Few questions from potential AL user

[dima1stg]

Hi All,

Thinking of using AL, but need to make sure I could handle it (and it could handle me lol) before purchasing.

I'm almost 100% music listener with interests in tons of different jenres, if that makes any difference. Using 2x8 miniDSP and REW for a while, tried MSO as well. Was prettu much happy but have moved into place with terrible room acoustic - not treated open floor extenfing to dining room and kitchen with open pathway, half sloping ceiling, open 2nd floor, brick fireplace, you name it...
Current config consists of bi-amped B&W 683 S2 for mains (later may be tri-amped), may add B&W 603 as auxiliary mains for compensatory role. Subs are SVS SB-12NSD, and HSU ULS-15 MK2. In addition to usual perceptional flatness and clearness, I'll need to calculate optimal crossovers for all components. Not really interested to minimize seat-to-seat variations.
I'm a regular guy who knows very well what phase, standing waves, cancelation, interference are, but have not used advanced things like impulse response, step response, minimum and linear phase, etc., even though theoretically understanding most of these terms, have no idea how they affect listening experience and how to read graphs and make conclusions based on them.
So,
1. can a guy like me make a full use of AL without learning above mentioned advanced things, reading their graphs and making educated conclusions
2. if not, how adequate response/help can I really count on considering I'll be asking more or less basic (read "boring") questions
3. what frequency range is AL really useful for
4. what else should I be aware of

Many thanks!

 

[juicehifi]

Welcome to the forum, dima1sta!

Yes and no. There is a pretty low getting-started threshold with Audiolense, but most people want to understand everything they see in the GUI. So there is often a steep learning-by-doing curve involved. But it can also be a rather fun journey.

You can lean on the community here for help with trouble-shooting and all sorts of advice, or you can use Mitch Barnett's consulting services. Mitch will guide you through your part of the process and will create very good correction filters.

Audiolense is useful across the whole frequency range, but is typically used very differently in the top, mid and low end.

You need a multichannel sound card or dac, ideally one that also has an input that you can use for the measurement microphone. And you need a measurement microphone. There are certain combinations here that has issues with glitches in the measurement streams etc, which will distort in the measured impulses. These problems seem to surface most frequently when two usb units are used (usb dac plus usb mic). I recomment that you ask on the forum before you make the investments, as these problems can be very frustrating and sometimes they do not have a long-lasting fix. If you have gear that produces unreliable measurements you will always be uncertain about whether your next measurement is OK or not. But if you have reliable gear, you can trust the measurement every time. And obtaining reliable measurements is really the only topic that can be difficult here.

At the moment, microphones from Isemcon seems to stand out as the most frequently recommended, but any individually calibrated measurement mic that is hooked into an analog feed of a sound card should do.

I hope this helps, and I also hope some of the other users will add their own getting-started experiences here.

 

[dima1stg]

Thank you, Bernt.

I indeed use UMIK-1 mic (USB) and some generic USB-TOSLINK converter for output. REW doesn't have issues with these.
I do all measurement via a lalptop, so there is not much I can do with regard to altering this combination. Is there a trial version of AL that I could use to see if this combo works? Alternatively, can AL import mesurement data from REW, either directly from mdat file or via REW's export (like MSO users do)?
Also, how do I get in touch with Mitch regarding his services?

Many thanks!
Jim

 

[jrobbins50]

My understanding is that no, you can’t import measurement data from REW into AL.

I cannot sufficiently underscore the warning Bernt offers regarding your use of the UMIK-1 microphone with a USB DAC, such as the miniDSP 2X8, to perform AL multichannel measurements. I have tried both the UMIK-1 and UMIK-2 microphones on repeated occasions with my Okto dac8pro, a fine MCH DAC indeed.

It’s been a disaster, as my multiple posts in this forum, and troubleshooting with Mitch Barnett, can attest to.

Because the mic and the DAC have separate clocks, it’s almost impossible to run frequency sweeps across six or eight channels in AL for even three seconds without getting glitches in the audio stream the microphone records. AL’s magic is the timing improvement, getting your subs to hit your ears at the same precise moment as your tweeters. I could not get there properly no matter how many measurement tries, and different measurement PCs, I undertook.

Beyond frustrated, and at Mitch’s urging (pleading?), I finally acquired one of those ISEMcon microphones that Bernt references and a Motu Ultra-lite mk5 audio interface — which has the microphone preamp and 48V phantom power for the microphone built in — to use for measurements. Now, I can run eight channel sweeps in AL for a full ten seconds each, with no glitches in the audio with fully repeatable measurements.

I would not have thought that this would make a difference in the filters AL can create. Wow, did I ever not get it before. So much more punchy bass and stable imaging. I wasted so much time before.

Take a look at my system and room, and the before and after AL measurements here: https://audiophilestyle.com/profile/4425-jrobbins50/?tab=field_core_pfield_3

Bonus if you invest in the Motu unit: you’ve got a great multichannel DAC to boot (using Motu’s Gen 5 ASIO driver for both measurements and playback).

Mitch Barnett is a great help and wealth of knowledge on AL and creating excellent filters. However, his terms of service for multichannel now provide that he won’t work with a client that wants to use a USB microphone and a DAC with its own separate USB clock for multichannel measurements.

You can reach Mitch at mitch@accuratesound.ca.

AL does have a trial. I’m not sure how it works for AL XO, which is the version for multichannel you would need. Bernt can fill you in, but perhaps it would allow you to try multichannel measurements with your UMIK-1, take a screenshot of the results (as I don’t think you can save the measurements during the trial) and see if you can very closely repeat those measurements three times in a row. If not, I’m not convinced you’ll be happy with AL for your system unless you use a different measurement system. JCR

 

[juicehifi]

Thank you, Bernt.

I indeed use UMIK-1 mic (USB) and some generic USB-TOSLINK converter for output. REW doesn't have issues with these.
I do all measurement via a lalptop, so there is not much I can do with regard to altering this combination. Is there a trial version of AL that I could use to see if this combo works? Alternatively, can AL import mesurement data from REW, either directly from mdat file or via REW's export (like MSO users do)?
Also, how do I get in touch with Mitch regarding his services?

Many thanks!
Jim

The measurement issues that sometimes arise are related to hardware and drivers. Mitch and myself have examined a bunch of REW measurements in the past. Several takes sometimes give significantly different results with regards to timing and shape of the IR's, which is a clear sign that the problems appear there too. There's no way to know whether a REW mesurement is 100% OK or not when all of them are a bit different. It's more a matter of figuring out, based on various additional information, that this measuremen is probably good enogh, but that isn't.

With stable and reliable hardware you will get glitch-free data streams in and out and repeatable results down to the sample. And it becomes Plug & Play to perform a measurement. Your microphone and dac combo may turn out to work fine, but UMIK is regularly in the mix when these problems arise.

Note that these measurement problems are not relevant for frequency correction, only for time domain correction.

Audiolense works in demo mode without a license, and allows you to correct 90 seconds' of any piece of music as long as it is on a wave-file. So you can try ibefore you buy.

Mitch have access to import measurements from REW to Audiolense. This approach is not ideal, since it brings with it a series of potential errors that many users are likely to ignore. Mitch is aware of them, though. And sometimes he consults me. In any case I believe Mitch can be a of good help for you here.

 

[dima1stg]

Is my understanding correct that the issue lies with the fact that both input and output are USB, no matter what mic' brand/model is?

Is https://juicehifi.net/?page_id=31 correct download place and ver. 6.21 is the latest?

 

[jrobbins50]

I believe your understanding is correct. Yes, that’s the download page and yes, v6.21 is the most current I’m aware of and it’s what I use. JCR

 

[Valerii]

Is my understanding correct that the issue lies with the fact that both input and output are USB, no matter what mic' brand/model is?

The root cause of issue is a separate clocking of input and output, a connection type doesn't matter.

As such you need common clock for input and output, it can be done with most of studio-grade interfaces (for ex. MOTU ULTRALITE-MK5 can be a good start and it doesn't ruin your budget), or through the combining of different input and output interfaces with the same World Clock Source (but I'm not sure you need to deploy a professional multi-interface studio for a simple time domain correction).

Of course you should use the same studio interface to listen your records.

I hope the above may be useful for you.

 

[Ofer]

You did not mention what is your source. Basically you will need a device to implement the filters created by AL. Your MiniDSP won't do. It is made to implement less CPU intensive IIR filters were AL creates FIR filters. If your source is a computer then you will be OK as you can run a convolution engine of any computer. I recommend HLC but there are other options. If not then your source music (TV/CD) will need to go through a computer. For that to work you will probably need a professional audio interface.

 

[dima1stg]

@Valerii: the way my system is built is that all sources (all digital, none is analog), including meaurement signal from laptop, go to receiver' digital inputs, selected source is output to digital output that goes to miniDSP. Up to this point everything is in digital domain. Then miniDSP transfers signal to analog domain feeding multi-channel input of amp. Computer (laptop) is mainly used for measurement and rarely for music. I'm not sure of these:
1. where should mentioned MOTU ULTRALITE-MK5 (or alike) be inserted in a path?
2. on reproduction, as soon as DSP is setup, what's the need of this device?
3. if I understand you correctly, ALL folks using AL have to spent over $1K for measurement equipment alone (not counting AL license)?

@Ofer: my miniDSP model, OpenDRC 2x8, has FIR filters per channel.

 

[jrobbins50]

What receiver are you using? I’m confused why you are inputting digital into, and out of, the receiver, rather than bypassing it. What are your sources for music if not from the PC? Are you listening to SACDs on a disc player?

If I follow your setup, an audio interface such as the Motu would go between your laptop and the amps, bypassing both the receiver and the miniDSP 2X8 box.

Again, if I understand your setup, the Motu could possibly replace your miniDSP. If not, you are correct that the Motu would not be used other than for measurement. The latter is my own use case. I use my Okto dac8pro for listening because I upscale all content to DSD64 or DSD128 and the Okto plays DSD natively. The Motu caps out at 24/192.

Not everyone using AL has purchased an analog microphone and a separate audio interface. If you are using AL for two channels, as I understand most users do, then a USB microphone will do fine. It’s multichannel applications where you seek correction in the time domain for an array of speakers around the room that an analog microphone and audio interface becomes almost essential. JCR

 

[dima1stg]

Using 2 receivers. One is only for switching sources: TV, DVD player, Blue-Ray player, bluetooth adapter, and laptop for measurement and playing (via USB-TOSLINK converter). It's digital output connectc to miniDPS, from which 6 analog outputs go to the second amp analog multi-channel input used strictly for amplification and volume handling.
Speaker setup is 2.2 - stereo with 2 subs. However main speaker are bi-amped. (hence 6 channels in total). No surround of any kind. I'd think that simple fixed delay (set in miniDSP) for sub would work. As an observation of experimenting with REW, I found that small delays between lo and mid-hi channels of main speakers do flatten frequency response. Do I need a time domain correction in this case?

 

[dima1stg]

In the mean time I just downloaded AL. Selected surround version because of subs. While trying to do the first measuremet for 2.1 config - hitting the error. Please see attached.

 

[Ofer]

From what I understand of your setup the miniDSP is also your DAC. In that case a multichannel audio interface could go instead of the miniDSP only caveat is that it is not standalone and needs to be connected to a computer all the time to implement the filters. Another option is to use the audio interface for measurement only. That is, connect the audio interface to the 2nd receiver using 6 channels as you do with the miniDSP. Run measurements and create filters. Implement the filters in the miniDSP and return it to its original setup. Once measurements are done properly then you don't need the microphone and interface to create different filters and try them.
As for microphones there are good measurements mics that don't cost 1K$. I use DatonAudio EMM-6 analog microphone and it works great, comes with a calibration curve for each mic.

 

[dima1stg]

Correct, my miniDSP model is digital input to 8-hannel analog output, DSP+DAC. It has PEQ filters as well as FIR filters. Audio from laptop goes via explained audio pipe through one USB port, while there is a separate USB connection to miniDSP for management via which FIR filters are uploaded and remembered, so, once uploaded, this connection is no longer needed. Audio from computer goes not via miniDSP but rather via audio pipe that I explained earlier. While measuring, 3 USB ports might be needed: input from mic, audio out, and DSP control. While listening from computer - only 1 USB needed for audio output.

I calculated the cost of measurement equipment based on earlier @Valerii post regarding required permanent presence of equipment with "common clock for input and output" and mentioned entry level MOTU ULTRALITE-MK5 (have no idea what this device does). I'm really confused with this as different people mention totally different things.

The problem with analog mic is that I'll need to get a desktop and somewhat decent audio card - this alone may run well into $1K if not more, and then, unless it's a really good sound card, mic calibration is irrelevant as calibration must be done on entire "mic+sound card" path.
Here comes the beauty of calibrated USB mic - no sound card, no additional calibration required, can work with laptop or comp without sound card. Its digital path doesn't get affected as analog path does. This is why I'm trying to achieve the task with USB equipment as once there is a calibrated output from digital mic - as long as signal stays digital, it goes unaffected throughout entire audio pipe.
Please don't read this as a critique of AL, I fully understand the challenges of clocking and synchronization in digital domain.

Now, the question I'm trying to get answer about is whether 2.2 setup needs time domain correction if DSP has an ability to simply have a delay between channels (in increments of 0.01 msec, up to 1500 msec).
And what's going on with that AL error popping up? I only selected input and output and didn't screw with anything else.

 

[jrobbins50]

Ah the dreaded too many channels. Well, note at the bottom of the measurement page that your WASAPI input/output arrangement is not supported.

Best always is an ASIO arrangement. Can you point your inputs and outputs to ASIO4all and then select ASIO4all as both input and output for AL? If not, you can also try Windows MME or Direct Sound as inputs. You need to enable them under Advanced Settings of the Measurement tab. I could not get WASAPI to work with my UMIK-1 or UMIK-2 microphones when I used to use those. If you don’t use ASIO4all, make sure you have 5.1 set as your default setup in Windows Sound Properties. JCR

 

[dima1stg]

I see. I read in help that WASAPI is fine, especially in exclusive mode. Thanks a bunch!

 

[dima1stg]

What's the right site to download ASIO4all? Looked at https://asio4all.org/ - looks like crap and feels unsafe to download from.

 

[jrobbins50]

Yes, the site is a mess, but the product is good. You need to download an older version to get the required off-line settings tool. Sweetwater will explain the procedure: https://www.sweetwater.com/sweetcare/articles/installing-and-using-asio4all-for-windows/ JCR

 

[Ofer]

Just a few notes. Audio interface is another name for sound card. The Motu ultralight mk5 that a lot of members here use goes for ~500$ and you can find many used. So interface+analog mic can total around 500-600$. It's isn't small change but less than you calculated. As far as I understand if you do not have a discrete amplification for every driver you can't achieve TTD correction. That does not mean that AL won't correct the time alignment between all your speakers to the millisecond and that could mean an audible difference.

 

[sleach]

Hi, I don't want to hijack OP's thread, but as another potential user I am just wondering how much computing power I will need to run this?

 

[jrobbins50]

2 channel or multichannel? Do you listen to DSD files? Do you use HQPlayer to upscale your content? And what is your content source?

All these impact the answer as to required computing power. JCR

 

[sleach]

2 channel or multichannel? Do you listen to DSD files?

Thanks JCR, 2 channel. I am running 2 way active speakers currently the crossover is in a Minidsp 4x10hd. I dont have much dsd content, a couple of albums. Mostly I stream or play from my library with Roon.

 

[jrobbins50]

Ok so most of your content maxes out at 24/192. What is your Roon core? JCR

 

[sleach]

Hi JCR, its an HP elitedesk i5 6500t, 8g ram, 120g ssd. Currently Roon rock installed.
Cheers
Sam