Distortion Emulation and Results

 

Gold Member
Username: Arande2

Rattle your ... Missouri

Post Number: 3051
Registered: Dec-06
Skip down to the observations if you just want to see what I heard.

Recently I have been doing listening tests on the effects of distortion by processing sound to emulate a system. I have tried different combinations of crossover frequencies (or full-range drivers), amplifier/driver types (specifically their overload characteristics and power limitations). The way I do this is split the audio into tracks representing each driver (e.g. tweeter gets 900hz+, woofer gets 90-900, sub gets <90hz) depending on setup, then I set the parameters on each track to model the amplifier/driver system.

For example, a conventional woofer's response curve drops faster and faster as the coil extends away from the magnetic gap (xmax being measured where the field strength is equal to 70%, in most cases), so increasing the voltage or lowering the frequency will cause a non-linear increase in cone excursion until either the amplifier runs out of power or the driver reaches its mechanical (bottoming out) limits. The system might also include the effects of a crossover and another driver powered by the same amplifier; there might also be an EQ, compressor, or limiter (especially on subwoofers or in high-power commercial applications where driver protection is a must).

A mathematical algorithm which represents the system is entered into the engine which calculates each sample value based on its current value and the values of those samples around it. The result is the waveform as would theoretically be measured in a nearfield test of the driver under the same conditions in a practical test. The result from each driver is mixed back together into the full spectrum which is played back through my calibrated system at a level well within the distortion limits of my system, allowing me to have an idea of what different combinations might actually sound like.

I understand that it is crude to listen to the results through speakers with their own issues (crossovers, specifically), but it really does give me a good idea of what certain conditions might sound like WITHOUT damaging my ears or system.

---The models are somewhat crude and do not include effects such as cone break-up, driver ringing, or enclosure type.. The effects of these details are likely lost by listening through a speaker system in the first place.

...................

Most of the test tracks were from Tool's Lateralus, an album I believe has very good sound quality suitable for such testing.

Observations (most of which are logical, known, and/or expected):

-An overloaded bass driver will cause distortion that covers up details in the mid-range regardless of crossover frequency, although lower crossover frequencies will reduce the severity of the masking since less of the signal is being distorted by the affected driver.

-An overloaded bass driver will cause the sound to be harsh due to shifting of the spectrum toward the high end, all caused by a continuing increase in the intensity of higher-frequency drivers while the bass driver has reached its output limits.

-Using a compressor or limiter on the bass portion of the spectrum will minimize the mid-range masking and distortion by keeping the driver within its linear region; however, the output of the bass driver will be limited slightly, causing the sound to become harsh at a lower overall level (bad), albeit with unmasked mid-range detail (good).

-Bi-amplification is synergistic. Two 50-watt amplifiers spread across a multi-way speaker will give more headroom than a single 100-watt unit, more so with Tool than the other test tracks I used. However, 3dB of headroom is lost for pure sine waves (which is not an issue considering that most music contains many frequencies at once).

-Use of a dedicated driver/amplifier (subwoofer) for the bottom frequencies (below 150hz) increases headroom and greatly decreases mid-range masking (good).

-As frequency increases, less power is used, but the peak levels remain the same. Higher frequencies are more dynamic, but have a higher crest factor. With all of the Tool tracks I tested, 2khz+ had the same peak level as sub-100hz. In fact, at some points the high-frequency energy peaked 6dB higher than low (specifically when there's a bass or palm-muted guitar played along with toms). This is a result of the "quiet intensity" characteristic of Tool.

-It definitely sounds better to have excess amplifier power pushing a driver into its non-linear region rather than under-powering the driver, causing harsh distortion.

-In all cases, using compressors or limiters, especially in combination with EQ (i.e. using limiting to keep boosted frequencies from over-loading), dramatically increases the maximum drive level before harmonic distortion or masking is heard, although dynamics are still reduced when peaks are limited.

-All processing can cause phase issues which don't sound good. A bit of correction fixes this.

That's all I have to say for now, mammoth post as usual. There's a lot more, but I'm not really in the mood to elaborate at the moment.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17251
Registered: May-04
.

Well, Andre, I'm not sure what to make of your post. It's good that you are curious and that you have the desire to test concepts yourself rather than simply accepting what has been fed to you in a textbook or by someone else.

But your results are, as you say "logical, known, and/or expected." You haven't wasted your time knowing these things to be (mostly) true. I would squabble somewhat over your conclusion, "It definitely sounds better to have excess amplifier power pushing a driver into its non-linear region rather than under-powering the driver, causing harsh distortion." But you and I see things differently when it comes to high power. I see too many trade offs which I find unacceptable and unnecessary in more than a few watts. I side with Paul Klipsch in his opinion what this country needs is a very good five watt amplifier. Therefore, while I would find either form of distortion - non-linear movement in the driver/motor assembly or inadequate drive capacity in the amplifier - to be unacceptable, I would first seek a more practical remedy IMO by looking to the efficiency of the driver/waveguide and how the driver loads the environment.

That leads me to say I'm not exactly certain what you have accomplished with your tests. But I think it's good that you wanted to know for yourself what was occurring.



.
 

Gold Member
Username: Arande2

Rattle your ... Missouri

Post Number: 3052
Registered: Dec-06
Aha. I tend to dabble in things I already know.

Efficiency, efficiency, efficiency. I keep thinking there has to be some completely different approach (missing the forest for the trees, lost in the details). I simply don't have the experience and understanding of circuitry and systems (including the room and psychoacoustics) to elaborate.

The end goal being to optimize the storytelling, captivation, and entertainment of the media such systems present, I keep noticing something's not quite right (both tube and solid-state, horn and conventional, etc.) with every system I've heard. Unfortunately divergent and innovative thought is hard to come by, making it difficult to see clearly.

I don't know.
 

Gold Member
Username: Magfan

USA

Post Number: 2751
Registered: Oct-07
Andre, the more you notice...the more you know....the more you hear... the more you test.....the more you notice....around and around.

At some point, you will probably have enough coorelated information to make novel predictions based on observations and any theory they lead you to.

I suspect you are smart and curious enough to eventually notice something fairly new or unusual and will also be experienced enough to act on this information.
 

Platinum Member
Username: Jan_b_vigne

Dallas, TX

Post Number: 17252
Registered: May-04
.

"Aha. I tend to dabble in things I already know."


Well, that keeps us comfortable, doesn't it? You get to affirm your own abilities, which is always of some comfort. However, if as a musician you only played those parts you already know you can accomplish, you would never move forward as a performer. Your playing would eventually get very stale and non-inventive before you finally got bored with the whole thing and moved on. That would be fine if that is what is meant to be in your life. There are many things I engaged in when I was young and had fewer responsibilities eating at my time committment which I no longer even much think about. I was not lacking talent in those areas or ability to acccomplish what I set out to do. But life got in the way and they eventually went by the wayside. If this is what you feel you need to devote time to, just as with practicing your music, make going over those things you already know - already have affirmed - only a small portion of what you do each day. Practice until you can do the task well and then move forward. If you don't have a direction in which to go, find someone who can provide a path.

http://meta-gizmoz.com/index.html




.
 

Gold Member
Username: Magfan

USA

Post Number: 2753
Registered: Oct-07
Please search thru this Wiki for a Louis Pasteur quote.

http://en.wikiquote.org/wiki/Louis_Pasteur

one common translation?

'Chance favors the prepared mind'.

Stay curious, be aware but have a good knowledge base...and especially know where it stops.
« Previous Thread Next Thread »

Add Your Message Here

Bold text Italics Create a hyperlink Insert a clipart image Add a YouTube Video
Need to Register?
Forgot Password?
Enable HTML code in message
   



Main Forums

Today's Posts

Forum Help

Follow Us