ProSoundWeb Community

Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => Audio Measurement and Testing => Topic started by: Frank Koenig on September 24, 2019, 06:25:12 pm

Title: Measurement Disaster -- M-Audio
Post by: Frank Koenig on September 24, 2019, 06:25:12 pm
I've been using an M-Audio "Fast Track Pro" USB interface with a Windows laptop for making speaker measurements since around 2011 or 2012. I've made 100s of measurements on various speakers, often involving loading gear, driving out to some suitable place, and hauling speakers up and down lifts repeatedly to change the vertical angle. In short, a shitton of work.

Now I'm crying in my coffee. I just discovered that when using ARTA the damn drivers sometimes operate the interface at Fs = 44.1 kHz even when everything is set to 48 kHz. This is very insidious, as it just shifts all frequency responses up by 48/44.1 = ~9%. This, apparently, is just enough to mess things up slightly while not so bad as to be obvious when doing an overall system check with Smaart. A loop-back test, which I do now and then, looks perfect. I had no clue until I exported some curves to Armonia and then measured the response of the DSPs to put the settings back into a model. I can convert the sampling rates of my past measurements to fix them but I have to determine which ones are affected. I can't find any clue in the files as ARTA treats them all as 48 kHz. All I can think to do (after I put the M-Audio in the driveway and run it over with the van multiple times) is repeat some measurements on each speaker and look for the frequency of any "signature" resonances. This really sucks.

I now need to buy a small interface with rock-solid ASIO drivers. I'm looking at Focusrite "Clarett2Pre" as, among other things, it has independent phantom power switches for each input. I'd prefer something with a 120 VAC power input instead of a wall-wart, but it seems like you have to get into a large rack-mount unit to get that. I very much appreciate any recommendations. Thanks.

--Frank
Title: Re: Measurement Disaster -- M-Audio
Post by: Russell Ault on September 24, 2019, 08:55:03 pm
I now need to buy a small interface with rock-solid ASIO drivers. I'm looking at Focusrite "Clarett2Pre" as, among other things, it has independent phantom power switches for each input. I'd prefer something with a 120 VAC power input instead of a wall-wart, but it seems like you have to get into a large rack-mount unit to get that. I very much appreciate any recommendations. Thanks.

Have you looked at the B-word UMC202HD? No wall-wart because it's bus-powered; never had any driver issues. Price is right...

-Russ
Title: Re: Measurement Disaster -- M-Audio
Post by: Dave Garoutte on September 24, 2019, 09:41:29 pm
Bummeroskies :-[
I've done similar things while machining parts.  Made a very reasonable assumption that bit me big time later.

I use the Roland Octa-Capture as used by the SMAART class instructor.
Title: Re: Measurement Disaster -- M-Audio
Post by: Barry Reynolds on September 24, 2019, 10:06:57 pm
Bummeroskies :-[
I've done similar things while machining parts.  Made a very reasonable assumption that bit me big time later.

I use the Roland Octa-Capture as used by the SMAART class instructor.

You guys probably know what happened to Van Halenís live show when the digital out of a synth doing backing track of ďJumpĒ was set to 44.1 and the PA was running at 48K.   :-\

The backing track played in between keys.  The musicians just continued in the record key.
Title: Re: Measurement Disaster -- M-Audio
Post by: Frank Koenig on September 24, 2019, 11:59:37 pm
You guys probably know what happened to Van Halenís live show when the digital out of a synth doing backing track of ďJumpĒ was set to 44.1 and the PA was running at 48K.   :-\

The backing track played in between keys.  The musicians just continued in the record key.

At least Stonehenge in "This is Spinal Tap" was off by a factor of 12. This is more like the Hubble telescope. Thank you all for the replies. --Frank
Title: Re: Measurement Disaster -- M-Audio
Post by: Martin Morris on September 25, 2019, 03:08:54 am
Have you looked at the B-word UMC202HD? No wall-wart because it's bus-powered; never had any driver issues. Price is right...

-Russ

I think Ivan uses the UMC202HD with a Mac.

I tried the UMC202HD with Systune - gave up getting ASIO to work on a windows machine. Lots of emails back and forth to AFMG .... Grrrr. Went with a focusrite 2i2 which has been rock solid.

Cheers
Martin
Title: Re: Measurement Disaster -- M-Audio
Post by: David Sturzenbecher on September 25, 2019, 07:11:34 am
Went with a focusrite 2i2 which has been rock solid.

Cheers
Martin

My 2i2 was rock solid until it became a rock. The analog bits of it worked, but nothing digital on it would... and therefor hit the trash.
Title: Re: Measurement Disaster -- M-Audio
Post by: Mark Wilkinson on September 25, 2019, 08:50:31 am
Uhg, sorry Frank.

How many mics? 
If just one, and for a bit of $$, the RME baby face pro is really nice. Precise preamp levels set in software. Runs off USB power so no wall wart.  (Not sure if the USB would handle a high current-draw mic like a M30 though, without a power supply.  I'd ask the Rational Acoustics folks.) Oh, and you could use it for two mics, but your reference feed from it will be a single ended headphone out instead of balanced. It doubles as a neat little mixer and DAC.
I was at an advanced practicum held at the RA office in CT last week and heard Roland wants to be done with the Octa-capture. It's a great unit but not sure I'd go that way now.
We were told to bring our sound cards...I think I saw more UMC404s than anything else, but it was a very small group  (5).
Speaking of the UMC 404HD...that would be my vote for solid and value.  Wall wart though. 
But I think best for field work would be a XR-18 (or X-32 rack).  Plenty of channels and recallable preamps, which is huge plus...especially if SPL monitoring takes off.
Title: Re: Measurement Disaster -- M-Audio
Post by: Nathan Riddle on September 25, 2019, 10:57:26 am
Uhg, sorry Frank.

How many mics? 
If just one, and for a bit of $$, the RME baby face pro is really nice. Precise preamp levels set in software. Runs off USB power so no wall wart.  (Not sure if the USB would handle a high current-draw mic like a M30 though, without a power supply.  I'd ask the Rational Acoustics folks.) Oh, and you could use it for two mics, but your reference feed from it will be a single ended headphone out instead of balanced. It doubles as a neat little mixer and DAC.
I was at an advanced practicum held at the RA office in CT last week and heard Roland wants to be done with the Octa-capture. It's a great unit but not sure I'd go that way now.
We were told to bring our sound cards...I think I saw more UMC404s than anything else, but it was a very small group  (5).
Speaking of the UMC 404HD...that would be my vote for solid and value.  Wall wart though. 
But I think best for field work would be a XR-18 (or X-32 rack).  Plenty of channels and recallable preamps, which is huge plus...especially if SPL monitoring takes off.

FWIW, I've used my UMC404 without wall wart for a long time, no issues. Im sure it depends on the mic, but my Behringer mics do fine even loaded up with 3 mics and a loopback.

I am enjoying the babyface pro though.
Title: Re: Measurement Disaster -- M-Audio
Post by: Erik Jerde on September 25, 2019, 12:05:10 pm
For single mic & loopback work Iím using a sound devices usbpre2 with a m30.  If I was buying today Iíd get something with software controlled preamps so that I didnít have to calibrate spl every time I used the interface for something else.
Title: Re: Measurement Disaster -- M-Audio
Post by: Mark Wilkinson on September 25, 2019, 12:12:24 pm
FWIW, I've used my UMC404 without wall wart for a long time, no issues. Im sure it depends on the mic, but my Behringer mics do fine even loaded up with 3 mics and a loopback.


Didn't know it would power via USB ! Thx
Title: Re: Measurement Disaster -- M-Audio
Post by: Daniel Levi on September 25, 2019, 02:34:35 pm
I know my ESI 1010e PCIe interface gives out ~12V un-powered with the full 48V needing a 12V PSU.
Title: Re: Measurement Disaster -- M-Audio
Post by: Don Ernst on September 25, 2019, 05:25:34 pm
I use a Focusrite Saffire 6 which is a 2 in 2 out interface.  It is USB powered and has been good until recently when Output 1 has started making noise and the signal level drops.  I switched to Output 2 and that works fine.  I may still replace it soon. 

I also have an Allen and Heath Zed10i mixer that is a 4x4 USB interface so I can use more than 1 mic for measuring.
It is a bit larger than the Focusrite, and is not USB powered.  It has been rock solid.
Title: Re: Measurement Disaster -- M-Audio
Post by: Martin Morris on September 25, 2019, 11:19:32 pm
I've been using an M-Audio "Fast Track Pro" USB interface with a Windows laptop for making speaker measurements since around 2011 or 2012. I've made 100s of measurements on various speakers, often involving loading gear, driving out to some suitable place, and hauling speakers up and down lifts repeatedly to change the vertical angle. In short, a shitton of work.

Now I'm crying in my coffee. I just discovered that when using ARTA the damn drivers sometimes operate the interface at Fs = 44.1 kHz even when everything is set to 48 kHz. This is very insidious, as it just shifts all frequency responses up by 48/44.1 = ~9%. This, apparently, is just enough to mess things up slightly while not so bad as to be obvious when doing an overall system check with Smaart. A loop-back test, which I do now and then, looks perfect. I had no clue until I exported some curves to Armonia and then measured the response of the DSPs to put the settings back into a model. I can convert the sampling rates of my past measurements to fix them but I have to determine which ones are affected. I can't find any clue in the files as ARTA treats them all as 48 kHz. All I can think to do (after I put the M-Audio in the driveway and run it over with the van multiple times) is repeat some measurements on each speaker and look for the frequency of any "signature" resonances. This really sucks.

I now need to buy a small interface with rock-solid ASIO drivers. I'm looking at Focusrite "Clarett2Pre" as, among other things, it has independent phantom power switches for each input. I'd prefer something with a 120 VAC power input instead of a wall-wart, but it seems like you have to get into a large rack-mount unit to get that. I very much appreciate any recommendations. Thanks.

--Frank

Langstons review of the RME Babyface Pro - scroll down a bit - Highly recommend !! https://forums.prosoundweb.com/index.php/topic,154105.msg1553613.html#msg1553613 (https://forums.prosoundweb.com/index.php/topic,154105.msg1553613.html#msg1553613)

Edit: this link shouldn't require any scrolling? - different browser maybe not ...


Cheers
Martin
Title: Re: Measurement Disaster -- M-Audio
Post by: Keith Broughton on September 26, 2019, 08:23:14 am
FWIW, I've used my UMC404 without wall wart for a long time, no issues. Im sure it depends on the mic, but my Behringer mics do fine even loaded up with 3 mics and a loopback.


Same here and no problems...so far. :)
Title: Re: Measurement Disaster -- M-Audio
Post by: Ivan Beaver on September 26, 2019, 10:29:26 am
I think Ivan uses the UMC202HD with a Mac.

I tried the UMC202HD with Systune - gave up getting ASIO to work on a windows machine. Lots of emails back and forth to AFMG .... Grrrr. Went with a focusrite 2i2 which has been rock solid.

Cheers
Martin
Sometimes I use the UMC202 with my windows machine.  I don't have a Mac.

For simple single mic measurements I use either the UMC 202 or a Sound Devices original USB pre.  It is more a matter of what I have handy or what is in the tool kit or on the desk



Title: Re: Measurement Disaster -- M-Audio
Post by: Frank Koenig on September 26, 2019, 01:36:56 pm
I wrote a little R program to convert the sampling rate of my saved impulse responses (DFT at old Fs, pad with zeros, IDFT at new Fs). It looks like it works and fortunately I used a somewhat systematic naming convention for the files so I have a way of salvaging measurements in bulk. I set about identifying the corrupted measurements by comparing a quick near-field, on-axis frequency response with the stored data and, in the case of the one speaker I've tried so far, that looks like it may work. All I need is to match up a couple of clear peaks or dips and it's not too hard to see if they're at the same frequency or off by 9%.

No new interface yet but I know how to trick the old one into working correctly AND I CHECK IT. Going forward I may just do everything at 96 kHz. Memory is cheap, computers are fast, and it should avoid some pesky artifacts in the top octave.

Thanks all for all the good info on interfaces.

--Frank