A shotgun mic is designed to ignore any sound arriving at the sides. This is how it focusses totally on the sound arriving from the front. It works great…
…unless side and front sounds are the same!
Read on to find out why…
are wide-open spaces. That seems so obvious right? But bear with me. Usually there are no reflective surfaces nearby. Nothing for the sound to bounce off. The main problem for sound outdoors is the ambient noise. On the other hand, indoor spaces are enclosed. You are surrounded by reflective surfaces such as walls, floors and ceilings. For us it is cosy, warm and sheltered :). But it can be a big problem for sound.
Let’s think about sound outdoors first.
You get in close with the mic using the boom arm. Great, we have nice clear dialogue. Now we will break the sound down into its component parts. The sound arriving at the sides of the mic (ambient) is completely different to the sound arriving at the front (dialogue). Put very simply, the shotgun mic has slots along the sides to break up the sound. Sound coming right in at the front goes straight down the tube. So it ends up being louder than any sound at the sides (known as”off-axis” sound). Incidentally, the longer the tube is, the better this works!
Now let’s think about indoors.
We have reflected sound made of the dialogue bouncing off all those surfaces. But indoors the sound arriving at the front of the mic is also dialogue! So it ends up canceling with (a delayed version of) itself. This creates what is known technically as phasing. What is that? It’s that metallic, harsh, whispery, un-natural quality that makes dialogue hard to take.
is to use a cardioid (or hyper or super cardioid) indoors. It will sound much more natural and engaging for your audience compared to a shotgun mic.
With lockdown I’ve been going back through the archive of old tapes to digitise them. Both stereo and four track tapes. The topic comes up often on the Sound On Sound forum, and I am grateful to Hugh Robjohns and James Perrett there for the advice and discussion over the years. Along with my own practice, I’ve whittled the process down to 12 steps.
1. You will need a decent cassette deck
Ideally a standalone hi-fi unit with a line output. These are usually on phono connectors. You can connect this to the mic/ line input of your computer. Even better, use an audio interface so you can optimise the level.
I use a Yamaha MT4X four track. This can play regular cassettes as well as the four track ones. It’s one of the best decks Yamaha made, with a very low noise floor and a frequency response all the way up to 18kHz, out-performing the Tascam or Fostex four track machines. The only machines that are better are the fairly rare Marantz four track recorders.
I would advise against the Walkman sized USB players you can get (pictured above). These will be wobbly and noisy with even the best tapes, so you will not get the best out of the tapes you have.
2. Any audio interface will do, even the one built in to the computer
The signal to noise ratio of the average cassette is around 50dB, far worse than any modern digital device. Best case for a recording would be 85dB with noise reduction engaged on a top flight deck (or the Yamaha above) and we’re assuming the cassette will never degrade over time. To put it simply: the quality of any modern digital device will be a lot better than most cassettes. The audio interface I use is a humble Focusrite Scarlett which is around 110dB! This is a very popular model for a reason: it’s reliable, clean and neutral sounding, and best of all Focusrite have very good long term driver support.
3. Change the shell
To reduce pitch wobbles, remove the reels from the shell and put them into something really good like a Maxell XL-IIS shell. There are plugins that can remove pitch wobbles (known as wow and flutter) such as Melodyne Capstan, and the proprietary Plangent Processes. These are expensive. This way is not- just make sure your hands are clean and grease free and work on a large surface with plenty of light.
4. Clean the heads
I think we all know about this step, but here are some tips. If the tapes are old and are ferric, you’ll need to do this between each tape. Wait about one minute for it to dry, and just clean ONLY the metal parts. Don’t use alcohol on the rubber parts of the transport because it will dissolve them eventually.
5. Adjust the head azimuth
We’ll need our ears for this! Listen to the playback with L and R summed to mono. Adjust for maximum treble. If the tape is really bad, boost the treble with an eq so you can hear what little high end there is. Super-tweaky tip: if it was a mono recording, do still record L and R, and combine them to mono to get slightly less tape hiss.
6. Don’t use noise reduction
Even if Dolby was used, consider leaving it off. Why? Here’s the simple explanation. You can actually hear the difference for yourself- playback the tape and switch the NR in and out.
The technical explanation: noise reduction is a good thing. But tapes lose their magnetism over time. This will cause the Dolby playback to operate incorrectly. The circuit will think it is hearing a quiet part of the tape, and apply too much noise reduction. Professional machines let you adjust the level going in to the Dolby circuit to compensate for this.
If that’s too technical for you, just leave the NR switched off and think of it as a free high frequency booster that you can polish with eq later. Most of the time, you will need as much high end as you can get from the tape!
For regular cassettes this idea works fine, but for most four track tapes it won’t work. These more commonly used dbx noise reduction not Dolby. Again, you can hear this for yourself if you switch the dbx in and out as the tape plays. It doesn’t just change the high frequencies, it changes the dynamics as well. It literally makes the music “suck”.
7. Contradicting step 6: do use noise reduction, but not on the tape deck!
This is a tip I got from James Perrett on the Sound On Sound Forum. An even better approach, is to, again, leave the NR off, but this time apply it with a plugin. The only plugin that can emulate Dolby A, B and dbx is U-he Satin. I’ve had great results with the Dolby.
The tone of the tape will be much nicer because you can adjust the level in the plugin, instead of trying to find a professional deck to do it on. Another big plus is if you denoise (which I will explain in step 8) BEFORE the Dolby plugin, it does a much better job of removing the noise than Dolby can, plus you get more treble back off the tape, with a smoother top end than no NR at all. In my experience this is especially noticeable with cymbals, and the esses in the vocals. I haven’t tried it with dbx.
8. Record all of the tape in one go
Record all of the tapein one go including the silent bits such as the blank leader.
I mean it. Play the tape, hit reecord in the DAW, and walk away for an hour.
It’s much easier to edit afterwards than sitting there listening to which parts of the tape you want. Make sure to keep the original file un-altered as a backup. I often find I need to go back to these later. Especially if better plugins come along that can improve the sound.
9. Noise removal
This is different from Noise Reduction. NR needs to be encoded in the recording, and decoded on playback (see step 6). Noise removal uses a computer to figure out which part of the signal is noise, and which part is actual audio. I use Izotope RX. If you use Reaper you’re in luck because you have Reafir: Kenny Gioa shows you how in this video.
If you can find a section of tape with no audio, just noise, the plugin can “learn” the noise profile, and then “intelligently” remove that from the audio. I usually do a profile for each song, if there’s a run in of no audio. This way, you can deal with noise generated by the gear used to make the recording as well as the tape itself. I also add the blank leader to the profile so Izotope can remove the noise the tape deck itself adds.
10. Use whatever eq you like, and as much of it as you need
Don’t be afraid to use several instances in a row because you will need severe eq especially to restore the top end. I also like to use a multi-band compressor, but it’s up to you what is needed to restore the sound to a useable quality, or enhance, or whatever you choose.
11. Often the top end is not there so you have to fake it.
For various reasons cassettes will not of the brightness and zing that we are accustomed to since the CD came along. Some simple processing can help with this. I find that Slate Revival is great for this. It’s a free and very easy to use aural exciter. Another good free processor is Voxengo GEQ it’s a combined graphic EQ and exciter with a bit more flexibility than the Slate offering.
12. Consider digitising the tape at 44.1kHz, 16 bit
This is not really a big deal but might be a consideration if you have a lot of tapes to digitise. 16 bit will save about 30% of space compared to 24 bit. The majority of cassettes had a dynamic range of around 48dB which equates to only 8 bits. Some combinations of deck, tape and very careful recording could get you a little more but it is rare. As for the sampling rate, only the very best tape decks can go to 18kHz although there are some rare machines that can even get to 20kHz so the standard 44.1kHz is more than enough in most cases.
But it’s probably more convenient to stick with 24 bit and if you are concerned about the quality of the filter in the ADC, a higher sampling rate might be preferable. Given the very poor specs of most cassettes though, I’d suggest it’s probably not worth arguing about!
I’ve personally found that an essential post processing tool is A-1 Stereo control. It will clean up unstable stereo images caused by slight wobbling of the tape as it passes across the playback head. The “Safe Bass” feature works by filtering the bass to mono, and you can adjust the overall stereo width as well if you like.
Here, for free is a properly looped mapped and tuned selection of most of the original Fairlight CMI III factory library. It also includes a version of the IIx library with a bunch more unknown user content, imported on the series III. This apparently is what anyone with a series III would have acquired over the years. Arguably, the II is the more interesting sounding machine because it changes the sound quality a lot. The Series III was always intended to be state of the art sound quality.
It is provided in the following formats:
There are WAVs associated with each format so download one of those if you’re not using one of the supported samplers. Bear in mind though, that there are a lot of files since the CMI III supported multi-sampling.
The Kontakt version is compressed in its own folder to avoid problems with moving the directory. This is the only version I have tested properly. The other formats probably need a little tweaking so if anyone wants to improve those feel free to share the files and I will update this folder.
We created this using Redmatica Keymap after extracting the WAVs from a CMI Hard Drive using CMIOS9. I further tweaked the Kontakt version after. So in many cases the loops and the tuning will be more accurate/ smoother than the original. In some cases where the voice is made up of a selection of dissimilar samples, the mapping might be a bit funky 🙂
My hope with this is to:
1. Help people make great music.
2. Bust some of the myths about the Fairlight. Now you can form your own opinion about these classic sounds.
3. I see a number of people trying to make money out of these sounds, selling them in Kontakt format on Ebay, or to Fairlight owners at inflated prices. This is morally wrong. Plus the sounds are not looped as they would be on the Fairlight.
4. There is some cork-sniffing in the Fairlight community, generating illogical arguments about having to own a Fairlight before you are entitled to use these sounds. Furthermore there is no valid copyright claim on this library (which is why so many people sell it on Ebay). It is provided here as free to use. Luckily I am friendly with some Fairlight owners who think this is silly, and my thanks to them for helping with these sounds.
I’ve started releasing an EP series called “Transitions” inspired by this time of year, when the birds prepare to migrate from wintery Ireland to sunnier countries. Each EP will have several short transitional pieces linking the main ones. I enjoy playing around with structure so this idea gives a nice excuse to indulge this 🙂
This track is called “Yay!” and came together very quickly. Around the time I finished it, a flock of swallows were gathering on our houses. Then the buzzards came along, for their breakfast. This was great because I’d been trying to photograph them all summer, out on the bike following them. They are quite timid and tend to fly off when you reach for the camera. So imagine my joy when I could just stand in the studio window and get great close ups! At one point she looked right at me. They also played around with each other, pouncing in mid-air. Spectacular. This video is made with my own photos.
My wife helped me with this one. The synth groove had been sitting around since 2012. The Pro Tools session fell out of the big pile of Pro Tools sessions of ideas, and she came up with a great orchestral line (violins, cellos and a touch of English horn) a piano part and the reversed bridging parts. The drums are from a Minnie Driver record. We tried to keep the mood of the synth groove while adding a cinematic build to the piece. We hope you like it 🙂
I played guitar on the other tracks, a Gibson Humming Bird. The final track “Sun Groove” came out of doing a cover version of Mike Oldfield’s version of Francisco Tárrega’s “Étude”. WIth the way Pro Tools works, I was able to keep all of the sounds of the original, and do a completely new piece of music. So this has some of the original old Fairlight percussion that Mike Oldfield used, taken directly from the original 8″ floppy disks. I also used a few Emulator II sounds in the same way. I love the atmosphere on this old sounds. The people that made them had to take great care in the recording to get them to sound so good with such limited sound quality. It’s nice to stand on the shoulders of giants!
I’m still using an Akai MX1000 76 key weighted master keyboard. This thing was made in 1991. I rescued it from a recycling centre in 2007. It has served me well, it’s even done the Body & Soul music festival! It’s got a comprehensive MIDI spec so I was able to set it up to run an entire show in Apple MainStage remotely- start, stop, patch changes and level controls. Its off-white colour (typical of Akai at that time) looked good with a white MacBook and white T-shirt. I’ve also helped out with a modern re-engineered memory card, so this keyboard would be good for several hundred patch changes with MainStage. Brilliant!
It is old though, so it’s time for some maintenance. Some contacts will have corroded, and some capacitors might be coming to the 30 year mark and so need to be replaced. So I intend to document maintenance for this machine, and share what official documents I have.
Problem: Aftertouch on my Akai MX1000 Midi Master Keyboard is not working
Possible cause 1: Aftertouch cable disconnected.
During transport, the flat plastic strip that connects the aftertouch pressure sensor to the internal printed circuit board may have shaken loose. Gently insert the strip back into the connector.
Possible Cause 2: Aftertouch cable worn.
When the cable has been jammed into the connector roughly a couple of times, the leads on the cable may have worn. You can try cleaning them with a qtip and some alcohol. If the leads are damaged, cut a few millimeters of the cable and reinsert it.
Problem: Aftertouch on my Akai MX1000 requires extreme pressure on the keys
Possible cause:Aftertouch pressure sensor strip corroded.
After some years, the leads “inside” the aftertouch pressure sensor strip will start corroding, forming a thin non-conductive layer that degrades aftertouch performance.
1. Open the Akai Remove the upper row of screws from the back of the Akai
MX1000. Remove the screws holding the top cover down. There
are three screws located on the right in a mirrored L formation and
five more on the left that are also in a (normal) L formation. You
may have to remove the two screws above the small rim
underneath the board too. You can now gently open the upper part
of the Akai MX1000 which will expose the internal circuitry and
keyboard springs etc. (I will add pictures of this procedure later)
2. Disconnect and remove the keyboard
Disconnect the flat cable on the mainboard. Rest the top cover
against something so it cant fully flip to the other side once we
remove the wire holding it. Remove the four screws holding the
small metal support in the middle of the Akai (that has a ground
wire on it holding the top cover). You need to remove this in order
to be able to take out the keyboard. Remove the large screws on
the bottom of the Akai that have rings around them. These hold
the keyboard itself in place inside the MX1000 casing. Remove
the remaining screws on the left lower side of the Akai that are
supporting the left side-panel of the MX1000. You can now gently
move this section (including the mod and bend wheels) a few
inches to the left, allowing you to lift up the keyboard from the
3. Remove the keys
Remove the springs. Gently insert a screwdriver in the back ring of
the spring and remove it Be careful. These little bastards will
easily hit you in the eye if you don’t pay attention. Now remove the
keys. Start with the white keys first, then do the black ones.The
first key on the left is the “E”, marked with a double “EE” sign to
signify the first key (last key is signed “GG”, all others have single
letters). You can remove the keys by slightly shifting them towards
you and then lifting the ends. If they get stuck, give the little white
plastic hooks inside a little push.
4. Store the stuff
Make sure you keep all the keys and springs together in the same
place. Parts are hard to find, so you don’t want to lose anything
5. Disassemble aftertouch strip
Once all the keys have been removed you will be able to gently
remove the upper layer of the aftertouch pressure sensor. This is
the white felt strip on the front part of the top. The top layer
consists of a felt strip glued to a plastic strip with some white
conductive material which is glued to the bottom layer with a sticky
Post-It like glue. You can easily remove this layer and press it
back on later.
6. Clean the pressure sensor
Once you have removed the top layer of the strip, you should be
able to see the two metal leads that acre causing the problems.
use a q-tip with some alcohol to remove the thin black film on top
of the grey/silver coloured leads. You may have to repeat this
process a couple of times until the strip stops colouring the QTips.
Don’t force it though. If you rub it too much you will damage the
strip beyond repair :-(. Gently rub the strip with some dry cotton
QTips to make sure everything is properly cleaned. Note the
difference between cleaned and corroded area in the picture. Wait
until everything is completely dry and free of any cleaning alcohol.
Now gently press the top layer back onto the pressure sensor.
Check the cable (see above) and reinsert the keyboard into the
Akai MX1000 case. Reconnect the flat cable, insert all the screws
into their original locations and close the cover.
7. Test it!
Select the System Menu and press Transmit. The display will now
show all outgoing MIDI data. If you press a key a little harder than
normal you should see the aftertouch messages scrolling by. You
can now assume your Mighty Marvel Pose #38 :-).
The Stanley Super 800 EP, edited, mixed and mastered by Tomás is on general release, distributed by RMG, so all good retailers should have it in stock now. Look out for the video on No Disco. It’s getting a lot of airplay, particularly the lead track “Moonlight”. Stan is topping the Wittness Rising Poll on Tom Dunne’s Pet Sounds radio show.