Hang Loose Convolver cheat sheet?

dotnet

Member
Thread Starter
Joined
Mar 30, 2020
Messages
44
Does anyone have a short list of what to do or to avoid when exporting impulse responses for use as filters (or filter sets) with Hang Loose Convolver (accuratesound.ca)?

I gather that I should export the measured impulse response in 32bit float format, or should I export the min phase version? I’m mostly after correcting the room response. How should I go about making a filter set for multiple sample rates? Take multiple measurements at different sample rates, export them individually as WAVs and zip them up?

Are there any other gotchas, or hints from people who have done this?

I’ve been using PEQ-based filters calculated by REW for some time now, but would like to try the HLConvolver’s FIR filters that let you switch between filters during playback, volume levelled and without delay.
 

dotnet

Member
Thread Starter
Joined
Mar 30, 2020
Messages
44
Pardon me for following up myself…

I think I may have got the service that AccurateSound provides (create a filter from submitted mdat files) confused with what the HLConvolver software does. After a bit more reading I think I need to feed the HLConvolver a filter calculated by REW and exported in WAV format, rather than the measured impulse response of the room. Have I got that right?

If so, I believe my questions about how to export and how to make multi-rate filter sets have answered themselves

I’d still love to hear from anyone who’s used REW generated filters with HLConvolver.
 

jtwrace

Member
Joined
Mar 12, 2019
Messages
139
Location
Orlando, FL
More  
Main Amp
(4) Hypex NC502MP
DAC
Topping D90
Front Speakers
JBL M2
Subwoofers
(2) Rythmik FM8, (1) UM18, (1) Custom DIY Ciare 18
Other Speakers or Equipment
BSS BLU 50
Streaming Equipment
RPi4 using RopieeeXL
Streaming Subscriptions
Tidal
Other Equipment
All Belden 1800F and 1313A
I’d still love to hear from anyone who’s used REW generated filters with HLConvolver.
HLC doesn't create FIR Filters. You would need to use Audiolense, Acourate or other to create them and then use HLC as the; well, Convolver. I'm sure @Mitchco can answer it much more eloquently.
 

dotnet

Member
Thread Starter
Joined
Mar 30, 2020
Messages
44
I was hoping to use REW to create the filters, because that’s what I have and know. Are REW filters not suitable, despite being exportable in WAV format, as requested by HLC?
 

Mitchco

Member
Joined
Oct 1, 2017
Messages
79
Hello @dotnet Yes, from the file menu, select "Export filters impulse response as wav" Select stereo and 32bit float. Don't select normalize samples to peak value.

You can either load the stereo.wav filter directly into HLC or you can select a number of samples rates and then bundle up the filters in a .zip file and load the entire .zip into HLC. If the "host" supports automatic sample rate switching, then the corresponding filter will be automatically loaded based on source sample rate.
 

dotnet

Member
Thread Starter
Joined
Mar 30, 2020
Messages
44
Thanks Mitch, much appreciated. So the REW filters are suitable for HLConvolver after all.

What I’m still struggling with (conceptually) is this: in REW, all the supported target devices implement IIR filters, at least as far as I can see. I wonder whether this is taken into account when simulating the resulting filter effect with respect to phase, resonances etc. And I wonder whether this affects using the filters with an FIR convolver? In other words, are filters as calculated by software like REW or Audiolense – in particular when expressed as impulse WAVs – inherently either IIR or FIR, or is that just a property of the device applying them?
 

Mitchco

Member
Joined
Oct 1, 2017
Messages
79
IIR filters contain no excess phase correction. Whereas FIR filters have the capability to contain timing/excess phase correction independent of the magnitude correction. It is the capability of the software creating the correction filters as to whether there is an excess phase correction or not. As far as I am aware, REW creates IIR filters only whereas Audiolense one can create FIR filters either minimum phase with no time domain correction or linear phase with time domain correction.
 

jtwrace

Member
Joined
Mar 12, 2019
Messages
139
Location
Orlando, FL
More  
Main Amp
(4) Hypex NC502MP
DAC
Topping D90
Front Speakers
JBL M2
Subwoofers
(2) Rythmik FM8, (1) UM18, (1) Custom DIY Ciare 18
Other Speakers or Equipment
BSS BLU 50
Streaming Equipment
RPi4 using RopieeeXL
Streaming Subscriptions
Tidal
Other Equipment
All Belden 1800F and 1313A
Gotcha, thanks!
I'll put my opinion and user experience here after going through the Audiolense experiment. You want/need FIR! Truly incredible.
 

dotnet

Member
Thread Starter
Joined
Mar 30, 2020
Messages
44
Thanks, I might consider it in the future, depending on available tools. Unfortunately, Audiolense requires a Windows PC which is a bridge too far for me :)

I’m only applying room correction in the sub-Schroeder and transitional range in an acoustically well-treated room, not HT surround or active cross-over stuff. REW has been good to me, I’m using the filters in Audirvana and Roon. The attraction of HLConvolver is the seamless switching between filters, and the option of stand-alone system-wide correction (as opposed to app plug-ins). The latter allows me to run post-correction RTA measurements in REW, while switching filters, without needing to play back recorded sweeps.
 

jtwrace

Member
Joined
Mar 12, 2019
Messages
139
Location
Orlando, FL
More  
Main Amp
(4) Hypex NC502MP
DAC
Topping D90
Front Speakers
JBL M2
Subwoofers
(2) Rythmik FM8, (1) UM18, (1) Custom DIY Ciare 18
Other Speakers or Equipment
BSS BLU 50
Streaming Equipment
RPi4 using RopieeeXL
Streaming Subscriptions
Tidal
Other Equipment
All Belden 1800F and 1313A
Unfortunately, Audiolense requires a Windows PC which is a bridge too far for me :)
Only to take the measurements. Then you're done with it. Borrow one if you can...it WILL be worth it. What AL does is far above though regardless of the frequency. The phase and timing is not to be dismissed.
 
Top Bottom