Dynamic Noise Reduction and the DNR System 911 Overview

You are here: greyghost.mooo.com > Features > Dynamic Noise Reduction/DNR System 911
Dynamic Noise Reduction and the DNR System 911

This page is about the National Semiconductor Dynamic Noise Reduction System, as well as the DNR System 911 noise reduction component and other, similar devices. It also gives an overview of competing noise reduction systems and how those systems operate. As such, it's fairly long and not all of it may prove interesting to you. Feel free to use the links below to jump to the section you'd like to view.

In case you aren't sure you should be here: DNR isn't another name for "Dolby Noise Reduction". This page has no information pertaining to a "department of natural resources" or "do not resuscitate" orders. A different Dynamic Noise Reduction process may be applied to video, but its use and method of operation are completely beyond the scope of this page. This video related process has nothing to do with the National Semiconductor system discussed here.

Table Of Contents:
Introduction
Dolby B, C, S, HX (Pro), How It Works
dbx Noise Reduction
High-Com / High-Com II
JVC ANRS / Super ANRS
Other Approaches
Dynamic Noise Reduction
What DNR Is and How It Works, How to Get
DNR System 911, How to Use, Effectiveness
Other Equipment with Dynamic Noise Reduction
Other Thoughts

Introduction

Noise reduction techniques are a relatively recent development in audio playback and recording. Many of them showed up around the same time magnetic audio tape started to become suitable for use in high fidelity music capture and reproduction. All of the most popular noise reduction systems work by pre-processing the source material as it is recorded. The goal of any noise reduction system is to maximize the quantity of desired signals while minimizing noise as much as possible.

The following discussions lend some insight into how noise reduction systems requiring preprocessed source material (unlike DNR) work.

Dolby B Noise Reduction

The most common noise reduction system (at least for cassettes) is Dolby B noise reduction. Practically every lower mid-range to high end piece of cassette tape playback or recording equipment used to be equipped with a Dolby B decoder. Even some portable equipment can decode Dolby B noise reduction. It is my understanding that Dolby Labs refuses to license this technology for use in any currently manufactured cassette deck or player. Some manufacturers have responded by including the still-available (and license-free) DNR system that is the focus of this page.

A few things helped to make Dolby B noise reduction popular. In particular, Dolby Labs worked hard to get it out there, and to make it attractive for tape transport manufacturers to include it with their products. Many single-chip Dolby B encoder/decoder implementations exist and this helps reduce the added cost of including Dolby B noise reduction. Yet most importantly, Dolby B encoding generally sounds acceptable--if rather bright--on equipment that cannot decode it. A slight reduction in treble can be applied to such tapes if the listener desires to do so, but most people won't and not all equipment has tone controls. In fact, some people prefer the slightly "brighter" sound.

Dolby C Noise Reduction

Dolby C noise reduction works by placing two Dolby B encoder/decoders back to back with one another. More audio frequencies are also processed by the Dolby C noise reduction system. This results in the elimination of more noise from a tape than Dolby B encoding manages, yet it sacrifices acceptable playback quality on equipment without a Dolby C decoder available. Some people claim that Dolby C encoding is alignment sensitive and really only works well on the deck where an encoded tape was first recorded. I tend to agree with this assessment of things.

Dolby S

Dolby S aims to solve the shortcomings of Dolby C and it's said to be tolerable on equipment with only a Dolby B decoder. This video gives a stunning demonstration of the possibilities, if you don't mind the fact that the deck used has some speed regulation issues. It's said that Dolby S encoded tape will give signal-to-noise performance rivaling that of a CD.

In recent times, I managed to acquire an Optimus (RadioShack house brand, built by Pioneer) dual well cassette deck with Dolby S noise reduction. My own tests have proven that Dolby S is extremely effective at combating tape noise, even on lowly normal bias cassette tapes. Playback on a deck with only Dolby B-type decoding is possible and the results were acceptable to my ears. Without a Dolby decoder of any type being available, you will hear undesirable artifacts of the dynamic range compression process.

Dolby HX

While on the subject, Dolby HX is not a noise reduction system per se. Dolby HX is a headroom extension system, built into the biasing circuit of a given tape deck that makes it possible to drive the recording level up higher than would normally be possible. Most tape decks work on a fixed "one size fits all" bias, and some have a bias adjustment that can be set (sometimes with the aid of built in calibration circuits) for each type of tape you use. You could think of Dolby HX as a sort of "dynamic bias" adjustment system. Better tape decks will also allow the tape type to be selected (or determine it automatically when a tape is inserted) so the bias circuit can be set in a roughly appropriate area for the tape type in use.

Bias, of course, is what makes recording audio to magnetic tape possible. The biasing mechanism applies a signal to magnetic tape as an audio signal is also recorded. This enables the tape to offer acceptable performance when capturing audio signals by altering the way the magnetic particles behave. Low end cassette recorders use DC bias, which while suitable for low fidelity or voice recordings, may produce a "gritty" sound when music is recorded. Better cassette decks and recorders use AC bias--an ultrasonic signal applied to the tape to effect the desired behavior from the magnetic particles.

How Dolby Noise Reduction Works

All of the Dolby noise reduction systems are companders, a combination word that means compressor and expander. A compressor limits the dynamic range (difference between the loudest parts of a recording and the softest parts) while an expander attempts to "widen" or even restore a previously compressed musical program to its original dynamic range. As an example, classical or orchestral music performances are almost always uncompressed when recorded and have a very wide dynamic range. Other types of popular music are typically more compressed, as are radio broadcasts. You can usually roughly "gauge" how compressed a given musical program is by how much you are tempted to adjust the volume control while the program plays. If you find yourself wanting to turn the volume up during a quiet section of the performance and to turn it down when a loud section comes on, what you're hearing is not compressed (or compressed very lightly).

The compression technique has been abused by some, leading up to what many refer to as the "Loudness Wars".

Dolby Noise Reduction works by compressing especially quiet passages on tape and boosting their levels, almost exclusively at the high frequencies, as that is where many tape formats need the most help. This is done while the tape is recorded. At playback time, a suitable Dolby decoder does the opposite--it cuts the high frequencies and compressed musical passages back down. Done right, this reduces tape hiss considerably. Done incorrectly, or on a machine that is cheaply made or suffering from problems, Dolby noise reduction can "crush" the sound and make it sound muddy. During louder passages, and with lower frequencies less affected by tape hiss, Dolby noise reduction backs off. (The point at where this happens on your cassette deck is marked with a "Double D" symbol on the recording level meters. It's usually around +2db or so.)

Compression in this sense doesn't refer to digital encoding of audio data to make it smaller. It refers to a reduction of the dynamic range (the difference between loud and quiet sounds) on a recorded medium or live source. This sort of compression is sometimes referred to as dynamic range compression. It can be done in both the analog and digital signal handling domains.

dbx Noise Reduction

Another noise reduction system is known as dbx. dbx (yes, it's not capitalized) noise reduction was much more aggressive than anything Dolby Labs ever offered to the consumer tape recording market. As such, it had one big problem--equipment without a dbx decoder would not give acceptable playback results when playing a tape encoded with dbx noise reduction. dbx did many things differently than the Dolby methods. First, dbx was applied to the entire frequency spectrum (low to high). Contrast this to Dolby's processing of only the higher frequencies. dbx also compressed the program material much more aggressively...often to the point of distortion, which will be revealed upon playing back an encoded tape with the decoder shut off. When played back, an expander restores the signal to what should be its original state, suppressing tape hiss and noise in the process.

dbx found more use than just with magnetic tape. Some records were encoded with dbx, and when played back through a suitable decoder, the results are spectacular. Pops and clicks can just disappear entirely. dbx was also selected as the noise reduction system of choice for the MTS ("stereo") analog television sound system used in the US and many other countries following the NTSC standards. Yet dbx never enjoyed the success that Dolby's noise reduction methods did in the consumer market, because it was more expensive to implement and very sensitive to deck alignment and quality.

dbx unfortunately also had a tendency to sometimes cause "breathing" upon decoding. "Breathing" refers to an effect generated when the compression process would change the volume level rapidly, as it could when a passage went from loud to soft or the other way around. This would cause the noise floor to "move around" and become more noticeable.

High-Com/High-Com II

A third type of system that enjoyed some popularity in Germany (at least) was the High-Com system from Telefunken. I've never heard these in action, though what I have read is generally positive. I have also seen claims that High-Com outperformed Dolby B in some cases. Various revision of High-Com were produced.

ANRS / Super ANRS

For a number of years, JVC marketed tape decks featuring a noise reduction system known as ANRS or Super ANRS. The ANRS system is basically equivalent to Dolby B type noise reduction, while Super ANRS was supposedly similar to Dolby C type noise reduction. It has been stated that the ANRS system was developed by JVC to avoid paying licensing fees to Dolby Labs.

I don't know much about this system, as I've never owned a tape deck made by JVC.

In light of Dolby Labs' seeming refusal to license their existing noise reduction systems for use in new cassette decks and players, I have wondered if JVC might still be willing to license either of the ANRS approaches. It seems likely that knowledge of ANRS within JVC (now JVC-Kenwood) has probably been forgotten over time.

Other Noise Reduction Approaches

There are other methods that can be used to reduce tape noise, without being considered true "noise reduction" systems.

Some of these are outlined in the following table, along with an explanation of potential advantages or drawbacks.

Approach	Advantages	Drawbacks
Wider tape tracks Higher tape speed	Better fidelity, lower noise.	Less recording time, any transport issue stands a great chance of ruining the tape beyond all repair.
Helical Scanning	Extremely high fidelity (as in VHS Hi-Fi systems). Tape runs slowly compared to other systems. (Helical scanning makes the tape appear to be moving at high speed from the perspective of the recording electronics.)	Transport complexity, imprecise editing, time for transport to start up.
Digital Recording	Extremely high fidelity. Fairly resistant to degradation.	Lack of popularity, potential transport unreliability, limitations of the sampling rate/depth, limitations of the ADC or DAC. Few or no new machines being produced.
High-Bias Magnetic Tape	Greatly reduced noise floor (and improved signal to noise ratio) as compared to any normal bias tape. Metal (type IV) tape can be nearly silent without need for noise reduction.	Lack of popularity, high cost, generally only compatible with decks that can adjust their biasing and equalization preamplifier circuits to compensate for the differing characteristics of the tape. High-bias tapes can be hard to come by. Type IV tape is no longer made.
Other Mediums (CD)	Nearly perfect sound reproduction, no degradation on repeated playbacks.	Not as easy to create as a tape recording.
Other Mediums (Mini-Disc)	Almost as convenient as tape, great editing functions, very good audio quality in most circumstancs. No hiss noise. Disc and track titling functions, reusable media, portable.	Never caught on, new players are not being produced and some have reliability issues. Audio data is compressed on most models using lossy ATRAC compression.
Computer Recording	Nearly perfect sound reproduction, capable of extremely high quality audio capture. Can record for a very long period of time or from many different inputs at once.	Uncompressed audio takes up a lot of disk space, and it isn't always convenient or possible to take a computer where the sound you wish to record is located. Computers are typically not as rugged as magnetic tape formats.

There are some people who find some or all noise reduction processes to produce intolerable changes to a recorded program. All the popular noise reduction systems do in some way change the music, and sometimes it may not work out well for many reasons--the type of program material, listener sensitivity, operator errors, faulty circuitry and no guarantee that a suitable decoder will be there at playback time...

Enter Dynamic Noise Reduction

Dynamic Noise Reduction Logo (a registered trademark of National Semiconductor Corporation)

What if you could devise a "perfect" noise reduction system that could work anywhere, with almost any type of source, even one that wasn't pre-encoded?

Dynamic Noise Reduction works differently from the other noise reduction systems described above. It is a playback only system, meaning that no pre-encoding is required or even possible.

Although Wikipedia's article on the subject claims that Dynamic Noise Reduction (hereafter referred to as DNR, at least most of the time) was based on the non-patented Philips Dynamic Noise Limiter system, I'm not sure how true this is. National Semiconductor produces an integrated circuit that offers the DNR system and its datasheet discusses two (now expired) patents: # 3,678,416 and # 3,753,159. Although the patent landscape is different today than it used to be (and subject to much more abuse in recent times), I'm reasonably sure that anyone who came up with a system as potentially game-changing as the DNR or DNL systems could be, would have protected it with a patent.

Various National Semiconductor documents discussing the DNR system claim that its use is royalty free, an interesting circumstance for a system that utilized patented technologies.

The patents, both issued to Richard Burwen, describe the system in "legalese".

What and How

As already mentioned, DNR does not require pre-processing or encoding of the source material. In theory, it can be applied to any audio source in need of noise reduction at any time. In its most common implementation, DNR functionality is supplied by the LM1894 IC. An LM832 (more suitable for low level signals as might be found in "portable equipment", though it could possibly be used in a phono preamplifier stage) was also produced.

If you have the time, I'd recommend stopping right now and reading National Semiconductor's application notes for DNR, specifically AN-384, AN-386 and AN-390. These are very well written and reasonably understandable documents that explain what is going on, as well as the issues faced and how they are mitigated. Some of this document is based on those, or on material contained in those documents, so please do read them if you can make the time to do so. (I was recently alerted to these links having decayed, and have since repaired them accordingly. Thanks, Kevin!)

What the DNR system does is to reduce certain types of noise, particularly hiss from tapes.

How it does this is very simple, yet extremely clever at the same time. There are a few different things going on here. Let's talk about each one.

First, and most simply, the DNR system is a variable low-pass filtering system with a fast attack time. A low-pass filter is used to filter out everything above its set point, and a high-pass filter does exactly the opposite. "Attack" refers to how quickly the system can disengage and cancel its filtering effects, after which the system will eventually release (resume its noise filtering activity). The attack time can be varied by external component selection, but should generally be set to a low value for quick operation. This process is known as noise gating and is commonly used in recording studios and live sound systems so that a microphone or other input will only accept sound under certain circumstances, such as a drum being hit. A hold time can be optionally specified to prevent sudden changes to the filtering action that might result in audible artifacting or distortion of the incoming signal.

The following graph shows how a noise gate would work when an incoming signal trips it, causing the gate to open. You'll see why this is relevant in just a bit.

Noise Gate Method of Operation (credit: Wikipedia, public domain)

(image credit: Wikipedia, image placed in the public domain by its author)

In cooperation with this "gating" action, the DNR system adaptively low-pass filters the incoming signal to reduce noise. It is tuned to work from around 1kHz up to 30kHz (well beyond the commonly accepted range of human hearing which ends around 20kHz). When the music is at a low level, or consists mostly of low frequency sound that would not be affected by its noise reduction process, the adjustable low-pass filter is implemented at lower frequencies depending upon whatever is currently being played as well as the "sensitivity" factor (of which more later). This blocks the high frequency components of the audio being run through the DNR system, along with the hiss (since the hiss typically has a lot of high frequency energy). If high-frequency program material comes along, whether loud or soft, the low-pass filter is dialed back to allow higher frequencies through the "gate".

As with many other noise reduction schemes, there are two signal paths in the LM1894 circuit. One is the audio path, which is affected by the low pass filtering process. The other is the control path, through which unmodified audio is passed. As its name implies, the control path is responsible for operating the filtering mechanism when it's appropriate to do so and deciding "how much" low-pass filtering to apply. (National Semiconductor is careful to point out that unmodified audio is all that should be fed to the DNR IC. Putting it after any tone controls in the circuit will alter its behavior in an undesirable way.)

A single control path drives both channels of noise reduction to keep the stereo image from "wandering" between speakers.

Those of you who are familiar with human hearing may have picked up on the other aspect of the DNR system's operation. DNR takes advantage of knowledge about just how it is that the human hearing system works, or the concept of psychoacoustics. Without our even knowing it, our hearing and brain perform a ton of processing on everything we hear. It's nothing less than amazing. Some of what we do every day with what we hear could not be accomplished externally without significant computing power and expense. Yet it is there all the time, working for us. Stop and think sometime about all that you hear and how it influences your actions without your necessarily even realizing what has taken place. Your brain can hear a sound and listen to the little differences in timing, phasing and other factors to determine where that sound might have come from, or what should be done when it is heard. It is also capable of "tuning out" interfering sounds to focus on a desired sound, even under particularly adverse conditions.

Like I said, it's nothing less than amazing.

The DNR system takes advantage of one particular aspect of psychoacoustics--that a louder, more desired tone will tend to cause the unwanted noise to escape your perception. Consider, if you will, loud music being played from a sound system or whatever--maybe even a live band. With a loud noise source in the area, you don't notice as many background sounds. Cars passing outside, people talking, the air handling equipment cycling on and off...stuff like this isn't even noticed unless its volume can match or exceed that of the dominant sound source.

Now let's move to a quiet house. You can clearly hear the trees moving in the wind outside, other noises the wind might make, cars passing outside, the cycling of your furnace or air conditioner and perhaps even the noises your own body makes. Whatever the noises are, you will probably hear them. With the loud music example earlier, those noises disappeared. They were still there, you just didn't notice them.

Tape hiss and some kinds of unwanted noise work much the same way. Compared to loud passages in a musical program, tape hiss takes a back seat and is overlooked by our perception. Therefore, during a louder or more intense passage of music, the DNR system can open up its low pass filter and increase the audio bandwidth. The hiss is there--and not being filtered at all--but you don't notice it! When the music dies down, of course, the hiss will be back and probably very present in your hearing. This is where the control path in the LM1894 reacts and closes up the high end audio response, where hiss and other unwanted noise typically lives.

Of course, this could cause a problem. Say you have a quiet passage in the music where there is a lot of occasional sound energy at high frequencies--a high pitched solo instrument, perhaps. The performance might have gaps of silence between notes. The LM1894 can deal with this as well, by exploiting more of the way our hearing works. Provided the low pass filtering of the sound is reversed quickly enough or prevented from fully kicking in again, this will escape our perception. The LM1894 also has the ability to leave its low pass filter somewhat open on a sort of delay or "hold" mechanism. Between these two systems, the result should be sounds that do not appear to have been "damaged" or that are allowed through the filter. A sensitivity parameter controls this mechanism, and adjusts the response of the LM1894 to a hopefully appropriate point for the type of material being played back. Most DNR systems have fixed sensitivity, something which I will discuss later.

This does mean that the types of noise that can be reduced by the LM1894 are limited. I'll talk about this more a little later. It does well with noise that might be incurred as a result of tape hiss, interference to a radio broadcast or the audio portion of a TV program. It really cannot deal with noise from clicks and pops in a vinyl record, unless they are concentrated in the high frequency range. It is also not likely to help a recording made with poor equipment--no noise reduction system can ever put back what wasn't there to start with.

Sounds Good or "I'd Like To Try It". How Do I Get It?

I'll start with the good news first. The LM1894 is still produced and available today. It's not terribly expensive. There are plans for circuits (warning: potential pop-up/under ads) online that you can build if you have some basic electronics experience. You might even have some luck with an eBay search. Maybe you could ask a friend who is handy with electronics (really handy, not just someone who says they are) to help you out in exchange for dinner or something.

If you want an off the shelf, ready to use product, I'm afraid that you may have a lot of looking ahead of you. There have been some mass-produced devices made with the DNR system implemented as a primary function or "extra".

General Motors (specifically, their Delco Electronics division) was one of the few companies to put DNR in a consumer product. Many of the Delco car radios produced in the late 1980s have a DNR circuit built in. It can be turned on readily with the push of a button. You might look at your own set to see if it has this feature. Some later sets (probably produced in the 1990s) have an "Auto DNR" designation on the front panel, implying that some circuit logic exists to enable the feature when it would help and turn it off when it would not. The AM/FM radio used in a 1994 Chevrolet pickup truck has a "DNR" character in the display (complete with fancy font!) but it has not ever illuminated in the vehicle's life so far and I doubt the needed circuitry exists for it. (However, the optional matching tape deck claims to have "Auto Dolby B Noise Reduction" functionality for which no indicator exists or illuminates, so anything is possible. The owner's manual never says anything about DNR at all.)

I'd love to know if your GM car radio has a DNR indicator in the display and if it has ever come on. You know what to do.

So, if you have a Delco car radio in your vehicle, look at the front panel and see if there is a DNR button or designation. Turn it on and give it a try if there is. I don't see it making much difference in a car environment, but you never know and it costs nothing to try.

Chrysler is known to have implemented the Dynamic Noise Reduction system in some of their car radios from the 1980s.

If you have varied source material played from different devices, you need a standalone implementation Dynamic Noise Reduction system. If you're handy with a soldering iron, that's not a problem. If you're not so electronically inclined, your options are rather more limited. Here are a few of the units I know of. Other standalone devices implementing DNR circuitry may have been made available, and there are devices which use a similar approach without using the DNR system or components.

The DNR System 911

This is the DNR Dynamic Noise Reduction System 911. It is a DNR system with some user adjustable parameters, a bypass switch to take it out of the audio circuit and an "enhancement" function.

Before I talk about this thing too much more, let me say that I would like to know more about this thing. The information I have is pieced together from the LM1897 datasheet, random web searches and experimentation. If you have some information on this--no matter what it is--I would love to acquire it and make it available here (if at all possible). Please, please contact me! I will gladly credit your submissions or you may remain anonymous.

It was made by Advanced Audio Systems International, Incorporated, seemingly once located at 4040 Moorpark Avenue in San Jose, California. They do not appear to be in business any longer. It bears serial number 08337 and appears to have been made in mid-1981, around the 30th week of the year or so. (If you know anything about this company, or can offer any information at all, please do contact me!)

I can say that the company was incorporated around September 25th, 1980 by Charlie Smaltz. Mr. Smaltz may have had some affiliation with National Semiconductor as their marketing manager for consumer linear ICs. AASI was not far from the National Semiconductor headquarters. I believe he may presently work or have worked for Micrel Semiconductor in more recent times.

I'm not sure when or why the company shut down, but my guess is that it failed, given that the California Secretary of State's office shows the business entity's status as "suspended", and defines this status as follows:

"Suspended or Forfeited: The business entity's powers, rights and privileges were suspended or forfeited in California 1) by the Franchise Tax Board for failure to file a return and/or failure to pay taxes, penalties, or interest; and/or 2) by the Secretary of State for failure to file the required Statement of Information and, if applicable, the required Statement by Common Interest Development Association."

I had no idea such a thing existed until I watched some videos on Youtube and stumbled across it. This set off a long running search for one on eBay. I finally found one. Of course, I paid too much and a much lower priced example showed up afterwards. (S'okay, though. I accepted this sort of thing a long time ago.) I'm told that other companies built devices including the DNR system along with other functionality. Let's have a look inside it.

Although this unit has a serial number and a company name on it, I can't help but think that it was originally assembled from a kit. There is just enough imperfection in how it's put together to make me wonder. If these were "professionally" produced, they were almost certainly hand-assembled. Assuming the serial numbers are consecutive and that these were factory assembled, anyone putting together eight-thousand plus of these by hand has my instant respect. Looking closer at the circuit board will reveal a few more interesting things.

Please click the picture above for the (large) original if you wish to see the component IDs. Even if you view only the small image, you can clearly see that the PCB was hand-drawn, due to the softly curved traces. It's also interesting to note that the word "component" (over by the LM334 current regulator) is misspelled and some of the letters in other words are rather unevenly applied. You will also note that for some reason, there are connections on the PCB for 110 volts AC. I'm not sure why this was done, as it brings the potential to introduce 60 cycle hum into the system. I don't think the power switch is directly connected to either side of the AC line, but I strongly suggest unplugging the unit before you disassemble it. The power cord is not polarized.

There are two LM1894 ICs onboard. As the LM1894 operates on two channels, I believe these have been cascaded for greater (up to 18dB) noise reduction. A 1K potentiometer from Mouser Electronics (another strong hint that this was a kit?) rounds things out. As for the front panel switches, the leftmost one is the power switch. The middle switch puts the DNR system in and out of circuit, while the last switch selects a high or low "enhancement" level. I believe this to put some "sparkle" back into the sound that might be lost by the DNR process.

Blinkenlights!

If you really must have or only care about the number of "blinkenlights" in your stereo system, the DNR System 911 can certainly help you.

That's right. No less than twelve LEDs are used to demonstrate what the LM1894's low pass filtering is doing at any time. The indication shows the overall audio bandwidth at any given time, as fast as the LM3915 lamp driver can go, from 800Hz to 30kHz. There is even a power LED.

Needless to say, the effect is striking, especially if the indicator happens to "scale" from one end to the other. Higher frequency percussion will usually produce a nice "scanning" display at each hit.

Taking It Apart

I suggest you don't, if you care about the health and wellbeing of your DNR System 911 device. If you simply must, there are five Phillips screws that hold the top cover on. After you've removed them, set the "Sensitivity" knob to either of its extremes (note which one you chose), undo the setscrew holding it to the control shaft and pull it off. Set it aside safely and carefully use a pliers or wrench to remove the nut that holds the control shaft to the front cover. Really, be careful. You will mar the finish if you're not. When the nut has been removed, the cover will slide forward to reveal the internals. The wood sides remain attached to the unit.

When putting the unit back together, take care not to crush or mangle the power LED.

How To Use The DNR System 911

Caution: this information has been more or less reverse engineered from National Semiconductor's documentation, as well as random experimentation to see what it did with various musical sources (all of them tape based). I don't guarantee that it is 100% accurate. If I should get a manual, I'll post the official instructions here. Although I don't see it as terribly likely, damaged audio equipment is not my problem. These instructions are presented in good faith.

National Semiconductor suggests setting the sensitivity control by playing a blank, biased tape. (Should you need to make one of these, just use your tape recorder to record a tape with no audio signal on it and the input level controls set to minimum.) While you are playing such a tape, turn the sensitivity control to the point where the tape hiss sound starts to vanish. You may want to use headphones for the most precise adjustment. Turn the control past this point until the hiss starts to brighten a little.

I set the system this way and it did work. Yet I found that setting it while music was playing tended to work better. You want to shoot for a fair balance between the fidelity of the music and reduction of the background hiss, all the while being careful not to "step on" the music or "crush" its high frequency components. You also want to avoid muddying the sound. In general, the higher up you turn the sensitivity control, the more high frequency program material will be allowed through the low pass filter.

If all you want is a "quick set"...I found that a setting between two and three (or slightly more) on the dial worked well for normal bias tape. Going as high as four to six was possible with a high bias, high quality cassette tape as the source material. Using another form of noise reduction along with the DNR system also allowed for use of an increased sensitivity setting.

Other Devices Using or Featuring DNR

There are other other devices which build in the DNR component, either as their primary function or as a "bonus" feature. There is at least one other device from Advanced Audio Systems International (the DNR System 450, mentioned on page 2 in the March 15th, 1983 issue of the Chicago Tribune and reviewed in "Videotests" from October 1981), along with a few entries from Radio Shack (who else?) and the Burwen/KLH DNF-1201A. Most all of these were produced in the early to mid 1980s. Some of them are much more common than others. The RadioShack devices are likely to be the most common of all.

I finally found a DNR System 450. It's a simplified, single input only version of the DNR system. That's really all I can say, since the one I have is pop-riveted together.

Various MTS capable television tuners implemented the DNR system, including Recoton's FRED and FRED II models. These aren't of much use to those wanting to apply DNR to anything other than a television broadcast, unless you were to modify the device. Many people won't even be able to use these devices for their intended purpose, with the retirement of full power analog TV broadcasts. That does mean you could pick one up cheap and try your hand at modifying it.

I don't know what the any of these units are worth on the open market. The ones I've seen have been all over the place pricewise...some reasonable and some out there in "you're dreaming!" territory. Be patient and you can get a good deal. I wouldn't give more than $20-30 for one in good working order.

I do not have any of these devices for sale.

Effectiveness

Everyone's experiences will vary. Some people don't like any noise reduction process or its effects on the music.

I played a wide variety of music from different tapes into the DNR system, using a restored Technics RS-M218 cassette deck as the source. Normal and Type II high bias tapes were used.

For the most part, the system did a good job shoehorning tape hiss into the margins, and it seemed about as effective as Dolby B noise reduction when used by itself. It only fell down badly once--on program material featuring a very wide dynamic range and a harp solo with silence between each of the notes. I also noticed this behavior with other plucked string instruments, like guitars. The DNR system just didn't open up the high end response quickly enough to avoid "crushing" the beginning of each note played on the harp. I could notice this and the effect was somewhat annoying. Adjusting the sensitivity control upwards did help, as did enabling the "high" enhancement level. This improvement came at the expense of more hiss being heard. I did not find the slight increase of hiss to be bothersome, and it was still much better than listening to the unaided normal-bias tape.

Otherwise, and with every other type of music I played (ranging from pop to rock to country), the system did an exceptional job of squelching hiss on tapes not treated with any other noise reduction process.

National Semiconductor also mentions that the DNR system can be used in conjunction with any other noise reduction system, such as Dolby B noise reduction. I tried this, using the Technics RS-M218's built in Dolby B encoder/decoder. Indeed it does appear to work very well. The two noise reduction systems in parallel pushed the hiss from a new old stock Sony HiFi 60 minute normal bias cassette tape to the point where I could not perceive it any longer, with no deleterious effects on the music.

Other Stuff

DNR did not do very well in dealing with the menace of HD Radio Self-Noise along with other corruptions of the FM stereo signal. The noise in corrupted FM stereo signals is simply too far down the frequency scale for the DNR system to do much with it without crushing any semblance of high end frequency response at all times. I did find an application where DNR did an amazing job. A record with significant groove noise and wear was restored to nearly perfect apparent fidelity.

DNR may also prove useful to the few but proud AM stereo broadcasting enthusiasts.

This page is still in a "beta quality" stage and is subject to improvements, adjustments and corrections. Constructive feedback would be greatly appreciated.

Go Back >

This page is Copyright © 2011-2019 by William R. Walsh. Some rights reserved. Reproduction of this page in whole is allowed, subject to terms and conditions available from the top level page of this server. Reproduction in part or reuse in other projects requires permission. No advertising material is to be displayed alongside this material. Page created August 2nd-5th, 2011, updated February 5th, 2012 and March 24th, 2015. Minor updates were made in January 2018 and again in late December of 2019.