Introducing Hang Loose Convolver from Accurate Sound

[Whoareyou]

Because in a First step i think to use the convolver with fir filter to XO and linearize the CV, and Dirac to correct the room.
I know that mitchco use also the convolver for the room, but i don't sure to be able to do this with optimum result, and i still don't understand if the best is a single measure or a multi measure for correct the room.

Interesting approach and I never would have thought of doing this.

BTW - With AudioLense, I get excellent results with just one measurement. I actually moved from Dirac to AL and enjoy not having to take multiple measurements (as with Dirac).

Of course, once you complete those measurements in Dirac its simple to generate the filters. Having said that, once I got the hang of AL, not much of a difference, although you can spend quite a bit of time on the correction filter if you are a perfectionist.

Finally, and most important, I feel the results with AL sound superior to what I got out of Dirac.

 

[Zapuan]

Approccio interessante e non avrei mai pensato di farlo.

A proposito, con AudioLense ottengo ottimi risultati con una sola misurazione. In realtà sono passato da Dirac ad AL e mi piace non dover effettuare più misurazioni (come con Dirac).

Ovviamente, una volta completate queste misurazioni in Dirac, è semplice generare i filtri. Detto questo, una volta che ho preso confidenza con AL, non c'è molta differenza, anche se puoi dedicare un bel po' di tempo al filtro di correzione se sei un perfezionista.

Infine, e cosa più importante, sento che i risultati con AL suonano superiori a quelli che ho ottenuto da Dirac.

Usare solo Audiolense è sicuramente un approccio più semplice che integrare anche Dirac, così come una singola misura per la stanza è un processo molto più veloce e ripetibile.
When you say that AL produces superior results, are you referring to measurements or is it a subjective opinion?

 

[mk1classic]

Not sure why you would use both AudioLense and DIRAC?!
AudioLense perform both crossover duty and room correction in one operation.
Room influence / correction depend on your gating values on frequency setting and TTD. Multi seat is also an option in AL.

 

[Whoareyou]

Usare solo Audiolense è sicuramente un approccio più semplice che integrare anche Dirac, così come una singola misura per la stanza è un processo molto più veloce e ripetibile.
When you say that AL produces superior results, are you referring to measurements or is it a subjective opinion?

Measurement and subjective. I have multichannel setup with passive speakers. For my configuration bass management is essential since I have a couple of room nulls that show up on all of my speakers.

First problem with Dirac multichannel is I couldn't properly integrate my 2 subwoofers. Easily validated objectively and subjectively.
2nd problem I have with Dirac (and this one is mostly subjective) is it simplifies stuff to the point where all you do is set the target curve, but what it does with its other correction algorithms is a mystery, and my frustrations with it.

No matter what I did with Dirac, either adjust curtains and/or change target curve, Dirac always sounded processed. Yes, subjective but easily distinguishable sound signatures between original and filtered. Instead of "fixing" room problems it changed my speakers to sounding different than the speakers I purchased. As part of this correction journey, I added room treatments, moved speakers around, etc. I couldn't fix it.

So best way to describe AL; With AL and its filter options and XO's, I was able to design a filter that corrected and left the sonic signature of my speakers intact.
Some of this is measurable, and some is A/B comparison, but my speakers sound like my speakers with room issues corrected.

Of course, this is my room and my speakers, so Dirac's algorithms may give you much better results than I was able to achieve with my speakers / room.

Also, I haven't tried the Dirac Live bass control so results with that could be better.

 

[Zapuan]

Non sei sicuro del motivo per cui dovresti usare sia AudioLense che DIRAC ?!
AudioLense esegue sia il crossover che la correzione della stanza in un'unica operazione.
L'influenza/correzione della stanza dipende dai valori di gating sull'impostazione della frequenza e sul TTD. Il sedile multiplo è anche un'opzione in AL.

I'm not sure Audiolense is better than Dirac for room correction. But I can try, I'll look at the multi-measure of AL, I didn't know it could be done.

 

[Whoareyou]

I'm not sure Audiolense is better than Dirac for room correction. But I can try, I'll look at the multi-measure of AL, I didn't know it could be done.

For my setup and room, I thought it was better. Of course, you may have completely different results for your speakers and room.
As you are saying, only way to know is by trying it out.

And yes, AL does provide multi-measurement ability, but I'd first try without it and see what you think. I was skeptical at first (having used Dirac), but it really works very well.
Much easier to do, and AL has a learning curve. I'd ask about the measurement process over at the AL forum on AVnirvana.

 

[Zapuan]

Ok guys , thanks to mitchco and all of you. These days I've read articles and forums on AL and I've understood that single measurement is most likely a good way. My doubt was because years ago I had tried to compare Dirac with Sbragion's DRC and the filter generated in Dirac from a single measurement always sounded too artifact compared to multi measurement. But I think it is because in Dirac there is no possibility of time domain control. I have to try AL.

 

[DWPress]

Yesterday I had some fun. After my last exchange with @mitchco he was able to find the issue within his HLC Host that prevented some AU plugins from loading on Mac. That's fixed now!

In the thread on Dirac Live Bass Control here I was able to integrate it into my already existing active XO with some interesting digital signal routing. Public thanks to Mitch for making a great tool even more versatile and better!

 

[HerbertWest]

FWIW I happily use HLC and Audiolense filters with movie content out of a Marantz AV7706 (Audissey disabled). HLC handles 11 channels (5.4.2) with no issue ; end to end latency measured by the avp is ~60ms IIRC (without linear phase crossovers ofc).
That is on top of a more traditional role for stereo music upmixed to Auro3D via penteo (no AVP involved, just my UCX2/ADI).

 

[mitchco]

A quick update: Hang Loose Convolver 1.1.4 is released which sports a new high quality filter resampler:

Hang Loose Convolver (HLC) - Accurate Sound

is a stereo and multichannel convolver designed to compare level matched convolution filters in real time with instant switching between filters. Bundled with Hang Loose Host.

accuratesound.ca

 

[DWPress]

Thanks @mitchco, great tool you've made!

 

[Whoareyou]

A quick update: Hang Loose Convolver 1.1.4 is released which sports a new high quality filter resampler:

Hang Loose Convolver (HLC) - Accurate Sound

is a stereo and multichannel convolver designed to compare level matched convolution filters in real time with instant switching between filters. Bundled with Hang Loose Host.

accuratesound.ca

Thanks @mitchco, great tool you've made!

The filter banks, allowing direct comparison between my filters are a great feature by themselves, but the addition of filter resampling / fine-tuning really adds to "why I should use this product".

This feature enhancement is a great improvement / idea for the product.

 

[Keith_W]

I tried Hang Loose Convolver today. I would like to give my heartfelt thanks to @mitchco who talked me through the setup. Essentially, I had written the .CFG file correctly but I had selected a rather small buffer size - so instead of sound, I was hearing a series of pops. After I increased the buffer size, the problem went away.

I then had a listen and there was zero latency when switching between filter banks on HLC. This allowed me to compare the effect of different target curves very quickly. Over the next few days, I will be doing more comparisons between time alignment, different strategies for room correction, and so on.

Unfortunately, the latency when playing audio is quite significant, about 3-4 seconds between selecting a track and actually hearing sound. With JRiver it is probably 1-2 seconds. This is enough to cause significant lip sync issues when watching a Youtube video. I will have to play with it a bit more to see if I can get the latency down without reducing the buffer size so much that I get dropouts.

A really unexpected finding was that HLC sounded better than JRiver! I could not believe my ears because both are bit perfect players, and I am using JRiver's output into HLC. Now, HLC has much higher gain than JRiver, so I initially thought it was that. So I played some white noise and matched the gain with an SPL meter. Doing back to back testing was not ideal, because it involved about 2-3 minutes of reconfiguring JRiver's output in order to do a comparison. Even after volume matching, the difference persisted - HLC has more bass, more top end, and just sounded clearer.

I was utterly dumbfounded. Surely it is impossible for bit perfect playback to sound different?!?!? If anything, my expectation bias was for both of them to sound the same!

Then I remembered that I had ISO226 volume control turned on in JRiver. Doh!! With JRiver, I was listening with the Vol turned up to 35. To obtain a level match in HLC, I had to turn down the volume in JRiver to 20. This means that what I am actually listening to is the different curve applied by JRiver to compensate for the lower volume! So I turned off ISO226, turned down the gain in HLC, and achieved volume matching with white noise and an SPL meter. The difference persisted, again in favour of HLC, but I suspect that I still have improper gain matching. Or maybe there is some secret sauce in HLC that might account for the difference.

I actually like the result with HLC having so much more clarity at low volumes compared to JRiver due to the differences in the ISO226 curve. Homework for tomorrow is to design more filters for testing and find a way to change the settings for JRiver's ISO226 reference point. I will crack out the voltmeter and measure the difference at DAC output. If it still sounds different, then i'll have to investigate further, maybe by running the DAC output through an ADC for analysis. It's 3am in Australia, time to go to bed. And you can laugh at another one of my silly mistakes.

 

[Sal1950]

Glad your happy, Mitch is a smart fellow.

 

[mitchco]

A couple of updates:

1. HLC Linux version available for download. Please send me an email if you wish to target a different distro than Ubuntu.

1. Most Digital Audio Workstations (DAW)’s support “automatic delay compensation” when using VST3 or AU plugins that have latency. The idea is that the plugin reports its latency to the DAW and the DAW compensates for the latency. So when tracking or producing audio for video postproduction, the tracks and/or video is delayed by the number of samples reported by the plugin so that the audio lines up perfectly with the video.

While HLC is a 0ms latency convolver, some FIR filters have inherent delays, like if using linear phase FIR filters for example. The latency changes based on; type of FIR filter used, sample rate, number of filter taps, how much excess phase correction has been applied, and if using digital xo’s, whether min or linphase. HLC now reports the FIR filter latency to the host for automatic delay compensation:

A 131,072 tap minimum phase FIR filter with minimum phase digital XO reports 0 latency samples:


A 131,072 tap minimum phase FIR filter with linear phase digital XO reports 5,540 latency samples:


A 131,072 tap linear phase filter with linear phase digital XO reports 37,863 latency samples:


And if the sample rate changes and HLC does not find a matching FIR filter for that sample rate, then the FIR filter is resampled to maintain its frequency resolution and reports the new latency samples:



Great for DAW’s, but what about consumer applications? Working with the folks at JRiver, the latest version of JRiver now supports automatic latency compensation. This allows one to use full tap length linear phase FIR filters with excess phase correction and not have any lipsync issues while watching movies.

If folks can think of other consumer (or pro) applications that support the plugin model, please send me an email. I can work with the developer to implement automatic latency compensation.

Happy listening!

 

[HerbertWest]

Good stuff @mitchco, these features are super useful!!!

 

[3ll3d00d]

The capability to level match digital correction filters to bypass level and switch seamlessly between filters in real-time has been designed into HLC. There is science behind the design, which can be read here: https://accuratesound.ca/products.html

fwiw the link above is a 404 now

 

[Whoareyou]

@mitchco
Great addition, but you "broke" my Jriver zones Of course just kidding. The zones with previously entered lip sync offsets are now handled automatically between HL and JRiver. Since that old value doesn't know anything about the new latency compensation the old value is now incorrect.

And FWIW, the value was simple to adjust as it just needed to be configured to what I have entered using JRiver's built-in convolution engine ( I assume this value represents Jriver's portion of the delay outside of filter latency).

Big benefit is that the number is entered once and if I make change to my filter, the delay is now handled automatically between HL and JRiver.

Perfect and wayyyy easier!

 

[mitchco]

Hang Loose Convolver (“HLC”) is now available on Raspberry Pi4.

Requirements:
Raspberry Pi4 4GB (2GB is likely to work).
64 Bit OS Debian version 11 (bullseye).

Performance:
Process 32 channels of convolution using 65,536 tap length FIR filters at 48 kHz sample rate.

Example playing a 7.1.4 Dolby Atmos (already decoded) music file using a VST3 plugin AudioFilePlayer. HLC is configured for 12 channels with 2 channel I/O being summed:

While this example is using a 7.1.4 (12 channel) file being played, the FIR filterset has digital XO’s and bass management built in. So, with 12 channels of direct signal and 11 channels of bass offloading means 23 channels of discrete convolution is being processed. With 23 channels of convolution processed there is still considerable CPU headroom and buffer size left.

Note: HLC is a zero latency convolver, meaning no signal delay is added by the convolution engine. Therefore, one can process 65,536 tap minimum phase FIR filters immediately with no added signal delay. This is good for situations where lipsync is required but you still may want high-resolution FIR filtering capability of 65,536 tap length filters.

HLC comes with HLConvolverHost, which allows you to plug in virtually any VST3 plugin for additional processing. The simple audio settings dialog allows you to easily choose inputs and outputs, sample rate, and buffer size so you can be up and running in minutes.

Updated Operations Guide.

 

[Keith_W]

That's great. BTW, does HLC still require VB Audio Cable? I heard a rumour that you were developing your own input for HLC so that VB is no longer required.