Recommend a good mic preamp for measurements

[lukeamdman]

your line in-out loopback looks like a mic input

 

here's my 6i6

 

cyan is line out to mic in (front combo jack)

red is line out to line in (rear TRS)

purple is line out to inst in (front combo jack)

 

comparison.png

 

the other measurements look like some sort of feedback. Can you show a pic of mixcontrol?

 

 

 

 

 

Besides the sampling rate this is basically all default settings. 

[3ll3d00d]

interesting, I didn't realise they had a new mixer app out (I'm still using mixcontrol). 

[lukeamdman]

interesting, I didn't realise they had a new mixer app out (I'm still using mixcontrol). 

 

Mixcontrol looks better than this.  New one is a very vague interface. 

[3ll3d00d]

actually now I see this is for the 2nd gen scarlett interfaces, is that what you have? mine is a 1st gen

[lukeamdman]

actually now I see this is for the 2nd gen scarlett interfaces, is that what you have? mine is a 1st gen

 

Yes, this is a 2nd gen. 

[lukeamdman]

Also, I RMA'd the first 6i6 I received after seeing the same results.  Focusrite tech support had no answer for it either and recommended the RMA.

 

I just got this one today and it's the same so...

[3ll3d00d]

Your mixer looks like you are sending a mix to the monitor output, the daw channel output is cut off the bottom of the pic though so not sure where that is going. Perhaps you can confirm?

 

I suggest you switch to direct playback and repeat your test anyway.

[lukeamdman]

Your mixer looks like you are sending a mix to the monitor output, the daw channel output is cut off the bottom of the pic though so not sure where that is going. Perhaps you can confirm?

 

I suggest you switch to direct playback and repeat your test anyway.

 

You're onto something here...

 

"Direct Routing" mode

 

 

 

Line out to combo in using TRS to TRS and "Line" setting:

 

 

 

Same thing but with the "Instrument" setting:

 

 

Line out to mic in with a TRS to XLR cable:

 

 

All 3:

 

[SME]

Nice.  Glad you got that cleared up.

 

I had to re-read this thread because I'm looking at needing that analog phantom power source soon to plug my EMM-6 into my Motu A16.  But I guess everyone here is using a USB interface.  I could just try buying something cheap from Guitar Center or whatever, but I'd at least like to be able to measure whatever roll-off it has.  Is that possible if the roll-off depends on interaction with the mic itself?  Is my mic calibration from Cross Spectrum useless?  I was disappointed to see a lot more low-end roll-off on the EMM-6 than the UMIK-1, but I gotta have an analog mic to use on the Motu.  Hmm.

 

Edit: If I buy a cheap phantom power interface and I run a sweep with the UMIK, then I replace the UMIK with the EMM-6 without touching the stand and re-run the sweep, I can use the respective calibration files along the calibration for the Motu input to back-out the roll-off of the interface.  Ugly and not very accurate, but it can hopefully at least get me in the ballpark for < 10 Hz where I'm likely to have problems.

[lukeamdman]

Nice.  Glad you got that cleared up.

 

I had to re-read this thread because I'm looking at needing that analog phantom power source soon to plug my EMM-6 into my Motu A16.  But I guess everyone here is using a USB interface.  I could just try buying something cheap from Guitar Center or whatever, but I'd at least like to be able to measure whatever roll-off it has.  Is that possible if the roll-off depends on interaction with the mic itself?  Is my mic calibration from Cross Spectrum useless?  I was disappointed to see a lot more low-end roll-off on the EMM-6 than the UMIK-1, but I gotta have an analog mic to use on the Motu.  Hmm.

 

Edit: If I buy a cheap phantom power interface and I run a sweep with the UMIK, then I replace the UMIK with the EMM-6 without touching the stand and re-run the sweep, I can use the respective calibration files along the calibration for the Motu input to back-out the roll-off of the interface.  Ugly and not very accurate, but it can hopefully at least get me in the ballpark for < 10 Hz where I'm likely to have problems.

 

My first 6i6 was definitely defective.  Output 3 didn't work, and even a line out to line in measurement looked crazy.  Also, trying to "reset to factory defaults" would cause a BSOD...

 

I recommend going away from the cheap USB mics or any mic that needs a calibration file for bass.

 

I sent my CSL calibrated UMIK to Josh to compare to his M30:

 

 

 

Even with a cal file it was still down ~4db more at 5hz than the M30 (no cal files).  Also, Josh had to offset the SPL for the UMIK because it was a few DB low as well.

[3ll3d00d]

If I buy a cheap phantom power interface and I run a sweep with the UMIK, then I replace the UMIK with the EMM-6 without touching the stand and re-run the sweep, I can use the respective calibration files along the calibration for the Motu input to back-out the roll-off of the interface.  Ugly and not very accurate, but it can hopefully at least get me in the ballpark for < 10 Hz where I'm likely to have problems.

 

which rolloff are you attempting to counteract there? that of the phantom power supply? 

[3ll3d00d]

@lukeamdman if you get a chance, could you run a sweep to measure the loopback and show the IR pls. I'm just curious if they've fixed the HF ripple seen in the gen 1 interface.

[lukeamdman]

@lukeamdman if you get a chance, could you run a sweep to measure the loopback and show the IR pls. I'm just curious if they've fixed the HF ripple seen in the gen 1 interface.

 

Do you have a screenshot of what you're looking for?

[3ll3d00d]

Do you have a screenshot of what you're looking for?

there is a pic in http://www.hometheatershack.com/forums/rew-forum/147730-soundcard-calibration-sanity-check.html

[lukeamdman]

there is a pic in http://www.hometheatershack.com/forums/rew-forum/147730-soundcard-calibration-sanity-check.html

 

2nd gen looks identical to me:

 

[Contrasseur]

Hijacking this thread for a measurement setup. I'm hoping to do the same thing, become an additional source of Data-Bass.com worthy measurements. Cost is a big factor, so what would you all consider the cheapest setup I could use to get sufficiently accurate results? In the short term, I intend to be measuring much smaller and more affordable subwoofers, so I'll only need maybe 125-130dB at 1m.

 

Which of the following would you consider suitably accurate for these purposes, and which ones will need an upgrade? If so, what's the most cost-effective option I can upgrade to?

 

Interface

Currently have Behringer UM2. Upgrading to Behringer X32 rack soon

 

Impedance measurement

Precision resistor bridge from interface headphone output

 

Amplifier

Behringer INUKE3000DSP

 

SPL Calibrator

None currently, just using Galaxy Audio CM-130

 

Microphone

Dayton Audio EMM-6

 

Software

REW

[SME]

which rolloff are you attempting to counteract there? that of the phantom power supply? 

 

Yes.  The phantom power supply provides +48V DC to the mic and because this supply voltage is presented on the same lines that the mic signal is, the power supply design must use coupling capacitors to block this voltage from the downstream line level input where it can fry equipment.  DC coupling caps don't brick wall at DC, so all such power supplies are likely to exhibit a roll-off of some kind.I

 

I may be missing something here.  It looks like phantom power supplies are designed to fairly strict standards, at least when they are implemented competently.  It could be that a compliant designs have consistent roll-off for any given mic.  In which case, a mic calibrated with and including a standardized phantom power supply may roll-off the same on a another standardized phantom power supply.

 

This could also be wishful thinking on my part.  I could try emailing Cross Spectrum to see if this is the case or if there's a particular phantom power supply they recommend to ensure that the calibration points down to 5 Hz are valid.  Heck, if the calibration includes the standardized phantom power, then that could explain the stronger roll-off I see vs. the UMIK.

 

In the meantime, I thought of a work-around that might allow me to keep using my UMIK-1 while still getting reliable automatic time synchronization of

measurements involving different sources.  It's based on the technique I've used in the past involving simultaneous sweeping of a chosen reference source at teh same time as the test source, and then analyzing the IR of the combined response together with the individual IRs.  This technique worked fairly well before, but it wasn't reliable enough to completely automate.  I think I just figured out a better way to do the analysis using a frequency domain multi-linear regression that will hopefully be nearly fool proof.  That's good because for my DSP optimizations, I'll be sweeping 16 sources in succession at each of several measurement locations.  One round of in-room measurements could easily involve 100-200 IRs.  There's no way I want to manually time align those.

[3ll3d00d]

Interesting point. I guess you could build your own to get a known rolloff - for instance http://sound.westhost.com/project96.htm

 

Perhaps I just don't have enough subs to care too much about a few dB here or there at ulf as I find my subjective preference is more important. Mind you there is also the question of whether a better microphone gives audibly better results at high frequencies (for full range correction) which, if such a mic also has more extended lf response, is another aspect to consider.

[Ricci]

A while back I did a quick SPL calibrator comparison for Luke and looked at his Umik a bit. The calibration file was pretty good down into the deep stuff. Was still a bit lower than the m30 with no cal file but close. SPL is limited and THD had some curious little artifact. Nothing huge but it was a bit higher. It however did not fit a 1/2" SPL calibrator at all. Just an FYI.

 

Contrasseur try to use the amplifier for impedance measurements of larger bass drivers for a bit more juice. Also try to save up for an SPL calibrator. My CM130 I had was not terribly accurate.

[Contrasseur]

Would the Chinese Special (something like this) be trustworthy enough?

 

What value and power rating of precision resistor do you use for power amp-driven impedance measurements?

 

Also, how do you calibrate 2.83v or 2v measurements?

[Ricci]

Would the Chinese Special (something like this) be trustworthy enough?

 

What value and power rating of precision resistor do you use for power amp-driven impedance measurements?

 

Also, how do you calibrate 2.83v or 2v measurements?

 

I still have 2 of those. They work as advertised. For the money I don't think you can do much better unless you find a killer used deal. Also have compared a third one. My big money ACO Pacific 511E mostly sits in the bag while the regular ND9 gets the wear and tear. It's within 0.5dB and I know that so I might as well let the cheapy take the abuse.

 

Currently I'm using a precision low tolerance 8ohm resistor from PE.

 

Voltage readings go like this

Plug in speaker to amp

Run 55Hz sine wave out of REW

Adjust gain until the correct voltage is read at the speaker terminals using a quality true rms voltage meter. 

Log all gains and voltage

Keep track of gain for each measurement throughout

Calculate applied voltage later

 

I use 10x voltage now. 14.1 volts for 1 ohm, 20v for 4 ohm and 28.3 volts for 8ohm nominal. Your DMM will measure with better accuracy with a little more voltage versus using the millivolt scale and attempting to hit exactly 2 volts or whatever. Also it is easier to get the gain correct at the higher voltage. I also use 10x voltage for sensitivity measurements now. It better represents actual signal strength during use and more importantly greatly increases the SNR against the background noise. This is then normalized back -20dB lower to match the "1 watt" measurement.

 

Also I highly suggest having an external unit with precision digital gain adjustment for fine tuning. I use a DCX2496. It allows 0.1dB adjustment of output gain. It can be impossible to reach the desired voltage using 1dB gain increments. Analog devices are also much harder since it can be impossible to be repeatable with the gains. My setup runs like this...I use an external sound card. There is no volume adjustment or any of that inside windows. Only the gain in the measurement software. I don't even use that. I set the gain in the various softwares I'm using to the same output level (strong as possible but not clipped) and also set the mic input gain to minimum for headroom. The input and output gains on the SC I've already done extensive testing on and I know they will not clip at their settings used. These are also left alone. The only gains I use to adjust the signal after setup are in the DCX2496 which has a range of -30 to +30 which is plenty. The amp is always set on it's least sensitive setting. I never touch the other various gains. Only the DCX2496 if possible. I do sometimes have to adjust in the software for active systems, which have a very hot gain structure usually. This doesn't matter though since the drive voltage isn't recorded. Once the initial "100w", 20v or 28.3v drive signal is established I log the gains and save this as a "key". From there all I need do is keep track of the gain setting in the DCX and I can calculate the applied voltage later based on the gain structure. Up until amplifier limiting or clipping anyway. That is the way it is done. I'm certainly not measuring the voltage manually for each and every measurement. It's not needed as long as the gain structuring is kept strict track of.

[lukeamdman]

I ended up sending the 6i6 back and exchanged it for the Motu Ultralite AVB.  Mojave has the Motu and a line out to mic in loopback shows its -1db point is just below 3hz. 

 

As far as I can tell the 2nd gen 6i6 has more LF roll-off than the previous version. 

[dgage]

Thanks for that info. I do like the Motus and like you, I owe that to Mojave.

[lukeamdman]

Motu Ultralite AVB.  Line out to mic in:

 

 

[dgage]

Interesting.  Not sure if it is coincidence but every set of measurements I've taken of my 24s in a room don't start to roll off until 6 or 7 Hz.  I believe the subs likely roll off around there but I wonder how much I have to attribute to my mic preamp.  Hmmm.  Will have to ponder ways to address.