RocketTheme Joomla Templates
     
Home arrow Blog arrow tagsarrow Vinyl tech Stuff

Dublog

Whats happening in the Dub Studio brain?

Tag >> Vinyl tech Stuff

20 Apr, 2008

There is a common belief that when cutting vinyl dub plates and masters the lower frequencies must be in mono, but this is only partly true. Its all a matter of amplitude. Its possible to cut very wide stereo at any frequency, but there is a limit to the amplitude of the wave. Below is a close up photo of a 600Hz sine wave cut in mono at 45rpm, magnified x40 (fig.1).

600Hz sine wave - mono

Fig. 1 (600Hz sine wave - mono) 

 
The next photo is of the same frequency but this time one of the channels has been inverted so its in opposite polarity to the other, creating a sine wave with maximum stereo width. This means instead of moving from side to side, the cutting stylus moves up and down. In this case the vertical motion of the cutting stylus was not too great, and the groove never becomes too shallow.

600Hz sine wave - opposing phase

Fig. 2 (600Hz sine wave - opposing polarity) 
 
 
 
In the next photo, a wave of similar loudness wave cut, but this time the frequency has been halved to 300Hz. In this photo one can clearly see that groove at its narrowest point, is much narrower than the 600Hz groove, and is at the limit of what a stylus can reliably track.

300Hz sine wave - opposing polarity

Fig. 3 (300Hz sine wave - opposing polarity)
 

But why should this be the case? Why should a lower frequency need to have a larger amplitude? Its because the loudness is related to the linear velocity of the wave on the record. A lower frequency means a lower linear velocity, so the amplitude must be greater to make up the difference. Greater amplitude means greater groove excursion, and so for example cutting the same loudness at 150Hz would mean even greater amplitudes, and therefore even bigger waves in and out of the surface of the disc. 

For this reason I use a 6dB per octave roll-off from 400Hz downwards on the stereo width of the recordings I make. This ensures that as the frequency gets lower, and the waves need greater amplitude, the stereo width gets narrower and the chance of the vertical modulations being too large to cut is reduced in accordance (see fig. 4).

 


Fig. 4 (Sonnox Filter used in M/S matrix on Side channel)


For more details on how to achieve this yourself before submitting material to the studio please see the blog entry on Stereo Processing.

 

 


20 Apr, 2008

Its common(ish) knowledge that large stereo differences between left and right channels can cause problems when cutting and playing vinyl (see: Bass - How Low can You Go?). There are varying opinions as to where to start 'monoing' the lower frequency, but the truth is its a compromise between what sounds good, and what's gonna stop the vinyl skipping, so its gonna vary depending on the track. Personally, as a rule of thumb, I use a 6dB per octave filter from 400hz downwards. This I find to be more pleasing to the ear, than say a steeper curve, and a lower roll off frequency.

But what do I mean by a 6dB per octave filter? Basically this means that I set up an M/S matrix, and in order to center the signal from 400Hz downwards, I apply a filter to the Side channel within the matrix, that cuts the Side by 6dB for every octave lower than 400Hz (fig.1).

 

 
Fig. 1 (Sonnox Filter)

 

In fact, once you have set up an M/S matrix, you can perform any sort of processing you like to the stereo width. You could for example use a limiter to ensure overall stereo difference between the track never goes over a certain level, or use an eq to widen the stereo difference at a certain frequency, and narrow the difference at another. All this can be done by simply applying processing to the Side channel.

Setting up a M/S matrix can be a rather complicated affair using a mixer, but it can be done. Personally, I prefer to use a simple plugin to perform the complicated wiring needed. There are several on the market, most noticably the S1 included in most Waves bundles, and more recently the excellent Brainworx series of software and hardware plugins. I use the S1 from Waves in the studio at the moment, but I am going to upgrade to the bx_control from Brainworx soon (see fig. 4). Its a better plugin, and Waves refused to let me transfer my S1 license to the laptop for when I am working on location, so I can't use it to demo at seminars. Plus the bx_control is cheaper.

To set up a matrix in Pro Tools, for example, you just need a stereo aux and two instances of the S1 matrix, or the bx_control. I usually place one instance of the S1 in the first space in the channel strip, and one in the last space. That leaves me with 3 available spaces to place say an EQ, some sort of dynamics processor, and perhaps a limiter (see fig. 2). Then if you want to use the matrix you just send the signal through that aux. As long as the plugins you use are in multi-mono you can simply unlink them and change either the Middle or the Side by clicking on the L or R button respectively.

 

  

 Fig. 2 and 3 

If the plugins you are using don't support multi-mono (they are either mono or stereo) then you can use a similar setup but using a series of auxes. First setup a stereo audio track, then 2 x mono auxes and one stereo aux. Send the output of the audio track to stereo bus 1-2, then the inputs of the first two mono auxes to mono bus 1 and 2 respectively. Send the outputs of the 2 mono auxes to mono buses 3 and 4 respectively, and set the input of the stereo aux to 3-4. Then put an instance of the M/S plugin on the audio track, and the stereo aux. You can then process the Mid and Side channels by placing plugins in either the first or second mono aux (see fig.3).

A similar set up can be used for most DAWs, but one that has a comprehensive suit of settings already built in is Samplitude. The recipe is as follows:

You will need:

One stereo track

One Submix Bus

Two Aux Busses

1) On the stereo track, set output to Bus 1, send to aux 1 at unity gain, send to aux 2 at unity gain. Load your track.

2) Mute Bus 1 (this bus may be used to compare the original mix)

3) On Aux 1, right click the pan pot, select preset "Mono (get mid signal from stereo source)". Rename this track "middle".

4) On Aux 2, right click the pan pot, select preset "Side signal (Stereo) (from stereo source). Rename this track "side".

You may now EQ the Middle and the Side independently of each other by using the channel strip of the corresponding aux bus. Or you can insert plugins on each of them. If you want to listen to either the Middle or Side, solo the corresponding Aux Bus.

bx_control 

Fig. 4 

A word on plugins... 

Just to show I am not biased towards the Brainworx plugin because of Waves licensing, I am gonna give a brief explanation of why I think its a better plugin.

a) the Waves S1 Matrix plugin is just a simple network that changes the left and right channels to Mid and Side. So L+R=M (left channel) and L-R=S (right channel).

b) the Waves S1 Matrix has no other features, all you get is a grey box, no knobs or meters. There are some extended features included in the S1 Stereo Imager but they are still somewhat limited.

c) the Waves S1 Matrix has the unfortunate side effect of reducing the overall signal by 6dB (this is because to give L+R without digital over the signal must be halved, and even then some samples do slip through).

d) the Brainworx plugin is cheaper

e) it has a wealth of features, but crucially it accepts L and R or M/S. This means you can set the first instance to accept L and R and it gives out M/S, then you can set the second instance to accept M/S and it gives out L and R. I have tried running a signal through the bx_control as shown above and then mixed it with the clean inverted signal, and the result - total silence, just as I hoped. Its impossible to achieve this with the Waves S1 Matrix as you have to gain the clean inverted signal down by 6dB so it is matched in level with the processed signal, but because -6dB is only approximately half the level, you still hear the signal come through, and no amount of adjusting can get rid of it.

f) the bx_control has meters showing not only the RMS and Peak of both the M and S channels but also balance and correlation meters.

g) the bx_control has numerous kill switches to solo L, R, M and S as well as a balance control knob, a mono maker knob (presumably this makes everything below the selected frequency mono?), and a stereo width knob. Not to mention an adjustable input gain. This means it can either be used to control various parameters of the stereo width, or alongside other plugins as well.

I guess I might be a little unfair on Waves here, because I know the S1 Stereo Imager can perform some of the tasks of the bx-control. But as a rule i would use the S1 matrix as a simple encoder/decoder and the bx_control for the full works. The Stereo Imager just wouldn't come into it. 


20 Apr, 2008

I have been getting a lot of questions from clients recently about file types, sample rates, bit rates and such like, so I thought I would clarify a few things here.

There are really no limitations about what you can cut onto a dub. As long as it makes a sound, it should record. Generally speaking, the higher the 'resolution' the better. That's to say, when it comes to sample rates, bit depths and data transfer rates, the higher the better. There are a few things worth pointing out though:

Sample Rate - vinyl is capable of reproducing extremely high frequencies. One need only look at the frequencies recorded onto quadradiscs to see how far the bandwidth can extend, so there is no need to limit the sample rate to 44.1 kHz unless you want to. I prefer to work at 88.2 kHz and I upsample all material to this sample rate before working on it. I don't downsample again before cutting because there is no need.

Bit Depth - even the finest audiophile vinyl is generally recognised to have less dynamic range than CD audio. Therefore it makes little sense to me to submit tracks to the studio that are deeper than 16 bit. I have been sent 32 bit tracks before, and they needlessly took ages to transfer over the internet and loads of hard drive space to store. I work in Pro Tools, and the highest bit depth I can currently run is 24 bit.

Bit Rate - a lot of the material I get sent is in mp3 format, and I cut from mp3s all the time. A lot of the time this is because the client only has mp3s, but sometimes its because they have purposely compressed the files to send across the net quicker. Generally I would advise against data compression, unless absolutely necessary. However, if all you have is an mp3, there is nothing to be gained from converting it to a wav file, or burning an audio CD - once the compression has taken the data out of the file, its gone forever. Having said that, plenty of clients have dubs cut from MP3 with pleasing results, especially those with a data rate of 320kbps.


20 Apr, 2008

No matter how much I test dubs before I send them out, one thing that seems impossible to avoid completely is the dreaded skip. There can be a whole load of reasons why a dub might skip, so I thought I would do an entry here to explain a few of them.

Grooves and Land - the space between the grooves on a record is called land, and if there is not enough of it, the grooves get too close, and can sometimes cross over each other, causing the dub to loop continuously. If this is the case, the only solution is to recut the dub. 

Too Loud - the louder the cut, the greater the chance of it skipping, and part of getting a dubplate nice and loud is understanding the limit to which you can push the recording so that it plays on most turntables. If the plate has been cut too loud, it may play on some turntables and not others, or worst of all it will not play at all. If it skips sometimes and others not, again its best to get a recut.

Stereo Width -  if the track has a lot of stereo width in the bass end, it can cause the tracking of the stylus to become less reliable. Although vinyl with a lot of stereo width may play fine in a controlled environment at home, the chances of it skipping in a club are much higher, especially if its been cut quite loud. I 'mono' the bottom end from about 400Hz downwards using a 6dB per octave filter, which is usually more than sufficient, but its always best to give the dub a thorough testing before playing it out.

Damage - vinyl dubs can get damaged just like any other record. If the dub is handled too roughly, is subjected to heat, or has scratches on the surface, it may stop tracking properly. If you discover a problem within the first week of the dub being cut, and it has suffered none of the above, its fair to ask for a recut.

THINGS TO DO BEFORE CALLING FOR A RECUT:

Check the dub on a couple of decks just to make sure its not a problem with the turntable. Each time you play the dub, make sure you have the turntable set up properly.  Three things which I always check are the stylus condition, the tonearm weight and the anti-skating.

Make sure the stylus is mounted inside the cartridge properly, and that its free from fluff, dust and grime. Dedicated brushes can be used to clean off any build-up, and its usually a good idea to brush in the same direction as the grooves, away from the cartridge (never brush sideways, or towards the cartridge).

The weight of the tonearm must be adjusted in order to reduce distortion and skipping. To find out the correct weight for your cartridge, see the manufacturers guidelines. For example the 'tracking force' for the Shure M35X is stated to be between 1.5 and 3 grams. This means that the stylus should weigh down on the record somewhere between these two weights. To set the weight, first make sure the counter weight is facing the right way (I am amazed how many times I have seen it the wrong way round in clubs). On a Technics deck, the numbers should be facing you. Then float the stylus above the record by turning the weight so it moves away from you until you reach an equilibrium. Now hold the weight still, and turn just the black band until you see the zero line up with the black line at the top. If the stylus is still floating just above the record and the counter weight says zero, this means the stylus, cartridge and counterbalance have been calibrated correctly. Now you can adjust the weight to the manufacturers spec. by turning the whole barrel until the correct weight lines up with the black line.

The anti-skating lever found on most DJ turntables can make the difference between skipping and not skipping, particularly on very loud dubs. Try turning it up to maximum to see if that helps.

If all else fails, it probably means there is a problem with the dub itself in which case I am happy to do a recut. 

 


Gift Vouchers

Want to buy someone a dubplate but not sure what music they want on it? Then a DubStudio dubplate voucher could be the perfect gift.

Dub Studio Dubplate voucher
Click here for more info.





Lost Password?
No account yet? Register