rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool

I can understand the concept and advantages of turning a PC into a headless linux box dedicated to filtering and possibly streaming only, but I must admit the idea of using a tweaked-to-the-bone "home PC" running windows and doing all sort of things (media player, video, web and apps, etc.), as well as acting as an active crossover is not something I feel confortable with :gasp:

I dont know why these guys are overthinking and overcomplicating it. A dedicated linux box with brutefir is about as simple as it gets for convolution.

maybe they are afraid of a command line only interface.

heck using it is easier than when I had 2x minidsp 2x4hd boxes.
 
...have audio hardware get direct CPU access, meaning not via the southbridge that share most of computers I/O's via 4x PCIe lanes connected to CPU. For example platform writing this message have 16 lanes with GPU direct to CPU but also have a 4 lanes slot direct to CPU where a PCIe one lane soundcard could fit or if its USB soundcard then place USB3.x 1 lane PCIe card in that slot, also platforms that run CPU's build in graphics there is often a 16 lanes PCIe free to try.
Using a (weak) video card via south bridge would free the x16 slot for direct connection to CPU. And HDMI can go direct pretty simply, of course. Lots of options if you don't need a beefy GPU in the same system.

Your additional x4 slot's a great option, but few CPUs offer more than 16 direct lanes. Even Xeon E3s fell to 16 lanes after Ivy Bridge.
 
I can understand the concept and advantages of turning a PC into a headless linux box dedicated to filtering and possibly streaming only, but I must admit the idea of using a tweaked-to-the-bone "home PC" running windows and doing all sort of things (media player, video, web and apps, etc.), as well as acting as an active crossover is not something I feel confortable with :gasp:

I dont know why these guys are overthinking and overcomplicating it. A dedicated linux box with brutefir is about as simple as it gets for convolution.

maybe they are afraid of a command line only interface.

heck using it is easier than when I had 2x minidsp 2x4hd boxes.

Feel me hunted here :p but okay deep respect for Linux users, also intend myself take the learning curve when time allows and actual in Christmas holydays had JRiver up running on Ubuntu Gnome 16.04 LTS and yes i know that was bit cheating because it had GUI but it was Linux world :D.

That said command prompt query "systeminfo" my win7 reports back 14 marts 2013 22:27 so whats wrong with stability it had always started up every day since :eek:. Okay this computer doesn't steer speaker system which have other Win7 dedicated computer but that's for another reason, it could do it and actual is used as player and daily workhorse but also it wasn't cheapest hardware at beginning.

Using a (weak) video card via south bridge would free the x16 slot for direct connection to CPU. And HDMI can go direct pretty simply, of course. Lots of options if you don't need a beefy GPU in the same system.

Your additional x4 slot's a great option, but few CPUs offer more than 16 direct lanes. Even Xeon E3s fell to 16 lanes after Ivy Bridge.

Good idea move to southbridge actual some graphics have only 8 real lanes and have tested GT610/GT640 2.rev/GT720 to run fine from 4 real lanes into southbridge. Also the other way have tried to see if old PCI audio card can be pressed to lowest buffer settings giving it southbridge all alone and install all dirty I/O direct to CPU via shared AHCI/USB3 card, and that also worked good under the short exercise at least.

Yes its E3-1270 V2 Ivy bridge and so sad they dropped those 20 lanes since then to 16 and think costs is high nowadays step up to E5 series.
 
Administrator
Joined 2004
Paid Member
I ran a laptop with Ubuntu, JRiver as a sort of media sever for awhile. I used the impulses made in rePhase for EQ inside JRiver. It wasn't headless, but the lid was closed and the display off. Since JRiver will allow you to use the interface on one computer while running the server on another. I can't remember if that worked between Windows and Linux, but it did work from one Windows machine to the next. I may have used the web interface to control the Linux box.

I had done the setup before using a little Windows netbook with Atom processor as the music server and controlled from another computer running JRiver. Worked like a charm - except - the Atom CPU choked on the convolution and would stutter. Never could optimize it well enough to run the convolver smoothly. Linux on an old laptop had no problem with the same rePhase impulse.
 
BYRTT Thank you for your answer. I will try it out. Im not sure whether or not the "program" or "background", changes the sound that much. I have tried it before - But epic with the explanation. I toogle between Jriver showing/not showing all the time.


ernperkens: Thank you, I will check it out.

mark100: Thank you. I will try it out. I think that it is the whole thing about measuring/locking that I cant seem to figure out (Ie the loopback). Right now I am using one soundcard for playback and another for measuring.


Also feel a little bit hurt here about my solution to audio.:D
I would say that my future system might be to do convolution on a separate machine, though both the convenient-factor and budget comes in to play. My purpose is not pure music, but also video. The inbuilt "adjusting for delay in Jriver when using convolution, is working really good. If I only use Jriver, the whole system is really stable. So it is actually only when streaming content, there is an issue.
Is it not at little bit silly when watching movies to send the audio to another computer, while keeping the picture/visual on another ?

The bottleneck for me is to find at good and relatively cheap solution to sending the audio out from the computer to another. As I see it now, I would end up with three soundcards. Esi Gigaport+, soundcard with digital input, soundcard with digital output.

Didnt knew that jriver has the possibility to control Jriver from one computer to another - thought it involved teamviever etc.
 
Member
Joined 2009
Paid Member
I can understand the concept and advantages of turning a PC into a headless linux box dedicated to filtering and possibly streaming only, but I must admit the idea of using a tweaked-to-the-bone "home PC" running windows and doing all sort of things (media player, video, web and apps, etc.), as well as acting as an active crossover is not something I feel confortable with

I am not too, but in my case ( vintage os vintage pc components) i know how to have a stable and working environment with XP and old pro soundcard. After all i used this kind of config in my daytime job for maybe 15 years... ;)

I never had other soft than the one dedicated to music on this kind of config and win XP was installed with the needed feature nothing more. Highly cleaned version.

No wifi, no internet, no game, no dvd or blu ray,... This lead to have multiple pc but once you managed to have some 'nearly silent' config it's ok.

AS 1201 said it is easier than ever to have a silent config this day and for me the challenge is to have an old os running flawlessly with this components (the ssd and sataIII controller may be troublesome to configure).

But this can be done.

I dont know why these guys are overthinking and overcomplicating it. A dedicated linux box with brutefir is about as simple as it gets for convolution.

maybe they are afraid of a command line only interface.

Yes if you want only convolution i think this is the best way. I' ll use my config as a dedicated software drived audio processor/generator (using same hardware, same VST Host but with other audio VST like FX and some Synth) so it needs to be flexible enough. And don't think a simple VST system won't work: the soft i linked run great and stable ( VSTHost and Voxengo Pristine Space VST).

It's in place since maybe 20 years now and once the hardware is choosen right this is rock solid configuration. Lots of studio run with Nuendo and pc windows based for post production duties. Even if protools is the norm in prostudios some other solution does exist and are as good (or even better, think Sequoia ) but native.

And as i already stated this will be an experimentation, this is for a second system and my mains already have a Dolby Lake for what they need (and as from time to time still works with musicians i need the lower latency possible).

If it work good i could use an headless running linux solution long term in my living room and maybe as a DRC with my mains.

I must admit that in my case the linux GUI is not what i would call appealing but i suppose once you learned how it works it would not be a problem. But for experimentation it'll be less time consumming for me this way.
 
Last edited:
As a former ICT system administrator I don't really see the problem using Windows once you get it running reliably. I wouldn't use just any Windows versions as Windows has had its hit and miss versions. Right now I'm using Windows 7 Pro as my Audio and Home Theatre PC and it doubles as my 3D design PC. It's workstation grade (Xeon quad core) with the needed graphics card for design duties which makes Home Theatre possible with the TV as secondary monitor. I got a slightly older model with PCI slot to accommodate my trusty Asus Sonar Essence ST which gets me favourable results.
Yes, sure I've got to store Blu-ray and even DVD as MKV, not really a problem. I just love the flexibility of JRiver combined with a couple of VST plugins. Latency in video playback is solved within JRiver. The freedom to use any processing task or routing I may come up with is liberating. Even with full audio processing, while running Blu-ray video it's only using less than 10% of my processing power. I've even played my son and his friends movies while doing 3D design work simultaneously. It has been rock solid like this for years.
When I upgraded my PC a few years ago I did think about using my old one as a dedicated audio box, but due to the better graphics card in my new Workstation I figured I might as well stick with one PC and use the extra power for some MadVR tricks within JRiver. My (9 year) old PC is doing audio tasks at work for my students, as well as 3D design. (also running Win 7 Pro)
 
Last edited:
I did not intend to hurt anybody's feeling, sorry if I did :D

What has me worried is having a PC (be it linux or windows, really) running a crossover together with other unrelated softwares.
I am sure it can be made stable, but that implies a lot of care and monitoring, and is not compatible with the way I want to deal with my home PC (installing software, testing stuff, etc.): too much constraints. I also want to be able to upgrade my home PCs at a different pace than the audio system.

That is why I feel having a dedicated PC for the crossover (or DSP hardware solution), and then using another one (or a chromecast, or whatever source might appear over the years) as a source is a better bet in the long run and offers more freedom (like doing random stuff on your PC without worrying about stability issues).
 
I did not intend to hurt anybody's feeling, sorry if I did :D

What has me worried is having a PC (be it linux or windows, really) running a crossover together with other unrelated softwares.
I am sure it can be made stable, but that implies a lot of care and monitoring, and is not compatible with the way I want to deal with my home PC (installing software, testing stuff, etc.): too much constraints. I also want to be able to upgrade my home PCs at a different pace than the audio system.

That is why I feel having a dedicated PC for the crossover (or DSP hardware solution), and then using another one (or a chromecast, or whatever source might appear over the years) as a source is a better bet in the long run and offers more freedom (like doing random stuff on your PC without worrying about stability issues).


in my living room I use one pc for convolution only, and another one for everything else.
 
All the talk about PC convolution, along with a helpful jump start from BYRTT, has moved me to try the JRiver route....

yesterday, I got a JR ID to run a 6 channel FIR filter, using the same FIR files I use in openDRC boxes, but converted to wav.

The ID is an Intel NUC running linux, that can be used as a standalone media computer, or a DLNA renderer or server. It can be run headless and controlled by phone.
I've been using it solely for 2 channel playback, either inside at home letting my Win 10 desktop stream to it, or when outdoors, as a standalone media computer.
In case you haven't seen it..JRiver Id

I really like the small silent form factor of the NUC, and the idea that ALL it is ever used for is playing media and convolution.
It would be great if convolution fully pans out.......

The ID's Performance reporting said the cpu was running 10x faster than real time, processing 6 threads of 6144 taps at 48 kHz. So it looks like there is room for the additional taps to handle lower frequency needs. And there are always faster NUCs if needed.

I can already see the lack of flexibility vs the openDRCs is an issue, as is how to use the ID with other signal sources....such as REW test sweeps.
But it's very encouraging so far...

How DO you folks who are using JR to convolve, use REW with your setup?
(I'm very unfamiliar with JR other than using it as a fancy IPOD...)
 
mark100,

Admit don't know anything about those nice JR ID's running Linux, but can tell under Windows platform go to Tools/Options/General/Features and tick "WDM Driver" active, then you get a new digital output soundcard route to use in REW so all the DSP correction is into sweeps, hint to avoid samplerate mismatch distortion remember to set right sample rate in Windows sound control panel for the new output device named "JRiver Media Center 22".
 
Last edited:
jriver linux (or mac) doesn't support any form of audio line in atm, which is a rather annoying gap, so you would need to use REW having exported it's sweep measurement as a wav and then selected that for playback by jriver. I think you then have to manually press "measure" on your REW machine at the right time in order to get a valid measurement. Is the id a general purpose linux machine or does it just run jriver? I thought it was the latter.
 
Administrator
Joined 2004
Paid Member
BYRTT Didnt knew that jriver has the possibility to control Jriver from one computer to another - thought it involved teamviever etc.
Yes, it's the "Ultimate Couch Potato Mode", tho I can never remember its real name. It' something you do once, then it always works. From the local computer running JRiver you control another instance running on a remote computer, and the interface is exactly the same. There are few tweaky things you cannot do remotely, but they aren't day to day needs. I used remote desktop to do tweaks.

Now with the new JR id, which I've never seen before (thanks Mark!) it should be even easier. It's a very cool step forward.
 
BYRTT, the ID has a WDM checkbox where you said to look, but my bet is 3lld00d is right about a no go.

3lld00d, I think it just runs jriver, with maybe some very limited general capability.
Tried seeing what startup options are available prior to jriver loading....this is what I got...means nothing to me :confused:
 

Attachments

  • linux r.jpg
    linux r.jpg
    55.6 KB · Views: 224
Administrator
Joined 2004
Paid Member
For my use, JRiver made sense. Back in 2008 I was looking for a player that had good remote control capabilities and found JRiver. I've stayed with it ever since. It is, by far, the best player I've found for organizing large libraries of files. That in itself is a great benefit.
Then you discover so many good features. Bit perfect playback if you want it, 64bit floating point DSP with built in routing, filters, convolution, channel mapping, sample rate conversion, etc.

Using rePhase and JRiver DSP you can do pretty much whatever you want, with one of the best interfaces and libraries around.