THE EWS STRESS CONTROL PANEL: Articles [subpage EWSoundZine]

How to control the Audio output

By Morten Egelund Rasmussen

The application AudioRoute, discussed in this article, is a great help for a better understanding of the EWS-routing. This article has been written some time ago, but lately the author, Morten Egelund Rasmussen, added the following extra information:

Finally I ran through the source code for AudioRoute. I also checked the graphics.
For those of you who don't know what AudioRoute is, I can shortly tell you that it is a utility I started to write more than 1/2 year ago, but never finished. Anyway: I'm just a user like the rest of you, so I wanted to make a control panel that clearly shows how the EWS works.... And it really does :-)
If you want to develop AudioRoute further (or just download the program as is), please visit the AudioRoute page. You can download the C++ sources and/or the graphics used in the user-interface. The URL is http://surf.to/audioroute
Please notice, that I am not on the mailing list, so any responses to me should be CC-mailed directly to me on silverjam@yahoo.com.

Note by Computer ConText: Those who are interested in more information about the EWS64XL routing should also download the file ewsrouting.zip from Etienne Brauns on the Documentation page of our Download Page.

Original article:

Get the picture...
How do you send the different audio signals to different speakers? I guess this question might be relevant to a lot of EWS64-users, so here comes a recipe.
First of all: get a decent picture of how the internal routings of the EWS-synthesizer work. Take a look in the EWS hardware manual, and get the program AudioRoute from my homepage. Note that you can't yet access all controls in AudioRoute.

Get Audioroute...
Here's a short description on what to do (all done in AudioRoute):

>	Set the Line-out 1 source selector to the synth.
>	Set the synth input selector (the one to the left) to the audio device you prefer.
>	Set the Master volume levels for the 4 different types of signals in the synth part (the synth is the big red part of the window).
>	Set the panning for each of the 4 signal types. That's done with the quadpads below the volume sliders. You can send the individual 4 signals to wherever you want between your 4 connected speakers (front/back - left/right).

Stop reading now if you don't use the Synth wave devices

>	In your wave-file playback program ('Cool Edit Pro' for example), you can set which wave device number (#) you want each wave track/file to be played in. Set track 1 to "Wave device #1", track 2 to "Wave device #2", etc...
>	You now have two different options:
	>	Option 1: Send each track through the effect units (chorus & reverb), but not 4 channel sound.
		>	Turn on both effect units (chorus & reverb) and set the effect parameters of each. (You can do that in AudioRoute).
		>	Go to Virtual Channels (one of Terratec's windows). Here you can set the left/right position for each "Wave device" and the amount of each "Wave device" you want to send through the effect units.
		>	Now you should get your sound where you want it.
	>	Option 2: Don't use effect units, but use 4 channel sound.
		>	Turn off both effect units.
		>	Go to "Virtual Channels". Here you can set the left/right - front/back position for each "Wave device".
		>	Now you get your sound just where you want it.

Please note that all the controls in AudioRoute are Master controls. All other controls work lower in the hierarchy.
To avoid confusion with each wave device (which you set in Virtual Channels), I would suggest you put the AudioRoute wave volume to maximum and center the wave signal between the 4 channels. Then make the settings for each "wave device" in the Virtual Channels window. As soon as that works, you can always experiment with the Master control in AudioRoute.
I hope this explanation is sufficient. To find more detailed information, I'd suggest you to download the manual to the Virtual Channels (see Documentation in Download section).

The SetManager: Priorities, Programs and Variations

By Claus Riethmüller

I've recently noticed that there is some confusion about the priority setting in the Terratec SetManager. You can assign a priority number from 0 to 127 to each Bankset (94B or TTS) and a lot of people don't know what this means and how they should use it.

First of all: Programs and Variations
The Dream chipset can access up to 127 programs in 127 variations (not banks!). So it can access theoretically 128*128=16384 instruments in total. To select a program do a normal MIDI Program Change with the number. To select a variation use MIDI Controller 0 (BankSelect).
Note: you need to do a Program Change after the BankSelect to select the instrument (the BankSelect only defines the variation for the instrument after the next Program Change). This means that you can only assign one instrument to each program/variation. Not more.

The purpose of Priorities
If you have two banksets loaded on different slots in the SetManager, and both contain a instrument with the same program and variation number, you have a conflict. The priorities have been invented to get around this problem. You simply give one of the two banksets a higher priority number and the instrument out of this bank will be used. Note that you can't access the other instrument with same program and variation now!
This has a lot of features: you can for example load different types of the same instrument for different styles of music... or different songs. Just increase the priority of one bank in the SetManager and you have selected the instrument. If you need the other, change it back.
This also means that you should place new instruments onto places (programs and variations) which are not occupied by instruments in the GM or GS Set if you have it loaded and you want to use them.

Mute and Solo
The Mute and Solo switches are also easy to explain: If you set one bank in a slot to Solo, only the instruments from this bank can be accessed - not caring about the priority. If you Mute a bank in a slot the instruments from this bank can't be accessed. You can simplify that by saying that all banks in the different slots behave as one single bank.

Introduction to this page

In the near future we will discontinue the Master Class section. Therefore we decided to centralize all the articles from the EWS Stress Control Panel: the original EWSoundZine articles as well as the Master Class articles. Of course you will still have the opportunity to contribute new articles, either for the EWSoundZine or the Master Class. Lateron we will also move the 'developers articles' to this page.

How to select an article from this page

First of all: you need Internet Explorer to view this page properly. Under the headings on the right you find buttons with a compressed title of an article. When you move your mouse over a button, a short description of the article appears. When you click a button, the full article will be displayed in this window. If for any reason you want to invoke this introduction text again, just click the button Introduction again.

Conditions and guidelines for contributing an article
Conditions and guidelines for contributing an article
>	All articles have to be written in English. Mistakes in spelling or style are no problem and will be corrected as far as our knowledge reaches (we are no native speakers either).
>	Regarding the Master Class articles: the idea behind them is to provide an opportunity to the 'avarage user' to learn from experts in a certain field. So please keep the target group in mind while writing these kind of articles. Let's put it this way: if we understand it, it's good :-). Also avoid (or explain) abbreviations and slang.
>	The author gets credit for the contribution under his/her full name. Sometimes we copy a contribution from the mailing list or our Discussion Forum. Of course we will ask the author's permission first when we would like to publish it on this site.
>	The copyright remains with the author. No rights from other parties may rest on (part of) the content of an article. Computer ConText (the publisher of this site) rejects all responsabilities in these matters.
>	Computer ConText reserves the right to edit the contributions to enhance the readability. If the author objects to the alterations the corrections will be altered according to his/her wishes. Computer ConText also reserves the right to expand an article by contributions, opinions, corrections etc. made by other readers.
>	Computer ConText is not responsable for any inaccuracies and mistakes in the contributions.

Think Twice before you criticize...

By Arnoud de Jong [issued: 1999]

The EWS64 Stress Control Panel exists for more than a year now. On the www.ews64.com location we have had almost a 100000 visitors. Together with the hit rate of the old location we already passed that figure.During that year there has been a lot of criticism on the EWS64, particularly on the software. With and without good reason. We are independent from Terratec, so it's not our job to defend their policies. But at the other hand we are in a better position to put things into perspective...

Almost on a weekly basis the frustration about the card and it's software pops up in the mailing list. We understand that frustration, but it's inherent on computer hard- and software. We won't repeat the cliche comparisons with other products on the market. We also won't emphasize the fact anymore, that even the BIOS of your mainboard needs a regular update to match the demands of the latest developments. In the same way Terratec people have to deal with all kinds of obscure other-party software that might sabotage their latest driver improvement.It's better to put things in a more historic perspective, so that you will realize what kind of audio/midi equipment you actually have in your personal computer. Here we go.

About ten years ago, musicians that wanted to work with synthesizers and computers, had to invest large amounts of money. I'm not good in 'inflation correction', so the prices I'm about to mention are Dutch guilders, converted to today's US dollars. As a midi computer we used the good old Atari ST, because it had a standard on-board midi interface. That was about $1000 to start with, if I remember correctly. Then we had to buy the early versions of Cubase. About $600 dollars. Now we come to the subject of synthesizers. At first I used to buy expensive synthesizers (like Korg PS3200 and Jupiter 6), each about $4000 dollars. No midi-interface at all or just a rudimentary version. Talking about frustration here: I had to change over my new Korg four times before I had a specimen that didn't die on the first day. The whole process from the first crash until the stable synth took about five months!

There was a huge 'inflation' in synthesizers. One year you spent thousands of dollars, the next year you were outdated by a newer model. If you were crazy enough to upgrade, you could trade in your 'old' synth for about one fifth of the original value. So I decided not to spend huge amounts of money on one synth anymore and I switched to a 'cheaper' thingy like a D10 (about $1300). Extra sound modules like the U220 and the first Roland Sound Canvas (together about $1800 I think) also were a better option. Those three I still have in my possession. They are good for a total of 384 different instruments (variations and drumsets not included). Although, different... Partly overlapping. Anyway, that's about eight dollars per instrument. Your EWS64XL (street price around $500 I think) costs about four dollars per midi-instrument. I guess they invented the buzz-word 'Total Cost of Ownership' for this, so let these facts get through to you for a moment...

The first generations of soundcards were no option for the serious midi musician. The sound of the Yamaha OPL chip was a laughter. Even the introduction of wave-tables wasn't the improvement that could match the soundquality of a real synthesizer. Above that the midi-interfaces of those cards were not 100% Roland MPU 401 compatible, which lead to all kinds of strange effects (yes, sure: with timing and regarding the transmission of sysex data if I remember correctly). With the introduction of the Turtle Beach cards some improvement was noticeable, but they were very expensive.

The first card that could match the synthesizer sound quality (to be precise: of a Roland Sound Canvas) was the EWSXL64. It's a public secret that the largest GS soundset that comes with the EWS in fact is a Canvas soundset.

With the principle of uploadable soundsets into RAM on standard SIMM-types (!), Terratec made a product that was very good value for money in comparison to what you had to pay a few years before. The musician got 2 midi ports (dig ancient Steinberg price-lists for multiple midi-ports for the Atari and compare), digital coaxial IN/OUT, optical IN for harddisk recording (again: dig old price-lists to compare with dedicated hardware in that field), the option for an extra synthboard and recently the MicroWave PC.

Even if you do not use all those options, the EWS64 is a very good musicians card. But you have to realize that -as other users mentioned earlier- that it's an 'all in one card', not a replacement for dedicated equipment. But suppose you just use it as a synth. It's still cheaper than a Sound Canvas. It's more versatile, because you can use multiple soundsets. Read our review on the Sonic Planet sample cd and you will realize that you have a potential in your computer that only a few years ago required a stack of very expensive synthesizers. The only thing you have to do is to sit and wait for the next high quality sample cd-rom. So you won't get tired on your EWS as quick as you would be on a normal synth. As long as ISA is supported on modern mainboards (which I doubt, but that's another story), your EWS64XL will not be outdated easily.

Of course there are reasons for criticism too. But that goes for every product. And with most products you do not have a forum like the mailing list to express that criticism, let alone that the manufacturer listens to it.

Let's make it clear once again that we are not 'defending' or 'promoting' Terratec. We just look from the consumer's and musician's perspective here: do we get value for money, does the manufacturer listen to it's customers, does he make efforts to correct mistakes, does he make efforts to improve the software, is there good support? The answer to all those questions is 'yes'.

The EWS64XL is not a product you will master in one day, one week or even one month if you want to explore all it's possibilities. You have to study things like midi, harddisk recording, sampling, basic audio techniques et cetera to have a good result. So do not blame or flame others if you're too lazy to RTFM or study the techniques you have to work with.

Ok, the user interface and underlying routing of the EWS is still a bit of a barrier for a more happy relationship with your card. At the other hand, while reading some of the hundreds of emails we get, or browsing the flaming in the mailing list, I can't help thinking: some people are so lazy and so spoiled... :-).

Non-Operating Systems

By Arnoud de Jong [issued 1998]

Is working with Windows NT indicative for someone's professionalism? That question occurred to me when I read some remarks in a discussion thread in the mailing list. The one about EWS-drivers, not being available for other operating systems like Windows NT and Linux. It kind of ended up in a discussion about which operating system was better. Maybe I'm gonna write some controversial things here, but remember: I can be such an ignorant, stupid fool sometimes...Let me put forward a few propositions.

>	Windows 95 has been developed for the 'consumer market'. That requires Plug and Play and multimedia support. At the same time the manufacturer can take the liberty of having a kind of permanent beta-product in the market (most consumers are dummies, you know). Windows 95 is a non operating system. Argument: any operating system that's not completely shielded against applications that can mould it to their own will, results in a 'non operating system'.
>	Windows 98 is the pupation from Win95 to Windows NT 5.0. Better protected, but still vulnerable. Anyway: it will keep the consumer buying.
>	Windows NT 4.0 is a NOS: a network operating system. Ok, an NT client can be used stand-alone, but it's still primarily made for use in a business environment. The EWS is (to my knowledge) not on the NT Hardware compatibility list, so in fact you can't 'demand' from Terratec or any soundcard manufacturer that it supplies more than a Soundblaster compatible NT-driver.
>	Linux (poor people's Unix) is also from origin a Network Operating System. That means: multimedia was not the first thing the inventors had in mind while developing it. Very stable, straightforward, no-nonsense, with a very direct relation to the hardware. In other words: it's a 'do it yourself' operating system. If you want to use it stand-alone, with a soundcard, no one will object. You only have to write your own drivers. Linux lovers will gladly do that, as long as you provide the hardware specifications. And here seems to pop up a small problem with that Dream chip...
>	Windows NT 5.0 has to be the synthesis between Windows 95/98 and the present NT version. By the time we got that, our EWS will be a bit obsolete.

My bold conclusion: a ROM-based operating system is the best. Some advantages: untouchable, stable, very fast and it saves RAM. In the past one disadvantage was the problem of upgrading: you had to change a chipset; a bug in the chipset was not easily correctable. But those arguments are no longer valid. For the consumer it's an advantage: the manufacturer has to make sure his OS is indeed a final release, not some kind of veiled beta. Furthermore the costs of replacement of such a ROM-OS don't exceed the costs of a Windows upgrade. If you are interested: have a look at RISC-OS, the ROM-based operating system of Acorn RISC computers. Then you will see how pointless a discussion about the quality of 'non operating systems' like Win95/98/NT is. And how much Mr. Gates nicked. We're just stuck with Windows because it dominates the market, but for the rest it's the VHS under the operating systems. And if you are still a fan of Win95 and completely disagree with me, ask yourself: do I know any other operating system that I have to re-install from time to time because it's polluted?

The fact that we have to live with Windows is dictated by marketing mechanisms. So everybody tries to make the best out of this situation. Some people choose for NT, because it's more stable and faster than Windows 95. Others choose Windows 95, because it's more tailored to the audio jobs they want to do. Both groups are in the right in some way. But either choice has nothing to do with professionalism, rather a lack of better options. There is no logic in the reasoning that you consider yourself an audio professional, so you use Windows NT, so it's crazy that a manufacturer has no NT-drivers for his soundcard (yet).
First of all: in my experience real professionals never sing their own praises. Second: if you need to earn your money in the audio business, you use dedicated equipment. The EWS64XL is a good soundcard, but it's a soundcard, if you get my meaning...

One dB

By Arnoud de Jong [issued 1998]

Last night one demonic Bird (dB) nearly finished me off. It was 3 AM. I had been working on this site for the whole day. Imagine: barely asleep and some stupid creature wakes you up with shrill and nervous whistling. I jumped out of bed and closed the windows, but the high pitched noise remained audible. Was this a bug in the software or the firmware? Mark how mixed up a person can get by maintaining a Stress Control Panel.
I stepped outside through the first door I could find. Luckily I have a balcony, otherwise I didn't live to write this down: our apartment is on the seventh floor. The dB was still busy provoking me. I started shouting the following statements into the darkness:"Ain't birds supposed to sleep as well at night?!" The dB turned up the volume.
"Do you realize that your S/N-ratio doesn't comply with modern regulations for audio equipment?" (Our apartment is situated along a rather busy road).
"Are you aware of the fact that your signal is 6dB louder than..."

By that time some men in white coats straitjacket me while the local television silenced me by sticking a microphone into my mouth. "Ahoo-i-em-ing-hooh??", I tried to say, meaning in fact: "Are you in sampling too?" But the men in white took me away. In the background I heard a last sound from the dB. It was a kind of crackling noise. I suppose the cat from across laid his claws on him. I had to finish this editorial in a small and (finally) silent room at the lunatic asylum.

So are you all satisfied now? Voilà: the new EWS Stress Control Panel! How did I end up doing all this? I just got the old one a bit on track. And now TT offers me a special domain for it. Had to start all over again! Without a little help from my friends. They are busy playing with some soundcard (EWS or something) or a girlfriend (blond or something), while I am trying to keep this site up-to-date. You know how this goes: you buy a new soundcard, it's too complicated for you, you read an email list, you can't remember all the solutions, you try to put some structure in it by making a deal with the other members of Computer ConText to work together in setting up a web-site, after a week one member meets a gorgeous Danish girlfriend, another one gets a midlife crisis and you end up doing things alone! Should I go on? And I don't even know how to write proper English!

Anyway, I am lucky that I have you, the other EWS-owners! You helped me out as long as the Stress Control Panel exists. And I beg you: keep doing that, please! 'Cause I don't even know how my EWS works, how Ed!son works, what you can do with it... I was too busy keeping you out of trouble with the d* thing! By the way, if you were wondering what caused the delay of Ed!son: that was me. I mailed Terratec almost every day to postpone it's release, 'cause I just couldn't take much more of all this..., how shall I put it... Never mind: enjoy the new Stress Control Panel and keep sending us (the cat from across and me) your questions and solutions.

Technical Information about Digital Interfaces

By Edgar Bültemeyer

In the email digest Sam Glass wondered if anyone had been able to get any hardware information from Terratec regarding their implementation of the coaxial digital in / out. For example, signal voltage levels, and in particular the voltage range."I was just reading about the AES3 specifications: the frame structure is the same but the voltage levels, (3-7v pk-pk), may be different. A company in America make cables to connect between the 7 pin connector on the bottom of my Sony DAT and the coax in & out. There's an active version which does all the signal level translation, and a passive that leaves the levels alone and apparently works with most equipment..."

Reason enough for Edgar Bültemeyer to elaborate on this topic. Reason enough for us to preserve his contribution for others who might need this information.

The overview
>	S/PDIF: Sony / Phillips or IEC598 Digital Interface. Peak-Voltage: 0.5 V Impedance: 75 Ohms
>	AES/EBU: professional Version of digital Interface used on Studio DAT´s etc. Peak Voltage: 5V Impedance: 110 Ohms
>	TCD7 and TCD8: the interface on SONY portable DAT´s. The signal FORMAT coming out is actually S/PDIF, but the voltage differs. It´s 5 V. This signal is electrically seen TTL.
>	TCD100: The same except the voltage. Because the TCD100 uses just 2 batteries, the voltage is 3V. This signal is electrically seen low voltage TTL.

The choices
>	Buy the optical cable
>	Buy the active coaxial cable
For the TCD7/8 (at least the ones I have found in the net):
>	The passive coaxial cable
>	Look for some TTL to Optical or Coax converter circuits in the internet and build them yourself (actually much cheaper).
>	What I did: Look for the DAT-page of Heiko Purnhage: http://www.fet.uni-hannover.de/purnhage/dat/dat.html (Note by CC: this link is dead nowadays).
Build yourself a connector for the 7 pin-plug and connect the signal-pins to the coaxial input of the EWS. I used small pins of PCB-connectors which are available very cheap at the local electronic store. Then some shielded cable , and 2 cinch plugs. All I can say is. that this worked for me (TCD100 - EWS). I don´t know if the EWS survives 5V (TCD7 and 8). Could someone try it out? ;-) PS: Please don´t blame me if you burn your DAT-Deck, Terratec Soundcard, Computer, Your house, your car or anything after trying out my solution. If anything like this happens or the whole universe is snoozed into the nose of a giant dog (like in Hitchhikers Guide to the Galaxy). It's not my fault!

The different file formats for Banks and Instruments

By Claus Riethmüller

The difference between the file formats
There are four file formats which are really useful and relevant for EWS users: TTS, TTI, 94B and 94I. There are also other file formats defined by Dream, such as 94K and 94L. Those are not really useful for the concept behind 94winst, Ed!son, Ed!son Micro and only partly for the Hoontech Editor (the formats are also invented as information for the developer of editor-software). These formats all have a different meaning.

You can divide the formats into two groups: TTS/94B and TTI/94I

The first group TTS/94B is a format for a set of samples (TTS=Terratec Set) which contains a whole bank (94B=SAM9407 Bank). It contains at least one, but theoreticaly up to 127*127 instruments in one file (127 programs on 127 variations). At the moment the information in TTS and 94B are exactly identical. I have shown this with my TTS->94B and 94B->TTS
convertor already (http://members.aol.com/cridi).

The second group TTI/94I is a file format which includes only one instrument (of course it can contain different samples/splits. The I stands for Instrument). This means that this file format is not bound to a program and variation number. I've never seen a TTI (will be invented with Ed!son Instrument Editor it seems) so I don't know if it will be possible to make a 100% equivalent convertor for this.

In short: One TTS/94B consists of at least one and up to theoretical 127*127 TTI or 94I files.

Reasons
Unlike 'crappy pseudo sampler cards' (like AWE series...) the EWS is not a card to edit banks and stuff with some samples and instruments in it . The EWS can and will be used as a real sampler. This means that you can edit every instrument directly if it is stored into your sampler's RAM.
This task will be performed by the Ed!son Instrument Editor. So Ed!son will primarily be a editor for one single instrument, like the 94winst-program. Ed!son will be different from the Hoontech Editor or the editors for Vienna and AWE cards. This is the main difference in the concept behind Terratec's and Guillemot/Hoontech editing. This all means that it will probably be useful for Terratec to store aditional data in the TTS file, which is not defined in 94B.
That's the reason why Terratec invented this format. Problems with filters, envelopes, clipping, portamento, splitts and so on have nothing to do with the file format. The basic parameters of the Dream are not changeable, not even by Terratec. If you have such a problem it can be firmware, software, driver, the weather, but not the file format!

Technical description
(If you're not a programmer you can skip this part )
The TTS/94B format is divided into different parts with a header each. Each header is declared with a 4 byte long RIFF-chunk. These are as follows (the file format contains a lot of pointers):

SAMP/TTSA	contains the sample data
PRIO/TTPR	priority data
RLOC/TTRL	parameter relocation
IMAP/TTIM	instrument table
PARA/TTPA	instrument parameters
( ^ / ^ )
(94B/TTS	this shows they only renamed it)

The 94I-Format is a format wich does not contain the sample (it's saved in a external file: either WAV or SMP format). I think Terratec will change this in TTI.

Conclusion
This means for us now: Ed!son Micro produces TTS files with one instrument. So if you want to edit/use it in the future together with other instruments, it must be possible to extract (convert is the wrong word here) this single instrument into one TTI or one 94I. The 94winst-program produces 94I files, which can be combined with other 94I files into one 94B set. So if you plan to use these instruments again in the future, we either need to have the posibility to edit the 94I directly (like 94winst does) or to extract the instruments from the 94B into TTI's or 94I's. The latter option is better, because we can edit files made by other people that way. The same goes of course for wavto94i-files.

This means for us in the future: the Ed!son Instrument Editor or a part of an updated Bank-/Setmanager will need to have the feature (and I'm 100% sure it will have it) to extract single instruments from a TTS and a 94B into at least TTI and maybe even 94I to be able to edit it.

DirectSound (DirectX5)

By Jörg Knitter

DirectSound is one part of the big Microsoft DirectX package. Aim of DirectX was to deliver a hardware independent interface for multimedia programming for the 32-bit Windows systems. Additionally, Microsoft wanted to replace DOS as the standard gaming platform.
Microsoft makes a distinction between DirectX Media and DirectX Foundation.

DirectX Media includes:
>	DirectAnimation
>	DirectPlay
>	Direct3D Retained Mode
>	DirectShow (formerly known as ActiveMovie)
These function libraries are based on the low level DirectX Foundation functions.

Apart from DirectSound the DirectX Foundation:
>	DirectDraw
>	Direct3D Immediate Mode
>	DirectInput
DirectSound delivers a hardware independent access to soundcards, i.e. it is not important if the soundcard supports 3D sound or sample ram. Only the CPU and memory usage shows the performance of the soundcard.

Static and streaming...
DirectSound distinguishes between static and streaming buffers. Static sounds are short sounds that can reside on the soundcard´s memory so that they can be played with low latency. If there is no or not enough memory on the soundcard, the sound is loaded into the computer´s memory and played and mixed using the CPU.
If the developer has a very time critical application, he can force the usage of soundcard´s RAM; if it is not available there will be an error message. Of course, it is also possible to force the use of main memory.

Streaming sounds are long sounds, like background music in games, that reside in the main memory. As they may be very large, only a little part of it is loaded into the memory. If a part has been played, the used memory area can be overwritten with the following part of the sound. So streaming sounds have a circular characteristic. You can compare this functionality with the hands of a clock. One 'hand' points to the play position, while the other 'hand' points to the starting position where the data can be overwritten up to the current play position. If the play position reaches the end of the buffer, it continues playing from the beginning of the buffer. So the play position can be before the write position of the buffer! Finally, playback without interruptions depends on the buffer size and the frequency of buffer-filling.

3D Sound
Besides these standard functions, DirectSound5 offers the possibility to add 3D sound. If this has to be done by the software (like with the EWS, as 'AudioRendering' is still not available), the CPU usage of an Intel Pentium 90 is at 5% per sound. The quality of this emulated 3D sound still has not the quality of a real HRTF, but the last one is more complicated and more CPU intensive. If you want to use it with Direct3D you can simply take the object coordinates and pass them to DirectSound without any conversion. Additionally, Doppler effects can be simulated with DirectSound3D. Delay or even reverb is still not implemented.

To get simple delay effects and high quality 3D sound and/or multiple channel output, you still have to wait for DirectX6 (planned for autumn 1998). You will be able to create something like channel maps, so that you can develop applications for multiple channels soundcards that also work on 'normal' soundcards. The support for streaming and recording sounds will be improved.

Additionally, DirectMusic will be introduced for the first time. The most interesting things are simple midi access and downloadable soundfonts.

If you want more information about DirectSound visit www.microsoft.com/directx<. There are two very interesting documents about how to program static and streaming sounds. Although both documents are based on DirectSound2, you will need the same code when developing for DirectSound5. There are still two points left that are not mentioned in the documents: Mixing with sampling rates other than the default 8Bit and 22kHz and 3D sound (introduced with DirectSound3).

Mixing with sampling rates other than the default
You normally do not have to create a primary soundbuffer. You simply create secondary sound buffers and send the play command. The sounds are mixed automatically with 8Bit and 22kHz. If you want a higher mixing quality, you have to obtain access to the primary sound buffer (do not create one!) and change it´s preferences. The following code shows how to achieve a mixing rate of 16Bit and 44kHz stereo. Notice that the buffer size of the primary buffer is set to 0 (dsbdesc.dwBufferBytes=0;)!

{
...
memset (&pcmwf, 0, sizeof(WAVEFORMATEX));
pcmwf.wFormatTag = WAVE_FORMAT_PCM;
pcmwf.nChannels = 2;
pcmwf.nSamplesPerSec = 44100;
pcmwf.nBlockAlign = 4;
pcmwf.nAvgBytesPerSec = pcmwf.nSamplesPerSec * pcmwf.nBlockAlign;
pcmwf.wBitsPerSample = 16;
ZeroMemory(&dsbdesc, sizeof(DSBUFFERDESC));
dsbdesc.dwSize = sizeof(DSBUFFERDESC);
dsbdesc.dwFlags = DSBCAPS_PRIMARYBUFFER;
dsbdesc.dwBufferBytes = 0;
if (DS_OK != lpDirectSound->CreateSoundBuffer(&dsbdesc, &lpDSB, NULL))
...
}

3D Sound
If you want to add 3D sound there is more to do. First of all, you have to add the parameter DSBCAPS_CTRL3D to the flags of the primary buffer (e.g. dsbdesc.dwFlags = DSBCAPS_CTRL3D | DSBCAPS_PRIMARYBUFFER). With the handle you get from the CreateSoundBuffer command (e.g. lpDSB) you can get access to the IDirectSound3DListener interface with the QueryInterface command (e.g. lpDSB->QueryInterface(IID_IDirectSound3DListener, (void**) &lpDS3DListener)). This interface has functions to change the position, speed and other preferences of the 3D listener. The lpDS3DListener from the last example must have the type LPDIRECTSOUND3DLISTENER.

Additionally, the initialisation of the secondary buffers has to be changed. The flag DSBCAPS_CTRL3D has to be set and the following line has to be added: lpDSB->QueryInterface(IID_IDirectSound3DBuffer, (void**) &lpDS3DB);. You have to make sure that you do not use DSBCAPS_CTRL3D and DSBCAPS_CTRLPAN as flags at the same time; otherwise you will get an error. Either you use 3D sound, or you use the panning capabilities of DirectSound.

Finally you must set preferences like lpDS3DB->SetMaxDistance(...)‘ or ‚lpDS3DB->SetVelocity(...) that are described in the DirectX SDK.

The DirectX SDK is free and can be downloaded for a limited time from www.microsoft.com/directx.

The difference between Decibels and Decibels

By Jorma Sauvala

The concept of dB (decibel) is a common source of confusion among audio consumers, mainly because dB's are happily used to describe two different things: voltage levels and power levels.

1.	If we are talking about electric voltage or digital sample values (i.e. what you could measure in your EWS64 line in/out connectors or what you would see in your Soundforge level display), the rule of thumb is indeed that a 6dB increase means doubling the original value: +6 dB = 2 x = 200%. Example: sample value 2000 is 6 dB higher than sample value 1000.
2.	If we are talking about electric power or acoustic power (as with loudspeaker output), the rule of thumb is that a 3dB increase means doubling the value: +3dB = 2 x = 200%. Example: a 200 Watt amplifier will give you 3 dB more acoustic power than a 100 Watt amplifier.

In case 1) the formula for dB difference between two values V1 and V2 is 20 * log (V1/V2).
In case 2) the formula for dB difference between two values P1 and P2 is 10 * log (P1/P2).

The 6 dB rule of thumb in case 1) comes from the fact that 20 * log(2 * V / V) = 20 * log 2 = 20 * 0.301 = 6 (dB).
The 3 dB rule of thumb in case 2) comes from the fact that 10 * log(2* V / V) = 10 * log 2 = 10 * 0.301 = 3 (dB).

As long as we are not dealing with power amplifiers and loudspeakers, we should stick to case 1.

Thus, for us:
+6dB = double the original value = 200%
- 6dB = half the original value = 50 % etc.

This is how SoundForge, CoolEdit and any digital editing software with dB scales work.

Assigning DirectX plugins to the midi-part of the EWS

By Oliver Hutz

This article is dedicated to the people who have fun with fiddling around with their EWS and Windows proggis (well, didn't we all have fun with it ... :-).

The way it's done...
This is how to assign DirectX PlugIns to the MIDI part of the EWS. This explanation uses WaveLab.

1.	Be sure you have a program that supports PlugIn-effects and process them to the 'live' audio input of a soundcard using an audio record device driver. E.g. this is done by Steinbergs WaveLab (you'll find a demo on your EWS driver CD-ROM). Check also if you have loaded a soundset to the EWS memory and start a sequencer or other 'MIDI-through' application.
2.	Connect only OUT-2 of the EWS to your mixing desk and set the output selection in the ControlPanel to B (to hear the direct out from the synth). Adjust also the OUT-2 volume to a normal level.
3.	Be sure you have at least 2 wave devices active in the EWS-synthesizer driver.
4.	The whole point of it is the internal routing of the audio INs and OUTs. This is done by the FX- and V-CHANNEL panels. First, turn all effects and sends off. The [Synth-RECORD] (upper right slider) has to be set to [MIX]. Then turn on only the [MIDI EQ/V-SPACE (RECORD) SEND] (upper left). Set the WAVE-part to [4CH.] but let the EQ/... SEND turned off. Open the V-CHANNEL panel and set the second device to the rear OUTs (OUT-2). BTW: you should save this configuration as a setting for the EWS-ControlPanel (.mix-file).
5.	Start WaveLab and assign the playback to EWS-WAVE-PLAY #2 and WAVE-Record. Depending on your overall system performance, set the buffersettings to the smallest possible values (note: use also the buffer latency feature of the EWS-Synth driver). The overall latency has to be adjusted to a minimum without dropouts of the wave stream, usually around 40-50 ms. BTW: The function Autostop if drop's out (German: Autostop bei Dropouts) should be turned off.
6.	Choose the plugin(s) you want to hear, activate the Live-Input in WaveLab and press the play button. Now you can hear the MIDI instruments with (depending on your plugins) cool effects!

This application should work with other programs, too. I didn't test VST's live monitoring, but it should work as well. The upcoming WaveLab 2.0 will support VST effects, too. Tip: download the brandnew Transmitter-plug in from Steinbergs website and use it (with a little delay and reverb) on pads.

The Firmware

By Claus Riethmüller

Firmware? Firm where?
First of all I'll explain the normal usage of the word 'Firmware'. Modern signal processors (DSP's) can be described (theoretically) as a small 'computer' or processor with an input and an output device. Some signal goes into it, is transformed (or not ...), and goes out of it again ... This 'transformation' is made by a small piece of software (a type of Operating System) which is called 'Firmware'. You will find firmware-data in a lot of electronic equipment such as in any SCSI-devices, telephone-controlling systems and even modern washing-machines. And oh wonder: also on modern soundcards.

The French Dream
The EWS64 has a Digital Signal Processor (=DSP) developed by a French company called Dream (http://www.dream.fr). Terratec calls this chip in his manual 'synth'... The complete name is Dream SAM9407 (I will call it 'Dream' from now on). There is also another DSP on the EWS, the Cirrus Logic Crystal (called 'Codec') but this is not really interesting for us now.
Back to the Dream chip: This DSP is a single chip that uses one single interrupt of the PC-ISA-BUS and it uses a MPU-compatible (MPU is a older Roland standard for a midi port in a PC) port-address to establish communications with the software. Internally the Dream is able to address up to 64 MB of RAM (it needs at least 1 MB to work).
You can connect different digital in- and output devices to it such as a MIDI ports and it has 2 stereo digital in- and outputs. On the EWS one of the MIDI-Port, the Codec and (partially switchable) the 18 bit Philips DA/AD-converters are in default connected to the Dream.
What we have until now: a chip build into a system that allows to get Codec signals, digital and analog audio signals and MIDI signals in and out of the chip and the chip is able to read and write from an two the memory (on the SIMM Module of the EWS). Not more.

The default Firmware
Now comes the part of the Firmware: the Firmware is a piece of software that processes the input and output of these signals. How this is done is all developed by one (or a lot ...) programmers at the Dream company and by Terratec.
If companies like Terratec (or Hoontech and Guillemot) decide to use the Dream chip on their soundcards they are getting a default Firmware with it's source code and some development tools. Maybe they are even getting a useable documentation...
The default Firmware of the Dream chip implements all the main functions: MIDI (Wavetable + FM synthesis), MOD, WAVE, AUDIO-IN, CHORUS, REVERB, 4CH, PITCH-SHIFTER, FILTER, SURROUND etc.

Modifications
Terratec made some (not too much ...) modifications to the firmware. For example they invented the compressor for the clip on/off function, they increased the amount of max. wave channels from 8 to 32, they are using 16 of the 32 internal Dream MIDI channel for additional things like the MicroWave PC extension etc. One of the more important things: they increased the max. size of looped samples from 256kb to 512kb. They are even planning to break the 512kb limitation (by getting rid of the FM-algorithm it seems). Most important change: they changed the complete MIDI-Controller structure. While in the original Dream Firmware nearly everything is controlled by NRPN's (Non-Registered Program Numbers), they changed the structure so that it matches the structure of other Terratec soundcards (Maestro for example). This seems to be the reason why there are up to today still some problems with real-time editing of parameters.

Not where you think...
The Firmware itself is stored in the RAM of the soundcard. Not in the EEPROM as a lot of people think. The EEPROM on the EWS64 only holds the PnP configuration data (IRQ's, ports, etc.). This means that the Firmware needs to be loaded into the RAM at every startup of the system. It's saved on the Harddisk in the File EWS64_OS.BIN - make your you always have the latest version of it on your system and no copy of an old one in C:\EWS64 or in the Windows-dirs.
Now, how does the Firmware perform it's task? The only connection from the PC to the EWS is the MPU-Port address (default is 330h). Over it the PC is able to read and write Data from and to the Dream chip. Some parameters are going directly through to the MIDI ports and some are accessing the Memory. Every Data that is used by the Dream chip (playing Wave-files, samples from Mod files, samples from soundbanks) must be stored in the RAM to be played. If you are playing a wav file the data goes through the driver to the MPU-port and is stored in the Dream RAM. Then it is played through the Dream Chip. You can notice this by playing small wav files while looking at the memory-amount display in Setmanager.
The driver is the interface between the application and the MPU-port (=Firmware in this case). All data from the driver needs to be sent to the port and answers from the soundcard are read from the port by the driver (same backwards ...). It's a bit strange that Terratec has implemented no way to access the MPU-port directly. This is able on other Dream based cards and there are some (not much) functions who would need that and only can be accessed by DOS(!) applications now ...
I wonder why they did not implement such a function (or is it just not documented?) since when they are releasing other Firmware files (such as the Vocoder Firmware) they need to update the whole driver to be able to make a front-end application (if it won't be a DOS application).

Future
In the future it will probably be possible to load and unload the Firmware while the system is running (not only on reboot). This will allow you to switch between different fx algorithms without rebooting.
The only thing I am (and others, too!) are missing now, is: a Firmware developers documentation with diagnose and development tools (come on Dream!).
This was long and some times a bit technical... But I hope that it made some things about the mystical 'Firmware' more clear.