UMIK-1 or any USB MIC unsuitable for timing measurements?

[levimax]

I have been happily using a UMIK-1 and REW for DIY speaker tuning and room EQ and it seemed like it was working OK. I am starting on a new speaker project and decided to get a little more serious and look at CAD help. When reading through the documents for VituixCAD I came across this in the "how to measure using REW section".

Note! Single channel measurement systems such as USB microphones (with latency variations by default) are not recommended for speaker engineering due to timing and phase variations and normalizations. REW should NOT be used without electrical loopback as timing reference or cal and timing reference for acoustical measurements to avoid timing manipulation by the program.

Is it really true that all USB microphones a fatally flawed for timing measurements? I notice the latest versions of REW has timing references before and after sweeps so maybe the USB timing issues have been resolved?

Assuming USB MIC's are fatally flawed for timing measurements I do have some interfaces I could use as recommended but I don't have a "regular" microphone and know nothing about them except there are a lot of different configurations and use cases. What would be a good "regular" microphone for speaker measurement purposes? (I doubt I would use it for anything else).

 

[staticV3]

Please don't use black text color.
Many people use ASR's dark mode, and black text is unreadable there:

 

[levimax]

Please don't use black text color.
Many people use ASR's dark mode, and black text is unreadable there:
View attachment 358215

I had no idea, I just changed red to black. I edited post and think I fixed it

 

[DWPress]

Your post color is fixed but I don't have an answer for you. I've seen this statement in the past though, hopefully John will chime in and have an answer. I still have my Dayton UMM-6 and a Focusrite mic preamp from the days before the Umik came to be, hope I don't have to dig it out again next time I want timing references.

 

[Keith_W]

The problem with USB mics is that it is not clock sync'ed to the DAC, which can give timing inconsistency. For this reason, Audiolense does not recommend USB mics although it will work. Acourate does not allow USB mics at all.

 

[Blumlein 88]

I assume you know with REW as mentioned in your post, that you can do timing acoustically or with electrical loopback. If you do it acoustically REW can fire a pulse from the opposite speaker as the timing reference (or the same one being measured). Which is good enough for most EQ purposes. Obviously a potential problem for design measurements. As others have answered on USB you have variable latency. I would think with a given piece of gear that latency might be quite stable, but I can see how even a few milliseconds would cause problems if you aren't doing simple EQ. 5 msec latency is fine for recording, but would allow sound to travel 5 feet in that time. If the latency is stable I don't see a problem, but the part in red implies REW itself responds with variable timing. Which probably as Keith_W said because the DAC sending out the signal and the UMIK don't have locked clocks between them and there will be variable amounts of timing drift.

I am a couple updates behind on REW so I don't know if the timing measurements before and after help the issue. I suppose they could compensate in software, but I don't know the answer to that. Maybe John from REW could answer for us. @JohnPM

Though 4 years old this PDF shows details of how to do some more sophisticated things with REW.

https://www.scribd.com/document/578793926/REW-101-HTS-Current-Version

 

[Rick Sykora]

Having the loopback reduces errors and, unless the errors are easily characterized, they end up in the input data (and thereby affecting the design). As it stands, you will have the latency caused by the USB interface that needs to be reconciled with the latency due to mic distance from the speaker. REW has been improved to make determining the air transmission latency easier but will be off if other parts of the chain are adding time errors. From what I have seen, am sure others have designed speakers without driver phase data but if you want SOTA simulation, you need it to be accurate.

Since @kimmosto wrote the USB mic warning, am sure he can articulate the issue more accurately than I.

 

[Sokel]

Depending on your interfaces you can start with something like this as @staticV3 has suggested lots of times:

Primo EM258 Mono module, with 3.5mm plug, 1.5m

A tiny (6 mm dia) high quality Primo EM258 capsule, wired as a complete working microphone module, ready to build into a housing or enclosure. Requires capsule and cable to be supported. These rubber holders fit the capsule: Primo 6mm holder and 6 mm Microphone Holder Good ultrasonic response...

micbooster.com

 

[phofman]

IMO the problem with timing is that USB mics have no second analog input to which the loopback analog signal from the DAC could be fed. As a result REW cannot measure the acoustic delay - the difference between electrical (loopback) and acoustic (microphone) signals.

 

[Jukka]

As far as I know, by following REW instructions you can make meaningful measurements with USB mics, but they are valid and comparable only for the session. Anything that involves a microphone depends on how the mic is placed, so if you swap a speaker you will have invalidated comparability, because it's difficult to place a mic perfectly as it was before, relative to the speaker. Timing data for crossovers has to be accurate to a millimeter. Otherwise... actually anyways you will need to make adjustments based on verification measurements.

Similar problem exists with USB-interfaces for analog mics. The channels may not be in a needed sync and similar drift can occur as with USB-mics.

That said, you should be able to get started with USB-mics and if you do it right, you most likely will not see anything surprising if verifying with a proper setup.

 

[Dumdum]

A Dayton xlr mic and a Scarlett solo or 2i2 works very well and is a cheap entry

I also make up an rca lead to double 1/4” ts jack to take measurements

I use twin rca as I have Dsp and can swap it for a stereos RCAs in cars where I use Smaart extensively for measurements

 

[Rick Sykora]

Forgot about my other thread and @JohnPM had addressed the USB phase question here:

Designing an active 2-way speaker crossover

https://kimmosaunisto.net/Software/VituixCAD/VituixCAD_Measurement_REW.pdf "Note! Single channel measurement systems such as USB microphones with latency variations by default are not recommended for speaker engineering due to timing and phase variations and normalizations. REW should not be...

www.audiosciencereview.com

 

[gnarly]

For the purpose of DIY speaker building.....
I don't have a big problem with the USB mic /sound card combo having potential timing sample differences when measuring driver sections.....

Because, those sample differences are so vastly swamped by the USB mic's need to use an acoustic timing reference, that sweeps from 5kHz up.

Heck, even a dang two-way most often doesn't have good measurement content above 5 kHz, for the lower section..
3-ways, and above.....forget it.
Subs alone? hahaha

 

[ernestcarl]

For the purpose of DIY speaker building.....
I don't have a big problem with the USB mic /sound card combo having potential timing sample differences when measuring driver sections.....

Because, those sample differences are so vastly swamped by the USB mic's need to use an acoustic timing reference, that sweeps from 5kHz up.

Heck, even a dang two-way most often doesn't have good measurement content above 5 kHz, for the lower section..
3-ways, and above.....forget it.
Subs alone? hahaha

I use another speaker for the acoustic time reference. However, even then, my results can sometimes be off or a glitch can occur in the middle of a sweep so I now do repeated measurements and discard measurements that look off -- by looking at the IR/step and phase.

 

[gnarly]

I use another speaker for the acoustic time reference. However, even then, my results can sometimes be off or a glitch can occur in the middle of a sweep so I now do repeated measurements and discard measurements that look off -- by looking at the IR/step and phase.

I don't get you mean by using another speaker?
How do you account for the difference in distance to the acoustic centers of the drivers' sections?
When you can't measure time below 5kHz?

I don't see how, but love to learn

 

[ernestcarl]

I don't get you mean by using another speaker?
How do you account for the difference in distance to the acoustic centers of the drivers' sections?
When you can't measure time below 5kHz?

I don't see how, but love to learn

As I said, I use completely different speaker (so not exactly the DUT) that is already setup nearby as my acoustic time reference. When doing individual driver measurements recently on my RX699, for example, I used my Neumann KH120 as the reference and told REW to automatically adjust the time offset for every new following measurement.

Center with good directivity for Revel M126Be?

This was my issue as well. I tried so hard to just stick with stereo, but many shows and movies the dialogue just gets lost in the mix. I try to avoid the dialogue boost function whenever possible considering it just boosts everything in those frequencies and can get boomy/muddy. Being able to...

www.audiosciencereview.com

I mute/unmute (or plug/unplug from the amp) individual drivers of the tested speaker using another external xo DSP device -- in this case, a miniDSP 2x4HD.

 

[gnarly]

I dunno here. For DIY speakers.
I've a made multiple thousands of dual channel measurements, studying fixed time delay to acoustic centers.
Studying the effect of IIR and FIR linear-phase high pass and low-pass filters, both electrically and acoustic on fixed timing delays.
(When acoustic, always measuring the same speaker from a fixed mic location)
For DIY speaker tuning, there's no way i can see that a surrogate speaker/driver could ever work for a timing ref.

Are your measurements geared towards aligning full-range speakers, already set up?
With subs, etc?
Then I can start to see what you are doing works
Other than how to get a decent sub-only timing....

 

[ernestcarl]

Are your measurements geared towards aligning full-range speakers, already set up?
With subs, etc?
Then I can start to see what you are doing works
Other than how to get a decent sub-only timing....

I suppose so, yes.

To be frank, I really do not do DIY speaker builds myself. If one were to move either speaker, the timing reference would no longer be accurate. I think it works well enough, for "testing" already finished speakers. But I did slightly re-do the xo EQ here which meant the timing reference has to work else the prediction will not match the verification measurements at all afterwards.

To be sure, I did verification measurements across different listening distances and angles and (at the very least) my "base speaker EQ" and xo adjustment works well enough:



Yet, for optimum "room EQ" I would not use the same post (base speaker) EQ correction for my desk and couch space. What sounds and measures best for the desk at less than 1 meter in the front half of the room does not work as perfectly for the couch (~3 meters distance) situated near the rear end wall of the room.

 

[gnarly]

To be frank, I really do not do DIY speaker builds myself. If one were to move either speaker, the timing reference would no longer be accurate. I think it works well enough, for "testing" already finished speakers. But I did slightly re-do the xo EQ here which meant the timing reference has to work else the prediction will not match the verification measurements at all afterwards.

Gotcha, big thanks for the clarity re your objectives/process.

And yep, for room eq with full range speakers, I can see how acoustic timing ref easily works.
Still have trouble with subs.
But I guess if you use a surrogate speaker colocated with the sub that has >5 kHz content, the worst error you have distance wise, is the difference in z-axis acoustic centers.

 

[Dave Bullet]

I have both a USB and 2 channel mic for measurements. I go to the USB one for convenience most of the time.

The issue with USB is you have no reference timing marker for measuring the "flight time" for the impulse, which will give you absolute phase information between different driver measurements providing for reliable phase integration for crossover modeling.

However you can work around this for USB measurements in a multi-driver system via the parallel driver approach and relying on drivers essentially being minimum phase devices (only works when using raw driver measurements as any components in the signal path may alter minimum phase behaviour).

The approach is as follows:
1. Measure tweeter
2. Measure woofer
3. Measure both tweeter and woofer in parallel

Make sure you don't change drive levels nor mic / baffle placement. i.e. ideally you will have both driver terminals exposed and just clip to the next driver to measure

4. Load measurement #3 into your simulator
5. Load both driver responses into your simulator and wire them up in parallel to the source / amp. Ensure you check minimum phase (derived phase) or extract this outside the simulator and load with the driver response
6. Increase woofer "Z" (acoustic offset) until the summed #5 response, aligns to combined measurement response (you loaded into step #4)

You now have a reliable Z offset for crossover simulation.

The downside is this is a single axis technique as Z will change off axis (the amount depends on the drivers / placement).