I was recently reading Paul Nagle's review of the Waldorf XTk [see SOS March 2000] and found it very informative. It is for this reason that I thought I should contact him for advice.
I am looking for a really decent synth from which I can take the majority of my bass and lead sounds. The type of sounds I need are quite big, unique and organic, and would suit the ambient, driving, progressive trance style I'm into. I quite like the Novation Supernova at times but find the sounds a bit round and not sharp enough. It is for this reason that I am looking at maybe a Waldorf instrument, but I would really appreciate some advice on which you feel would be best for my needs.
SOS contributor Paul Nagle replies: Choosing a synth is always a difficult matter, and personal taste is more important than all other factors. Over the years I've found that some synths click with me and others do not (for example, I sold my Minimoog because I simply failed to find my own 'identity' when playing it). Some synths appeal at once, others only through time and patience. Auditioning in a shop doesn't always reveal everything, either. I fell in love with the Access Virus when I first played one and, with daily use, I like it more than ever. In contrast, I didn't like the Supernova so much after a short shop audition but — after spending time with one — I now rate it very highly. And again, I disliked the user interface of the Waldorf Q keyboard when I first tried it — and I still do (although it does sound amazing).
There are so many facets to each instrument that advice from others should always be tempered by your own experience. Waldorf synths are certainly "big and unique" but when you say "organic", I think instead of an analogue synth — but I'm guessing your budget doesn't stretch to an Alesis Andromeda? The Waldorf Pulse is a superb lead and bass machine if you can find one.
I personally find the Waldorf Microwave 2/XT series ideal for spiky textures, dirty, aggressive leads and ambient soundscapes. I wouldn't describe its general character as "organic", and I have never succeeded in making the warm, analogue‑style solo patches that I personally like. Perhaps it's something about the detune of the oscillators that always sounds a little harsh, but remember, this is a digital synth with its own strengths. It does have an impressive modulation matrix so that subtle, varying timbres can be created, but it takes time to understand and fully exploit.
I also think that, with programming, the Supernova series can sound very sharp and hard, and if multitimbrality is important to you, they take some beating. Oddly, one of my own favourite 'Virtual Analogue' synths is the Yamaha AN1x, which has impressive synthesis power, and you could pick one up cheaply second‑hand. Not the best user interface, perhaps, but it sounds awesome and is much underrated.
So I'm afraid there are no easy answers here. There are so many good choices out there, and you should be able to make almost any kind of music with almost any synthesizer.
Does anyone know how I go about creating a reverb from an algorithm and a microprocessor, or do you know where I could find out? I'm building a complex guitar effects pedalboard for university and I wanted to do away with the usual stamp‑and‑squint editing, so I am incorporating an embedded PC with TFT touchscreen. At the moment I am in the very early stages but I need to get a good grasp of the concepts involved. I would be grateful for any information.
Editor Paul White replies: Klark Teknik publish a book on digital reverb, called Audio Systems Designer, which explains the basis of the algorithms, so it would be worth calling them to get hold of a copy (it costs under £10). However, every company uses their own interpretation of this basic concept, as reverb algorithm design is as much art as science.
A typical reverb algorithm starts with a tapped delay line or FIR (Finite Impulse Response) filter, to produce the early reflections, which can number anything from dozens to hundreds, depending on the hardware power available. These are randomly spaced to simulate a real room response and tend to decay in level at longer delay times. Some taps feed the right channel and some the left, to create the impression of stereo.
The outputs from these taps are then fed into networks of recirculating comb and all‑pass filters, to build up the reverb density, and some EQ is included within these networks to produce the high‑frequency roll‑off associated with natural reverb. Duplicate networks are needed for the left and right channels. The outputs from these recirculating networks are then summed to produce an audio output. How these networks are arranged and how they are fed (from individual FIR taps or from summed groups of taps) is different for every reverb design, and once you have got a workable algorithm there's apparently much tweaking of the parameters needed, to get the decay time to be nominally even across the audio spectrum.
I have a SoundBlaster Live! card and read with anticipation your recent review of the new Audigy card [see November 2001 issue]. I was considering trading up when I first heard about it. I'm interested in your comments about its sound quality — your comparison with the Mia card was very revealing. It was obvious that you perceived a definite difference in quality. I'm definitely looking for a noticeable step up from my current Live! card, both for recording and playing back samples, but having read your article I'm now not so sure that the new Audigy is what I should go for.
I've amassed a good collection of Soundfonts, so I'm reluctant to abandon these for a new sample set, such as Gigasampler or Akai (although I perhaps would if it meant a significant step up in quality). I like having an onboard hardware sampler that gives me 64 voices, as the current Live! card does. I understand that the HALion software sampler can only send out over a maximum of 16 MIDI channels, and I have a possible 64. (I use Cubasis VST) I do think that the Audigy is huge value for money, as was my Live card when I bought it three years ago. My question is: do you think that the Audigy is a significant step up for me? Is it a genuine semi‑pro solution? I've come to the place where perhaps quality is preferred over feature sets.
The other options I've looked at include the Marian cards, or perhaps a Creamware Powersampler (by the way, I find the Creamware options and software/hardware combinations completely baffling!). The Mia would be great except for the lack of MIDI, which would mean shelling out for a USB device.
PC Specialist Martin Walker replies: You've obviously already thought through your options carefully, but it can still be tough making the final decision when it comes to choosing a soundcard. It certainly makes sense to carry on using the SoundFonts you've amassed over the last few years, especially as having a hardware SoundFont player gives you guaranteed 64‑voice polyphony.
Since you're primarily interested in improving audio quality, you could retain your SB Live! solely for SoundFont playback duties, and add a second soundcard (subject to a free expansion slot still being available) alongside this for recording and playing back audio tracks. I had no conflicts running the SB Live! and Audigy alongside my Echo and Yamaha soundcards when I reviewed them, so this is one way forward, and will give you noticeably better audio quality for recording and playing back audio tracks, although your SoundFont quality would be identical. If your new soundcard had a digital input, you could also send the fixed 48kHz digital out of the SB Live! into this to improve SoundFont audio quality, although this would force you to use a 48kHz sample rate throughout — not ideal for any musician whose final destination is a 44.1kHz CDR.
The Audigy would give you noticeably better audio quality for your SoundFonts if you want to replace your SB Live! altogether, but according to my measurements most soundcards designed from the ground up for musicians will give you lower background noise on your recordings, by between 5dB and 12dB, as well as providing true 24‑bit performance for both recording and playback. Of course, you won't get all the Audigy's DSP effects or its SB1394 ports, but if you're primarily looking to improve audio quality this trade‑off is something to bear in mind.
If your quest for better audio quality makes you decide on a different make and model of soundcard, there are now quite a few options for £200 or less, including models from Echo, Marian, M Audio, and Terratec. Don't be too worried at having to abandon your SoundFonts, since with Cubasis VST you could run a SoundFont player as a VST Instrument. There are various shareware models available at bargain prices, such as SamplerChan at $69 (www.samplerchan.com).
The only thing to bear in mind is that running 64 sampling voices in software will consume a significant amount of native processing power. I haven't tried any of these VST Instruments out personally, but you can download free demos to see how you get on, and this should give you a feel for how much of your CPU they will gobble up before you make a decision and part with any money.
If you decide to buy a mainstream commercial software sampler, like Creamware's Volkszampler ($99), splash out on the Creamware PowerSampler soundcard (bearing in mind its more limited 32‑voice polyphony), or a professional stand‑alone softsampler such as GigaStudio, you'll still be able to use your existing sample collection, since these all have SoundFont import facilities. HALion isn't currently an option for you, since it won't integrate with Cubasis VST, only with the more expensive Cubase VST 5.0 series, or Nuendo. However, it also imports SoundFonts, and you don't need to worry about a 16‑channel multitimbral limitation — if you need any more, you just open another HALion and you get another 16, and so on.
Finally, even if you decide on a different software sampler that doesn't directly import SoundFonts, you could convert them using a utility such as Amazing Sounds' CDxtract (www.cdxtract.com) or Chicken Systems' Translator (www.chickensys.com).
I'm a London DJ (currently of the bedroom/house‑party variety), and very much new to electronic music creation, although I've been writing music for years. Now I'm finally getting into it properly. I've been playing with a demo of Acid Pro 3.0 and when Kylie's new track came out I felt it lacked oomph in the drum department and was also a little bit too pedestrian in tempo. So I grabbed an MP3 of it, and an MP3 of another house track (just the beats from the first eight or so bars), imported them both into Acid, and proceeded to muck around, overlaying the beat over the top. It worked well, except that I found that every one or two bars of drum sample I had to resync it to the Kylie track by altering the offset value, whilst auditioning the whole thing to see if the beats were in sync. I've tried another drum loop against it just to be sure, but it appears that the Kylie track's tempo alters slightly throughout the entire track. Is this common practice? If so, why — to stop people trying to do what I just did? Or to gradually increase tempo in order to provide a build‑up which is subtle enough so the listener doesn't consciously notice, but subconsciously the buildup is there because the song is gradually getting faster?
Producer Steve Levine replies: I don't think it is generally the practice on 'dance music' to intentionally speed tracks up. (It's more common in other musical styles to program subtle shifts in tempo into a chorus, to try to add some emotion — bearing in mind that doing this means that 'flying in' vocals becomes very hard, as chorus one may not fit on chorus two.) What is more likely in this case is small timing‑resolution shifts.
I am not aware of what gear was used to generate the drum track in this case, but it is well known that many dance and R&B producers use the various models of Akai/Linn MPC drum/sampling workstations for their drum tracks, and all of these machines have an inherent 'feel' — small timing delays in the exact quantise position of the beats. In fact I have made a set of MPC grooves (from recording these delays) which I use in Logic, and it perfectly recreates the MPC feel/effect. Looking at the MPC grooves in the edit window of Logic, you can see that they are moved by 6‑18 ticks at Logic's high resolution when quantised, so aligning new beats to this groove would cause a flam or drift unless exactly the same groove was used.
Also, in the case of the Kylie track I don't know whether the final mix master has been speeded up using a timestretch program. This would almost certainly mean that the overall tempo would not then be consistent (due to the way these programmes work, by snipping out small amounts of waveform). The tempo variation would not be noticeable to the ear, but if you tried aligning something to it, those small errors would suddenly show up.
I have a graphite 466MHz Apple G4 for which I wanted some more RAM. I saw an ad in a magazine for a mail‑order store offering 256Mb for £34 but when I asked them to email me a quote it was £49 plus VAT. I rang back to ask why, and they told me it was best to use the more expensive Kensington RAM if I was doing any serious audio or video work (I do both), as the cheaper RAM was prone to system crashes in such cases.
I then rang another outlet to get a quote from them: £22 for 256Mb. I told them what the first store had told me, but they seemed to think that it was not true and the type of RAM wouldn't matter. Do you know the answer?
Assistant Editor Sam Inglis replies: PC music experts, including our own Martin Walker, recommend that you buy good‑quality branded RAM from manufacturers such as Kensington or Crucial, rather than cheap no‑name stuff. As the RAM used in PCs is identical to that used in modern Macs, I think this advice would hold across both platforms. It's quite likely that you won't have any problems with cheap RAM, but for the sake of £2O it isn't really worth risking it. If you do happen to get a faulty piece of RAM, it can produce very annoying, non‑repeatable crashes and glitches which can cost you a time and money in the studio, and be very difficult to track down. Having said that, £49 seems expensive even for branded RAM, and you should be able to find a better deal elsewhere.
I have a Boss DR660, which I have had for about four years, on which the velocity pads appear to be wearing out. They have become so sensitive that anything above the most feather‑light tickle produces a MIDI velocity of 127. I also have a Roland PC180 which, on certain keys, is exhibiting the same behaviour. What exactly happens when these things wear out out, and is the same fate waiting for my Nord Lead? Is this an expensive repair job, or do parts need replacing eventually?
Assistant Editor Tom Flint replies: Everything will wear out eventually, and I'm sure the Nord Lead is no exception. Pressure pads and mechanical moving parts like keys are prime candidates for going wrong, and you may also find that buttons which are used a lot will give out eventually too.
If you decide to get the keyboard and drum machine repaired professionally, it's either a case of sending them back to Roland for their service department to fix, or finding a specialist repair centre. If you take them to a Roland dealer or shop they will probably just send your gear back to Roland or to a specialist repair centre anyway, and will charge you accordingly. Bear in mind that any company repairing your synth will charge you for expenses, labour and VAT, which could easily add up to the second‑hand value of your equipment. A Boss DR660s can now be bought for about £120 (see the SOS Readers' Ads at www.soundonsound.com/adverts). I'm sure you can get a second‑hand PC180 very cheaply too, but Edirol currently sell a PC180A MIDI Keyboard Controller for around £89 including VAT, and that is bundled with a copy of Steinberg's Cubasis AV MIDI + Audio sequencer. At that price it might just be worth buying one new; buying second‑hand gear can take a lot of time and effort and there is always the chance that the thing you buy is wearing out too.
If it turns out that the repairs on your gear will be more expensive than buying second‑hand (or new) replacements, you might want to try fixing the equipment yourself. I'm sure your old gear won't have a warranty to invalidate, so it's certainly worth opening it up and taking a look inside. One possibility is that dust, dead skin, and general grime could be gumming up your keys and pads and something as simple as that could affect MIDI triggering. It is also possible that some moving parts are out of alignment or that a wire or two has worked loose and your problems may be fixable by pushing something back into its correct position. While you have your gear open, check the keys and pads to make sure they are not wearing underneath in any way. It is likely that you will be able to replace parts like these yourself, and a call to Roland (Roland Spares +44 (0)1792 515020) will soon enable you to find out if the bit you want is available. Roland UK's service department stocks an impressive range of spare parts.
If the keys and pads show no obvious signs of wear, then the trouble might be wearing circuit contacts which the pads and keys press on. Personally I would switch on the keyboard or drum machine while it is open and carefully test the contacts to see if I could identify any obviously fixable problems — but be warned: by doing so you risk blowing the circuitry or possibly electrocuting yourself!
I have a couple of completely unrelated queries for you. First, I am aware that certain CDRs will not play on certain hi‑fi systems. Obviously, commercially released CDs don't suffer from the same problem. Is this because the burning process for commercial releases is different, or because the discs themselves are different? If I send a master to a duplicator (such as the ones who advertise in your magazine), how can I be sure that the resulting CDs will play on any hi‑fi? Or is it possible to overcome the problem with my own duplication?
Second, I'm interested in getting into guitar synthesis, but it seems limited to dedicated units such as the Roland GR33. What I really want to be able to do is control any synth module or keyboard with my guitar. Can you suggest any devices that would help me do this? (By the way, I want to do it live, so software wouldn't be suitable.)
Assistant Editor Sam Inglis replies: For an explanation of why CDRs won't always play on hi‑fi systems, and of the manufacturing differences between CDRs and commercially released CDs, see Q&A SOS November 2000. As far as I know, there's no way to guarantee in advance that CDRs will work in any given audio playback system, but you will probably find that different brands of CDR, burned at different speeds, will make a difference. Duplicators should be able to advise you on the best choice, but if you want to guarantee that your CDs will work on any system you will probably have to get them glass‑mastered and pressed properly. If you want to get 500 or 1000 copies pressed, this is often the most efficient and cost‑effective way of doing it in any case.
To control MIDI syths with your guitar, you'll need to add a MIDI pickup and an interface. These convert your playing into MIDI note data, which can then be sent to the synth or sampler of your choice (provided it's MIDI‑compatible, obviously). Playing synths in this way needs some practice, and can require a change of technique, but can yield good results. Two MIDI pickups to look out for are Yamaha's G1D/G50 system (reviewed SOS December 1996) and Blue Chip's Axiom AX100 (reviewed May 1998).
I've converted my loft so that I can practice my sax. The walls, floor.and ceiling are all padded with rockwool, and the floor has good‑quality underlay plus carpet.The two doors are solid fire doors, but still the sound is travelling through the house (a three‑bedroom semi). The walls and ceiling are plasterboard, skimmed and painted. The room I'm useing is aproximately 26ft by 8ft by 8ft high. Is there anything else I can do to prevent annoying my very unhelpful neighbour? I've even considered building a room within a room!
Editor Paul White replies: Putting Rockwool on the walls may improve the acoustics but does little to help with soundproofing. The fact that you have plasterboard walls and ceilings is probably the biggest problem, but the only way to improve these is to add more mass. Providing the structure is strong enough, I'd be inclined to add a layer of lightweight fibreboard and then put the thickest plasterboard you can buy directly on top of that. Then you'd need to get the whole thing reskimmed, but it should cut down on sound loss through the walls quite significantly. You might even want to add an extra layer of plasterboard to the wall facing the neighbour.
If you think you're getting too much leakage down through the floor, consider making a simple floating floor by putting chipboard flooring on top of a layer of rockwool. You'll need to use two layers of chipboard screwed together, to get enough strength and rigidity, but it should be a lot better than just relying on underlay.
The other thing to be aware of is that your fire doors may still leak sound around the edges, so check that the seal touches the door all the way along all four sides when the door is closed. Paying attention to these details will undoubtedly improve the situation, but whether that is enough depends on how bad the situation is at present. Sound travels through houses in mysterious ways, so the practical results aren't always quite as calculated!
Is it necessary to record in stereo, when stereo reverbs, spatial panners, and such‑like could just be applied on monitoring and mixdown to add a stereo spread?
Editor Paul White replies: To reproduce true stereo, a signal must be recorded in stereo using two microphones in one of the recognised stereo configurations, such as coincident XY or as a spaced pair (SOS has covered stereo miking at various times over the years; you can check our web site for relevant articles). This stereo signal is recorded on to two tracks, which must be panned left and right in the final mix, using a stereo mixer channel or two mono mixer channels. However, most of what you hear on records is artificial stereo, where a mono sound source is passed through a mono mixer channel and then positioned between the left and right speakers using the pan pot.
The impression of space is usually added using reverb; it is common to feed the reverb from a mono signal, from which is derived a pseudo‑stereo output. The reverb output is then fed into a stereo mixer channel or aux return, where the two sides are panned hard left and right. Other artificial processing, such as panning the original sound to one side and adding echoes panned to the other, or stereo chorus or flanging, can also be used to create stereo effects.