There have been many millions of words penned about the technical limitations of the microgroove record over the years. Here are some more…
Doesn’t vinyl sound better than CD or digital?
Right, straight off and we’ve hit a problem. “Better” is almost a meaningless term when talking about audio. Everyone’s tastes are different, everyone’s hearing is different, everyone’s equipment is different. You can get into facts, figures, talk of dBs and frequency ranges and cm/s and bit-rates and ‘warmer’ and heaven knows what else, but sound is ultimately subjective.
What vinyl won’t do is give you an exact duplicate of a digital audio source. It often won’t give you an exact duplicate of an analogue audio source, either. Sticking a sharp point into a moving groove and getting sound out of it goes right back to Edison and his wax covered cylinders. The basic idea was refined and improved and changed massively during the following century… cylinders to shellac discs, big acoustic horns to microphones and amplifiers, shellac discs to polyvinyl chloride, 78 to 33 and 45rpm, mono to stereo, quadraphonic, half speed mastering etc. But you’re still sticking a sharp point in a moving groove, like Thomas did when he listened back to himself reciting Mary Had A Little Lamb for the first time in 1877.
So what does this all mean? Why will my music sound different?
The first thing is that you can’t get the same sort of high frequencies out of vinyl as you can with a digital source. You can cut those frequencies into the disc to some extent, but you can’t always get them back again. Obviously the higher the frequencies, the faster the cutting stylus has to move to create that pattern in the groove. Then when it comes to playing it back, the stylus on your record player has to try and follow that groove, which is a lot more difficult. When it can’t follow the groove correctly, you get nasty distortion. A lot of this is due to people’s playback equipment, cheap record players (and badly set up expensive ones) just can’t ‘track’ really high frequencies properly. You then might get into a world of protractors and anti-skating weights and hyper-elliptical styli and cork mats and snake-oil cables in search of the ‘perfect’ sound, which might never have been there in the first place. Just take it as read that a lot of people are going to be playing records on a horrid plastic thing that cost them £30 from a supermarket.
In fact, there’s some distortion everywhere in a vinyl groove. But it’s not always unpleasant… think of a square waveform from a synthesizer oscillator. It’s physically impossible to cut an actual square waveform into a groove, the cutting stylus just can’t move that quickly from side to side, and even if you could cut it you’d never be able to play it back. So the sharp edges of that waveform get ’rounded off’ to some extent. Some people might call that ‘harmonic distortion’. Some people might call it ‘warmer’…
So no matter what you do (unless you were cutting a pure sine waveform, which could be a bit dull to listen to) the audio gets changed in some way or other. There’s just no way of getting around it. We’ve been spoiled with digital sound somewhat over the recent past. You put in a square wave, that’s exactly what you get out. Now, you’ll get some going on about sample rates, 44.1khz isn’t enough, you get steps in your waveforms, it sounds harsh and all that. This is where personal opinions get in the way of any science, and it’s a bear pit I’m not jumping into. Suffice to say, records can sound really great, and they’re nice things to have. Heaven knows, I’ve got thousands of the things here… And you must fancy having your sounds on a record rather than a CD or digital file, or you wouldn’t be reading this. Right? Right…
So what are the other limitations?
Many and varied.
I’ll try and cover the basics in as short a time as possible, as I’m in danger of boring myself here. When cutting something to a vinyl disc, there are three underlying factors that you have to take into account:
- The length of the material
- The amount of bass present
- The overall volume of the disc
Here we have a classic case of ‘pick two’. The length of the material will determine how much of the physical space on a blank record that the groove takes up, no matter what speed it rotates at. Bass frequencies require the groove to move more from side to side than higher frequencies. To cut the material ‘louder’, everything is magnified. When you magnify everything, the room needed for those bass frequencies grows more than anything else, and that reduces the amount of time you can have that wide groove spacing going on before you run out of space on the disc at the middle.
So, imagine we’re trying to make a 7″ single. If you want it really loud, and it’s got lots of bass in it, then that limits the time the music can play for. If you want it to play longer and have lots of bass in it, then that limits the volume that the record can be cut at. If you want to get something quite long onto that 7″ and have it loud, then you need to cut the bass back.
Is that it? Bass vs volume vs running time, then?
Oh, dear me no. There’s lots more to it. Remember me going on about high frequencies before? Well, there’s more on those, for starters. But first, a cartoon… (apologies to Bill Watterson)
Yeah, I still lie awake occasionally like that. However it’s totally relevant. At the outside of a disc, the surface moves faster beneath the stylus than it does when you get nearer to the middle. What that means in practice is that there is less and less space to fit all the information in the closer you get to the middle. Think of it as bandwidth… When you try and pack loads of treble (especially) into that slower moving groove, distortion levels go through the roof because the cutting stylus gets to a point where it’s almost making right angles into the plastic, but all the frequencies suffer. If you’ve decided that you want that 7″ single we were talking about to rotate at 33rpm rather than 45, so you could get more onto it, the problem gets even worse…
Basically, the faster the disc rotates, and the further out from the middle the groove is, the better it’ll sound. So if you had a nice bright track which lasts about 6 minutes, it’d be better to cut it into a 10″ disc at 45rpm than try and squash it onto a 7″ at 33… It’ll sound so much better. If you’re not sure, ask. It’s why I’m here, apparently.
So how much can I get on a disc, then?
Well, as you can hopefully see, it varies greatly depending on the size of the disc, the speed it rotates at and the length/content of the audio. To play safe, here are some rough guides on maximum playing times, assuming the cut is to be at a decent (rather than very loud) volume and the low end is kept mostly intact.
- 7″ at 45rpm – 3 to 4 minutes
- 7″ at 33rpm – 4 to 6 minutes
- 10″ at 45rpm – 5 to 9 minutes
- 10″ at 33rpm – 9 to 12 minutes
As I say, this is only a really rough guide. Certain material might allow for longer playing times, other material might have to be less. For example, I can get 5 minutes onto a 7″ at 45rpm, but it’ll be quieter, you won’t have any floor shaking bass in it, and there won’t be a lot of room to scratch a witty message into the run-out groove area, which I’m sure you’ll agree is the most important thing… the world of records would be a much duller place without George ‘Porky’ Peckham’s little inscriptions, eh?
I’ve got records that last longer than that…
Undoubtedly. However, one major factor in that will be that many pro-cut and pressed records were made on variable pitch lathes. What that means is that the lathe had a system where it was fed a copy of the audio signal before the signal actually went to the cutting head/stylus. It could then adjust the distance between the grooves (yes, it’s one groove but you know what I mean) depending on the volume/bass content of the signal. So if there was a section of really quiet music, it could pack those grooves (yes) closer together, then when it sensed that things were kicking off in a revolution’s time, it could open things up again. My lathe doesn’t do that. Yet. So for the moment, I have to pick an appropriate pitch/spacing to accommodate the loud bits, and stick with it for the duration of the track/side.
But I need my records to sound really loud!
Do you really? Are you sure about that? I’ve mentioned ‘the loudness wars’ elsewhere on this site, but it’s worth perhaps having another rant about it here. This all goes back to the 1950s and 1960s, when the two main outlets for rock and roll were radio stations and jukeboxes. That’s when radio stations really started employing compressors and limiters to make their signal appear to be louder than the competition, so when Bobby and Angie were out cruising the strip in their dad’s Oldmobile (or in the UK, driving around the town centre in mum’s Morris 1000) and they twiddled the AM radio dial, they’d stop on the station that jumped out of the speakers.
Likewise when down at the cafe, pumping coins into the jukebox. The amp volume on those was always pre-set by the owner, so the punters couldn’t mess around with it. So, cutting a record ‘louder’ would mean that there was more chance of it being heard above the chatter, coffee machines, clinking of coke bottles and the sounds of fighting. This also meant that these records would absolutely blast out of the little portable record players that were commonplace at parties in the ’60s and ’70s, despite their amps only putting out a couple of watts at best.
Anyway, that’s continued to this day. Setting aside the squashed dynamic range of most digital recordings these days, there’s a desire to still have vinyl records cut at the loudest volume possible. Now there is the factor of signal to noise ratio to consider, i.e. you want the music to be heard above the surface noise that is an inherent part of the vinyl format (no matter how or where it’s cut). But once you get past that, you don’t need everything to be cut at +9db slamming into the red all the bloody time. Amplifiers have volume controls. DJ mixers have gain knobs.
The louder you cut a record, the more distorted it will sound overall (and the more strain it puts on my cutting head/stylus), the more problems there will be with sibilance (see below), the more compromises you have to make with frequency range, and the less time you have to play with. I’d much rather you had to turn the amp up a bit than make a record that didn’t sound good.
Okay, okay, calm down… What about all this stuff about bass having to be mono?
Indeed, and there’s a good reason for it. To get to the bottom of the matter, hopefully not getting too technical about it, and ignoring tons of history, a mono groove just moves from side to side (laterally), and thus does the playback stylus. When the first stereo discs were introduced in 1958, they’d adopted a system known as the Westrex 45/45… so called because there are two drive coils, mounted at 45 degrees to the disc.
In this system, the groove not only has lateral movement (side to side), but also vertical movement (it gets deeper and shallower). But it’s not that one plane carries the left channel, and one the right. The mono part of a stereo audio signal, i.e. the stuff that’s dead centre in the stereo picture, is cut into the lateral plane of the groove. The stuff that’s panned to one side or the other gets cut into the vertical plane. Mid/side for those who know. I’m starting to wish I’d never started typing this.
To cut things a bit shorter, when you’ve got low bass content panned over to one channel or the other, you end up with large modulations in the vertical plane. Effectively the groove becomes a very tricky rollercoaster for the playback stylus. If you then start adding stereo effects to that bass content, echos or phase-widening gizmos etc, the problem gets even worse and the playback stylus literally ends up being thrown off the rollercoaster and out of the groove. In fact, if I tried to cut that sort of thing, it could even end up with the cutting stylus trying to move in two directions at once, which isn’t good for anybody… least of all me, they’re bloody expensive things to replace.
Upshot: the low end absolutely has to be in mono, so it gets cut into the lateral plane of the groove. There isn’t a hard and fast crossover point, but for safety I’d recommend keeping everything under 170hz dead central. The more alert amongst you may have just made the connection between this and the need for more space on the disc when there’s lots of bass… it’s all connected. Circle of life etc.
Oh, and just to finish the section on bass, try and get rid of everything below about 35-40Hz. It takes up loads of space, muddies the sound and again you could easily end up with severe tracking problems.
So, bass in mono, keep the top end under control, stay away from the middle, no sudden movements, don’t walk on the cracks in the pavement… anything else?
You need to keep your audio nice and open. By which I mean, don’t shove it through a brickwall limiting plug-in thing in an attempt to make it ‘louder’. We all know about the loudness wars over the last 20 years or so, arguably started in earnest when digital tools like Waves’ L1 started turning up. Again, trying to translate hyper-squashed audio into the physical movement of a tiny cutting stylus just won’t go well. The results will sound awful – distorted, muddy, tiring – trust me.
That’s not to say you can’t use compression. Far from it. Where would rock and roll be without compression? I love it, when it’s used right. Also you need to compress things to a certain extent because of surface noise. Yeah, I know… “life has surface noise” as John Peel said. But it’s always there, and despite the fact that I consider my discs to have unobtrusive levels of noise*, it is yet another problem to be overcome. Vinyl doesn’t have the usable dynamic range that digital files / CDs do.
So don’t squash the life out of your music, leave some headroom there (-6db would be a good amount), make sure there’s no digital clipping going on. Make it pump and breathe by all means, make it loud, but don’t turn it into a hard-limited sausage. You won’t like the results, and neither will I.
* Again, subjective! But I’ve heard enough records in my life to know what I personally consider acceptable and not acceptable.
I’ve read that I should de-ess vocals?
You should. As much as you can. And not just vocals. Sibilance is a major cause of distortion on records, and it’s not always where you think it might be in the frequency spectrum i.e. way up at the top end. You can get problem frequencies much lower down, 6k to 8k, maybe even lower than that. It’s not just ‘sss’ and ‘ch’ vocal sounds either, as the likes of hi-hats and cymbals (both acoustic and from drum machines) can cause problems if they’re too bright. And tambourines. Oh, most certainly tambourines. Really listen out for that sort of thing when you’re mixing/mastering.
The last thing you want is me having to smother the whole final mix with de-essing and drastic EQ-cuts because it’s too bright to cut successfully. Those excessive mid-highs can also easily pop the fuses on the cutting head, which will nark me right off, I can tell you. I’ve only got so many spares.
Well, if you know everything, can you master my track for me?
No, I’m really not a mastering engineer… I’m just outlining the things which I know will cause problems at my end. I don’t even master my own records most of the time. I’d really recommend getting a true pro to do the job, preferably one who has actually properly mastered for vinyl before and knows the pitfalls. It needn’t be as expensive as you might think. The better the source you send me, the better I can make the results and the closer to what you’re already happy with.
I have another question…
Get in touch. I think I’ve typed enough above for the moment, but if I can think of any more important factors, I’ll put them up here in due course.