Hi All,
Thanks to Gary for sharing such a great song!
Here's my mix submission:
WAV:
https://drive.google.com/open?id=1-wwdh ... 4EHUvFiD_C
MP3:
https://drive.google.com/open?id=1mXXWv ... uhqHIG7jYX
Overview
While it has a lot of the early 70’s vibe (e.g. Creedence Clearwater Revival, Pink Floyd style parts), it also has rock sensibilities (the EGTR Gain guitars) that can be used to turn the song into something huge and wild.
After listening to a few songs from the band Wilco (that Gary listed as references), I thought that this song could use some of the close-up intimate room sounds for the quieter acoustic guitar and vocal sections but go spectacularly bigger and wider in the choruses.
Mostly I've tried to keep the vibe of the song much the same, keeping the dynamics that are present whilst bringing out the beautiful sounding acoustic guitars, the awesome guitar solos, and making the most of Gary's vocal performance with his strong, deep and sometimes gritty voice.
While there was repair work to be done (particularly to the drums), it wasn't all as bad as I expected.
DAW
Reaper v5.963
Screenshots
Channel List 1
Channel List 2
Channel sends
Repair work
I spent quite a bit of time both identifying the groove and then figuring out what elements were causing the groove to falter in places.
I didn't initially align the tracks to the grids in Reaper as I wanted to see what was actually in the tracks first, find the groove and then do the alignments. However, after aligning the tracks to the 72 bpm tempo and bars, the time-alignment of tracks went smoothly.
Once I had time-aligned the drums and bass, I time-aligned the EGTR Gain tracks and guitar solos. In many cases, particularly the vocal, I chopped the audio into pieces and made very small (millisecond) adjustments to get them to better fit the groove (see the screenshots for cuts in the audio files).
I experienced phase problems with the drum tracks when attempting to EQ them. After adding the extra stereo drum room, I was able to get the drums into reasonable shape. I didn't do any sample replacement on the drums and I left the bass and drum tracks as is.
Essentially, I time-aligned the other parts around the groove created by the drums and bass.
Minor changes to the effects
I wasn’t keen on the long intro for the song with all of the Pad1 and Pad2 sounds, particularly in conjunction with such beautiful acoustic guitar sounds. So I shortened that section to make it more listener friendly.
In fact, I ended up only using Pad1 (heavy HPF and LPF to push it way back into the mix) and then automated volume and panning to give it some motion. I pretty much did the same with tremolo GTR in various places to avoid the repeated and overpowering nature of the sounds.
I then used a section of the first Pad1 section in the Verse 3 and I've dropped the drums at that point both to create a more intimate and softer Verse 3 but also to give the new Chorus more impact.
Initial levels
I adjusted the item gain of many of the drum parts and other loud sections down to around -18 dBFS for proper gain staging.
Generally I aimed for a max peak level of -24 dbFS for each channel ("staying in the green and sometimes flickering into the yellow").
Vocals
I did a lot of automation on the lead vocal, both pre-fx volume and volume (see Channel Screenshot 2). The pre-FX automation was used to even out the vocal level (avoiding excessive compression) and also bringing the vocal to life, particularly at the start and end of words and phrases.
I also automated the channel volume for verses and choruses.
On the main Vocal track (VOX Lead) I used:
- TDR Proximity to bring the vocal to the front of the mix without sounding too compressed
- TDR Nova was then used to remove any resonances and do some very slight de-essing duties
- UAD Antares auto-tuning to help subtly fix any "pitchy" parts in the performance. With more time, I could have done more manual correction but I think the Antares plugin does a pretty good job without sounding obvious. ReaTune can also be used for this too
- UAD SSL Channel E which was used for compression and EQ
The compression was light (ratio: 2:1, around 2-3 dB gain reduction).
The EQ was for beefing up the vocal, removing some nasal sounds and adding some "air" [/list]
I used a lot of processing on the vocal using Reaper's flexibility with folder groups and sends to 6 other channels (See Channel Sends screenshot):
- VOX CMP - Pre-fader send to UAD 1176 AE (2:1 compression, 1-5 db gain reduction, standard "10-to-2" attack and release settings) into a UAD LA2A set to around 50 with the needle barely ticking 1 dB gain reduction)
- VOX Grit - I found some of Gary's vocals became a bit "pretty" sounding particularly in the last choruses. To add more energy and some grit, I used a trick from Warren Huart and set a UAD 610-B into serious distortion (3 o'clock setting, +10 dB gain) and blended this in using automation during the choruses. It also had the effect of lifting a line in the last chorus ("Denver says, it's too late now boy, you're dead") where Gary's voice has a really awesome gravelly quality to it
- VOX Thick - Native Instruments Reflektor XT Delay with straight 1/32 note tempo-synced delay and tape saturation to thicken up the vocals
- VOX Delay - NI Reflektor XT Delay with multiple tempo-synced delays (1/8 dotted, 1/16 straight blended together)
- VOX Reverb - Ducked reverb using ReaCmp and Reaper's brilliant multi-channel sends to feed both the signal and sidechain with LiquidSonics Seventh Heaven Pro reverb (RT 1.40 s, large hall setting, 10 ms pre-delay)
Acoustic guitar
I wanted the acoustic guitar to sound natural so I didn't apply any overall compression to it.
TDR Nova dynamic EQ to remove resonances at 264, 541 and 1116 Hz. Added a low shelf (-3 dB @ 265 Hz, Q = 0.67), removed some mud with a narrow band (-3 dB @ 264 Hz, Q = 1.86).
After mixing this in, I noticed that it didn't quite have the right feel, so I fed the original signal into a UAD Neve 1073 to add some transformer colour (no EQ, just the circuit) and then into the dynamic EQ
Drums
I time-aligned the drums (mentioned in the Repair work above) and then passed the original signals into
UAD SPL Transient Designer on all the drums to help recover some more thud and smack. I tried to get away without it but it just sounded so much better to re-assemble the transients.
- Kick - UAD dbx 160a (ratio: 2:1, GR 1-2 dB), ReaEQ (HPF @ 48 Hz, -6 dB/oct to protect woofers,
band +1.5 dB @ 75 Hz to get some oomph, band +1.7 dB @ 2.6 kHz for beater smack, LPF @ 3.7 kHz, -4 dB/oct to cut highs from the mix)
- Snare - TDR Nova to remove resonances (ringing) and do EQ duties (HPF @ 101 Hz, 6dB/oct, band +2dB @ 221 Hz for more body and high shelf +3dB @ 12.5 kHz), UAD dbx 160a (ratio: 2:1, GR 1-2 dB). Normally I would compress before EQ but using the corrective EQ seemed to be a better idea.
- Toms - Only SPL Transient Designer
- Room + OH Bus - I mixed all the overhead and room channels together and compressed them with TDR Kotelnikov (ratio: 2:1, 1-5 dB gain reduction). I originally tried using EQ on these tracks (usually just HPF) and that failed miserably due to phasing issues. I decided to leave all the bleed in and do a better job of time-alignment instead.
All of the drum and room bus channels were then sent into a bus with TDR Nova acting like a de-esser on a 5 kHz band (1-2 dB reduction usually just snare), into UAD SSL E Channel with 2:1 compression and EQ to fatten up the kit, remove some boxiness and give it some highs). Then into the TDR Kotelnikov acting as a drum bus compressor (ratio: 2:1, 1-2 dB gain reduction).
Rhythm Bus
Standard NY-style blending of both Kick and Bass together with TDR Kotelnikov compressor (2:1, 1-2 dB gain reduction).
Electric guitars
I just panned these left and right (around 60%) and used both low and high shelves on them to push them back into the mix.
Tremolo guitar
Ducked using with the lead vocal with ReaComp compressor (around 1-5 dB gain reduction). As previously mentioned, I used automation of volume and panning to get it moving in conjunction with Pad1 (see the screenshots, look for the zig-zag red panning automation lines)
Lead guitar
Original signal sent into UAD Neve 1073 (circuit only, no EQ for transformer mojo), then into TDR Nova dynamic EQ for LPF and high shelf with resonance removal.
EGTR Lead track signal passed into another track (panned 70%R) with Native Instruments Guitar Rig 5 (Tweedman SC Solo Lead preset) and Reflektor XT delay. Both original and FX tracks are then blended together in a bus (track folder)
EGTR Gain
Similar treatment to Lead guitar (using NI Guitar Rig 5 with same preset) and Reflektor XT delay but original EGT GAIN and EGTR GAIN FX panned hard left and right (80%), blended together in the EGTR Gain Bus, then into a stereo field widener (width 115%). ReaEQ used to create a low shelf for EGT Gain original track.
Background Vocals
Panned hard left and right (84%), blended into bus with TDR Nova for EQ and dynamic deresonance, then into NI Reflektor XT with Choir Shift preset (adds ducked delay and chorus).
Reverbs and Delays
Ambience (Seventh Heaven Pro, RT1.40 s, Small Ambience, 10 ms pre-delay) and Size reverbs (RT1.40 s, Medium Hall, 10 ms pre-delay) were used with a Short Delay (Reflektor XT, 1/64 note tempo-synced delay) and standard 1/4 and 1/8 note delays (TS4 and TS8).
See the Sends Screenshot for details of which channels are sent to these Reverbs and Delays.
I would normally use a shorter ambience reverb time but 1.40 s with a pre-delay seemed to suit the sound I was after.
Dynamics info
Measured with MAAT DROffline Mk2
44.1k/24bit WAV
Channel columns are listed in order of: Left Right Joint
Max. TPL (True Peak Level - BS.1770): -1.53 -1.59 -1.53 (ITU)
Max. SPPM (Sample Peak Program Metering): -1.59 dBFS -1.60 dBFS -1.59 dBFS
RMS: -19.23 -18.24
Max. Momentary Loudness (BS.1770): -11.31 (non-ITU) -11.42 (non-ITU) -11.36 (ITU)
Max. Short-Term Loudness (BS.1770): -12.99 (non-ITU) -12.78 (non-ITU) -12.87 (ITU)
Integrated Loudness (LUFS - BS.1770): -18.20 (ITU)
DR (PMF Dynamic Range): 12.52 12.61 13 (PMF)
Min. PSR (SPPM-to-Short-Term Loudness Ratio): 7.99 (AES)
PLR (TP-to-Integrated Loudness Ratio): 16.67 (AES)
LRA (Loudness Range - BS.1770): 14.62 (ITU)
Please feel free to ask me any questions if something isn't clear.
Cheers,
Anto
I also use a pretty bright Room for ambience (SP2016) and of course a 1/8 delay (Echoboy jr) for the "cloudy cushion". I used have 1/4 delay throws, but that p#!ce of sh1te Mixbus decided to get rid of it. I even had to automate the send fader with my mouse, because apparently you can not automate send bypass/mute in Mixbus... so I am torturing myself with a software that can't do the simplest things and is your master that decides what to keep and what not. 