Intentionally Coloring the Sound

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Is it possible to 'improve' the sound of FM Radio or streaming audio by intentionally coloring the sound of the speakers?

Example One: you may have read a review of a speaker or source component that is intentionally forgiving, making it seem that bad recordings sound better. Forgiving could mean soft treble or a less detailed sound.

Example Two: Perhaps a speaker and/or source component could be designed to sound rich and smooth, like an old jukebox, to add some 'glow' to sometimes thin sounding mp3s and radio.

I want to build a new system for my workshop area and I was thinking it would be an interesting exercise to create a system that made FM sound 'better'. Since the sound is often competing with saws, drills, sanding, pounding, etc, it's not like I need deep bass, or imaging, or detail, I just want it to sound.....um.....fun, something that makes you want to dance along with the music while you work.

I could choose a soft dome, instead of a metal one. Perhaps I need to choose a paper cone instead of plastic cone. Perhaps there is filter design that can make the treble sound 'sweet'. Or perhaps I can use a smaller box to boost output around 200Hz to add some warmth or richness to the sound?

Any thoughts?
 
Administrator
Joined 2004
Paid Member
Interesting idea. For me, a boost around 70-100Hz sounds warm. Also a gentle roll-off above 400Hz helps. Be down about 6dB by 20Khz. A dip at 3K and 6K can take a lot of edge off, tho you lose some vocal clarity.

Wooden boxes that vibrate and paper cones can also add warmth. That's how they did it in the old days. ;)
 
more dsp (minidsp http://www.minidsp.com/ ) and compression is about the only thing that will make it through the ruckus of those power tools.

Bass is what you need for fun sounding boomboxyness.

Think like you are in a car, which means a boost in the midrange area.

Forget about treble in such a noisy environment, 5khz is about as good as it gets, so consider intentionally rolling off high frequencies a bit to save your ears when its turned up loud over the power tools.

Get a equalizer and play around, consider floor standing open baffles to fill the room with audio.

http://www.hi-fiworld.co.uk/loudspe...udspeakers-open-baffles-and-bass-part-15.html
 
Last edited:
Soundwave,
The FM band is itself already band limited to about 50hz to 15khz. So there is no use in having a metal dome tweeter there as anything above 15Khz is pure noise to begin with. I would stay with a soft dome. Also most plastic cones are made of polypropylene and it does have a higher damping than paper. This may account for somewhat muddy and slow response in the lower frequencies. I would try and find a light weight paper cone for listening to radio. Don't forget that most radio consoles in the olden days had tone controls and with those you could warm up the bottom end by the bass control and also add a little sparkle to the top end that there was. I can't say what streaming radio bandwidth is, don't know but it would probably be comparable to MP3 quality I would think. Highly compressed to get as much music to you with the least amount of bandwidth. Another option would be some of those headphones with built in radio that would also protect your hearing from a table saw or whatever you are using in the garage.

Steven
 
Good point Pano, thin wooden boxes do add to the sound, I got plenty of 1/4" plywood laying around....

As far as DSP or EQ goes, I am using a leftover Pioneer car headunit for my source/amp, and it does have a 7 band EQ. This relates to my question of what fq to play with. I mentioned 200Hz as a possible area to boost, but Pano suggests 70-100Hz. Perhaps with a ported box at 70 plus some very minor EQ boost will give me that extra warmth? I obviously don't want too much EQ at the port otherwise I'll blow the speaker!

Freax - if you suggest rolling off above 5KHz, wouldn't that sound really dark? Maybe a steeper rolloff above 10Khz would be better? If I only used a full-range driver, they can get peaky around 5-8 KHz and make it sound hard, which I don't want....so definitely a tweeter with a steep rolloff up top I think...

I found a paper coned goldwood GW-204 on parts-express that claims it can go down to 70hz...maybe I'll model that in the next few days and see what I can theoretically get out of it....
 
If you can find an old 60's Tube stereo hi fi at a flea market, antique store, or garage sale, that would be good. I've had a few of them, and the tube FM section is very warm and musical, that fills a room with sound. Avoid the 70's "Sears" Solid State transistor ones don't sound like what your after.
 
I modeled the Goldwood GW-204 in WinISD yesterday and got some interesting results. Almost no matter what box size you put it in, sealed or ported, .1 cu ft or 5 cu ft, this driver has a big hump starting at 130Hz and ending around 300Hz with a total of a 5db peak around 160Hz. This would be horrible for a normal speaker but since I am looking for a warm, round sound, this looks almost perfect. What are the chances that I find the perfect speaker the first one I model?

According to the Ultimate EQ Guide pdf, the 100-250Hz range adds fullness and roundness.

Is there any other free software I can use to double check my results? Would anyone be willing to use their software to double check my results? I just used the mfg published specs on Parts-Express. Here's the link: Goodwood GW-204 I am using a .2 cu ft box myself, but it almost doesn't matter with this speaker.

This speaker is rated at going as low as 70Hz, but it's almost impossible to get that lol. Even a ported box is barely going to have usable at that point.
 

GM

Member
Joined 2003
Using MJK's original freeware MathCad software, a 66.57 Hz end loaded TL made from a 4" diameter piece of pipe stuffed at 1.67 lbs/ft^3 with polyfil [blue trace is its I.B. response].

GM
 

Attachments

  • Goldwood GW204-4S 66.57 Hz end loaded TL.gif
    Goldwood GW204-4S 66.57 Hz end loaded TL.gif
    6.6 KB · Views: 201
Are we just going to consider only the bass department or ...maybe...oh yes, tweeters :
I would suggest the use of a waveguide ( you can make it with unused speakers, broken ones that you don't need )
It may produce some funky sounds!
I gave a chance in trying to pair a little 2" fullrange with a 8" WG, together with an 8"
woofer. Also paper cone tweeters work good. Just model the box so that the tweeter is external but very near to the woofer ( of course ); also you can put it below the woofer.
 
I appreciate the effort GM, but my intention IS to have that upper bass hump in the response. I am intentionally going for a warm, rich, round sound. The blue dotted trace is basically what I am looking for (I think...). I was commenting on the chances of finding just the right speaker as the very first one I modeled for my experiment. But you did verify that this speaker does have an upper bass hump!

Yes Pico, I am doing a tweeter, but just a regular one....a Dayton ND20 rear-mount. I am probably going to do a band-pass on the tweeter, between 3.5Khz on the low side and maybe 8 or 10Khz on the upper. This way I can roll off the upper treble and it won't be harsh when I crank it with the power tools going. Also, with a 2nd order crossover of 3Khz on the woofer, I'll also end up with a notch in the response around 3-3.5Khz - this area typically sounds bright to me, so this will also help with keeping the sound smooth when I turn it up.

I did think about a paper cone tweeter, but I like how that Dayton looks flush mounted. I am also going to bottom mount the goldwood so it looks flush.
 
I appreciate the effort GM, but my intention IS to have that upper bass hump in the response. But you did verify that this speaker does have an upper bass hump!

You're welcome! Understood, it's a sim I did for myself, I posted it to show the blue trace to confirm. At least a century of testing tells us that a ~1.4 Qt is the upper limit for acceptable 'warmth' and even then this is down in the bass where our hearing acuity is pretty bad, so am curious if this one works for you in the mid-bass thru the lower mids without mass quantities of stuffing.

GM
 
Yes Pico, I am doing a tweeter, but just a regular one....a Dayton ND20 rear-mount. I am probably going to do a band-pass on the tweeter, between 3.5Khz on the low side and maybe 8 or 10Khz on the upper. This way I can roll off the upper treble and it won't be harsh when I crank it with the power tools going. Also, with a 2nd order crossover of 3Khz on the woofer, I'll also end up with a notch in the response around 3-3.5Khz - this area typically sounds bright to me, so this will also help with keeping the sound smooth when I turn it up.

I did think about a paper cone tweeter, but I like how that Dayton looks flush mounted. I am also going to bottom mount the goldwood so it looks flush.

So, you've just described what a WG does naturally without any electrical notch : it forces the membrane ( of the tw) to couple to the air in the cavity
or boundaries of the WG ( maybe Geddes would...:eek:) depending on the geometry and, for low frq. cut-off, on the diameter of it, which is related to
the freq. to be reproduced. So it raises the sensibility ( it adds gain) of the driver, thus combined with directivity control, on the lower range ( the highs
remain unchanged), which is good when pairing with a big woofer ( 8" is gigantic) for many reasons (no: only one :eek::p sensitivity matching).
Using new style fullrange (such little multimedia speakers) as a mid tweeter
sligthly puts the power acceptance a little higher, so no worries about crakles and fuzzs that may suddendly destroy ...domes.
I'm not pursuing anything, just to make you notice that there are some solutions.
 
GM - Yea, we'll see. I never built a speaker that intentionally sounded very warm. I'll have to check the Qtc when I get home - if I recall, in the .2 cu/ft box I was planning on it was 1.85!!! I'll double check to see if in increase in box size makes a noticeable difference in the Qtc value.

Pico - I didn't know that a WG acted as a natural bandpass filter......I'll remember that for future projects! :)
 
All boxes, even an OB, is an acoustical band-pass filter.

FWIW, back in the bad ol' days we used electronics, drivers ripped from old radios, TVs, consoles, which typically were very high Q systems, so that they would perform good enough in the heavily vented cabs of the tube era, so for individual shop speakers we would use milk crates or similar open containers that we wrapped with cloth or similar to protect them from dust, insects, occasionally vacuuming them off to allow them to 'breathe'.

For small workbench systems, small, leaky cardboard boxes lightly stuffed with fiberglass insulation worked well enough for AM and even early FM. Kleenex tissue boxes were popular for one particular size oval driver since the cutout seemed like it was designed for it.

GM
 
Founder of XSA-Labs
Joined 2012
Paid Member
I think that the passive vibrations of the cabinet, if not built out of dead sounding mdf can add warmth to music like a guitar body. Also horn loaded cabinets can have additional resonances that give character to sound. A speaker with lots of character is the Cornu spiral horn. Made of very thin panel or foam core, it definitely colors sound but for the better.
http://www.diyaudio.com/forums/full-range/223313-foam-core-board-speaker-enclosures.html
 
The trouble with thin wall enclosures is that they will resonate frequency selectively. Seems like a crap shoot to me. A better approach based on my experience is to build the box out of MDF, but put a passive radiator made out of a driver in it. By shorting the terminals on the driver that is acting as a passive radiator, you can to some extent tune the mechanism. I put two 8 inch speakers in a box about the size of an AR4x (10 x 19 x9 internal?), and put an 8 ohm resistor across the terminals of the secondary 8 inch, and it sounded very good and warm, without audible ringing at bass frequencies. As for the tweeter, I'd rec. going with a conventional tweeter that goes to at least 15kHZ, and using the treble control to roll off the high end above 10-12kHZ if you actually want to. The tweeter you mentioned, the ND20 (?) is one I like and use, and it's resonance is at around 4kHZ, so shouldn't be used below about 6kHZ in my opinion. I bring it in at 7kHZ with a one pole crossover cap. Zaph tested it and claims it's ruler flat from 5kHZ to beyond 20kHZ. Often the bigger issue is the off axis response that will have a big jump at the crossover frequency, where the woofer has gotten rather directional, and then a small diameter tweeter comes in with it's much wider dispersion.

Open baffle speakers have intentional coloration that many people like, but they have to be placed at least 3 feet from any walls to be a plus.

Tube amps sound warmer because of soft clipping, 2nd harmonic distortion, weak damping (1+ ohm source impedance allowing the woofer to ring a bit), and according to Bob Carver the speaker simultaneously acts as a microphone, picking up the acoustic signature or the listening room, and coupling that into the negative feedback path of the amplifier, which then amplifies that on top of the program source. With power tools running, this may not be a good thing.

Bottom Line: Tone controls and perhaps the passive radiator is my suggestion.

If the speaker is going to be a 2 way, and you're using that ND20 tweeter (3/4 inch dome), the woofer should be small diameter (6 inch or less) so the off axis response at the crossover point doesn't cause harshness (big step up in the off axis upper midrange freqs.). A 6 inch woof with an 8 inch woof acting as a passive radiator with shorted or resistor across its coil could be good.
 
I made my first prototypes this week and have been listening to it for a couple of days....but its not quite what I am hoping for. First, as expected there isn't anything you could call bass. It's got a 7db hump around 150Hz-200Hz, but instead of 'warm' sounding, its sounds....um.....exaggerated - like it's resonating. Second, as mentioned in my other posts, I tried creating a bandpass filter on the tweeter to roll off the upper treble, but it sounds....grainy...like sandpaper tweeter domes. I'm thinking I maybe used too cheap of parts on my tweeter...or I just did it wrong...or my soldering is weak/dirty....

I am wondering if I need to go back to the drawing board? Since I have no idea what freq's make for a warm, rich sounding speaker....perhaps my choice of 150-200Hz is wrong?
 
I am wondering if I need to go back to the drawing board? Since I have no idea what freq's make for a warm, rich sounding speaker....perhaps my choice of 150-200Hz is wrong?
A 7 dB hump at 150-200 sounds like mud to me, and even worse in the usual 8 foot ceiling rooms.

I like my shop speakers to be as flat as possible above about 100 Hz, as I wear deadphones a good amount of the time (they mellow out the top end), and have a "loudness contour" lifting the subs response from 20Hz to about 100 Hz.

My little Sony receiver (from a dump in Alice Springs, Oz, took about 2 months on a slow boat to make it to the USA) has a 5 band EQ, then the 12 volt car sub crossover has a low boost (sometimes engaged).

Having 8 small 2 way speakers and four subs spread out around the shop allows the speakers to play over the level of the loudest tools with only a few watts. They are a motley crew ranging from Radio Shack cones to ancient Acoustic Research midrange, at least 10 different types of transducers, 34 count total. I only actually paid for a few of the drivers, most were curb rescues ;^).

At any rate, I'd advise going for a flat response, and EQ to taste, you can pick up EQs for cheap and insert them in the tape monitor of any old receiver.

Have fun, use hearing protection...

Cheers,

Art
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.