safeandsound123@googlemail.com
0044 (0) 7810 271371
The techniques of producing and mixing retro and vintage DAW music tracks
As a dedicated mastering engineer who is comfortable working with any and all genres please do not “typecast” me as a dance music mastering engineer. I am very comfortable masetring all genres as you can see in my extensive pro engineering BIOG page. In spare time I like to make some music as well, it keeps my production skills in shape and allows for some personal creativity.
Making a retro styled vintage sounding track is great fun, technically useful and this entry is about one I produced recently. I wished to explain for others what I did in the digital domain to emulate the sound of signal paths, synthesis and drums for the early 2,000’s. The following information can be applied for music of any era with a little thought. After all what else is a mastering engineer likely to do with their leisure time ?
Here is the completed track on YouTube
The sound equipment :
I based the signal path loosely on the commonly used 24/32 channel “project mixing console”. So, think Mackie 8 bus, Soundcraft Spirit Studio, Studiomaster P7, SoundTracs Topaz and maybe Allen and Heath GS series. The Mackie was arguably the better specified of those listed (with a fully parametric band and a HPF which worked on both mic and line sources) so I based my approximate emulation on this console.
Firstly, I suggest doing a little research online related to the production time period of interest. I would possibly type in some notable artists names with “in the studio” appended so you can get an idea of the type of equipment in common usage at the time. An image search can be especially useful so you can see the equipment in use in the studio.
For my track I knew I would be making something in the vane of an artist/s working from a fairly well specificed project studio. So an 8 bus mixer, 4-6 effects processors (reverbs, delays, phaser, flangers etc.) Names that spring to mind would be a Lexicon Alex/Reflex, Digitech Studio Quads, Zooms etc. Most project studios would only have had a few channels of compression/gating such as from dbx, Alesis (3630), L.A. Audio, Behringer composer’s, possibly a BBS or Drawmer. I used 8 in total which is slightly more than most studios of the time would have had. (a couple of these were not in use in the final mix)
Outboard EQ’s were not so common back then. (If you were doing well you may have had a Focusrite RED EQ or possibly a Manley Massive Passive) with a Mackie you had a decent set of EQ on each channel. The basics of the Mackie channel EQ’s are as follows (more on that, with a TOP TIP later) Low shelf at 80Hz +/15dB – High Shelf 12kHz, 1 semi-parametric band 45Hz to 3kHz and one fully parametric band with a range of range 500Hz – 18kHz – a HPF at 75Hz with an 18dB/Oct slope.
The signal path elements I tried to emulate :
I emulated some extremely subtle 3rd (and 5th/7th) harmonic opamp distortion (these mixing consoles were filled with 5532/4560/TL072 DIP8 package opamps) not exactly well known for euphonic saturation, however they do a low distortion, reliable job of amplifying signals within their headroom specification.) The odd harmonics are due to them commonly being biased Class A/B. For this I chose Hornet Plug ins, Analogue stage MK II which has an opamp harmonic profile. It is now confirmed this plug in does not model crosstalk. (in testing it myself I could not detect crosstalk using metering). So I am pleased I added Slate VCC “Mix Buss” for this characteristic. This software also added a little noise (hiss) as would be inherent in all analogue mixing consoles. I felt the aging effect of this plug best left at 0 years “aged”.
*Just to add the Mackie 8 bus did gain (pardon the pun) a reputation for having the line input gain maxed and fader dropped for some distortion. It was convenient crunch, and crunch it did. (though this was common on any of these consoles and not the sole preserve of the Mackie… I recall commonly overdriving input gain on my P7 line input gain on my Roland SH-101 to great effect)
Crosstalk – most likely not a major contributing influence on analogue sound but a detail worth employing nonetheless. Crosstalk is inherent in analogue mixing consoles. It is when a signal from one signal path makes its way into an adjacent circuits signal path channel. It is often capacitive or inductive in nature and typically is worst at around 10kHz. It also has ambiguous phase relationship and it is likely to be unknown in phase relation, relative to the desired signal. For this I added multiple instances of the Slate VCC (Virtual console collection) “Mix bus” plug in. I used the Channel plug in first but as far as I understand this does not model crosstalk so an instance of the VCC “Mix bus” was added after Hornet Analogue stage MK II. I chose Trident model (80B) which one could argue as being the closest choice to something like a Mackie 8 bus.
One can quite easily approach any era of music if you take into account the prevalent equipment used. Be it tape machines, specific mixing consoles and production processes based on equipment capability. Of course it will be difficult to emulate a world class 1970’s recording space but that is rather beyond the scope of the article.
For those interested, more on the Trident 80B here:
https://www.historyofrecording.com/Trident_Series_80B.html
There was nothing other than Slate VCC Mix Buss and Hornet Analogue stage MKII on the master bus:
An important EQ difference
For equalization I made a special version of the Mackie channel EQ in Fabfilter Pro-Q 3. Namely, 2 shelves and 2 mid bands and a HPF fixed at 75Hz 18dB/oct. However, and this is key to emulating the desk channel eq on stereo sources (i.e. stereo output synths). I right clicked each band indicator dot and every stereo instance was “Split” into L and R and for the left and right channels I added a visual and component tolerance inconsistency.
A project mixing console is a series of mono channels (almost all analogue mixers are), for a stereo synth source 2 channels are required panned hard left and right. As such the EQ settings would never be the same for each channel. Component error would be to the order of 5pct. Visual error could be the same or worse. So an intended EQ cut of -3dB at 330Hz.. would be something like -2.78dB at 310Hz for the left channel and the right channel -2.93dB at 328Hz. This should change the phase response between left and right channels a little. And as a consequence the stereo imaging precision. (or lack thereof)
I think this is most likely one solid reason why analogue desks have the oft cited and reputed wide sound. It is at least one reason and a reason rarely mentioned.
Just to add as point of interest crosstalk can also be generated by IVGI and SDRR by Klanghelm (free and low cost respectively). I found CPU usage to be a little high so did not use them this time. I was quite surprised to find my CPU (an old Intel 2600 i7 which generally does exceptionally well on a high track count projects) was feeling the pressure running at about 80pct ASIO in Cubase (RME 9632 PCI card).
As I understand the Hornet Analogue Stage MK II dynamically oversamples internally to 192kHz which may well add some CPU/ASIO overhead. I had about 60 tracks and 8 effects returns in the project but of course not all playing sound sources continuously. In addition, I had some phasers etc. inserted as could be typical of onboard synth effects.
My attention was brought to MackEQ from Airwindows. However each instance switched on seemed to bring gain to my channel that was unwanted as I had a few synths and sounds in the mix at the time I became aware of it. (Dropping down gain for every channel was not in my plan) The suggestion was nonetheless welcome as it made me think carefully about the EQ choices. This freeware plugin is worth considering. (given the gain applied insert before starting work on synths and getting gain structure set up)
The chosen sound sources
You need to know your genre and its history. Or at least do some basic investigations online, which is quite easy. So for my trance track I used the following synths, Viper, Access Virus Snow (no wavetables so it was like a model A and B), Tal U-No-X, NI Massive original (as it sounded old enough), Cakewalk Z3TA+2 which sounds rather similar to previous versions. reFX Vanguard 2, Disco DSP Discovery Pro (a Nord Lead a like) Rob Papen Predator 2 as this sounds rather like Rob Papens Albino synth. The bass pulse was a filtered Moog Taurus saw wave courtesy of Spectrasonics Trilian. (and shock horror, I added reverb to the bass !)
The above sound sources are likely to exhibit some aliasing in different parts of the synth architecture, often the filters. The envelopes may not be the tighest and so serve the purpose of vintage digital synthesis ideally.
I know this genre at the time used more breakbeats as percussive elements, edited, mixed and effected as per the era concerned and typically had a few voice samples. A few releases of the time had guitar so I added this courtesy of my Steinberger Spirit headless and a Line 6 POD XT Pro Rack 2004. It suited the tracks mood. I used a sample as the kick drum as it was convenient and sounded just like the right kick for the time. (a slight departure as I usually synthesize my own kicks in Kick2 or ISM Bazzism)
For effects I used Lexicon PCM plug ins (approximating semi pro Alex and Reflex units of the late 90’s) and for delay I used my favourite delay of all Waves H-Delay. On H Delay I had the analog setting set to position 1 for each instance which adds some noise (hiss). I set up 4 reverbs and 3 delays.. of short, mid and long (and a variant) RT’s and feedback. All FX return channels had the analogue emulation plug ins inserted before the processor. As these would also be subject to being returned to either dedicated FX returns in the master section or via 2 mono hard panned desk channels.
It would be unlikely that returns would have High pass filters engaged. Though not beyond the realms of possibility some low end would be shelved out. If returned to full line input channels as opposed to dedicated effects returns in the master section.
My noise floor hovered around -70dBFS as can be seen here…
I used various samples for the percussion that sounded right. I recorded the speech myself and added some basic eq, compression and effects.
I did not over drive any of these mixing console emulation plug ins, I was running “the desk” as if I had correctly PFL’d each channel (correct gain structure) at circa 0VU rather than overdriving anything for saturation effect. Any distortion was added selectively per sound, per synth. (there was very little in this mix)
I used 3 reference tracks of the era so I could keep on target producing a decent track (I hope !) and keeping a fairly authentic take on the sound of times gone by.
Choose your techniques for the era with care
Much comes down to synthesizing sounds of the time for this track, so it would be no use using modern advanced synthesis techniques or sounds that would not sound as if they were made in the relevant time period. Wavetable synths were not really commonly in use at the time so I kept with mainly VA and a little FM synthesis. Interest is maintained by using interesting MIDI programming and unusual melody/rhythmic sequences.
The same can be said for recording techniques of any period, was it recorded through a vintage desk? Onto tape ? Was it multitracked or predominantly played by musicians together live with some multitracking afterwards. How was the stereo image produced (if any) ?
For this track I did not do any kind of highly precise EQing that could not be performed on a Mackie 8 bus. So removing harsh resonances was not really an option. (something I routinely do with some sound sources) However as the lower mid range largely remains intact I personally feel this is why a gritty upper mid range and top end can still work as it is balanced against a warm lower mid. I call this opposing yet sympathetic areas of the spectrum. The perception of each is interdependent on each other. No dynamic eq, no precision super high Q digital EQing, and certainly no Mid-Side EQ.
Dynamic control was minimal. In fact I recall using only 1 compressor on the vocal element. A very simple work horse compressor DC8C from Klanghelm. More than adequate at doing the job.
In audio there are no rules. Well actually there are… but I know the rules well and so am confident in breaking them to useful effect. There is a lot of chat about use of high pass filters in modern EDM. Personally, I do not limit myself by a rule, when I need an HPF I use an HPF. I make a professional judgement based on sonic results not Youtube-isms. However, in precision we both gain and lose something. We gain control, detail and possibly power… however we lose life, a sense of freedom and wildness, drama. Precision can be superb it can also remove the life from music. So with this track I did not apply precision and ended up with something quite organic sounding, yes it is a little gritty*, yes the lower mids are not scooped out. The bottom end is loose and fat. And so arguably the track then sounds like 2001-2004 which was the goal.
*Synths of the time got away with their slightly hard sound because people were quite possibly not routinely putting high pass filters on many of their synths (or at least not a very high turnover/cutoff frequency), and so the sounds had the lower mid range to balance the slightly hard upper mid and high spectrum out. Undoutedly the sound is ragged and rough around the edges but age never really made much difference of whether I liked a track or not. Good tracks can be old or new.
As far as automation goes there was some, which a DAW of the time would be capable. However, I also considered in mind that some of the movement in tracks such as panning would be done manually by one or two pairs of hands when recording a mix, possibly to a DAT machine. So I did not do more than I thought was realistic at the time.
Period style mastering
Mastering was gentle, a few broad EQ moves and I felt early 2000’s mastering would not be complete with the use of the classic limiter which would have been heard on 1,000s of tracks of the era. The Waves L2 Ultramaximizer.. The track sits at the same perceived volume of masters of the era.
Conclusion
This was immense fun and rewarding, thinking about, planning and executing the process was a blast and you get an end result as well. What is not to like? It might be a bit niche but I like to challenge myself from time to time and the process can be really smooth and enjoyable with some planning. I had all the synths except Vanguard, 2 outlay was small. I treated myself to Waves L2 Ultramaximizer and thought it would be the icing on the cake. Preparation and thought about the era is key to success.
By the way this is my spare fun leisure time activity. If you are interested in having 20+ years of professional audio engineering experience applied to your tracks in the form of my dedication to mastering my rates are here (1 Track is £30.00):
Footer
Please feel free to “social media like” the article top right on this page and spread the article to those who may be inpsired in how to make retro tracks.
Copyright Barry Gardner 2022