First Galaxy 140 results.

**Chrisbee** · **03-22-07, 05:01 AM**

Hi

Had a quick spin with REW with my new Galaxy 140 this morning to see how it performs.

Sub is a 4 x 15" AE1B15 True Infinite Baffle line array (as seen in my signature).

I am getting a low readings warning from REW on calibration but I continued to calibrate to 80dB and ran a few sweeps.

Listening position: Galaxy 140 SPL meter. 80Hz crossover with BFD boost +16dB @ 20Hz. C weighting ticked. No meter or soundcard correction files.

Here's the 140 without C weighting or correction files.

Now I just remembered that I could use the ECM8000 correction file with the 140 but there isn't a correction file in the downloads area.

**Chrisbee** · **03-22-07, 07:05 AM**

Two more graphs using ECM8000 calibration files:

C weighting ticked.

C weighting not ticked.

They look the same to me.

All measured:

**PeteD** · **03-22-07, 07:56 AM**

How does this look compared to what you thought you had with the RS meter? I don't recall the dip in your response and I thought you had a house curve dialed in with the +16 dB of boost, no?

Pete

**Sonnie** · **03-22-07, 08:56 AM**

Make sure you have the CM-140 set to "Fast" and that should eliminate the low reading warning.

**Chrisbee** · **03-22-07, 04:33 PM**

Quote:

Sonnie wrote:

Make sure you have the CM-140 set to "Fast" and that should eliminate the low reading warning.

Thanks Sonnie,

I set to Slow from habit. I'll have another try tomorrow.

Pete,

The RS meter showed the same dip at around 16Hz.
I think my usual filter set looks slightly more humped around 40Hz now.
More quality time with the BFD is indicated.
Perhaps I'd better wait until Sonnie posts a 140 calibration file.

**brucek** · **03-22-07, 06:26 PM**

Quote:

C weighting ticked

Quote:

C weighting not ticked

Quote:

They look the same to me.

As they should. The ECM8000 calibration file extends from 10Hz to 46KHz.
The C-Weight curve will only compensate outside the range of a calibration file.

The C-Weight curve is not to be checked when you use the ECM8000 calibration file because it will try and compensate below 10Hz, which would be meaningless for the ECM8000 as a flat mic.

Quote:

No meter or soundcard correction files

Why no soundcard file. You need to compensate for its response. If you don't the results aren't accurate.

brucek

**Chrisbee** · **03-23-07, 02:54 AM**

Thanks, brucek

You have a way of cutting through the swirling mists of doubt to bring clarity where none formerly existed.

I once did a soundcard calibration but it was so flat I decided it wasn't worth worrying about. (Soundblaster Live! and a 1 foot long cable)

As I use a 10 meter cable between the SPL meter and the soundcard and another out to my preamp I had better include those in the soundcard calibration. 60 feet of cable might have some visible effect in REW. We shall see.

**Chrisbee** · **03-23-07, 04:32 AM**

IB, 140 (fast) ECM8000 cal file, soundcard cal, BFD active.
REW still complaining about -20dB low readings on test.

Here's the soundcard cal image. (violet) The ECM8000 cal curve is also shown.(black) Do these look reasonable?

**brucek** · **03-23-07, 08:29 AM**

Your soundcard cal file is not good. The upper end ringing can be caused by too high a level used when setting up for that test or from soundcard features turned on. The low end response isn't very good, but not uncommon. You need to get a good soundcard file cal file before proceeding.

Once you get one, take a measure of the loopback cable (before removong it). It should measure perfectly flat because the soundcard cal file is inversly compensating for its poor internal response. Once you get a flat response on a loopback cable, you're ready to take measurements.

brucek

**Chrisbee** · **03-23-07, 12:56 PM**

Thanks, brucek.

By careful adjustment of level I have managed to improve my Soundcard calibration. I have reduced the Live! mixer to 2 speaker (rather than 5.1) and set input to Line-in rather than default input. No effects. No bass or treble. The bass response is further extended but the ringing persists. This is with my two foot long cable using two solid RCA to mini-jack adapters. No other programs running except REW.

**brucek** · **03-23-07, 01:18 PM**

Yeah, sounds like you're doing everything correct.

The low end response is quite good, only down 0.5dB at 10Hz, but the upper end ringing is awful.

I thought the Live card was good? Sure looks like an effect. Definitely selecting 5.1 wasn't correct. Stereo is the correct selection. Hopefully there's no equalizer on. Your long cables shouldn't be a problem, especially since you're including them in the calibration. I include my cables and my mic mixer in my soundcard calibration file. You want to eliminate all the effects from the test equipment as possible, so you're only measuring the equipment.

I have no ideas for you other than I wouldn't measure above 200Hz.

brucek

**Chrisbee** · **03-23-07, 03:08 PM**

I have been through everything in the soundcard controls. Nothing is enabled so there can be no active effects .

Here's the soundcard curve with 60 feet of cable in the loop. I really can't see any difference from the earlier 2 feet test.

What do you make of the skyscraper effect on the ECM cal file at the same point as the soundcard goes into spontaneous oscillation?

**Sonnie** · **03-23-07, 03:15 PM**

That sound card response doesn't look right at all... some setting is off there.

Quote:

Chrisbee wrote:

REW still complaining about -20dB low readings on test.

Try increasing your Input Volume from the Settings button >> Soundcard tab. Notice the Peak Meters on the right. The Input meter should be reaching at least -18. I generally have it match my output which is around -12. Right now it is only reading -20 and thus the warning.

**Chrisbee** · **03-23-07, 03:37 PM**

Hi Sonnie

I've tried lots of variations using the level button you have highlighted.

If I go too high I get a clipping warning and the soundcard curve goes all wiggly. If I go too low I lose the bottom end. The recent soundcard curves I posted showed about 6dB of headroom during the sweep and seems about optimum. This equates to approx -10 on both scales. I've tried ticking the 44 and 48 kHz buttons in turn without any change.

**Sonnie** · **03-23-07, 03:42 PM**

Something is haywire in your setup, not sure what it could be, but something is off some where. You have from -0.1db to -18db that should give you no warnings... that's an awful wide range. For you not to be able to make that range without warnings ... and for your sound card response to be the way it is... it's all suspect to me.

**brucek** · **03-23-07, 03:53 PM**

Quote:

it's all suspect to me

The same sort of results would occur if you happened to make a mistake and use the microphone input jack on the soundcard, rather than the line-in jack..... ?

brucek

**Chrisbee** · **03-24-07, 03:47 PM**

I've tried assorted input and output options without any improvement.

How about this?

1/3 octave smoothing and slight trace adjustment.

**Chrisbee** · **04-02-07, 04:14 AM**

I'm back!

I have downloaded the new 5Hz Galaxy 140 cal file from the downloads page. Many thanks to Sonnie and brucek for their efforts (and anybody else involved)

I am still suffering (?) from low levels on "Check Levels". Despite the 140 showing 75dB (C-fast) and the graphs looking very reasonable at this level I get a warning message each time saying that input levels are low. -32-34dB on average. Everything possible has been double checked and carefully matched to the illustrated REW help files.

Nevertheless I am getting incredibly consistent results:
Here's an "all measured" graph of my IB at the listening position:

Here's a single trace for confirmation of the shape of the new Galaxy 140 5-200Hz cal file. C-weighting unchecked in REW.

**brucek** · **04-02-07, 06:58 AM**

I guess we can't tell you to move your IB around the room to tame that peak at 12Hz eh?

That's fairly consistent with Sonnies results when you get below 7Hz it's hard to get a repeatable measure. The output is really down in the noise, so 7Hz seems to be about the real bottom of a useful IB measure...

brucek

**Ilkka** · **04-02-07, 07:52 AM**

Quote:

brucek wrote:

That's fairly consistent with Sonnies results when you get below 7Hz it's hard to get a repeatable measure. The output is really down in the noise, so 7Hz seems to be about the real bottom of a useful IB measure...

brucek

He could try bumping the level up to say 100-110 dB in 20-30 Hz range. That way the <7 Hz noises wouldn't affect that much. Switch the meter to 80-130 dB scale.

**TonyCo** · **04-02-07, 10:55 AM**

The ringing you are getting looks a lot like what i got before i remembered to turn off CMSS, did you check the light on the sounblaster was off?

Tony

**Chrisbee** · **04-02-07, 04:42 PM**

Quote:

TonyCo wrote:

The ringing you are getting looks a lot like what i got before i remembered to turn off CMSS, did you check the light on the soundblaster was off?

Tony

Thanks, Tony. But I always check the CMSS light is out.

Any idea why my input levels are so low guys?

Ilkka, if I do a full sweep at 110dB my house will fall down!

I think the REW sweep starts at 2Hz.

The doors already shake as the sweep starts with only 75dB as the target level!

Advertisement