Asio(4all) locked at 44.1 kHz for measurement

[Salt]

With Asio4All no matter if REW, Arta or HFD is in use: the sampling rate remains at 44.1 kHz, and the need for buffer for a single sweep is almost max.
The environment is RME ADI 2/4 pro and Umik 2, and all these would go for at least 96, if not 192 kHz.
What is the elephant I ignored?

 

[staticV3]

Use native Wasapi Exclusive wherever you can (e.g. Java EXCL in REW) and FlexASIO instead of ASIO4All anywhere else.

 

[Zinda]

I hate using REW! So many things like this constantly creating issues, having to reset it up each time to do a quick pass is a pain. Plus it ends up with a flat line as the result. I've found only 1 app that works if you have a good microphone in you're phone, I compared a bunch of phone mics against a Dayton test mic and found several phones that I saw had more accurate results than the Dayton. The moto phones have excellent mics and even Blu phone for $25 brand new was on par with the $30 Dayton test mic. The 1 app I found to work well is named "Spectrum RTA" (exactly like that) written by a Russian guy. Leonid. He's a nice guy that will reply to direct emails. His latest version works on my tablet but not on my new phone, I had to locate an older version to see results. His app is no longer supported on Play store but there is a version posted there. I had to find alternative sources for old versions like on Aptiod and other Android app sites.

What's nice about using an app besides the obvious convenience is his app has all tool built in. No rewiring an no computer fan noise while testing. The results are in a percentage not db but the results show repeatable exact results from test to test and show peaks and dips in a big way, smoothing is infinitely variable there's a pretest background test that compensates for found floor noise, takes 2 seconds to complete.

Just a suggestion for those who understand testing but don't care to put up with the hassle of REW just to see a flat line as the result. The tests are only for setting up your speakers and does not have other options but it is complete in that sense. It does allow external mics and Bluetooth as the signal output if you dont have an aux type audio output jack or don't want to use it. There are no extra cables needed to use it. Your mic will determine how accurate the results are. My tablet has a very poor mic and is unable to be used as a reading device. The size of a phone mic is exactly the same as what's used in test mics, I can see a possibility of replacing a phone mic with a test mic very easily. I, considering this for my tablet since it's much easier to see what's going on while running the app. Real time adjustments are possible when using white, pink or other noise generators, single tones can be produced with 0.X precision, useful for finding resonance and vibrations. Takes a few minutes to run through the settings but is simplistic in design and the functionality is spot on.

 

[staticV3]

@Zinda what OP faces is not an REW issue, it's an ASIO4All issue.

Simply use REW's built-in Java EXCL input type and all is good. Stable, bit-exact, consistent, mixed bit depth capable.

 

[unpluggged]

The environment is RME ADI 2/4 pro and Umik 2

It's an atrocity to use an emulator/wrapper when you have native best in class RME ASIO drivers.

Besides, as already had been said here, there is no need to use ASIO in REW.

 

[Zinda]

It seems like every time I go to set it up there's some volume control that's missing in Windows or there's an incorrect setting in REW and I spend my evening reading through tons of posts for nothing. That's why I searched for an easier method to basically just check where I'm at when I start to hear something out of balance. I'm more worried that both sides are putting out the same exact info to hold the imaging and staging steady. It's nearly impossible to guess . which out of 31 bands it is and white noise tends to be the best way to flatten response since it can be done in real time. I can run a sweep 5 times before my computer even shows an image on the screen and be done adjusting before windows is in ready state not to mention firing up REW and changing cables for everything.

 

[Zinda]

@Zinda what OP faces is not an REW issue, it's an ASIO4All issue.

Simply use REW's built-in Java EXCL input type and all is good. Stable, bit-exact, consistent, mixed bit depth capable.

I'll bet that if you tried the Spectrum RTA (it's free with NO PRO or paid versions) app you'd like it as long as you have a decent phone mic and replacing a mic in a phone or better yet a tablet, could make it a nice packaged set up that would be used just for speaker set up. It has no waterfall chart or phase info but those can be found for free in other apps that are basically useless for calibrating speakers.

I'd like to start comparing white noise tests against each other. I've always wondered how they can be trusted to even contain the correct noise patterns and how they keep the tones at the correct volume. I've always felt test programs have all been flawed by how they produce the test waves. Since it seems as though each program shows a slightly different plot how do we know which one is correct. Until I see 2 programs show me the same exact plot, when I use the same mic and same position for each test, I'll never trust any programs to be the one to trust. One thing I can do with Spectrum RTA, is see exact duplicate plots when I run multiple tests, that's far more than I see with other RTA programs, most can't even give similar results and never exact duplicates (or within reason).

I have ran a few white noise test tones through the app and I can confirm there's a huge difference between them on the RTA plots, also while running tone generators I've seen (and heard) some frequencies are noticably louder than others rendering their test unusable. There's even a slight difference when running through Bluetooth and a cable connection while running the same exact tests using the same equipment. I haven't tested different Bluetooth devices yet but I presume there could be a large differences between devices receiving and decoding the Bluetooth signal.

 

[MAB]

I'll bet that if you tried the Spectrum RTA (it's free with NO PRO or paid versions) app you'd like it as long as you have a decent phone mic and replacing a mic in a phone or better yet a tablet, could make it a nice packaged set up that would be used just for speaker set up. It has no waterfall chart or phase info but those can be found for free in other apps that are basically useless for calibrating speakers.

Seems like a lot of work and hacks just to avoid learning REW, or another speaker measurement package. With few capabilities. And dramatically limited microphone performance. ‘Useless for calibration’ seems an appropriate summary.

I'd like to start comparing white noise tests against each other. I've always wondered how they can be trusted to even contain the correct noise patterns and how they keep the tones at the correct volume. I've always felt test programs have all been flawed by how they produce the test waves. Since it seems as though each program shows a slightly different plot how do we know which one is correct. Until I see 2 programs show me the same exact plot, when I use the same mic and same position for each test, I'll never trust any programs to be the one to trust. One thing I can do with Spectrum RTA, is see exact duplicate plots when I run multiple tests, that's far more than I see with other RTA programs, most can't even give similar results and never exact duplicates (or within reason).

I can get identical results across different mics, interfaces, and analysis platforms. Here is a comparison of a vintage mic compared to a USB measurement mic.

Umik 1 Vs umik 2

I'm looking to REW/calibrate a couple Funks coming my way soon so below 20hz is essential.....right? No clue what you mean by "a couple Funks", sorry.

www.audiosciencereview.com

The traces were captured on two different interfaces. With the mic’s calibration files, the results are line in line matched.

I can do this all day, across many mics and measurement hardware and software. I get results that are the same. I’m not the most experienced speaker measurement person, but I do not observe the same issues you are feeling.

I have ran a few white noise test tones through the app and I can confirm there's a huge difference between them on the RTA plots, also while running tone generators I've seen (and heard) some frequencies are noticably louder than others rendering their test unusable. There's even a slight difference when running through Bluetooth and a cable connection while running the same exact tests using the same equipment. I haven't tested different Bluetooth devices yet but I presume there could be a large differences between devices receiving and decoding the Bluetooth signal.

It seems you are struggling. You would need to provide data to allow us to help, but random differences in white noise (or even better, pink) indicate you have some really wonky generator and/or measurement issues. Or need help getting proper measurements.

So without data, no.

 

[Ellebob]

Use pink noise not white noise. White noise will be bright with elevated higher frequencies and not used for acoustic measurements. It is for some equipment but not speakers or in room measurements.