quick FR from speaker workshop?

[jimbo1968]

I downloaded SW weeks ago but didn't get far with it. Now I've opened my project again and don't know where to start.

All I want is to measure the FR so that I can then design my EQs and LT. Can that be done relatively simply/quickly?

If someone could say: "yes it's easy. Just do A, B, C" that would be appreciated.

Thanks.

[Nuutti]

[list=1][*]Use loopback-cable from soundcard output to line-input x (right or left). This is reference channel.[*]Connect your mic to line-input y (left or right, not x ) If you can't use line-input for mic (no preamp) then you can also measure without reference.[*]Settings: Sample rate: 48000 (or whatever you want, limits the upper frequency and affects the resolution). Sample size 256k (max). Acoustic -> Channels: Microphone: y, reference: x, if you can not use reference, set to "none"[*]Create new driver to your project.[*]Activate the driver's window you want to measure[*]If measuring near field or gated you may do it directly from the measure-menu.[*]If measuring gated response, you have to set markers to positions where signal first arrives to the mic and where first reflection occurs. Another complication is caused by some soundcards (at least usb-types), where delay is not constant. With these cards you have to first measure pulse response, manually set markers and calculate frequency response from there.[/list=1]
When measuring make sure the left-right maximum and minimum values do not exceed about 32k. Set the volume(s) accordingly.

[jimbo1968]

Thanks, i'll give this a try tomorrow.

My mic is a RS spl meter. Is that output from that amplified?
I will try without reference first, because I would have to make up a loop-back cable.

[Nuutti]

If you get sensible signal levels from your meter, it should be ok. Don't know about the linearity though...

[Vikash]

Originally posted by jimbo1968
All I want is to measure the FR so that I can then design my EQs and LT. Can that be done relatively simply/quickly?

With SW, I'd have to say no really. It's something you really have to spend time familiarising yourself with if you want to get useful and acurate results.

Here are some pointers that might help in future:
- After doing a full calibration (follow Wallin or Claudio's website) save your audio mixer settings with Quickmix. This will save you having to reclaibrate every time! Then the next time you fire up SW, just reload the mixer setting from Quickmix, and confirm a few passive components (the reference resisitors on the jig) and you're ready to go.

- Create a new SW project specifically for FR measurements and use it as a base when measuring any speaker. You will have all your graph scales set up how you like, and mic calibration file loaded etc.

- I usually measure the nearfield, then port response if any, then gated.

- Gating. Since I'm not concerned with absolute SPL, I use a measuring distance as close to the driver as possible depending on its diameter to minimise potential room interaction. Then run the impulse response and set the second marker to just before the second pulse (leave the first marker right at the beginning, not at the first pulse). Obviously switch the markers on so gating is activated, then go ahread and meaure the on-axis response.

- I highly recommend going through Claudio's website once at least.

- IIRC Wallin's site has info about modding the RS SPL meter in case you haven't come accross it.

- If you plan on doing this more than a few times in the future, the Wallin Jig will make your life much easier.

[jimbo1968]

Thanks Vikash, but maybe you've answered a different question. I'm not that interested in the driver's response by itself. I've already built the speakers so I'm interested in the 'in cabinet in room' response.

For the FR125s I want to see how much baffle step to add at what frequency. It would also be nice to see whether my Heggs have a nice smooth baffle diffraction as intended. And if the FR125 peaks from your measurements are present I'd like to EQ them away and see the result.

For the 10" I have incomplete driver data, but was hoping to just measure the in cabinet response to then design the linkwitz transform.

I'm pretty clueless on what/how to measure. I was thinking I could just put the mic at my listening position and fire up the pink noise in stereo?

[jimbo1968]

ok, so this is my first meauring attempt.

this is the nearfield response with the mic 150mm in front of the driver. I did no calibration (because I haven't worked that out yet).

what does this tell you about my speaker and/or measuring system?

Being optimistic I'd say it was pretty smooth, but the 7k peak is quite a bit higher than the other fr125 plots i've seen, and the drop to 11k is deeper. It looks the test gear give up around 13k.


i'm using the mic input on a standard dell laptop with the RS sound level meter as mic. The lead from mic to pc is a bit flimsy also. I'm wondering whether I need to buy a mic and preamp from Vikash or a decent usb soundcard, or both.

[Vikash]

Nearfield is taken at single digit mm, not tens or hundreds It's only good for up to a few hundread Hz. Above that you need to augment it with gated farfield measurement. For farfield you need to measure it far enough away to let all the frequencies merge before reaching the mic (especially impotant for multi driver speakers). I use > 30cm when I measure small FR driver.

http://www.claudionegro.com/swacoust...nearfield.html

[Vikash]

Built-in laptop sound devices are usually rubbish. Even an entry level dedicated USB device will be far better.

[thalis]

...that it is an enclosed somewhat small mid-woofer of rather good quality, probably in a closed enclosure, that you need some work in the crossover, and that you have to smooth the curve (smoothing is mostly done in 1/3 of an octave);-)))