Choosing a Virtual Pipe Organ
"Never underestimate the joy people derive from hearing something they already know"
Posted: 28 May 2014
4 May 2017
This article addresses the problem of deciding which virtual pipe organ (VPO) to go for out of those currently available. It describes what a VPO is, and then deals with a wide range of technical and playing issues which affect the choice. A number of VPOs available at the time of writing (2014) are briefly discussed. The article contains links to many others on this website which augment the detail associated with the points raised.
(click on the headings below to jump to the desired section)
Over many years I have published various articles in public domain print journals and elsewhere, including this website, on the subject of choosing which type of conventional electronic organ to buy (see for example those referenced at ,  and  at the end of this piece). More recently I have also received many requests from those seeking similar advice as to which virtual pipe organ (VPO) to go for out of the several which are currently available - although they have similarities, VPOs are not the same as conventional digital organs as this article will demonstrate. Unfortunately I never feel able to answer the question posed in this form, as it is rather like having the conceit to tell people what car they should drive. With cars the answer is simple - try before you buy, advice which applies equally well to digital organs of the traditional sort from an established manufacturer or franchised dealership. But with VPOs things are not quite so straightforward because you cannot just walk into a dealer's showroom on the high street and ask for a demo. There are no dealerships or franchises to speak of on the same scale, because VPOs still account for a relatively small part of the overall digital organ scene. Therefore this article tries to fill the gap by suggesting what people might look for when deciding which one to choose, while leaving the final choice up to them. Although some of the material will be familiar to many readers, I thought it would be useful to bring all of it together in one place because this does not seem to have been done before, and my inbox clearly demonstrates a need for it. So do some of the letters in organ magazines. Another novel feature of the article is the extensive amount of detailed information available elsewhere on this website, which can be accessed immediately by following the many links referred to in what follows.
However, the article does not go deeply into technical issues specific to particular VPOs because these must be sought from the supplier and/or the user community of those which attract your interest. In any case, they change all the time. It also tries to avoid negative criticism or excessive enthusiasm about one product or another because we are speaking of musical instruments here, thus the ultimate choice depends on subjective factors as much as anything else, and everyone's requirements are quite rightly unique to them.
The material here is believed to be up to date at the time of writing (May 2014). However the scene is constantly changing and for practical reasons I cannot undertake to revise the article frequently and in every particular. Nevertheless it is hoped that readers will be able to take some general guidance away with them having ploughed through it. Nor is the article an advertising platform for VPOs, so further details of all of them must be obtained from their providers. But first we need to define what a VPO actually is, because the overlap with the traditional digital organ business is of course large though somewhat ill-defined.
Before answering this question, let us look briefly at conventional digital organs. These instruments are made by many manufacturers in many countries, some of which (such as Allen) are large enterprises which have been well-established for many decades. They are complete and self-contained in the one console (apart from external loudspeakers when offered), and they will work as soon as they are plugged in and switched on. Little setting-up on the part of the purchaser is required beyond that done by the manufacturer when he installs it. Two types of digital music technology are used in modern instruments - sampled sound and physical modelling synthesis (the latter only by Viscount which is the only firm offering both types). These technologies were compared and contrasted in reference .
VPOs have arisen more recently, since about the turn of the millennium to be more precise, which coincided with the appearance of sufficiently powerful personal computers for the first time. Although these look pretty laughable nowadays, running as they did under ancient operating systems such as Windows 98, they nevertheless provided a sufficiently capable technical platform for a different breed of digital organ to emerge - the VPO. In the early days of their evolution I wrote an article which can be consulted for more detail . A few VPOs are huge and elaborate custom built products, one of the earliest being Marshall and Ogletree's celebrated opus 1 at Trinity Church, Wall Street, New York City. This was intended to be a temporary replacement for the Aeolian-Skinner pipe organ there which had been destroyed in the cataclysm of 9/11, though now it seems to have become a permanent fixture. More recently, this year (2014) the same firm unveiled a five manual touring organ for Cameron Carpenter. More information about these and similar products can be obtained easily from the internet. However these spectacular commercial instruments are not the subject of this article, which is concerned with the much smaller and cheaper ones populating the opposite end of the market, and in particular those aimed mainly at the home user (though some have been installed in larger venues).
VPOs, although using much the same type of digital technology as conventional digital organs (but restricted to sampled sound synthesis only at the time of writing), are different from them in several respects. Firstly, the 'sound engine' in a VPO is an ordinary personal computer (PC) rather than special-purpose hardware and software built for the job, though the adjective 'ordinary' might not apply in every particular to some of them which require a custom-built computer having a very high performance. Thus one will always need a dedicated PC of some sort to run any VPO. I say 'dedicated' deliberately because it is inadvisable to use the general purpose workhorse PC already sitting on your desk at home. Although it might be possible to get some VPOs going in this way, multiple problems can arise especially if the computer is connected to the internet, running antivirus software or groaning under the burden of the other application programs most of us load into our PCs. A VPO needs to be able to call on the full resources of a computer, and it is most annoying to find that it suddenly stops working when Windows decides to update itself from Microsoft's website, your antivirus software starts a scan, or whatever.
The second difference between ordinary digital organs and VPOs arises from the first. The use of a modern PC with its blistering speed and enormous memory allows each note of each stop on the simulated organ to be separately voiced and regulated. Moreover, each sampled pipe sound can be several seconds in duration, enabling the 'liveness' of real pipes to be reproduced. This contrasts with many conventional digital organs which do not offer these advantages, though others claim to do so. In fact it was this capability which first gave rise to the name 'virtual pipe organ', because it enabled every pipe on a pipe organ to be simulated separately and independently from all the others.
A third difference, already mentioned, is that you cannot just go into a retail showroom and try a range of VPOs. Not only do the showrooms not exist on a large scale because of the small size of the market, but each VPO needs to be built up for (or by) each customer from various components and sub-systems, of which the PC is but one. For instance, assuming you want to play the VPO yourself rather than just fire MIDI music files into it, some sort of console needs to be interfaced to the computer . You will also need a more or less elaborate audio system consisting of multiple power amplifiers and loudspeakers, unless you are content to use headphones. Then you need to decide how many simulated organs you envisage playing, because a separate set of pre-recorded sound samples (each containing, desirably, one sample per pipe) will be required for each one. The choice here is vast for some VPOs, though a few can only accept a single sample set. Thus you need to confront how to assemble the complete system early on - some VPOs can be bought ready-made from a range of third party suppliers whereas others are definitely of the DIY variety with little support available beyond their particular user communities on internet discussion forums.
This system configuration and assembly issue is important from another point of view - overall cost. It is often bandied about that VPOs are inexpensive compared to a conventional digital organ, and when considering only the cost of their individual components this may be so. However if you are thinking of purchasing a complete turnkey product assembled by a small VPO supplier (of which there are many), then the total outlay could be comparable with, or it might might exceed, that of a conventional digital organ in the low to medium price bracket within the product range of most manufacturers. So you should keep this fact in mind. VPO consoles can be particularly expensive if you opt for high quality keyboards rather than cheap plastic ones, for example. (This is not a criticism, as it merely reflects the truism that good consoles for any type of organ are not cheap!). Some sample sets are also expensive, so if you intend to run several you must be prepared to pay for the privilege
A fourth difference between conventional digital organs and VPOs also relates to their sample sets. If you will be using more than one, the issue of how to control the stops arises somewhat forcefully. Labelling them is but one difficulty, and it has resulted in many VPO users opting for touch screens rather than traditional tactile stop controls such as tabs or draw knobs - which are also more expensive. With touch screens, the stops and their names are written by the computer onto the display(s) depending on the sample set selected. Only you can decide whether this is acceptable, though it is unfortunate that the needs of the visually handicapped are not well addressed by touch screens.
Continuing the theme of sample sets, the one chosen will either have been recorded 'wet' (to include the reverberation of the auditorium when the notes are released) or 'dry' (excluding it). Opinions are strongly expressed on this matter - some find the somewhat 'distant' effect of wet samples attractive whereas others do not. There is also a range of detailed technical issues to take into account, and I have discussed the wet versus dry issue extensively in another article on this site . Not all VPOs support wet samples, though all can reproduce dry ones. With these, reverberation is usually added to taste using effects processors implemented in either hardware or software.
A fifth difference might seem trivial at first sight but it is worth mentioning. Whilst a conventional digital organ is ready to play the moment it is turned on, a VPO cannot be played during an interval whose duration can become increasingly irritating once the initial excitement has worn off. Not only do you have to wait for the PC itself to boot up but also for the chosen sample set to load, and this can take several minutes in some cases. I get annoyed by this myself sometimes - real, as opposed to virtual, pipe organs are not actually like this! I have come across more than one person who no longer uses their VPO for this reason alone. Thus the problem is without doubt a disincentive to even switch the thing on for these disillusioned users.
Somewhat less trivial is a sixth difference - noise. Many computers are noisy because of their fans and hard disc drives, and this also can be irritating in a room where music is to be played. The smaller the room, the worse the problem. Like the boot-time issue, noise can seem to become more intrusive and annoying with time. However solutions to the problem are not beyond the wit of man and completely quiet computers can be obtained, but they tend to be expensive simply because they are custom built and because they use exotic components such as solid state disc drives and fanless cooling systems.
It is appropriate that we now look at the various components of a VPO touched on above when they are connected together, because a picture is worth many words. So a block diagram of a typical VPO set-up is shown in Figure 1, using pictures of some actual items which I use myself.
Figure 1. Block diagram of a typical VPO set-up
The console or keydesk sends signals, usually in the form of MIDI messages, to the computer. Each message corresponds to something you do at the console - none are sent otherwise. Thus there are the Note On or Note Off messages initiated by keystrokes for example. Other types of message are sent when stops, pistons, swell pedals, etc are operated. MIDI is a complex subject which you need to know about if you want to assemble a VPO yourself, and some articles elsewhere on this site might be helpful if you are new to it , . Unless you are an expert and can make it yourself, you will therefore need to buy some MIDI encoder hardware which is connected to all the electrical contacts in the console and which sends the appropriate messages to the PC when they change state. There are several sources of such hardware on the internet, but if you decide to leave this to a third party turnkey supplier they will do it all for you (at a price of course!).
The computer runs one or more vital pieces of software to convert the MIDI messages it receives into organ sounds - these are the synthesisers. Some synthesisers are in fact hardware-based, using the extremely complex chips on high quality sound cards for example. In this case, all the synth software has to do is to decide which synth module to activate for a particular MIDI message. Other VPOs use a technique called software synthesis in which the sounds are generated entirely by the VPO program itself (or a separate piece of synthesiser software running in the same PC), and in this case the software is much more complex. It also needs a computer with a much higher specification than the hardware-based method. In either case the computer also contains the entire sample set for the organ currently being simulated in its memory, which is why computers with lots of memory are needed for VPOs. In the diagram only the PC itself is pictured, and some VPOs can indeed run 'headless' in this manner. However others require a monitor, mouse and perhaps even a keyboard as well. Although some VPOs can run on a laptop, a desktop PC is more convenient in some ways because of the extra space available within the case which enables as many add-on goodies as necessary to be installed. These include sound cards, memory and disc drives to suit the VPO you are building. However a potential medium-term problem is looming here because the day of the traditional PC, either laptop or desktop, seems to be drawing to a close. Some firms no longer make them, the choice is dwindling and prices are no longer coming down at the rate they did a few years ago. Whether this might spell the end for the VPO after its short but exciting lifetime is an open question, unless the processing power of phones, tablets, etc can be harnessed to do the job instead. Therefore it will be interesting to see what happens. In the meantime, you should therefore not delay too long before purchasing a suitable PC before it is too late.
The audio signals generated by the computer then have to be amplified and fed to loudspeakers. As a general rule, the more audio channels the better for reasons which I described in detail in reference  (select the section entitled "Technical Matters"). Therefore you will need a separate power amplifier and loudspeaker system in each channel. However, be aware that some VPOs only support two channels. You may also decide to use additional signal conditioning in each audio channel, such as a graphic equaliser (often very useful for a VPO) and a digital reverberation unit.
So you can see that a VPO of the highest quality is not going to be cheap. At this top end of the market by far the majority of the cost is swallowed up by the console and the multi-channel audio system, particularly the loudspeakers. Of course, significant economies can be made by using console components of lower cost such as commercial plastic keyboards and touch screens, and by minimising the number of audio channels or even using headphones. It all depends what sort of playing experience you want, which is directly related to the depth of your pocket. But please don't go away with the belief that VPOs are always inexpensive, as some people claim! As a rule of thumb, their price will not be too dissimilar from that of a commercial digital organ using a console and audio system of comparable quality if you buy the VPO ready-assembled from a third party supplier. However most VPOs will then offer you the additional advantage of several simulated organs rather than being restricted to just one. But if you just want the experience of playing an organ at home, and getting very good musical and audio results from it, there is no doubt a VPO is the cheapest way to go if you are not too bothered about the quality of the console and other components. This is particularly true if you put the system together yourself.
There are many criteria by which to judge a VPO, but two of the most important from a player's point of view are polyphony and latency.
The polyphony figure is the number of simulated pipes which can speak simultaneously. In an ideal world this ought to equal the number of pipes on the organ being simulated, but other than for the smallest instruments this is impractical, and nor is it necessary. Polyphony is limited by the finite number of sound-generating (synthesiser) circuits in a hardware synth, or by the power (largely speed) of the computer hosting a soft synth, and for most instruments its numerical value lies far below the number of simulated pipes. Most of the time this does not matter provided there is enough polyphony to cope with ordinary musical demands because even on a pipe organ we do not play on all of the pipes at once. Thus we can arrive at typical polyphony requirements quite easily by considering the maximum number of notes which can reasonably be played simultaneously. Consider two coupled manuals with the pedals coupled to both of them, and let each hand play a chord of four notes and each foot one. This represents pretty 'heavy' music which would seldom be exceeded in terms of the number of notes, unless you routinely play Ligeti's Volumina. Thus there is a total of 10 notes physically keyed, a figure which has to be doubled on account of the couplers, giving 20 notes. If 6 speaking stops are in use on each manual, this means a polyphony figure of 120 is required (6 times 10 times 2) because the VPO will need to process 120 waveform samples at once. Since polyphony is often a multiple of 64 for historical reasons, one of those magic figures (integer powers of two) which crop up frequently in digital systems, the conclusion is that a polyphony figure of 128 would cope comfortably with the situation just outlined. It would also allow some independent pedal stops to be added. Thus a small to medium size two manual organ needs a polyphony figure of at least 128.
If there are octave and suboctave couplers however, this figure can reasonably be doubled to 256. Such couplers gobble up polyphony for obvious reasons. If a wet sample set is in use, it needs to be (typically) doubled again, because when a chord is released the synth still has to maintain all the notes sounding until they die away into inaudibility. If another similar chord is then keyed before the decay is complete, there must be enough polyphony remaining to cater for it. So a polyphony figure of 512 is not unreasonable for a medium sized two manual organ with octave couplers and a wet sample set. Larger instruments need correspondingly more. Therefore to keep the polyphony demand to the minimum in a VPO of modest performance, it is advisable to restrict the use of octave and suboctave couplers (which is why many commercial digital organs have none at all!), and use dry rather than wet sample sets. Reverberation can be added to dry samples by using external effects processors, which do not consume polyphony. Given these restrictions it is then quite possible to run a large two manual organ or a small three manual one, each with around 35 speaking stops, using a polyphony figure as low as 192, though this is admittedly getting a bit marginal. It assumes not all of the stops will be used at once, which is realistic (you would not use a celeste with full swell for example, nor would you usually need a salicional or dulciana with full organ). You soon know if there is insufficient polyphony because notes either will not sound, or they will be dropped, or the sound of the instrument will break up unpredictably. As another benchmark, a large three manual organ offered by a well known digital organ manufacturer (not a VPO) has a polyphony figure of 512. This uses dry samples and has no octave couplers.
Latency is the time delay between keying a note and hearing the sound, and obviously it has to be kept as low as possible. It does not have to be zero because the latency of a pipe organ is seldom less than about 50 milliseconds , but the latency performance of some VPOs is nevertheless unacceptably large - at least to me. It generally means the computer is not powerful (fast) enough. However, if the VPO is only going to be used to replay MIDI music files rather than react to the real time demands of an organist at a console, its latency specification can be considerably relaxed.
What follows is little more than a list of some VPOs and where to find them, included purely for convenience to reduce the need for a lot of surfing on your part. It is emphasised that further details of any of them must be obtained from the respective suppliers and/or user communities, some of which are active in the form of discussion forums on the internet. I cannot get involved in such matters myself. The fact a VPO appears here does not necessarily constitute a recommendation for it, as this article is in no way an endorsement or advertising platform for any of them. The list may not be complete as VPOs tend to come and go, but it is sufficiently representative of the scene as it existed at the time of writing. If you are interested in trying any or all of these you should download and save them now while they are still available even if you will not be installing them straight away. Otherwise it will be no good complaining they have vanished when you want to use them! The entries are listed in alphabetical order. Note that you will be well advised to take the usual antivirus precautions before opening any files downloaded from the internet. In particular you should immediately scan the download with up to date antivirus software before doing anything to it at all, such as unzipping it let alone installing it.
(Note to Webmasters: the links here are posted initially in plain text only. If you would like to see a clickable link, please refer to my links policy).
See genpo.sourceforge.net/index.shtml (accessed 25 May 2014)
GENPO stands for General Purpose Organ. Its sample sets are in SoundFont format and it is free 
See www.grandorgue.de (accessed 25 May 2014)
GrandOrgue offers builds for various platforms (Linux, Windows and Mac). It uses software synthesis and it is free .
See www.hauptwerk.com (accessed 25 May 2014)
Hauptwerk requires the purchase of a license though there is a free version with restricted capabilities. It uses software synthesis and it can run on different platforms. It comes complete with a sample set but many others are also available from third party sources, ranging in price from free to GBP three figures. There is also a wide range of third party suppliers who configure and build complete systems to your specification in turnkey form .
See sourceforge.net/projects/jorgan (accessed 25 May 2014)
jOrgan is so-called because it is written in the Java programming language. It simulates an organ console and acts as a software 'relay' for it. The actual sounds are then rendered using one of a range of hardware or software synthesisers. It works on various operating systems and it is free.
See www.miditzer.org (accessed 5 July 2016)
As its name implies, MidiTzer focuses exclusively on simulating the WurliTzer theatre organ, such as the Style 216, rather than classical organs. Its sample sets are in SoundFont format and it is free. Like jOrgan, MidiTzer is a software 'relay' sitting between the user and the simulated organ pipes. In common with some of the other systems mentioned here, it also simulates a theatre organ console on-screen which can control every aspect of the simulated instrument. A real console can also be used.
See : dl.dropbox.com/u/35253096/MyOrgan_1_0_6.exe (accessed 25 May 2014)
MyOrgan is no longer available from its originator, though it does pop up in various places on the web from time to time and it is not difficult to find a copy of the software. This might be because it still has an active user community. At the time of writing the URL above is a cloud file source which downloaded OK when I tried it. It used software synthesis and it was free .
Prog Organ stands for Programmable Organ. It was developed by myself and further details can be found here. It is no longer available but still has an extant user community. One reason for including it here is that it is described in detail on this website (see the link just given), thus it complements the material in this article with more technical information, much of which is relevant to VPOs more generally. Several hours of mp3 recordings made on many different simulated organs can also be auditioned by following the link.
OK, so having read this far, how do you decide which VPO to go for? There are several criteria which it might be useful to consider, and you might think that an important one is cost. However all of those listed above are free or they have a free evaluation version, so the cost of the VPO software itself is actually not all that relevant in the first instance if you are starting from scratch. Nor is the cost of the sample set(s) a show-stopper because free ones are also available for all the VPOs listed. Setting up a basic console or keydesk can cost money though not necessarily a lot of it, and some references were given earlier which might be useful . There are also several constructional tutorials or videos on the internet, such as on YouTube, which can be located by a bit of targeted surfing.
Perhaps a more important issue which has cost implications is whether you feel confident enough to assemble the VPO yourself. Much information is available on how to do it on the supplier websites, their user forums and elsewhere on the web, and it is probably fair to say that the DIY route is not as daunting for most people as it might appear at first sight. However you can use the services of third party system suppliers instead, though then the cost will escalate of course. Hauptwerk is particularly well served by such firms, though some of them might configure other systems also.
It is important to decide which type(s) of organ you wish to simulate, and thereby ensure that the appropriate sample set(s) will be available for the VPO you are focusing on. A related issue is whether you wish to run wet or dry sample sets because not all VPOs (particularly SoundFont-based ones) can handle the release transients of wet ones properly. This does not mean that dry samples will actually sound dry in your listening room, because normally you would add reverberation to them using external effects processors in either hardware or software form. Also check that the VPO software itself will run on the computer and operating system you wish to use - Windows, Mac or Linux.
Be aware of the need to read the small print relating to the licensing situation, both of the VPO software itself and that relating to any sample set you wish to run on it. Some licenses limit very tightly what you are allowed to do. This advice applies equally to free products as well as those you have to pay for. For instance, you can only play some VPOs, even some free versions, to yourself in your own home - no public performance in churches or elsewhere is permitted, including youtube. This applies regardless of the sample set being used. Or a VPO with no restrictions of itself will still be subject to the license relating to the sample set you wish to use, and again some of these are only released for private use only. I have had people complain to me that they have spent hundreds of pounds on a system only to discover that they cannot demonstrate or otherwise use it in a non-domestic environment (though I'm not sure what they think I can do about it!). So do read the small print carefully, especially before parting with your hard-earned cash.
A related issue with paid-for products is your rights if they do not work, go wrong, or simply if you do not like what you hear. It is little comfort to have paid a lot of money for a cathedral sample set to find that its sounds are not at all to your taste. So again it is necessary to read the small print here as well before parting with your money.
So to summarise, the issues raised in this article which can affect the choice of a VPO are (in alphabetical order):
Configuration and assembly - turnkey product or DIY
Console - touch screens vs stop keys/drawstops; must be MIDI-capable
External effects processors - hardware or software; more important for dry sample sets
Number of audio channels and loudspeakers required
PC host platform - operating system; speed; memory size; noise; bootup time
Polyphony required - depends strongly on size of simulated instrument(s)
Sample sets - wet vs dry; number required; type (theatre/classical); size of instrument
This might not be a complete list for your purposes but it should provide a reasonable starting point. If you really want to drill deeper into the decision-making process you could use these as column headings in a spreadsheet, with the various VPOs running down the side to form the rows. Then you could put a score (e.g. from 0 to 5) in the boxes of this matrix. The best match to your criteria would then be that VPO which ranked highest when the row scores were added together.
Although this might still seem daunting, bear in mind that all VPOs have much in common - they all need a computer, a MIDI-capable console and an audio system with at least two channels. Therefore, having invested in these, in some ways it does not matter much which one you go for in the first instance provided your PC is up to the job. All of them are free or they have a free evaluation version, and most have more than one free sample set. So you can always move to a different VPO subsequently without losing very much, if anything, in cash terms unless you will need to upgrade your PC or expand the audio system. If you still feel stumped, maybe just take a deep breath and go for it - build the simplest of set-ups with a single MIDI keyboard in the first instance, together with an ordinary computer such as one you might have lying around, and download one of the simpler VPOs. Once you get it going - and eventually you will - you will feel overjoyed (promise!), even though it might not fully meet what you had in mind originally. But having got a long way up the learning curve by doing this, you will then be in a much stronger position to refine your set-up as the next step.
Mention should be made of a major downside which could affect any VPO. It arises because they all rely on a PC to drive the system. Unfortunately, the PC is probably the most ephemeral and short-lived consumer item known to mankind. To illustrate this, ask yourself how many people are still using PCs which are 10 years old or more. A 10 year old PC today (late 2016) means it will quite likely still be running Windows XP (what's that?) because Vista had yet to appear, and it will be unable to access more than 4 GB of memory unless the motherboard was one of the rare examples which could accept more. Also the then-uncommon 64 bit version of XP would have been required as well. For peripherals such as sound cards, only old-fashioned PCI expansion slots will probably be available in these old computers, and these have long been superseded. The CPU will quite possibly be a single-core device running at less than 2 GHz.
Another topic worthy of an airing concerns certain shortcomings sometimes found in sample sets. For instance, not all of them comprise an independent sample for every note of every speaking stop. If the sample set is contained within a SoundFont, this can easily be checked by loading it into a SoundFont editor and observing how many notes of a given stop are associated with the same sample. If a particular sample is found to be 'stretched' across several adjacent notes (often called a keygroup), this does not mean that they cannot be played simultaneously because the VPO ought to use a rendering engine capable of pitch-shifting the sample in real time to suit each note which is keyed. But there will obviously be some loss of realism if sample-stretching is used. However the problem is more insidious if there are separate physical samples for each note which are not, in fact, independent. This can occur if the sample set producer only sampled, say, three or four pipes in each octave (instead of twelve) to reduce the amount of effort involved in generating the complete sample set. In this case s/he might have generated a separate sample waveform for each intermediate note by pitch-shifting the same sample beforehand in a wave editor. This type of shortcut can be difficult to detect, and it can be argued that if you cannot detect it, then it is not a problem. Nevertheless, there are some tell-tale symptoms which can reveal when this practice has been employed, such as a particular keygroup having identical chiff characteristics or an identical characteristic timbre (tone quality). I can assure you that if you come across unnatural effects such as this when playing the VPO, you will become progressively more irritated by it thereafter!
Another problem arises when similar-sounding stops do not use independent samples. A case in point concerned a Great Organ division which had two diapasons labelled Large and Small. In this case, the sample set producer probably had not bothered to sample one of them at all, but merely served up the same set of samples for both. They were probably made to sound superficially different merely by adjusting the relative volumes, or perhaps by applying a low-pass filter to one but not the other. As in the case mentioned above, such tricks do not fool a discerning ear for very long.
Yet other problems can occur, one of which continues to mystify me but it illustrates the sort of thing one encounters. Particular notes of certain stops in one sample set varied noticeably in frequency and amplitude during their sustain phases in a rather disconcerting way - the sound seems to 'lurch' in a most peculiar manner as the key is held down. No real organ pipe in my experience has ever exhibited such behaviour, so I examined the offending samples in a wave editor and discovered that there were indeed corresponding artefacts clearly visible to the eye. I can only assume that somebody was creeping around close to the microphones while the sample was being recorded, thereby impressing amplitude and frequency changes (via the Doppler effect) on the sample.
Perhaps such shortcomings can be forgiven in sample sets which are free, but they are obviously unacceptable in a product one has to pay for. Unfortunately there will sometimes be little redress available in such circumstances.
1. "Choosing an electronic organ", C E Pykett, The Musical Times, January 1987.
Also available on this website (read).
2. "Electronic Organs", C E Pykett, Organists' Review, August 1998.
Also available on this website (read).
3. "Digital Organs Today", C E Pykett, Organists' Review, November 2009.
Also available on this website (read).
4. "Digital organs using off-the-shelf technology", C E Pykett, 2005. An article on this website (read).
5. Not all users need a physical console because they are content simply to fire pre-prepared MIDI files into the VPO, much as a pianola used a paper music roll. However it is assumed in this article that you want to be able to play the VPO in real time from a more or less standard organ console. Some articles elsewhere on this website cover console-related topics for VPOs. See:
"A simple console for a virtual pipe organ", C E Pykett, 2014
"Re-using old stop tabs in Virtual Pipe Organ consoles", C E Pykett, 2014
"Console scanning, MIDI and PIC microcontrollers for Virtual Pipe Organs", C E Pykett, 2012.
6. "Wet or Dry sampling for Digital Organs?", C E Pykett, 2010. An article on this website (read).
7. "MIDI for Organists", C E Pykett, 2001. An article on this website (read).
8. "A MIDI pedalboard encoder", C E Pykett, 2004. An article on this website (read).
9. "The End of the Pipe Organ?", C E Pykett, 2005. An article on this website (read).
10. "Response Speed of Electric Actions", C E Pykett, 2001. An article on this website (read).
11. SoundFonts are widely used in digital music. They exist in the form of a standard file format for inputting a sample set into a synthesiser. Details are widely available on the internet, such as in the Wikipedia article at:
en.wikipedia.org/wiki/SoundFont (accessed 26 May 2014)
A wide range of SoundFont-based synths exist, both as hardware (e.g. SoundFont-compatible sound cards) and software (e.g. one of several software synths).
Probably the main disadvantage of SoundFonts is that they are not well suited to handling the extended release transients (reverberation tails) of wet samples in a highly reverberant environment. However the use of dry samples with separate effects processors is often just as effective, and it can be more flexible as far as the user is concerned because s/he can then adjust reverberation to taste.
12. There is some historical convergence between Hauptwerk, MyOrgan and GrandOrgue if one goes back in time. MyOrgan (free) used a similar if not identical file format for its sample sets to that of Hauptwerk version 1 (shareware) which had appeared a few years earlier. However MyOrgan was then subsequently withdrawn by its originator and support for it ceased, though it can still be obtained from sources such as that mentioned above. GrandOrgue (also free, and meaning the same in French as Hauptwerk does in German) then appeared to take its place. It is still available and supported, and it still seems to use a similar file format to that of the two previous ones.