DBX RTA-1 - repair and restoration

[abbey road d enfer]

Did you know Les is a talented pianist, and Gary is a kickass drummer?

I think most of the guys at dbx at the time were musicians. I remember, my first encounter there was with Joe Lemanski, he had a Les Paul in his office...

 

[DonDueed]

For those unfamiliar with the RTA-1, I thought I'd write up a description of its unique capabilities.

Like other pro spectrum analyzers, the RTA-1 can accept a signal input from a variety of sources and produce a display of the signal's spectrum in real time. The RTA-1 uses 1/3-octave filters, so it breaks down the audio into around 30 bands. So far, nothing too unusual.

The RTA-1 was targeted toward acoustics professionals such as touring sound companies and room acoustics consultants. For example, in concert venues, the RTA-1 could be used to provide the information needed to equalize the room -- and it could do so without the need to drive the room with pink noise. Instead, it could provide an accurate room-response spectrum using music as the signal source!

How did it do that? It had the hardware, and the smarts, to keep track of two signals simultaneously. So you could feed your music to one input of the RTA-1 (and to the PA system), and "listen" to the room with a mic connected to another RTA-1 input. As your intro music played, the analyzer would compare the two signals, note the difference in the responses, and display that difference -- i.e. it would show the room response curve.

This means the sound engineer could not only get a room curve without blasting everybody with pink noise, but could also run the process repeatedly even as the audience filled the venue, so the effect of the audience itself could be corrected for.

 

[DonDueed]

A second major feature of the RTA-1 was aimed at acoustics consultants. An important factor in room acoustics is reverberation time -- i.e. how quickly a given sound will damp out. This can and does vary across the spectrum due to various factors (room dimensions, reflective surfaces, absorbent materials, etc.)

The RTA-1's RT-60 function makes taking reverb time measurements simple. It has a precision pink noise generator that can be connected to a PA system or other sound source. When triggered, the RTA-1 turns on the pink noise to drive the room with broadband noise, then shuts off the signal and takes a series of "snapshots" of the spectrum as the sound decays. It then displays the result as a series of decay curves. This information can then be used to guide efforts to fine-tune the room's acoustics.

Obviously, this process is not done with an audience present.

 

[DonDueed]

Nowadays, some or all of this functionality can be provided with digital analysis (FFTs), but the RTA-1 was designed in the 1985 time frame. At that point it was still very expensive to do real-time FFT analysis over the full audio band. The necessary number crunching power just wasn't there.

The RTA-1 is an analog instrument, using switched-capacitor 1/3-octave filters and analog level detectors (two RMS detectors per band). The digital domain is entered only at the point where the level-detector outputs are read by an A/D converter. After that, of course, the tasks of calculating and displaying the data are handled by the computer.

 

[david moran]

Coming in very late, I am so glad to see the designer weigh in here. He is a fine smart engineer and a helpful colleague with whom I happily worked in dbx engineering for many years. I wrote the first user manual for the dbx RTA1, short and sweet and to the point, and the week I was laid off, as dbx went through endstage conniptions, I got a TW / rec eng friend to write a longer doc, which also became the ST4000 manual (these latter manuals are sometimes available as pdfs online ). I still use my unit to assess loudspeakers, indoors and out- (half-anechoically). DB and I know the ace tech who is a font for everything else about the product, and I keep up with our other db-exers.
It is a marvelous device, and was used for many years by designers of EAW, BA, Allison, and other speaker lines. Roy Allison in fact said even as a technology dinosaur it was so handy and so much faster and more reliable for driver and crossover iteration than anything in the world of impulse-based tech. (Now smartphone RTA apps can come close but are clumsier to use post-measurements.) Partly due to my advocacy (whingeing) over the decades every smartphone RTA now has temporal averaging, an essential feature uniquely implemented by DB in the dbx unit, setting it above everything from Ivie and AudioControl and the rest.

Online a pro sound user wrote it was the only RTA gear he knew that accurately and reliably corresponded to what he heard in the house. Other dbx engineers wrote a graphing program (has to run under dosbox) which is extremely handy for saving and displaying curves, and which now has updates (not fully finished) for modern Windows PCs. I can provide these programs to anyone interested, but limited support. More to come in case there are questions.

 

[user 133392]

I think most of the guys at dbx at the time were musicians. I remember, my first encounter there was with Joe Lemanski, he had a Les Paul in his office...

FWIW THAT's building in Milford was at one time a jazz speakeasy in addition to being an electric motor manufacturing facility. Throughout the building are pictures of the jazz greats that played there. The basement break room is a performance/rehearsal area - when I was there Gary had his drums setup.

 

[abbey road d enfer]

It had the hardware, and the smarts, to keep track of two signals simultaneously. So you could feed your music to one input of the RTA-1 (and to the PA system), and "listen" to the room with a mic connected to another RTA-1 input. As your intro music played, the analyzer would compare the two signals, note the difference in the responses, and display that difference -- i.e. it would show the room response curve.

This means the sound engineer could not only get a room curve without blasting everybody with pink noise, but could also run the process repeatedly even as the audience filled the venue, so the effect of the audience itself could be corrected for.

This was at a time where all live SE's were expecting and relied upon a flat system.
Today, a PA system is tuned with what they call "bass contour", which is basically a shelf EQ resulting in +12 to 14 dB boost @50Hz. Plus they have separate control of subwoofers level, which results in an utterly inflated perception.
And when there was a single SE, now you have a sound system designer, a system tuner (the one that boosts bass) and a FOH engineer.
Of course, 1/3 octave resolution is not enough for these guys; they have all sorts of software in their i-thingy that allow them to track micro-issues that may or may not be significant.

 

[david moran]

oh, 35y ago the smarter system and house engineers knew to boost the bass, I believe I recall.

Yes about resolution, a ridiculous outcome abetted by nominally FFT-based systems of doubtful accuracy and utility (significance).

 

[abbey road d enfer]

oh, 35y ago the smarter system and house engineers knew to boost the bass, I believe I recall.

They certainly did, but not in the ridiculous amounts we see today.

 

[Timbered]

Coming in very late, I am so glad to see the designer weigh in here. He is a fine smart engineer and a helpful colleague with whom I happily worked in dbx engineering for many years. I wrote the first user manual for the dbx RTA1, short and sweet and to the point, and the week I was laid off, as dbx went through endstage conniptions, I got a TW / rec eng friend to write a longer doc, which also became the ST4000 manual (these latter manuals are sometimes available as pdfs online ). I still use my unit to assess loudspeakers, indoors and out- (half-anechoically). DB and I know the ace tech who is a font for everything else about the product, and I keep up with our other db-exers.
It is a marvelous device, and was used for many years by designers of EAW, BA, Allison, and other speaker lines. Roy Allison in fact said even as a technology dinosaur it was so handy and so much faster and more reliable for driver and crossover iteration than anything in the world of impulse-based tech. (Now smartphone RTA apps can come close but are clumsier to use post-measurements.) Partly due to my advocacy (whingeing) over the decades every smartphone RTA now has temporal averaging, an essential feature uniquely implemented by DB in the dbx unit, setting it above everything from Ivie and AudioControl and the rest.

Online a pro sound user wrote it was the only RTA gear he knew that accurately and reliably corresponded to what he heard in the house. Other dbx engineers wrote a graphing program (has to run under dosbox) which is extremely handy for saving and displaying curves, and which now has updates (not fully finished) for modern Windows PCs. I can provide these programs to anyone interested, but limited support. More to come in case there are questions.

Hi David!

I'd definitely be interested in those programs you mentioned! I have the Sound Technology version of the RTA-1.

Thanks!

 

[david moran]

Hi David!

I'd definitely be interested in those programs you mentioned! I have the Sound Technology version of the RTA-1.

Thanks!

I wonder if this site will let an exe be posted

we'll see

nope

I am at [email protected]; pls send me your info

 

[NOON]

I'm interested in the software as well. You can change the extension to something the site will accept and add a note saying to change it back to .exe after downloading.

 

[david moran]

k, change back to exe

an6 is a graphing program of the old type, meaning runs under dosbox only under W10, I think; it used to require early Win and an ega monitor

48 is an update but incomplete, I believe, and runs under W10 mostly okay

I will somehow reward anyone who 'decompiles' any of these to clean them up, esp 48, and make them more complete

an6_c1 guarantees output to the right com port, iirc; actually, use it in lieu of an6

I am hazy on most of this at the moment, but w help of a colleague will be able to reconstruct everything, I believe ... eventually

use paint to invert colors after graphing

however crude, they are wonderful programs to use w the rta1, and over the decades I published many speaker test reviews for CD Review, Digital Audio, SpeakerBuilder, $ensible Sound, Car Stereo Review, Linear Audio, and the BAS Speaker using the dbx and an6

you need to make a special cable from the rta1 to the pc; will try and find the specs for that

 

[david moran]

here

 

[david moran]

View attachment RTA-1 serial cables.jpg

 

[david moran]

Hi David!

I'd definitely be interested in those programs you mentioned! I have the Sound Technology version of the RTA-1.

Thanks!

awesome, want to hear more

at one point at CES they were showing a color post-EGA version which never materialized

 

[Timbered]

Thanks David!

I'll take a look at the software and see if I can "decompile" any of it (I've been a software developer for over 40 years...) or at least try and figure out what they are doing, maybe rewrite them? Dunno, but I'll look.

Do you know what programming language or system that software was created with?

I think I read in my manual that the internal firmware is color capable (by attaching an external monitor). So, I'm thinking of replacing the internal mono CRT with a color LCD...

T.

 

[david moran]

basic? I will check. it runs under dos and the machine has to be ancient in order for it to graph properly

the rta1 can drive an ega monitor, yes, but that's it

maybe you can figure out how to add on a modern video card

DonD here will know all of this, assuming his memory (hands-on) is better than mine (user only)

 

[Timbered]

If, by some miracle, you have access to the original program source code, I could certainly write something that runs under Windows, if that would help you.

 

[david moran]

no, no source, but see what you can do w an48

I will inquire about source for that

another rta1 owner and expert has gotten plotting done using excel and wrote me

What I did find was two good tutorials on YouTube, one showing how to graph one set of data and the other showing how to graph two sets of date. Links below [[your task would be to get them to do the diff / normalization functions, which is one of the cooler things about the rta1]]