A microphone and some software sheds light on my speakers

[exhilaration157]

My dad just bought a Yamaha RX-v1900 AV receiver which I set up. I used the optimizer microphone which was really cool. It got me thinking about finding some software for my computer to measure frequency response and things. The Yamaha sounds great BTW.

I found a 30 day trial of this software that worked quite well. http://www.ymec.com/products/dssf3e/index.htm

I tried setting up the omni-directional mic that came with the RX-v1900, but I couldn't get it to work with any of my mic preamps. I suspect it requires some type of non-standard phantom power. So, I used my best cardiod recording mic it has a wide flat frequency response on-axis. I hooked that into my ART mic preamp and set it to a flat frequency response.

I measured the frequency response of two speaker models:
the recently purchased (for $45 ) ENERGY Point 1e
-circa 1990
-bass-reflex rear ported
-5.25'' woofer
-55-20,000 hz +-3dB 50 hz @ -10 db anechoic
-sens. 86 dB @ 1 meter anechoic

and the

Advance Speaker Korp V-II
-circa 1980
-replaced woofer surrounds
-replaced old japanese electrolytic cap with solen fast cap
factory specs are
-45-20,00 hz +- 3dB
-sens: 92 dB on axis one meter with 1 watt input
XO freq: 1800 Hz
-peerless tweeter

OK... so, I don't have an anechoic chamber. I figured the 30 cm distance would minimize room effects (not eliminate). I also measured at my listening position at 200 cm from each monitor.

Ok first up: The close measurements

Wow these little energy speakers are quite good I thought. The specs I got from the back of the cabinet seem to be quite accurate. The drop off at 20 khz is exagerated by my microphones response too. I guess I stole them at $45. They sound great as near-field monitors.

So the V-IIs aren't perfect but these results make sense. They have a brighter sound, with slightly scooped mids. I wonder if it has anything to do with the Solen fast cap I put in there shifting the XO frequency a bit because of the lower resistance. I could actually hear the dips around 1200 hz and 1800 hz when I listened to the sweep.

The Long distance measurements: I tried to emulate my listening position with the mic placement. I only played one speaker at a time to avoid phase cancellation effects. The curves I got from the left and right speakers were similar. A curve with both speakers playing at the same time yielded more dips and peaks (comb filtering due to phase cancellation I suppose).

The room effects do mask the flatness of the frequency response, but the character of the two speaker models can still be heard. The V-IIs have deeper bass and a more overstated high end, but still neutral compared to some other things I've heard. The ENERGY speakers are quite neutral sounding. Both are very detailed sounding.


I also measured THD. both speakers came in at around 0.5% THD, the microphone THD would play into that too.

The microphone frequency response is flat between 50 Hz and 5 Khz, with a wide 2-3 dB hump between 5-16 khz centered on 10 khz and then a 3 db roll-off at 20 khz. The -3db point on the low end is 40 hz.

I ordered solen caps for the Energy speakers before I did these measurements, assuming the crossovers would need updating. I'll be careful to save the orignal caps, and try one speaker first. I'll measure it and if the response isn't flat anymore I'll put the original ones back. It will be interesting to see what happens.

 

[exhilaration157]

The THD graphs, microphone response, and some extra pictures.

 

[JonL]

Excellent post! I wish I was able to do that in my room... I'm really curious to see my speakers' response in the room. Thanks!

 

[bobrown14]

What source material are you using to measure with? Are these test tones at specific frequencies or music??

Cheers,

Bob

 

[exhilaration157]

The software plays a sine wave sweep from 20-20,000 hz over a period of 20 seconds.

I'm thinking the energy speaker has a nice near-field response because of its compactness. The V-II have more distance between the drivers so they probably have more phasing issues when measured near-field.

The v-IIs might look better measured at 1 M in an anechoic chamber, but I don't have one of those. Each line on the y-axis is 2 dB. So even on the Energy speaker we are seeing a variation of 8 dB (+- 4dB) in the region where it is supposed to be within 6 dB (+-3dB). The advance shows 20 dB variation in the near-field (+-10dB). As I said this is probably in part due to its poor near field performance, since it is a much larger speaker.

Once you go to the 200 cm distance in the listening position you get 20 dB (+-10dB) variation above 100hz on the ENERGY speaker, even worse if you count that large dip around 80 hz. The Advance V-IIs in the listening position have a similar variation (20dB or +-10dB) but without that crazy dip at 80 hz. At the listening position there isn't the same performance difference there was in the nearfield, suggesting again that the V-IIs aren't for near-field listening. The v-IIs are a great sounding high quality speaker with a peerless tweeter and a Genesis type woofer (Advance was started by Engineers from EPI Genesis and Advent). Maybe speakers have improved a bit since 1980 as well. Anyway, enough speculating.


Just goes to show that in a real room setting the frequency response does not stay in the 6 dB window that is measured in the anechoic chamber at the manufacturer. That said, a speaker with better specs may still have better specs in your room. For some reason your brain can pick out the character differences in the speaker even with all of the room effects. Your brain has some idea of what the room sounds, and can kinda separate that from how the speaker sounds. I'm listening to the Energy 1e near-field right now (~60 cm, they're on top of my desk) and if I go sit in my chair the sound doesn't change too much. It's just not as direct.

 

[guiller]

Nice post! Thanks for sharing your results...

 

[71Challenger]

My favorite posts are pics of audio gear and or rooms and techie posts like this one.

I always wanted to spec out my room and different speakers.

 

[exhilaration157]

Thanks,

There is only one problem with these Energy Speakers. They smell like an ashtray. I've been airing them out, but when played loud the kick-drum puffs bursts of ashtray smell out the ports.

I'll post some more comparisons when my solen caps get here. We'll see what they do to the Energy's response and I'll try to to a subjective AB as well.

 

[dokblues]

Great post with excellent and in depth comparisons. Thanks for the software link!

 

[Mark W.]

I use a Spectrum Anaylzer in both my main systems. It puts out a PINK noise signal through the speakers and then with the dedicated mic set up at the sweet spot I adjust a 10 band dual channel (one set of sliders adjusts both channels) to bring the signal to as close to flat as possible.

It's amazing the difference it makes to have your whole system tuned to output the frequency response suited to the room it's in.

And the cool 10 band green climbing display looks way cool.

 

[exhilaration157]

Hmmm, the software I got will do pink noise and will display the spectrum in real time. So I can try doing that. I already know from these curves that my room reinforces at 50 hz and sucks sound at 80 hz. My sound card will do parametric EQ, so I can probably fix some things.

 

[Wornears]

"There is only one problem with these Energy Speakers. They smell like an ashtray. I've been airing them out, but when played loud the kick-drum puffs bursts of ashtray smell out the ports."

Stuff some Febreze dryer sheets inside them and swap them out every so often.

Neat post and info.

 

[exhilaration157]

Update, capacitor swap results

I ordered Solen metalized PP caps to go into the energy speakers. They came today.

As a preamble I will explain what the electrical difference really is between bipolar electrolytic capacitor and a metalized polypropylene capacitor really is, to my understanding.

A perfect capacitor would not have any added resistance other than the selective resistance a capacitor applies to certain frequency ranges. The voltage and current waves would be 90 degrees out of phase (volatage is 0 when current is full and vice versa) therefore there would not be any power consumed since power=Voltage x current. This isn't really the case in real life. Real capacitors dissipate heat. How much heat depends on the dissipation factor (DF). The dissipation factor is directly proportional to the equivalent series resistance (ESM), which is how much resistance the real capacitor has over the perfect capacitor. Bipolar electrolytic caps have a high DF (~5%) relative to the Solen metalized PP (~0.0002%). Using metalized PP caps translates into better dynamics and efficiency when used in a speaker cross over. But, passive speaker crossovers are complicated, so the reduction in ESM could affect the crossover frequency. Especially if the cap is in a part of the crossover that normalizes the impedance of the drivers (zobel network). I don't know enough to design it. I just know you could potentially wreck the balance, by putting a better capacitor in the wrong spot.

So, I was a little bit worried about how the ENERGY POINT 1e speakers would respond. Since they already had a good flat response, I didn't want to mess that up. So when the Solen caps came I did one speaker first. I remeasured the frequency response using the control speaker, then I measured the frequency response of the solen capped speaker. The respective frequency response curves are posted below.

The horizontal lines represent 2 dB

In both curves one can see a 4 dB dip below -50 dB at 3.5 kHz and a 4 dB peak above -50 dB at 15 kHz. If one looks at other points on both graphs they are very similar. Any variation is probably due to the fact that I had to move the control speaker out of the measurement position and place the experimental speaker into the measurement position, I used a measuring tape, but I wasn't that exact. The different capacitors did not effect the frequency response of the speaker in this particular case. (judging by the number of components in the XO 2 inductors and 2 capacitors I might say the XO in this speaker is a second order XO, maybe butterworth, I might get the initiative to draw a schematic later).

So, I did a little listening. The main thing I noticed was that the solen capped speaker seemed to have less grain on the sound. Like the difference between ISO 100 and ISO 400 film (that's my analogy and I'm sticking to it). To see that I wasn't just biased and crazy I recruited my 17 year old sister and my 54 year old dad to do pick the better speaker. My sister was oblivious to the fact I had done anything to them, and my Dad knew what I was up to. Neither of them knew which was which. I played some music that they had heard before, with center panned vocals. I played each speaker separately or both together via the balance control on my mixing board (not perfect, but ok) multiple songs were played. I let the listener tell me when to play each speaker or both. after about 3 minutes I asked them to pick the one that they liked best. Both my sister and my dad picked the Solen capped speaker. My Dad described it as having a piece of paper removed from infront of the speaker, and sister said it sounded less tinny.

So there you have it. That is what you can expect from replacing still good electrolytic caps with better PP caps. That is if it doesn't somehow shift the XO frequency or slope.

.

 

[ryuuoh]

What an age we live in~

30 years ago, I could only dream of testing my gear like this. Very cool, and thanks for the effort.

 

[kfa888]

I have the behringer RTA auto EQ unit with their measuring mic. Semi cheap and relieving info. Fun to play with and a cool tool to have. You can EQ your listening area flat automatically. It's kinda cool.

 

[JonL]

Great post. Thanks. I want to do something like this in my room with my rig someday.

 

[Clmrt]

I have the behringer RTA auto EQ unit with their measuring mic. Semi cheap and relieving info. Fun to play with and a cool tool to have. You can EQ your listening area flat automatically. It's kinda cool.

How often do you prefer the sound when auto-eq'ed?

 

[chadnliz]

Anyone ever try the REW program, its free and downloads at HTshack.com
http://www.hometheatershack.com/roomeq/

I did some tests with my Behringer unit and Mic and it helped find some peaks I tamed down, its fun to use and if I had one complaint its that you can start to hear the charts, in that once you see a problem your mind pays attention to something you never would have sweated.

 

[markus]

Thanks,

There is only one problem with these Energy Speakers. They smell like an ashtray. I've been airing them out, but when played loud the kick-drum puffs bursts of ashtray smell out the ports.

I'll post some more comparisons when my solen caps get here. We'll see what they do to the Energy's response and I'll try to to a subjective AB as well.

awesome post exhilaration ! do the Energy's have polyfill in them ? maybe some fresh stuffing will help the smell - it'd also give you a chance to measure with 'no stuffing' , and 'over stuffed' to see how much the bass is affected

once I get a dedicated laptop (please santa?) I'm getting this software:

http://www.rationalacoustics.com/pages/Smaart_Landing_Page

my buddy is the programmer for the company, and we were messing around in his basement taking speaker measurements a while back. I would LOVE to optimize my room, and my big ass line arrays with real test equipment :yes:

 

[exhilaration157]

Actually the smell has gone away on its own. They sat by an open window for a week. They have fiberglass filling.

I EQed with the Advance speakers (8 bands parametric so far) to compensate for the room modes. The difference was astounding. It's well worth the time. My soundcard is a pro-audio card (EMU-0404) It can support pretty much an unlimited number of parametric EQ bands.

The energy speakers are on my lab bench now. They're being driven by a cheap kenwood receiver, not quite as nice as the Behringer A500.