tigerleo

November 21, 2009

Behringer FBQ2496 EQ and Room EQ Wizard

Filed under: General — steve @ 4:38 pm

Background

I have two old TEAC speaker boxes that I replaced the speaker drivers in about 8 or 9 years ago with Jaycar 12” CW2130 polycone woofer, Jaycar/Response 5” CM2085 polycone midrange driver and Jaycar/Response 1” CT2010 dome tweeter. While the bass response was adequate in the lounge room where I originally put the drivers together, and in the following lounge room where they were placed, in my current lounge room, when driven by a DSE A2760 stereo amp the bass has been lacking. As I use these speakers for mixing computer created music a reasonable flat response is required. Presently I have been checking my mixes between these speakers and a 5.1 Logitech Z5500 speaker system to check the bass levels – as mixing takes me the longest in my music production switching between speaker systems to remix the compositions is slow and painful. So I decided to check the Jaycar’s response patterns using the free Room EQ Wizard

Equipment

Amp: DSE A2760 50w stereo amp

Soundcard: External Edirol UA-3FX 24bit 48K 2 channel

Microphone: Zoom H2

Sound level meter: Radio Shack 33-2055

EQ: Behringer FBQ2496 “Feedback Destroyer”

Speakers: Jaycar CW2130 12” woofer, CM2085 5” midrange driver and CT2010 1” tweeter [as above] with 3-way 4th order cross-overs in  45 litre bass-reflex enclosures

Testing the Speakers

Prior to purchasing the Behringer FBQ2495 I downloaded the Room EQ Wizard (REW) and read the help files – setting up the software to your sound card takes a while, however the instructions are well written and the software guides you through the process quite well. [My soundcard has an internal loop-through function so I did not need to route the output to the input channel to test the soundcard’s levels with REW.] Using the Radio Shack Sound level meter and the Zoom H2 microphone I managed to calibrate the REW software to my soundcard and then took my first measurement of the speaker’ frequency response [below – click to enlarge image]: 
original

Looking at the frequency response it was apparent the mid-range was well and truly above the bass and treble frequencies. I remembered that when I put the speakers together all those years ago I changed the phase of the mid-range speaker – this was to suit the room the speakers were in at the time and also due to 4th order cross-overs possibly affecting the phase shift of speakers. So I swapped-over the wires to the mid-range speakers and measured the new frequency response with REW [below – click to enlarge image]:
phased mid range

Now the bass and mid-range was more level however the top-end was down, so I swapped the polarity of the tweeters and measured the frequency response again [below – click to enlarge image]:

phased treble and mid range

It would appear that the woofer’s phase was out having swapped both the mids and tweeters around, however in the end I was quite happy with the overall improvement to the balance of the speakers – bass was present when listening to music and the levels were also more comparable to the overall response from the Logitech 5.1 set as a 2.1 system. I was so happy with the improvement with the 12” driver’s speaker response that I wondered whether I would need to purchase an EQ to correct the frequency response, however, mucking around with the frequency correction in REW it appeared that I would be able to get a very flat response out of my speakers [below – click to enlarge image]. Also. when I found out I could get the Behringer FBQ2496 EQ at $240 from Kosmic [Behringer’s list price being $349] I thought why not…

filtered

Behringer FBQ2496 EQ

I chose the FBQ2496 as it has 20 bands per channel compared to 12 bands per channel in Behringer’s DSP1124P. Both of these EQs accept MIDI data; meaning that the EQ settings can be sorted out on the computer and then downloaded to the EQ itself – much more convenient that twiddling with a knob and pressing buttons on the EQ’s front panel. With the REW software, EQ corrections entered are simulated on the computer screen so you get an idea of how your settings will affect the frequency responses – much easier than purely relying on your ears to tell whether frequency levels are matched. So I purchased the FBQ2496, installed it inline between the soundcard and amp, and measured the speaker response pattern again [below – click to enlarge image]:

measured

It was evident that having the EQ in the signal path affected the signal even when the Behringer was in bypass mode. So I had to go through manually and re-correct the frequency responses which, According to REW would look something like this [below – click to enlarge image]:

corrected

The FBQ2496 can boost by 15dB or cut by 36dB – I used the full 15dB boost in places to level out the frequency response. to make it nice and theoretically flat. I downloaded the EQ settings [see below for notes on getting the Behringer to take the settings] and measured the new response [below – click to enlarge image]:

3. tested

It was evident that the simulated frequency response was different to the actual measured frequency response, so I tweaked the EQ settings in REW and tried and few different versions to reduce the treble response. After a few goes I listened to some music…

While the bass was certainly more pronounced, the treble huffed and puffed – boosting certain frequencies by 15dB was not a good idea – while theoretically possible, the actual sound was atrocious and I preferred the sound of the speakers without the EQ on – I wondered whether I had wasted $240 on the Behringer, however, I decided to reduce the boost across the frequencies to no more than 6dB and downloaded the frequency corrections again. Overall I made quite a number of attempt ‘fixing’ and downloading different frequency correction before deciding to settle on the following measured response [below – click to enlarge image]: 

final

While the treble range looks very poor on the above chart, the sound of the speakers is much improved – the bass and mids are level and the hiss from the tweeters has been removed. I will probably try tweeking the EQ settings again sometime to see whether I can raise the treble frequency response, however at this stage my main need was for a more even bass response with the mid-range. Again, here is the frequency response of the speakers prior to installing and using the Behringer EQ [below – click to enlarge image]:

phased treble and mid range

Notes on the Behringer FBQ2496 EQ

Behringer’s manual is not clear on how to enable the EQ to accept all 20 bands as programmable EQ settings. At first I was only able to get 11 bands pre channel until I pressed and held-down on the Learn button and was able to turn off the 9 slots reserved for the feedback feature of the unit.

Downloading the setting from REW to the Behringer was bit of a trial – frequency settings need to be downloaded one channel at a time {i.e, Left and Right channels). Additionally, it would take at least two attempts per channel to get all the settings to load, sometimes more, therefore making changes in REW and downloading the settings to the EQ before measuring the actual audible speaker response was a slow process – most of this time was the actual downloading of the settings. As one could not be certain whether the EQ settings had downloaded I found it safer to clear all settings on the EQ before downloading new settings.

I guess the reason why using any +15dB settings on the Behringer created noise was due to the headroom available from the soundcard’s output – REW gives an indication of how much headroom is available and therefore I assume that exceeding that headroom would end up driving the amplifier too hard creating a degree of noise. However, this is just surmising what might be happening – it might also be that the Behringer EQ also has difficulties boosting signals at such a rate not to cause an increase in the noise floor…

Conclusion

Most people appear to use either of Behringer’s MIDI enabled EQ products along with the REW software for sorting out subwoofer responses. REW automatically resolves frequency peaks, however this is up to 500 HZ, so it is not much use for full-range speakers in that regard. However, being able to see a visual representation of how frequency corrections might affect a speaker’s frequency responses in REW is certainly a benefit than just relying on ears alone. I was going to get new speaker 145 litre boxes made for the drivers to sort out the bass response from the woofers, however spending the $240 on the Behringer looks like it will now be a much more economical, not to mention space-saving, solution. I am much happier with the sound from the speakers now – it is amazing what sounded OK one day sounds rather poor when your ears are presented with a better alternative – ears are fickle things!

2 Comments

  1. At first I didn’t see much difference between the before and after graphs, but once I viewed them at full size it became more obvious – the peaks are much more tightly grouped around the 90DB line in the after graph.

    Comment by Jim E — November 21, 2009 @ 6:34 pm

  2. It has more to do with the flatness of the overall horizontal line – not so much the individual peaks, though it would be great if they didn’t exist at all. I was aiming to get something consistently between the 85 dB line [the thickish black line] and the 90dB line… As it is it works out as an average of around 94dB, however from 4k the level drops dramtically. I aim to get any drop after 5 kilohertz, and a smoother drop than what is currently a very stepped-down response.

    Comment by steve — November 21, 2009 @ 6:40 pm

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

Designed by TigerLeo - Powered by WordPress