Dynamic Range Expander

[systemerror909]

Okay, couldn't figure out what forum to post this in, but figured here would be best, this isn't exactly a diy project but most of you guys are by far more knowledgeable than me so here it goes:

Listening to a decent amount of pop radio, poor quality .mp3's and even a couple of really badly recorded CDs, I'm really starting to notice compression in the music I hear, especially when compared to an even "average/mediocre" CD on my player. I realize that there are tons of other problems besides dynamic range in this type music from these type sources, but its really starting to bother me when I hear modern music that is intentionally compressed so that it can be played back louder over the airwaves. So before you criticize me for my taste in music ( I am a college student, I'm entitled to listen to crap, and in most social situations obligated to play it for others), just if some of you "old timers" could give me some more background on these things called dynamic range expanders.

Sadly, I've never laid eyes on one before. Only by the magic of ebay have I become aware of their existence.


Also, before I forget I would like to ask for some advice on getting a new fm tuner. Currently I have an early 90's model Pioneer integrated amp with tuner that is *good* but am guessing could be better. My friend has a 70's pioneer integrated amp with a very smooth analogue tuner with some cool looking gauges including a strength gauge and wide/close signal selection, features mine lacks. So any suggestions? I'm a big fan of Sony ES, which has a pretty good consistent quality, but most of their tuners still sell for a couple hundred which is out of my range. Any other brands/specific models I shoud look for as a relatively cheap improvement from the built in digital tuner?

[soongsc]

There are basically two types of expanders, one that tries to linearly expand the whole signal, and one that expands only the peaks of large signals (sometimes called the peak unlimiter).

The peak unlimiter assumes that some clipping occurs in the original recording, and is adjusted such that it expands only when some signal peaks occur. I have tried this and the results are very pleasing, especially percussion type instruments (pianos really are dynamic.

The full range expander make small signals smaller and large signals larger. The gain sort fo varies with the signal content. I have not tried this before, but is has been reported that there are "breathing" effects due to the varying gain which might be audible if you system produces good low frequency.

[Original Burnedfingers]

Check out a Pioneer RG-1

[1/137]

dbx 3bx is another

[Jonathan Bright]

As this is a DIY forum you might think about making one. There are a series of chips made by Signetics/Philips (I think) that are numbered NE570/NE571 and NE572. The 571 and 570 are probably not much use to you but the NE572 is worth a look. It is a stereo compander. IE it can be configued as an expander and/or compressor and is marketed as a "hi-fi" component so the specs aren't too bad. It is a chip that can be built with variable attack and decay times and can be adjusted for minimum THD and tracking error. Now, you can use one as a wideband model i.e. handling all the frequencies or you can copy the dbx3* that 1/137 has previously mentioned. This would involved three (or more) chips where the signal is band split and when done in this way you can get quite good performance. It avoids the "breahing" effects that have been mentioned. The NE572 has been around for a good while but is still available in Australia for about $7 U.S. so its not expensive. If you wanted to pursue this track I'd be happy to send you both theory and some practical DIY circuits. But I have no scanner so email me and we'll sort something out with snail mail. I've also just realised that the chip may be sold as a LM572 or SE572 but try NE572 first.

* BTW 1/137 is correct in using "dbx" in lower case. If you Google them or are shopping around you'll find that is their name.

[Jim Karr]

I have been the proud owner of a RG Dynamics DRE for over 25 years. The unit is extremely quiet, no "breathing", does wonders for FM reception and getting rid of turntable rumble.

LP's also benefit from the expansion. On this particular unit, it offers a "downward" expansion which contributes to the quietness.

I met the designer/CEO at the Chicago CES at the RG Dynamics display in probably 1981 or 1982. His name if I remember correctly was Robert Grodinsky, hence the name RG Dynamics. The RG-1 by Pioneer was a OEM unit built for Pioneer by Mr Grodinsky, and was a 14- or 16-dB unit.

RG Dynamics had two models, the Pro-16 and the Pro-20, and in the early '80's, retailed for about 400 bucks US (the Pro-20).

Internally, it's really simple. Not more than six IC's.

[soongsc]

Originally posted by Jonathan Bright
If you wanted to pursue this track I'd be happy to send you both theory and some practical DIY circuits. But I have no scanner so email me and we'll sort something out with snail mail. I've also just realised that the chip may be sold as a LM572 or SE572 but try NE572 first.

* BTW 1/137 is correct in using "dbx" in lower case. If you Google them or are shopping around you'll find that is their name.

Digital cameras ofthen to a preety good job could I get some info too?

[pinkmouse]

Expanders are fine for older material, but with stuff mastered or broadcast over the past five years or so they will fall down big time, due to the use of multiband compression, where compression is used in different amounts over different frequencies, often changing over time. Unless you know the algorithm used, then you have no hope of getting anything out that resembles the original signal.

[soongsc]

Originally posted by pinkmouse
Expanders are fine for older material, but with stuff mastered or broadcast over the past five years or so they will fall down big time, due to the use of multiband compression, where compression is used in different amounts over different frequencies, often changing over time. Unless you know the algorithm used, then you have no hope of getting anything out that resembles the original signal.

How do tuners replay them? The old tuners still work.

[Jonathan Bright]

soongsc. Sure just look me up in "members". Or my email address bright@gil.com.au But it will take a week or two for me to find the material and send it . Okay?