If you have spent a few minutes looking around the web to find suitable products for a fax server, you will have been presented with several options. These options dramatically vary in price, and seem to have very different approaches to the problem. It can be difficult to discern what the true good points and bad points are of a particular solution, and what it actually is you are paying for!
This will take you through what each type of product gives you, and relate it back to what you actually need from your solution. We make no apologies for the fact that this document favours our products, as this describes the reasons why we have designed these products!
What you have found probably falls into 3 main categories:
Before we go into these different types of solution, we will take a look at what your requirements are likely to be:
We are assuming that you need more than one telephone line for your fax service, and that you are going to be providing this for multiple users in some kind of server-based system.
There will be a set of things you know you need the board to support, and this may involve things like V.34 fax, colour, MMR and/or JBIG compression. This can be determined from looking at a product datasheet or a table in the products section of a website, so it isn't difficult to work out. If you don't fully know or understand the features you need, take a look at our 'How do I?' section for some assistance.
When you install the board(s), you want it to be uncomplicated and quick. When the system is running, it should stay running with little or no maintenance. You don't want to have to worry about compatibility when making your decision, and you want something that keeps on working without issues.
There is nothing worse than discovering the downside to a particular solution after the first week of operation. You want products that are designed to do what you want them to do, and live up to their promises.
You want to pay for a system that works, rather than lose time trying to fix it when things go wrong. You don't want to spend more than you have to, but will spend money if there is a real benefit.
Sometimes we gravitate towards the most expensive solution because of the confidence that engenders, but we may be paying for things we don't need, a brand, or simply an overpriced product. Vendors don't usually put that information in the brochure!
Overall, you want a low 'total cost of ownership'. It's unlikely that installing and running fax servers is really the core of your business. The products you buy should just work, leaving you to run your IT department or business.
Having told you what we think you might be looking for (which presumably you already know!), let's look at the possible solution in that context.
These are the most widely available products for PC-based faxing, and they are usually priced cheaply. Most of them are what is known as 'software modems', because they have the minimum hardware you can get away with, and simulate the modem functions (including fax) in software. A lot of fax software will work with these boards, as they generally use the Class 1 Fax interface. This is a standard from the International Telecommunication Union (ITU) that has been very widely deployed and tested. Installation is usually simple, as these products have to work for home users.
At first glance, this appears to be a very attractive solution. It is, after all, the lowest possible initial outlay and that makes it compelling. However, the initial outlay may represent a small percentage of the total cost of ownership. These products are unlikely to support the latest fax standards, or even some of the most fundamental standards in general use. Sending a fax with these boards may take as much as ten times longer, so your telephone bill by the end of the year could dwarf the entire cost of the system.
That isn't the end of the story. Several of the leading vendors of fax software refuse to support their products when they are used with single modems. The software will probably function with these modems, so this may seem an odd decision on the surface. However, the motivation behind this is because they know the problems their users will have from a reliability standpoint, and they don't want to be a part of that.
These modems have been designed for a single user environment, and to minimise the manufacturing costs as far as possible. These products do what they were designed for very well, but that doesn't include server operations. If you use these in a server solution, you will discover that they simply won't work reliably. They will drop calls, lock up themselves or the system, and result in a constant stream of support problems for whoever is responsible for it. This is for several reasons, including questionable implementations of the standards, cheap components, insufficient processing power, but mainly because the manufacturers simply don't care about server applications their devices were never designed for.
If the software vendors don't like them, then it's likely you won't either. We regularly come across customers who have chosen cheap single modems, struggled on for a while, and then come back to buy one of our boards.
Our Overall rating for single port modems:
| Area | Comment | Rating |
|---|---|---|
| Performance | 'Software modems' have limited processing capabilities, and so may not be able to cope with some fax scenarios. | 2/10 |
| Features | Will not support V.34 Fax. May not support even V.17! This results in unnecessarily large telephone bills. | 3/10 |
| Reliability | Not designed for reliability to levels required in servers | 1/10 |
| Installation and compatibility | Use of standards is good, and installation is simple. Compatibility may not be close to standards, as they are often poor implementations. Limited manufacturer support for this type of application. | 6/10 |
| Price | This is the lowest cost solution. | 9/10 |
| Overall | They do a very basic job, but create problems. It is a false economy to try and save money by using these in a fax server. | 4/10 |
The products that were designed specifically for this type of application are known as 'intelligent' boards because they contain a lot of processing power. This enables them to handle faxes in a stable and efficient way, using the latest protocols. There are elements of a fax call that are ultra-sensitive to time delays, and making sure they are handled by dedicated hardware is important. They are certainly more expensive than the single port alternatives, but are well worth the extra money when compared with the savings on support, uptime and your phone bill.
These boards will support all of the features of a high-end >$1000 departmental fax machine, such as V.34, JBIG and even color! V.34, for example, increases fax speed by 2.3 times, and won't be found on a low-end device. The savings you can make on your telephone bill with a V.34 device may be the same as your entire system over a year!
Within the 'Intelligent' fax boards, there are also two distinct types of product. The primary difference between them is whether they use a standard interface or a proprietary one. The type of interface won't affect you directly unless you are a developer of fax software or systems, but there are knock-on effects that will affect you (such as price!).
These products definitely sit at the most expensive end of the spectrum. They include sufficient processing power, usually in a few DSPs handling several channels each. They have an Application Programming Interface (API) specific to each vendor. They use large customized drivers for installation that provide this high-level interface for the fax software. Everything to do with a fax usually happens on the board, and that makes it harder for software vendors to add new features.
There is a lot of history to this type of board. The architectures for these boards were decided when most of these companies started over a decade ago. When this occurred, the standards were not well established, so the choice was to go for proprietary APIs that were strictly under the control of a particular vendor.
Today, things are very different in several ways. The standards are well-established, and the interfaces haven't changed because they don't need to. That doesn't imply no new features, but that the new features can be added either in hardware or software as required, without the interface changing! The fax software vendors can add new functionality very quickly, and continue to provide value through upgrades and patches on the same hardware.
Another important change is that PC Processors are in the region of 100 times faster than they were 10 years ago. It makes sense today that the software takes on certain tasks that were not appropriate when the proprietary architectures were settled upon. The fax hardware manages the real-time aspects of the fax call and related processing, but the pre-processing such as compression works very well in software.
The companies behind the proprietary boards have had to work with their APIs and build on their existing product designs. This means that APIs change regularly in order to add new features (the APIs weren't designed like the standards: to be extensible), and software vendors have to rework their support for these boards at every new release. In contrast, the fax standards used by Mainpine have been unchanged throughout the company history, despite the successful introduction of new features. Proprietary interfaces create a nightmare of compatibility issues and support problems in the field. Once a system works, it usually keeps working, but every time you upgrade your software or add a new board you may face issues.
The design legacy that these boards have means that architecturally they do not conform to modern methods of production or operation. The use of standards in design is what has given us the PC and communications architectures of today, and it is a vital design philosophy for reliability and compatibility. The right levels of performance and functionality can and should be accomplished using standards.
Overall these boards are very expensive, and come from vendors that built their success on product designs from 10 or more years ago.
| Area | Comment | Rating |
|---|---|---|
| Performance | Large DSPs running several channels provide sufficient power (usually around 10 to 25 MIPS per channel), even if this creates points of failure that affect multiple lines. | 8/10 |
| Features | Usually a good set of features as this is what they are designed for. However, there is often a lag behind the latest. Support for new features usually require hardware and/or API changes. | 7/10 |
| Reliability | Designed with server operation in mind. However, lots of components and sometimes even fans for cooling can produce reliability problems, as well as single points of failure. Several large vendors are also suffering from quality control issues as they try to lower costs by moving manufacturing offshore. | 7/10 |
| Installation and compatibility | Custom APIs, large and complex drivers - all this can make installation tricky. 'System releases' that have complex operating system and product combinations can also be cumbersome and confusing. A moving target for the software vendors | 6/10 |
| Price | This is the highest cost product. Non-standard interfaces and components push the price up. The boards are expensive to make, so expensive to buy. | 2/10 |
| Overall | They do the job they were designed to do, but the world has moved on to better things. | 6/10 |
These boards support Class 1 and 1.0 interfaces from the ITU, and use standard operating system drivers for maximum compatibility/reliability. This is similar to the single-port modems, but that is where the similarity ends. They have 150 MIPS of DSP and co-processor power per channel for the fastest fax protocols available - roughly equivalent to an Intel Pentium* processor running at 90MHz per channel! The implementations of the standards are much more rigorous, and tested with the leading fax software packages, specifically for use in servers.
They are expensive compared with the single port modems, because the component quality is higher and they aren't 'software modems'. However, when compared with the proprietary boards they are excellent value for money. They are designed with server applications in mind, without compromising on performance, reliability or functionality, but combining architectures and standards that make sense with today's technology.
Standards are powerful! While we try to test individual software packages ourselves, there are a lot out there, and we can't test them all. However, almost all fax software packages support either a Class 1 or 1.0 driver, and if it is a mature implementation, it will work without ever having been tested specifically with our products! Standards make designs simpler and more effective, and that reduces costs of design, manufacturing and support that we can pass on to you.
Because our products support standard interfaces, they also work with standard drivers which themselves have a very high degree of reliability. For example, in current versions of Linux* and Microsoft Windows*, the 'drivers' you use to install our boards are just INF files that tell the operating system how to connect up the standard serial drivers. That makes installation totally simple, and the use of standard serial drivers available in every copy of those operating systems means there is a lot less to go wrong, both during installation, and in operation.
Another advantage of the standard Class 1 and 1.0 interfaces is that they split the functionality between hardware and software in a way that is extensible, and that suits today's systems. The boards handle the standard operations of a fax, but they allow new features to be added in software. For example, JBIG compression and colour faxing are both added in software and require no hardware or interface change to work on these boards. Vendors of proprietary boards have attacked Class 1 or 1.0 interfaces as 'not suitable for server use', which simply isn't true. It is true that the implementation in single port modems isn't up to the job, but the standards themselves are not the issue. In fact, proprietary APIs cause problems with compatibility, and are a pain for the software vendors, so less suitable than the standards!
The software vendors like the Mainpine boards because they let the software do what it does best. For most users, once you have installed the board, there is little difference between the proprietary and standard boards; the standard boards will be more reliable and easier to upgrade, but both do the job!
| Area | Comment | Rating |
|---|---|---|
| Performance | A dedicated 100 MIPS DSP and 50 MIPS co-processor per port gives more than enough intelligent processing power, and scales indefinitely. | 10/10 |
| Features | The latest Fax standards, and a good split of functionality between hardware and software means the most up-to-date features are available now and in the future. | 10/10 |
| Reliability | Designed for reliability, long MTBF and a focus on quality. Dedicated processing hardware per channel minimizes the single points of failure. We are very proud of our products and our record! | 10/10 |
| Installation and compatibility | Use of standard interfaces and drivers. Wide testing with fax machines and fax software both in-house and in close co-operation with vendors. | 10/10 |
| Price | Costing more than single modems, but dramatically less than the proprietary boards. We can offer the product at these prices because we have the features of maximum benefit for you; no more and no less! | 8/10 |
| Overall | These products have been designed for this application: Designed using modern methods and high quality components and procedures. We know they do the job, which is why we offer a 30 day satisfaction guarantee and 3 year warranty. It doesn't cost us to do that if we've got the right products! | 9.6/10 |
When buying any kind of IT system, reliability is always important to us. However, we often end up making a judgement based on our gut-reaction to a solution because it is so very hard to get objective data to give us a feel for that. However, there are several issues that govern the reliability of a solution, and objective measurements we can make for them.
MTBF is the objective figure of hardware reliability that can be applied to almost any mechanical or electrical device. We use the Telcordia SR-332 "Reliability Prediction Procedure for Electronic Equipment" standard for generating our MTBF figure. This is the most widely used standard inside the USA and Canada, and accepted internationally.
The MTBF gives an indication of when you can expect a component failure based on established failure rates for the components in the situation they are used. Component failures usually occur in the first few hours of the life of a product, and then towards the end of the life as specified by the MTBF figure. This is why we undergo soak testing on all product before it is shipped to a customer. We catch any early component failures so the product should then operate without failure until the end of it's life.
Our calculated MTBF figures are different for each product, but the lowest figure for any of our products is over 25 years continuous operation (224 thousand hours), which we believe is a good figure for a server product. This can be objectively compared with other vendors that publish a figure. We can also supply more detailed information on the MTBF calculation when requested.
The design of the drivers and API can have a major impact on reliability, as well as the board itself. The amount of testing done and the general design ethic inside a company affect this, but you can't measure that externally and have to take a lot on trust or reputation. However, at Mainpine we use the standard operating system drivers that are far more mature than anything the entire fax industry could produce because they are so widely used and tested. We also use standards in our hardware design that are mature and objectively tested. Fundamentally, we have chosen routes for maximum reliability because that benefits us as much as you. We also undertake rigorous testing ourselves of course, but our design philosophy is something you can examine objectively, and compare with others.
Quality control procedures in manufacturing and testing are also a key area that affects reliability. There is no objective method of measuring this, but there are standards that apply to these processes. These standards make sure that quality procedures are in place and adhered to, and also make sure problems can be traced and fixed. The ISO-9001 and ISO-14001 are the standards that ensure this in manufacturing, and all our manufacturing plants conform to these standards and are regularly inspected by external agencies to assure continuous compliance.
This background on reliability should help you to evaluate a company and their products based on facts rather than just a general impression.
It may not surprise you that we state a position that favours our own products. However, in our defence, this position is exactly why we have designed them! We have a lot of confidence in our products, and that's why you can try them on a 30 day satisfaction guarantee. You can also listen to what our customers say. Here is a comment from one of our partners (a market leader in Fax) over why they work with us:
"For our customers, Mainpine products provide the features they need at the right price. For us, we don't have to worry about reliability or support issues. The RockForce range provide the perfect platform to showcase our software."
- Art King, President, FaxBack Corporation
FaxBack are experts in fax server software, and they recommend our products because they know that they represent value for money, and that they won't generate support headaches. This is a common reaction across all of our partners, and why our resellers are so eager to sell our products.

The Mainpine RockForceT multi-port boards for fax, data and voice applications are the only products available that meet all of the following requirements:
Prices start from $499 MSRP.