Welcome! Need support, you got it. Or share your ideas and experiences.
In Reply to: RE: Magnepan LRS posted by pictureguy on February 21, 2021 at 21:11:03
"... How did FIR filters get into this? ..."
Digital speaker and room correction is not uncommon (DEQX has been around for 2 decades) and PC audio has embraced it for quite a while.
I started with Klipsch, 101dB/1 watt/1 meter before going to Maggies. Had to buy stronger amps.
Started with 1.6's, then went to 3.6's. Had to buy stronger amps, AGAIN, then to 20.1's.
If you are familiar with making digital speaker and room correction filters, you will know many use the "cuts-only" method to avoid clipping. The cuts-only method can easily exceed 10dB attenuation to do speaker and room correction. That means your output is 10dB lower (unless you have special premamp hardware to compensate for this and drive your amp(s) accordingly, read: certain pro-gear with more than 4V output).
Thanks for explanation. I was worried for a second until you went for the 'DIRAC solution....
I'm not doing any form of speaker correction.
If you want to REALLY get an eye / ear full? Go to Emotiva Lounge and find the DIRAC thread which is not up to OVER 150 pages. Have a bottle of Aspirin handy!
I did metrology for a living for years. And can assure anyone that the devil in DIRAC and similar sytems is repeataility of measurments. Doing it a different way than you did today is a recipe for confusion and disaster. Those on the DIRAC thread on EMO webiste are really in a state.
For example? Some settings of LFE would appear to apply a 10db BOOST to those frequencies. Some people are hung-up on what I'd consider (for music) massive SubWoofer Overkill.
I get it going from 101db speakers to something in mid / low 80s. Makes PERFECT sense. I have 1.6s and find a PAIR of Parasound A23 suit me fine.....One per speaker but NOT yet an active crossover. So while I am passively biamping, i do NOT get the full and potential 3db benefit were I o go full ACTIVE.
Without all the electronic aids I could EASILY see needing more than 10x the power when going from highly sensitive KLIPSCH to Maggies!
Too much is never enough
I started with an old DEQX unit, still have it. I have ARC DRC in my Anthem Statement D2V Pre/Pro but never liked what it did so completely disabled it.
DIRAC has a PC software version that I have not tried, nor their embedded versions.
MarcL uses DIRAC for full blown home theater to good effect. I have seen his REW plots and his D50 clarity ratings are exemplary.
I have tried the Audiolense XO demo, but can't get consistent delay measurements with it on my PC. If I am going to use something for time and phase alignments, I need it to create consistent repeatable results.
I am using REW and RePhase. I get much better consistency/repeatable measurements with REW. I am also experimenting with REW on Realtime Linux to see if handles timing latencies better than Windows (with respect to measurements and playback).
emailtim, what version of Linux you running, I'm using Mint Cinnamon 17.3 and 19.2 Thanks
I am testing Debian (10) Buster with the latest 5.9 buster-backports vanilla and realtime kernels on an Intel Atom D525 and i7-3770K.
Kernel 5.10 hasn't been back-ported yet debian-buster.
I am also using the RedHat "tuned-adm" profile package, lightweight display manager and XFCE4 desktop, JRMC 26 and REW via ALSA direct to hardware.
Here are some of the latencies I have measured using cyclictest during 2 channel playback (X-axis is in microseconds).
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: