Article: Audiolense Digital Loudspeaker and Room Correction Software Walkthrough

[mitchco]

View full article

[mitchco]

On 05/12/2017 at 11:52 PM, Bob Stern said:

Thanks, Mitch, for another superb article!  Obviously this was quite time-consuming to write up, and your efforts are appreciated!

 

Can you provide some guidance regarding the relative strengths and weaknesses of Audiolense vs. Acourate for a passive, 2-channel system?  After adjusting both systems iteratively to achieve what you considered the best sound, were the results essentially identical, or do the two algorithms tend to produce different audible results?   How do they compare in the effort required to achieve the very best results?

 

Bob, thanks for your kind words. The simple answer is both products provide excellent results . For two channel, passive speaker correction, the workflows are similar. Beyond that is a point of departure, and comes down to ones’ specific requirements and preferences. There is enough info in this article and previous articles I have written where a keen reader should be able to sort out any differences.

[mitchco]

On 06/12/2017 at 4:39 AM, dallasjustice said:

Awesome job Mitch!

 

WRT using a pair of time aligned subs in AL:

 

You can individually time align each sub in AL. You can also use them as either mono or stereo subs. However, AL cannot measure a pair of mono subs at the same time. The only situation this sort of measurement would be needed is when the subs are used on opposing walls in opposite electrical polarity and time aligned for axial mode cancellation. I’ve talked to Bernt about measuring subs at the same time and it doesn’t appear that he wants to include this feature in 5.0. It’s the only thing holding me back from getting it. 

 

In your room, you probably would not need/want to set up your subs in this way. 

 

Ive been using the flat to 1khz filter now for two years and haven’t wanted to change it. It’s very versatile for all types of music, especially electronic bass heavy music. 

 

I do love how easy AL is for transferring filters to Jriver. 

Michael, good to hear from you and understood. Your use case is very specific and may have arrived at the best result. However, Audiolense does account for low frequency room reflections, it may be that you would get similar results with a traditional subwoofer setup. My subs have arrived and I will be able to confirm that sometime soon.

[mitchco]

23 hours ago, aj said:

Thanks Mitch for this elaborate piece of work.
Maybe I missed it, but will Audiolense (or Accurate) work with panel/open baffle speakers?
Thanks

Hi aj, I don't have personal experience using Audiolense with panel or open baffle speaker designs.  You could post your question to the Audiolense user group. Anecdotally,  I know others have with success. If you look at my eBook reviews, I believe reviewer Xan has a pic which looks to be open baffle design. You could send him a note as well...

Kind regards,

Mitch

[mitchco]

19 hours ago, dallasjustice said:

Mitch,

The delay values are automatically generated by AL5?  Is this a new AL feature? I don’t remember the old AL showing the delays calculated. Is it possible to manually adjust the delays?  Do the delay values correspond to the delays you may have already manually calculated using Acourate?  I’m just curious to know how accurate you think the delay values are. 

Thanks again,

Michael. 

Michael, the last time I used AL was for a passive design, so I can't say for sure, but I believe yes, it was in previous versions. I have my subs running and time aligned. As you know, getting sub time alignment is very difficult due to the natural low pass of the sub, as it takes time to peak. This leads to additional delay - typically 3 to 4 ms. Additionally, if you run a sub sweep that ends at 300 Hz, the sub will appear to have more delay than if the sweep is run to 1 kHz.

 

I am really impressed with Audiolense's ability to time align my subs, even with a low XO point of 40 Hz. My subs run from 12 Hz to 40 Hz.  I have repeated the measurement a few times and it has been exact each time. I can't say that for all of the other methods I have tried to date. I plan to write another article going into the gory details of this and a few other advanced features.

 

Hope that helps,

Mitch

[mitchco]

13 hours ago, number8 said:

What a super article. In the introduction you refer to electroacoustic speakers. I run a pair of Martin Logan electrostatic speakers. Shall I expect the same improvement level?

Thanks @number8 I don't have first hand experience with this, but @aj asked Bernt about this on the Audiolense forum and Bernt responded by saying that Audiolense does work with open baffle and panel type speakers. On another forum, I know someone using DSP with Martin Logan speakers with excellent results, so I would expect the same or similar results for yours.

Kind regards, Mitch

[mitchco]

Thanks @bibo01 My understanding from Bernt is the max sampling frequency is 400 kHz. 

Kind regards, Mitch

[mitchco]

@cjf it looks like @dallasjustice answered most if not all of your questions (thanks Michael!). I have been running with software/digital volume since 2011 using JRiver and have not had any incidents. Make sure Windows Sounds in Playback devices, Sound tab, are turned off, as well as Play Windows Startup sound is disabled. With the PC running, I can reboot the computer without any sound issues. If using JRiver, always good to have volume protection enabled...

 

Michael's time aligned results from using the cascaded subs method and Audiolense are excellent. My results are similar using stereo subs, to be seen in a follow up article in the New Year.  Cheers, Mitch

[mitchco]

@cjf I use an older microphone kit from http://www.isemcon.net/ashop/index.php?cPath=2_23_28 that is no longer available. It was a combo calibrated mic and preamp kit and still works well and relatively inexpensive.

 

There is a shootout of measurement mics here: http://www.blouder.com/affordable-measuring-microphones-shootout-part-3/  Also one from Ethan: http://realtraps.com/art_microphones.htm  As you can see, most of the measurements mics are similar in performance. Some folks will only use http://earthworksaudio.com/measurement/ using a high quality mic preamp http://earthworksaudio.com/products/preamps/zdt-1021/  

 

Not sure what the gain issue is, but my kit uses a mic preamp and RCA line level outputs with plenty of gain back to the Hilo ADC... 

 

In my experience, the most important aspect is that the mic is calibrated and comes with a calibration file. We are talking about fine tuning the perceived frequency response (i.e. tonal balance) arriving at ones ears. We can discern a 1 dB difference in frequency response, so accuracy and precision counts.

 

Happy New Year!

[mitchco]

On 03/01/2018 at 9:30 AM, vitalii427 said:

Mitch, thanks for a great article! Waiting for the next one.

Which subs do you have?

Hi @vitalii427 Thanks. Subs will be disclosed in the next article...

[mitchco]

13 hours ago, R1200CL said:

@mitchco

I’m in a process of understanding what the different room correction SW does or doesn’t. Which I think is almost a science of it’s own ?

 

Except from UI, good guides, support etc. 

Are there any main difference between Audiolense and Acourate ?

Can they in principle do (exactly) the same ?

 

 

 

 

 

 

@R1200CL I think @dallasjustice comments earlier cover some of this.

 

For sure, designing custom FIR filters for one's speakers in a room is a huge subject area. My eBook is one of many on designing DSP FIR filters, along with a number of articles here on CA. In principle, they can produce similar results, but there are nuances, like different psychoacoustic filtering, for example.

 

Both are excellent products  with state of the art DSP functionality, that can work with virtually any loudspeaker and room configuration. Loudspeakers in rooms can be fully optimised both in the frequency and time domain to your ears.

[mitchco]

3 hours ago, R1200CL said:

 

I probably buy that book. But I understand it’s quit linked with Acurate. I don’t know if Audiolense need a step by step guide. But If so, any plans of incorporating Audiolense in that book ? Or a new book. 

 

If needed at all. It seems to me that maybe Audiolense is perhaps a bit more user friendly. 

 

I appreciate your support. While the eBook walks through using Acourate, the DSP principles applied are the same. There are several sections in the eBook that are not tool specific and apply to understanding why we hear what we hear in small room acoustics, acoustic measurement techniques, understanding some parameters of good loudspeaker design (e.g. controlled directivity) and understanding industry guidelines and standards for achieving accurate sound reproduction in critical listening environments.

 

If you click on the Amazon link to my eBook and click on Look Inside, you can peruse the table of contents and read the first few chapters for free. I have plans to update the book on several aspects (e.g. sub integration plus some topics here) and cover more DSP tools, including Audiolense. I don't have an ETA for that yet and is a ways out.

[mitchco]

Glad you enjoyed the article. The Hilo has 6 analog outputs or 3 sets of stereo outputs. The line out L/R go to the midwoofers, the monitor out L/R are going to the tweeters and headphone out L/R are going to subs. Both the monitor and headphone out have analog volume control to match the different sensitivity levels of the drivers if required. Scroll down to the analog performance section to see  the tech specs. I have had no issues and sounds great.

 

I have a pair of subs hooked up for the past 3 months and about 3/4 the way through the next article on integrating and time aligning subs to mains. Should be out soon.

[mitchco]

Yes I think so, but I have not tried it myself. I believe @dallasjustice is using the Hilo this way and could confirm.

Thanks for purchasing my eBook. Feel free to PM me with any questions.

[mitchco]

33 minutes ago, nefilim said:

Not finding a lot of documentation on Audiolense' site, I'm wondering if anyone know what this means: 

 

"Clock drift correction in measurement is now practically perfect."

 

Does this imply I can measure on a different device than the playback chain? eg one USB device for ADC and another USB device for a DAC?

 

Yes, hard to find, but from the help file attached at the bottom of the article: "If you know that you’re using separate audio devices for recording and playback, or if you’re unsure, check the “separate play and recording device” option under advanced settings. That will prevent any clock drift related problems. When the clock drift correction is engaged, one of the speakers / tweeters will be measured twice."

[mitchco]

Hi @kravi4ka sorry for the delay, the notification did not come through. Re: RME Adi Pro while I don't have any personal experience with the product, it's specs look excellent and has the requisite 2 ch analog in and 4 channel analog out if you are considering biamping. Maybe you can outline your scenario? There is a thread here about it on CA: 

 

[mitchco]

Hi @ronkuper, I am glad you enjoyed the article.

 

1.0 left + 1.0 right. Simply let the sub handle the full LF part from both speakers. As if it were two subs spaced together.

 

If the SPL and the sound quality to your ears is sufficient, good to go!

 

Kind regards, Mitch

[mitchco]

Hi @Jesal, 3 way active, nice! Yes, I was confused on this point when I first started. There is a specification for the config file so JRiver and other convolvers can read a common file format.

 

The good news is that Audiolense automatically builds the config file, that references the filters, so that all you need to do is load the config into the Convolver. In JRiver, you need 6 channels, so you set the output format in DSP studio to 5.1. I.e. Tools, Options, DSP & Output format, Convolution, Channels,5.1 That’s it. No upmixing donwmixing.

 

A couple of tips. Look at my Audiolense triamp setup in the sub article. See the Speakers->Crossover configuration, note the Channel and Channel ID’s in your setup. Then map it accordingly in the Measurement Window. Also in the measurement window is check speaker connections. Click on that and a steady midrange chirp is sent to each channel/speaker combo one at a time to ensure proper channel mapping before you measure. It’s not very loud.

 

Cheers, Mitch

[mitchco]

Hi James, yes, I have experimented with near-field setups, even a set of computer desk top speakers. Sounds great! The process is indeed no different than what is described in the article.  Kind regards, Mitch

[mitchco]

@H-space Welcome to CA! Thank you for your kind words!

 

I had a look at your Gearslutz link - you have been busy! I think you will find Acourate or Audiolense to be a big help here. Given the size and dimensions of your room, I would give Audiolense a try as it does cancel low frequency standing waves and room modes. Also, I think the time alignment would be easier with your setup.

 

To answer your questions, I believe on the surface, yes and yes are the answers. But there are some caveats. @dallasjustice has experience using a 2nd DAC... You can find some info a few comments back in this thread. Yes, you should be able to presum the bass, but you may be shortchanging yourself, as one sub may need different correction than the other sub, as mine did in this article, if you look at the measurements. One sub had a wicked maximum phase reflection and the other did not... Also, I am a believer in time aligning subs to mains...

 

Another option is Mult-Sub Optimizer (MSO). The link to my integrating subwoofer article in the paragraph above has user comment information where a CA member is using it successfully with Acourate.

 

Note, if you plan on tracking along with mixing, then the linear phase FIR filters used introduce a 3/4 of a second delay. However, there is a minimum phase mode that can be used and only sacrifice a little bit of the time alignment magic. Audiolense is able to produce a minimum phase filter that is almost indistinguishable from it's linear phase counterpart.  Bert has done a really good job here. If interested, you will get more feedback from other members if you join the support forums. Audiolense is https://groups.google.com/forum/#!forum/audiolense

 

I think you will find that DSP will make a major improvement in the bass response in your room. Short of stuffing the room with bass traps, and I do mean stuffing, this is the only other way to get an even bass response to your ears.

 

Kind regards, Mitch

[mitchco]

Hi @TooSteep Thanks for your kind words.

 

That’s a great question! Before getting into the specifics of a particular loudspeaker, there are three approaches:

 

Audiolense can apply an overall time correction (using TTD) to an existing loudspeaker that is not time aligned. Assuming a 2-way design, the caveat is that the z-offset between the drivers needs to fall within the time correction window. It is likely for most desktop speakers that this would work. One would have to measure the step response and evaluate the sonic result to see if it is time aligned/phase coherent enough… You could download a trial version of Audiolense XO and try it on your existing speakers. I think the trial version corrects for 90 seconds.

 

A second approach is to biamp a two-way so that one could use Audiolense linear phase digital XO and the software would automatically time align the drivers and apply an overall time correction. This is the approach I chose for the article and my preference.

 

Finally, acquire a set of time aligned speakers for the computer desk. I have not looked into this much Vandersteen makes a time and phase coherent smaller speaker the VLR Wood. I have not seen any measures or listened to the product. Green Mountain Audio is another… Plus there are dozens of near field studio monitors, some of which are time aligned/phase coherent...

 

Again, great question/suggestion. I will start looking into this more and see what I can come up with…

 

Kind regards, Mitch

[mitchco]

On 9/16/2018 at 10:02 AM, TooSteep said:

 Thanks Mitch.

 

"Plus there are dozens of near field studio monitors, some of which are time aligned/phase coherent" 

I had no idea about this! It isn't obvious from searching or reading literature. Reading the rave reviews of the Dynaudio LYD48, there were some implications about this, but I couldn't quite tell for sure. I will try to find some.

 

I actually own a set of Green Mountain Audio Europa's, which I really enjoy. But the sound is not quite right up close, 2-3 feet away. I feel I need to be at least 7-8 feet away from them before I can benefit from their time aligned magic.

 

When you biamp a speaker, how does it not still run through the internal crossover? Or does it just effectively HPF the tweeter and LPF the mid?

 

Several near field monitors are active and typically already have digital XO's and if lucky, time aligned/linear phase operation. Genelec, HEDD, Adam,  Neumann, PMC, Fostex, Amphion, are just a few brands that come to mind.

 

JBL larger speakers like the M2 and Cinema speakers like I have require external XO and amps. So there is no passive XO network. For other speakers that say they are biamp-able, it usually is indeed still passing through the legs of a passive XO... One has to look really carefully as a "real" biamp scenario is where the amps connect directly to the drivers.

 

I am a diyAudio guy and have built/modified speakers and have either unhooked the passive XO and ripped it right out, depending on whether reselling or not.  That's another possibility...

 

As soon as I get some time, I will start looking into this deeper...

[mitchco]

Hi @ianmc Send me a PM with what you have and want to accomplish, as I am not sure about the context of your question.

 

Thanks, Mitch

[mitchco]

Hi @nefilim Yes, that's odd. It should be the same as this article. Maybe Chris @The Computer Audiophile has an idea as to what is wrong with the images? Chris, anything I can do to fix this?

[mitchco]

Hi @omid Welcome aboard! And thanks! Perhaps you can briefly describe your measurement mic, gear and signal path for taking measurements? I doubt you are missing anything, but always good to have a 2nd pair of eyes have a look. Having said that, if you are happy with the SQ, then stick with what works. But lets have look anyway 🙂