FR Measurement :-(

ani_101

So, i tried taking a FR measurement today... quite a few things are different - i have a new laptop and hence a new install of Holm Impulse - needless to say i don't have any of the settings. So between struggling with the settings and the usb mic and speakers and amp... half the time the sound was playing from the laptop speakers, the mic on the laptop was trying to measure... 

so after all that i think i got the the speaker drivers to play and the usb mic to capture (i think) and then I get this FR. All three are the same drivers, but i spaced them out to make it easier to see.

The first impulse seems to support gating at 300, but i get the weird dips  which seems to go away at higher or lower gating and the FR itself looks wonky quite a bit of a rise from 600Hz all the way to 12khz.

But i am still wondering if my settings are wrong or something off with the mic / amp / etc.... 

I am going to retry again over later... I still didn't hook up the tweeter, would after I figure this out. I also want to try measuring one of my existing speakers. I don't have the old laptop, and i don't trust this setup yet...

6thplanet

The blue impulse looks wacked, definitely try one of your other speakers to get a base line. The red looks ok, green is hard to see

dcibel

I don't see a problem. Is the response supposed to be flat? If so, did you load the mic calibration into Holm and check the box to use it?

Green includes some reflection, as you can see in the impulse there is a reflection at 3.8ms and the window ends after 4ms. Blue and red don't include that reflection, red is more smoothed in the 1kHz and below since the window is shorter.

I don't recall if you can change the window type in Holm, but a more aggressive window shape may be of some benefit, in the green trace the window is gradual from before 3ms to after 4ms, which may be good or bad depending on what you're looking for. To avoid all reflections, the gradual window is not the best, but will allow for more useful low frequency data. A brick wall or rectangular window is great to see where the reflections are in the response as you adjust the window.

rjj45

Sometimes Microsoft windows has a default for sound recording of "enhance vocals", which raises the treble. Dig down into sound settings and make sure that the recording device is flat.

Tom_S

Did you get this figured out Ani? I agree with Don - it's probably a setting somewhere. Did you get a new Dell machine? We got new top of the line Dell workstations last year for our edit suites. Very expensive powerful boxes, but every time they do an update, it turns on MaxxAudio to "enhance" the sound. I always think there's something wrong with the amp or speakers at first - videos sound horrible. 

ani_101

Not yet, i did dig deeper into the setting and set the mic volume to 100 (it was way low), turned off spatial sound - there was a weird setting playback sound from this device - the microphone which was causing it to echo, turned that off.

this is today's results.... the top end represents the peak/breakup i was expecting (5k -10k+), but the mid/upper bass is still off i guess... now there is no bass the 5k to 7k and lower is supposed to be way more flat than this slope

ani_101

the measurements are at 1M, tweeter axis

rjj45

You can increase your gate to about 4.6 ms and pick up some mid bass before you hit any reflections.

Tom_S

Something is sure out of whack! Does the output sound OK with music or does it have the 40dB slope you're seeing on the measurements? 

And computers were supposed to make our lives easier!

rjj45

Ok, maybe a little change in direction. If you already have a mic and preamp running for HolmImpulse, it's very easy to switch to ARTA and see if the results are the same. You can get the sw here: http://www.artalabs.hr/
Charlie Laub wrote a super easy tutorial to get started. If you need it, I have it somewhere on my computer. 

ani_101

rjj45 said:

Ok, maybe a little change in direction. If you already have a mic and preamp running for HolmImpulse, it's very easy to switch to ARTA and see if the results are the same. You can get the sw here: http://www.artalabs.hr/
Charlie Laub wrote a super easy tutorial to get started. If you need it, I have it somewhere on my computer. 

If you have the tutorial, then it's great.... would like to have it!

ani_101

ok, so there is something definitely up with the line out of my laptop and it's a brand new one... It kept shutting down holm impulse - it would take a couple of measurement, and then it would say starting audio stream and then just stops... but it plays music ok, so it's fine i guess

today morning, used one of the el cheapo usb audio dongle which are typically used with phones, nothing fancy, just a basic DAC and measured the penguins.

That's not too bad... the sweep sounded as it measured... 

ani_101

next i measured the new speaker....

Nice curves... cross around 3k and should be good right?

ani_101

did i fix it? No... cause there's a so called intelligent human behind all the hi tech tools...

here's the complete picture... guess what happened!

ani_101

So I did the battery test with a 1.5 volts AA cell. One woofer went in and one went out. What is the correct polarity - it should pop out???

so inverting it the measurement looks like this.... But now, why does the top end goes whacky??? it's outside the usable range... See the red curve, from 4k - 20k - what's happening here?

ani_101

Tom_S said:

Something is sure out of whack! Does the output sound OK with music or does it have the 40dB slope you're seeing on the measurements? 

And computers were supposed to make our lives easier!

Sometimes the biggest issues are caused by the dumbest of things!!! 

rjj45

ani_101 said:

rjj45 said:

Ok, maybe a little change in direction. If you already have a mic and preamp running for HolmImpulse, it's very easy to switch to ARTA and see if the results are the same. You can get the sw here: http://www.artalabs.hr/
Charlie Laub wrote a super easy tutorial to get started. If you need it, I have it somewhere on my computer. 

If you have the tutorial, then it's great.... would like to have it!

Happy to share. The wonder is that I found the doc!
Charlie_Laub-FR measurement using ARTA-sept-2012-update.pdf

Wolf

The impulse response initial spike tells you whether the measurement of the driver is normal or inverted phase. Up = normal, down = inverted.