Moving a file from one system to another has become a relatively simple operation in today’s imaging workplace. The incompatibilities once prevalent between Macs and PCs no longer exist to any significant degree. Large graphic files can be quickly transferred from one computer to another via removable storage media, as long as the recipient has the appropriate hardware and software to read what the sender provides. And if the two systems are separated by many miles, which is often the case in our virtual, global workplace, next-day service is certain to meet the most demanding deadlines, or is it?
As file sizes increase exponentially, deadlines tighten, and the volume of international business swells, a whole new set of file transfer issues and communication challenges must be confronted. Clients want to see layouts today, not tomorrow. Prepress houses and printers need files immediately, if not sooner. They often can’t wait for the delivery truck, especially if they have several million dollars worth of equipment standing idle, waiting for a job.
Straight modem-to-modem transfer may be the simplest way to move a file from one system to another. But while it works well enough for text files and smaller graphics, it doesn’t quite make it when file sizes of 20-, 50-, 100-, or even 300MB are transferred. It could take hours to transfer one 300MB file, and a small glitch on the analog line during transfer could bring the entire process to a grinding halt.
Since many companies already have an Internet presence, the logical solution, or so it would seem, would be to transfer images via the World Wide Web. Net transfers are being done every day. However, this method has many drawbacks for firms that frequently send and receive large graphic files. For example, there are security issues and transfer speed problems. Too many hackers have the means to break into Internet access systems, causing havoc and often compromising the confidentiality of the materials being transferred. And even though the sender may have a high-speed connection to the Net, once the file is routed to its destination, there’s no telling at what speeds it will be retransmitted. The sender has no control over the ultimate transfer speed.
Part of the problem with current high-speed data transmission is that many of the issues involved in this area of communication are still unresolved. Are there going to be broad-based standards? Is it going to be possible, and affordable, for smaller businesses such as independent photographers and graphic artists to transmit images at high speeds?
There are currently two alternatives available for transferring large graphic files from one location to another electronically: open systems and closed systems. The open approach relies on broad-based standards for hardware and/or software, while the closed approach requires proprietary hardware and software.
One of the most popular open solutions to high-speed image file transfer is based on the Integrated Services Digital Network (ISDN). Some vendors hail ISDN as the future of high-speed data transmission, while others criticize it as obsolescence waiting to happen. But in reality, all computer equipment is only one generation away from being obsolete.
ISDN has been accepted as a telecommunications standard worldwide. Some 50 countries have either implemented it or are in the process of doing so. There are an estimated 4 million ISDN lines installed around the globe, most of which are being used for voice, telephony, and data transmission. It’s particularly strong in Europe, where it has been well established for the last five or six years.
First introduced in the United States in 1984 in Boston, ISDN didn’t go anywhere in this country for quite some time. Now it’s gaining a foothold in the United States, and with the acceptance it has received so far, it’s most likely that ISDN will continue to evolve, and even-tually, be replaced by an entirely new technology.
ISDN is available in two types of lines: basic rate interface (BRI) and primary rate interface (PRI). BRI lines have one pair of transmission lines called B lines. PRI lines can have between 23 and 30 pairs,
depending upon which part of the world the line is being laid (in the United States, PRI lines have 23 pairs).
Because the lines are paired, it’s possible to handle multi-session file transfer. BRI ISDN lines are about four to five time faster than standard 28.8kbps analog modems. Transfer rates for BRI are 1MB per minute, while PRI lines can carry up to 10MB per minute. Regardless of how many lines are being laid or where they are, in theory, they can all communicate with each other.
There are two ways of working with ISDN lines, either on a station-to-station basis (sometimes called peer-to-peer) or through a wide area network (WAN). Station-to-station connections are the common way of communicating with a wide range of clients, vendors, and service providers. WANs are more commonly used for file transfer within an established set of computer users, either within a company or between members of a project work group. Users have more control over WANs, but peer-to-peer connections are faster. Either way works and it is possible to use a combination of both.
ISDN lines may be compatible, but the devices used to transmit and receive are not. It’s somewhat like the incompatibilities that used to exist between modems before Hayes became a standard. Until very recently, it was necessary to install similar hardware devices (usually plug-in boards) to transfer files over ISDN lines. And most of the dozens of ISDN devices available are engineered to communicate only with a similar device on the other end. While that’s not a problem when
ISDN is used for communications within an organization, it does present a problem when trying to send files to multiple vendors, clients, or agencies outside the company.
Board manufacturers were not always interested in supporting open architecture standards. As they saw it, the more wide open the architecture, the greater the competition. That was particularly true when the number of manufacturers was few and the vast majority of devices was NuBus-based. Now that PCI is becoming as much a standard on the Mac as it is on the PC, competition can no longer be slowed, and manufacturers are beginning to see the advantages of compatibility.
One of the companies leading the way in developing off-site standards is the British-based 4-Sight L.C., which has its U. S. headquarters in West Des Moines, Iowa. The company developed a software bridge between many noncompatible ISDN devices, called ISDN Manager. The program is available for both Mac and PC platforms. According to 4-Sight President Lyndon Stickley, “We’ve spent the last few years focusing on making our software drivers communicate with all the leading boards on the market.”
The company specifically focuses on the graphic arts and prepress environment, so it’s very familiar with workflow issues, system security, and other requirements involved in the graphic production and approval process. According to Stickley, there’s a lot of interest in ISDN, but most people don’t know much about it. This was apparent by a long standing joke that ISDN actually stood for “I still don’t know.” There continues to be a lot of confusion about ISDN.
“The whole world is saying, ‘We’ve got to go digital,'” said Stickley. “It’s the flavor of the moment.” But there is a number of misconceptions about what “going digital” actually means. The most important thing to understand is that ISDN and digital are not synonymous. “ISDN isn’t unique in the fact that it’s a high speed digital link . . . there are a bunch of digital links available. But there are some things that should be known about ISDN that make it the world’s mass market medium for file transfer. ISDN is absolutely the only dial-up digital technology in the world that is so available, affordable, and fast.”
Some other high-speed digital links include T1, switched multi-megabit data service (SMDS), asynchronous transfer mode (ATM), cable, and satellite.
“ISDN is finally becoming as revolutionary as the fax machine,” Stickley said. The fax machine changed the printing and prepress industry by enabling service providers to send a true representation of a job over a phone line instead of sending it in the mail. ISDN does the same; It buys you time.
Fast turnaround times are important in today’s instant gratification society. “You can be at a sporting event like the Olympics and syndicate high-res color images anywhere in the world to newspaper picture desks,” he pointed out. “You can do it in an instant with ISDN lines, rather than heavily compressing the images or sending them by courier. Sports Illustrated, AllSports, and many other news and stock image agencies are using the 4-Sight ISDN solution to facilitate image file transfer.”
ISDN is also very dependable, Stickley said. “You can send a 400MB file on one of our systems, and you know from connection to downing the line that it’s perfectly error free.”
Stickley’s company is doing with ISDN Manager what Adobe did with PostScript: Providing an interpretive standard that facilitates communications between different hardware devices and platforms. “Our software application makes hardware devices conform,” he said. As part of that program, 4-Sight has come up with its “4-Sight compatible” program, which certifies third-party vendors as being 4-Sight compatible.
While 4-Sight is not trying to compete with board manufacturers, it does bundle its software with two boards: the Harmonics and the SCII. It also supports more than a dozen other boards. Transfer speeds and capabilities of the available hardware vary. The Harmonics board, for example, can transfer up to 3.5MB per minute when tied into the maximum four dual-channel lines it supports.
The latest release of ISDN Manager Version 4 includes a useful soft proofing capability that allows both sides of the communications link to see and annotate a file, while allowing the generator of that file to maintain complete control. It’s a very effective way to work on graphics or images at remote locations.
So far, 4-Sight has nearly 10,000 sites installed worldwide. The company is licensing and OEMing its software, which is another step closer to global standardization. The company recently signed agreements with Linotype-Hell and Crosfield. There are also another 17,000 compatible sites installed by other vendors, and 4-Sight is adding 700 sites per month. “In the coming 12 months,” Stickley said, “we expect to install approximately 3,000 basic rate systems in the U.S.”
The price of ISDN Manager for a BRI is $2,795, which includes one adapter board. Prices for multi-line setups go as high as $12,000. There are also software options available for sites that already have the hardware but want to make it compatible with a broader ISDN base.
Advocates of proprietary file transfer emphasize the advantages of speed and security in closed systems. Companies like NetCo, Minneapolis, are establishing a network of proprietary sites–primarily in the United States, but soon to be available on a worldwide basis. NetCo markets the WAM!NET proprietary file transfer network. “There are plenty of people who will sell you a T1 (connection), a server, and a router. We provide a complete solution,” said John VanSickle, director of channel development for NetCo.
One of the major advantages is that there is no up-front expense for the client. “We’ve taken out all the up-front charges. You aren’t charged a monthly line fee (although a minimum usage fee is required), nor are you charged for the hardware,” said John Kauffman, marketing director for NetCo. “We’re going to drop $10,000 worth of hardware in your shop, have the local telephone company run a line, and hook you up to WAM!NET.”
Charges are accrued by the number of megabytes being transferred. The per-megabyte transfer fee can run anywhere from 37 cents, where the volume is relatively low, to just a fraction of a cent for high-volume users. The only requirement is a minimum one year contract, which partially amortizes the cost of the installed equipment.
One advantage of a proprietary system is speed, particularly for firms that frequently need to transfer very large image files. “Once we hook clients up to that backbone, their files fly across the country,” said VanSickle. Because it is a private network, WAM!NET subscribers can communicate with only other net members. “We don’t have remote log-on. If you want to talk to someone on our network, you have to both be on our network.”
NetCo uses ISDN, T1, and a number of other connections. The type and capacity of the phone line being installed depends upon the average file size to be transmitted. The rule of thumb is that the average file should be transmittable within one hour.
NetCo’s WAM!NET proprietary file transfer network saves clients up-front expense by not charging for hardware. Charges are accrued by the number of megabytes transferred. The system transfers fast, particularly large image files.
Every line and connection going into the system is custom engineered to the individual customer’s transfer requirements. “The type of line depends upon what’s available locally and what works best. We size the line to fit the need,” said VanSickle. For instance, “If the average file size is 600MB, a T1 line will be installed. If it’s 10MB, it will be something different.” The company offers ISDN as an option, but most of its solutions, at least 70 percent or more, are T1 and up.
According to Kauffman, ISDN is not a particularly popular option with clients. “I don’t want to knock ISDN, because we do use it at times, but it’s limited in its applications. It’s already seen as becoming obsolete in the very near future. Not many people want to invest $10,000 in a technological solution based on ISDN, because they see that there’s other technology coming online that will make ISDN look like a turtle.” He pointed to ATM as an example. “That’s on the horizon. It will speed things up to 1,000 times faster than T1.”
Going with a setup like WAM!NET has benefits. “With the company providing the hardware, not only is there no up-front cost, but the ability to keep current with technology is greater. You don’t have to continuously reinvest in new equipment.” It’s to the company’s and client’s advantage to have the most effective network possible installed.
Kauffman used ATM to make his point. Transmitting files 1,000 times faster would certainly make the client happier, but making the upgrade investment would also help the company. “If ATM can go a thousand times faster, that means our network has a thousand times more capacity. That means we can make a thousand times more money. So, if our cost of the network doubles, but we can charge it out 1,000 times more, it makes us more money. The advantages of technology will more than pay for the upgrade. In fact, the sooner we can upgrade, the more money we make.”
Kauffman conceded that WAM!NET is currently not cost-effective for small operations. Independent commercial photographers and graphic artists most likely can’t justify the minimum base price of $500 per month, which includes 1.3GB of file transfers. The $1,000 option includes 9GB of transfers.
NetCo can service customers anywhere in the United States and Canada. It should be available in Mexico shortly. Europe is planned for the first quarter of next year and the Pacific Rim for the second quarter.
There is a number of companies providing piecemeal solutions to high-speed file transfer. Several companies will lay lines and offer suggestions on hardware and software vendors. Others sell hardware and make recommendations on software, while some companies specialize in software and help guide clients to contracting the lines and hardware. It’s generally best to find one vendor that is solely responsible for the proper installation, configuration, and operation of the system. This approach, which may be more expensive in the short term, could end up saving countless headaches and a great deal of money.
Ron Eggers is a contributing editor with PHOTO>Electronic Imaging and a senior editor with NewsWatch Feature Service. He can be reached by e-mail at firstname.lastname@example.org. Go To Top
© 1996 PHOTO>Electronic Imaging
The behind-the-scene race is on between the cable companies and the telecommunications companies (telcos) to bring high-speed throughput to your door. For the moment, though, big cable companies seem more concerned with implementing telephone service to compete with local Bell companies than to provide a new means of high-speed digital access.
In the red-hot California telecommunications market, where more than 60 companies have obtained state approval to provide local phone service, TCI has already announced plans to launch its own telephone service for San Francisco’s East Bay suburbs by fall–perhaps in cooperation with Sprint, perhaps not. Meanwhile, companies like @Home are just now bringing cable-based 10MB Internet access to test markets, while long-distance companies like AT&T, MCI, Sprint, and the baby Bells (including Pacific and Atlantic Bell) are gearing up their own Internet access offerings.
Yet, at the bottom of this no-holds-barred feeding frenzy, initiated by this year’s Telecommunications Act, the challenge for the telcos is clear: How to get broad-band speed to customers over the existing plant of ubiquitous, voice-grade copper wiring, better known as plain old telephone wire.
The trump card several telcos are getting ready to pull from their high-tech sleeves appears to be a new technology called Asymmetric Digital Subscriber Line (ADSL). This modem technology is able to transmit more than 6MB per second, using existing twisted-pair telephone lines. Positioned as a technological fix until the arrival of the fully implemented Asynchronous Transport Mode (ATM), which may take decades to install, ADSL can literally transform the existing public information network and increase capacity by a factor of 50 or more, without having to install new cabling.
ADSL is an ANSI standard developed by the T1E1.4 Committee, which defines high-speed modems using twisted-pair wire. An ADSL circuit connects two ADSL modems on each end of a twisted-pair telephone line, creating three channels: a plain old telephone service channel (POTS); a medium speed bidirectional channel; and a high-speed, downstream data channel. Current offerings show high-speed channels ranging from 1.5- to 6.1MB per second, while the bidirectional channel provides duplex signaling at rates of 16KB to 640KB per second. The POTS channel is split off to maintain secure telephone service, even if the high-speed channels are rendered inactive or incapacitated. Products with downstream rates as high as 9MB per second, and duplex rates of 640KB per second, should be available by the end of this year.
Distance is critical, as current ADSL products can span anywhere from 9,000 to 18,000 feet, depending upon data rates. ADSL proponents insist that these capabilities can cover as much as 95 percent of a telco’s loop plant, depending on the data rate. The vision for ADSL deployment is to connect users from their door or curb to fiber-quality transmission by using ADSL to connect the “last mile.”
When compared to ISDN, not only is ADSL faster, but it is also capable of providing simultaneous telephone and data transmission service over the same single pair of copper cabling. It is faster, potentially cheaper, and requires no special lines for installation.
More than 30 telcos have successfully tested ADSL devices. The players include Ameritech, Atlantic Bell, Sprint, and U.S. West, while leading developers include AT&T Paradyne, PairGain Technologies, Motorola Semiconductor, Ericsson Telecom, Samsung Corporation, Newbridge Networks, U.S. Robotics, Tut Systems, and Harris Semiconductors. According to the ADSL Forum, an industry group promoting the technology, telephone companies around the world will have “little choice” but to use ADSL technology once the devices reach suitable levels of integration and cost. And that is happening now.
At the integration level, development agreements between the major players are helping to shape a robust, interoperable platform. For example, Ericsson, a leading provider of access network solutions, and Motorola Semiconductor have partnered to jointly develop an ADSL access solution based on Motorola’s transceiver, which supports STM and ATM transmission speeds up to 8MB per second. Aimed at the high-speed Internet access, remote LAN access, and interactive video services markets, Ericsson will market the product as the COBRA AT 8 (Copper Broadband Access).
Currently, semiconductor companies have already introduced chip sets that include off-the-shelf components, programmable digital signal processors, and other customized chips. Meanwhile, Motorola is also working with Amati, licensing that company’s Discrete Multi-Tone (DMT) line code technology to provide a single-chip solution to OEMs. Ongoing technology deals like this will certainly increase functionality and reduce cost and power consumption–paving the way to mass production. And while current ADSL products offer T1/E1 and V.35 digital interfaces for signal transmission, future versions will offer standard LAN interfaces as well, to accommodate both continuous and variable (ATM) bit rates. It won’t be long before ADSL is integrated to the point of being a PCI, USB (Intel), or Firewire (Apple) card that one can simply install into a computer.
More recently, PairGain Technologies Inc. announced that it has signed a licensing agreement with AT&T Paradyne Corporation. Under the agreement, PairGain will develop ADSL solutions based on AT&T Paradyne’s Globespan carrierless amplitude/phase modulation (CAP) transceiver technology. With this agreement, PairGain is expanding its comprehensive line of megabit access products for Internet and telecommuting applications. PairGain will round out its product mix by expanding its Megabit Modem and EtherPhone. This is just one example of the cross-pollination the ADSL market will need in order to ensure that a broad base of products is available to serve various customer needs.
Field trials are now taking place. In Edmonton, Alberta, the local telco, ED TEL, is using Performance Telcom’s CopperAccel ADSL products as part of a customer trial of a proposed high-speed Internet access service to be offered in the Edmonton metropolitan area. According to James Freeman, ED TEL’s network planning project manager for the trial, “ADSL offers the prospect of rapid, flexible deployment of broadband capabilities.”
Meanwhile, !NTERPRISE, the data networking arm of U.S. West has recently conducted technical trials of ADSL technology in the Boulder/Denver and Minneapolis/St. Paul markets, using PairGain’s Megabit Modem products. These models make use of the existing copper wire infrastructure and operate at speeds of 768KB per second bidirectional. That’s more than 10 times the performance of single-channel ISDN and more than 26 times faster than state-of-the-art 28.8 V.34 modem technology. The technical trial will be followed by a market trial in the fall, before more widespread deployment will take place in 1997. U.S. West currently serves approximately 25 million customers in 14 western and midwestern states.
It is clear that ADSL will play an important role in telecommunications over the next couple of decades. It’s going to take that long for the broadband cabling to be deployed. In the meantime, ADSL will be able to zip video, remote CD, live television, and group conferencing into our businesses and homes. The critical factor remains that telcos offering new ADSL services will need to reach as many new subscribers as possible during the coming introductory periods. Otherwise, cable companies installing bidirectional cable may gain a significant foothold that could slow down ADSL’s penetration. This will most likely result in hyper-aggressive, take-the-beach marketing and pricing against the telcos’ failed implementation of ISDN.
Copper vs. cable? The war has just begun.