Mark Wingfield is a mixing and mastering engineer at Heron Island Studio. He has three DownBeat Masterpiece albums under his belt, a Jazzwise Best Release Of The Year, and over 300 rave reviews from around the world for albums he has worked on. In this series of articles Mark gives advice on how to avoid common mistakes made in the recording studio which will hold back your album from sounding great. In his final piece he covers the other things to consider when in the recording studio:
Most people these days know that recording at 24 bits is essential. But we still get a lot of recordings coming to us which are recorded at 44.1 or 48 kHz rather than the higher rates of 96 or 88.2 kHz. You may not be able to tell the difference between 44.1/48 kHz and 96/88.2 but experiments confirm that many people can. This is important to many audiophiles out there, and many jazz listeners are audiophiles. They are also the part of your audience who are happy to pay for your music. Don’t neglect them for the sake of a little extra disk space when recording.
Admittedly the difference in sound between 44.1/48 kHz and the higher rates of 88.2/96 kHz is extremely small, to the point where most people on most systems will not hear a difference. If the audiophile listeners are not a concern to you, you might legitimately wonder why it is worth worrying about if the difference is so small. However there is a much more important reason to record at 88.2 or 96 kHz.
These days most music is mixed either entirely digitally or using a mixture of digital and analog processing. If any digital processing is being used and introduces any kind of saturation, (such as an analog emulation, any kind of saturator, many compressors and limiters) sample rates of 44.1 or 48 kHz will cause a problem. This is an issue known as aliasing, which is a kind of digital distortion (not a nice kind). Aliasing is the kind of distortion which is usually not immediately obvious to the ears, but it’s impacts are pervasive none the less. The effect aliasing has on the sound is often described as “cold and digital sounding”. A mix free of aliasing is often described as sounding “warmer” more “analog” and “natural”. Aliasing can also partially collapse the sound stage causing people to say that the music has less “depth” than music without the distortion. In other words it is to be avoided at all cost. This is where 88.2 or 96 kHz come in. These higher rates allow software processors used during mixing to completely remove aliasing. There is specialist software available which can show you exactly how much aliasing your processors are producing. When I first tried this I remember being shocked at just how much aliasing I was seeing at 44.1 or 48 kHz, it’s not subtle (even in plugins from the best manufacturers). I was also amazed at how completely the aliasing disappears when you move to 88.2 kHz.
This is why I recommend recording at 88.2 kHz. I don’t think there’s any need to go up to 96 kHz as there is no audible difference between 96 and 88.2 kHz and at 88.2 all the aliasing is already removed. Moreover recording 96 kHz will mean the mixing computer’s CPU is that much more stressed, which may limit some mixing options on many systems.
In rock and pop, a lot of saturation is used, sometimes on every instrument. Saturation is a kind of mild distortion which can also reduce the dynamic range (difference between loud and quiet notes). In jazz very little deliberate saturation is used and only in very specific ways. Engineers recording rock and pop often use microphone preamps which add a lot of saturation. This is not ideal for jazz, which is all about the details and the subtlety of playing. If you have saturation on every microphone, it fills in the gaps between sounds. By the time you’ve added all the saturation together in the mix, you will find it hard to hear the details of what everyone is playing. With too much saturation things can end up sounding muddy, cloudy, “closed in” and even thin and harsh.
If your engineer opts to use vintage microphone preamps, a vintage mixing desk or other vintage type gear for recording, keep in mind that these will add saturation to your sound (sometimes hefty amounts of it). With jazz you are safer using very clean but very high quality preamps. If the studio doesn’t have enough clean preamps for all the instruments, reeds, brass, guitar or vocals handle blanket saturation best. Piano I suggest you want to keep as clean as possible. Pianos have an enormous range and there are often multiple harmony lines and registers crossing over each other. Here saturation can really cause a mess, leading to a dull, muddy sound that’s fatiguing to listen to for long periods.
Saturation can turn cymbals into sand paper and ruin the tone and low end of a kick drum. Or on the positive side, saturation can make cymbals sound like silk and the kick drum deep and punchy. Used correctly, saturation can add sweetness, smoothness and what some people call “warmth” to the sound. It’s all a matter of using the right saturation in the right way. But it’s best to add it after recording, so it can be used selectively and applied to just certain frequencies. Think of mixing with saturation (or any other effect) as like cooking. You might add some garlic to the sauce right at the point where you’re frying the onions in oil, before you add the tomatoes, to get that perfect flavour. You wouldn’t want to add garlic to the celery before you chop it, to the pasta before you boil it, or to every other ingredient before you’ve even started cooking. For a great tasting meal, it’s important what flavouring you add, how much, and exactly when. This is why I suggest not adding saturation from hardware (like mic preamps and mixing consoles) during recording if you can avoid it. It’s very difficult to know how much and what type of saturation you’ll need (if any) before you start to mix. Adding saturation during recording is like deciding how much garlic to add before you’ve even decided on which meal you’re cooking.
Processing during recording
Back in the day engineers had no choice but to apply processing during recording – it wasn’t practical or sometimes even possible to do it all during the mix. Some recording engineers still like to use compressors and EQ on the signal as it gets recorded. With jazz there are problems with this. If you apply compression to an instrument as it’s recorded, you will never be able to undo it. You will be stuck with that sound. If all the sounds being recorded are world class and they all fit together to sound amazing, fine. But this is extremely difficult to get right during recording, and tends to call for the best engineers who specialise in this technique, in the finest studios. You are in effect partially mixing the track as you record it, and you have to know exactly how everything will sound together on every track. You have taken away the mix engineer’s ability to shape the sound later.
Rock is a different story: recording and mixing can be part of the creative process. In a rock recording, compressing or distorting the vocals during recording might actually inspire the band. A producer might add processing during recording because they have a vision for how they want the song to sound and they intend to influence the performance. This is not an approach often taken in jazz, and I advise against doing any processing while recording.
Guitars and synth keyboards are an exception as processing is sometimes an integral part of the sound and the player may have spent years honing processors to create their sound. But I would warn fellow guitarists off using stomp box or amp reverbs on your main guitar signal. Stomp box and guitar multi-effects reverbs are almost never in the same league as high end studio reverbs. Even if you are using a high end rack reverb for your guitar or keyboard, it doesn’t make sense to print it onto your main recorded signal. This can cause real problems in the mix. If you particularly like your reverb, record it on separate tracks so that the amount can be mixed later and so you have the option to use the high end reverbs in the mix studio.
Using vintage gear
“It’s vintage so it sounds great”. This is a myth and is responsible for many a spoiled recording. The truth is that some pieces of vintage gear sound good and some don’t. This was true when the gear was new. It’s more true decades later when many internal components may no longer be functioning well. If the studio you are recording or mixing in has some vintage gear, ask yourself: Do you know this piece of gear well? Do you know how it effects the sound in a variety of playing situations and musical contexts?
Certain famous pieces of vintage hardware, if they are expertly maintained, can sound wonderful on the right source in the right context. But old hardware tends to colour the sound pretty strongly. Do you really want that particular colour on every track? If you are processing an instrument with some vintage gear, do you know whether the sound will fit well with all the other elements in the mix? Unless you’ve worked with that particular piece of gear for years, it’s difficult to know.
I’ve seen people respond, in the spur of the moment, to an enthusiastic engineer’s suggestion that they use a “prized piece of vintage gear”. The results are sometimes regrettable. Suppose you use a vintage bass amp which happens to be in the studio. It might sound good when you plug in your bass and play on your own, but how is it with the whole band in the mix? Vintage bass amps often have very uneven responses, so certain notes will resonate much louder than others. This can cause absolute havoc in a mix, but it may not the sort of thing you’d notice over headphones or listening in the control room, in the heat of the moment.
Remember that vintage gear is used because it changes the sound, it adds saturation, it changes the frequency response, it may change the dynamics of what you play and it may affect the clarity. So the same applies here as to processing. If you record using it, there is no going back.
If on the other hand you record as cleanly as possible, you still have all the choices at mix time. Perhaps you’ll want a coloured sound on one tune but not on another or on the solo but not during the whole tune. Also keep in mind that most famous pieces of vintage gear have now been reproduced incredibly accurately in software form. Numerous grammy winning engineers, who own the prized hardware, say they can’t hear a difference, or that the software sounds even better than the hardware.
Vintage mixing desks are a case in point. Every channel on an old mixing desk sounds different. Some channels will sound better, some worse. Some will have more low end, or less clarity, or more saturation. And vintage hardware in a small studio is probably not well maintained, because to do would cost more than most small studios can afford. A software version, modelled on the finest example of a perfectly maintained piece of vintage gear, will often sound a lot better than the particular example of that hardware the recording studio happens to have. A well stocked mixing studio will have many software versions of all types of hardware processors. So if one doesn’t sound quite right, they can flip through several options in seconds to find the perfect sound. This approach is limited or impossible with hardware. My own feeling is that the days of hardware processing are over for most mixing purposes. Software sounds as good or better, is far more versatile, and far quicker to set up. Staying clear of colouring your recording with the vintage odds and ends the studio happens to have, means you are leaving the door open to better options at mix time.
I mix a lot of jazz records, but I also master many recordings which come already mixed. Most mixes will have areas where improvements can be made during mastering. These vary from adding the final polish to improving major frequency problems.
Any musician will have some idea what mixing is, but mastering is often less well understood. Various important things happen in mastering. The first thing we do is check the music, in a special acoustically designed environment, for any overall frequency issues that may not have shown up in the mixing room. Unless a room has professionally designed acoustic treatment, frequency issues in the mix are pretty likely. The mix engineer in a room without correct acoustic treatment hears a skewed version of the music. Some bass frequencies will seem louder than they actually are, others quieter. This will lead to incorrect mixing. The same holds true for all the other frequencies in the spectrum. This is the most common problem I hear when mastering. If it is minor, it can be fixed. If it’s more pronounced, it can be improved but never completely cured. So do look for a mixing studio which has professionally designed acoustic treatment. That’s not cheap, which is why so many mixing rooms go DIY, hanging a few small panels or putting foam on the walls. Although this might help the mid and high frequencies a little, it will do nothing to help the low frequency issues found in virtually all small or medium-sized rooms.
And finally… Reverb
The quality of the reverb is possibly the most important thing in a mixing studio. This is especially true for jazz. Some engineers like to use “classic” or “vintage” reverb units. They sounded great in 1987 or 1995 so they must sound great now, right? The truth is that reverb has come a long way in the past two decades, and even in the last ten years. Modern hardware and software reverbs are in a completely different league. They sound much more like real spaces than old reverbs do, and they make mixes sound less congested and more like you are playing in a great sounding space.
If you want that big, warm, detailed “audiophile” sound, use a modern world class reverb. If you are after an old time vintage sound, there is software which can reproduce the sounds of the classic EMI or EMT plate reverbs (used to record the classic jazz records of the 60’s and 70’s) that completely indistinguishable from the hardware.
Avoid digital reverbs or plugins which are not the latest state of the art units. Although high quality analog gear from the past can sound great if it’s well maintained, old digital reverbs are not the same. They work the same way modern hardware or software reverbs do. Digital reverbs digitise the sound using A to D converters and then they create the reverb sound algorithmically. But old reverbs had tiny amounts of computer memory, primitive computer chips and low grade digital converters. Today you have very high grade converters (or software that bypasses the need for converters all together), powerful processors and many gigabytes of ram. They give you much more realistic sounding spaces, smoother reverb tails and cleaner, clearer, reverb. This level of realism and detail simply wasn’t possible two decades ago.
In a jazz context, modern reverbs simply wipe the floor with the old ones. To my ears old reverbs can still sound great as a sustain effect. But they do sound more like an effect than a real space. The famed Lexicon 480L reverb was state of the art 30 years ago, and can sound nice as an effect, but it’s not up to today’s standards for creating realism and space. The same will be true of any older reverb unit, because the technology for creating real sounding spaces was simply not available 20 or 30 years ago. This was less of an issue when the old reverbs were in their heyday because most people were recording in larger studios which had their own natural reverb. Because the ambience in the larger studio rooms already sounded great, digital reverb could be used more as a tone enhancing or sustain effect rather than a realistic room sound. These days, recordings made in small deadened rooms need modern high quality reverbs to create a great sounding realistic space. I have heard many mixes spoiled by the use of old or cheap reverbs. Make sure the studio you mix in has world class modern reverbs. Examples are the Bricasti M7 or the Lexicon PCM 96, in either hardware or software form.
Mark Wingfield, Mixing and mastering engineer at Heron Island Studio
LINKS TO PREVIEWS SOUND REASONING ARTICLES: