broad sweeps alone wouldn't solve it, I started to take a notched approach again, taking away from nodes centered at 3.4kHz (right in the middle of the critical vocal range - naughty guitars!), and 5kHz, which had some fairly abrasive fizz left over from what the cut at 6.2kHz originally didn't get.
After all this was balanced, the low-end started creeping back up, so I set up a low-shelf at 230Hz to allow me to attenuate low-end conveniently as it rose up with each new stage of midrange processing.
What's important to realize here is that I was always going back and forth with the balancing. I didn't immediately start with a 10dB cut at 6.2kHz. It would've been closer to 5dB at the start, but as I worked out the midrange and certain frequencies started becoming more prominent than others, these cuts need adjustment. Each new stage of processing usually necessitates more adjustment in prior stages.
All subtractive
Please note that the EQ processing here is 100%, completely subtractive, exactly as outlined in the original guide. Not a single frequency was boosted at any point. This approach centers solely around removing frequency junk, or 'poking holes' if you will.
It's not about desiring to alter the tone itself, but rather trying to reveal all the desirable aspects of the tone, hidden beneath all the frequency junk.
Remember that no matter how much you process, you will not change the inherent character of the tone. Your best option is always to track well, and then smooth out the rough edges using your processing.
In this circumstance that is remarkably difficult to achieve... impossible in fact. The tone itself will never be 'perfect', and that's why it's good to show you where the limitations lie.
Further processing
The tone you hear is only about 80% complete.
It still requires a multi-band compressor across the mid-bass - that's something which only makes sense to tweak with the final mix rolling along. Without the final kick and bass sounds for reference, there isn't much point tweaking guitar lows.
It also has some funny abrasive stuff happening across the upper mids. It has to be accepted, however, that some element of this will always live on in the tone. It is incapable of sounding 'soft' without removing all riff clarity in the process.
Finally, I'll likely try an L1 at the end of the chain, chiefly to see whether it will help glue the guitars into the mix further and gain some free headroom, or whether it will cause the punch of the tone to fall apart. It's a simple 'try it and see' situation. If it works, it stays, if it doesn't, it goes.
-----
Anyway, thanks for bearing with me. I hope that despite the less than ideal starting circumstances this guide serves as a useful reference to some of you.
