A how to for a PC XO.

[14G-Dutch^]

I am a newb, sorry if this question bites. =)

I have a Pc with an Athlon 64 3500, 1 gig ram, and a Soundblaster X-fi. I haven't seen much in this post other than u might need 2.

Is the X-fi viable for this application? It is supposed to be quite powerfull as far as its processor goes. So do u need 2 cards, and if so, is that even possible?

Thanks

Jon King

 

[ShinOBIWAN]

I am a newb, sorry if this question bites. =)

I have a Pc with an Athlon 64 3500, 1 gig ram, and a Soundblaster X-fi. I haven't seen much in this post other than u might need 2.

Is the X-fi viable for this application? It is supposed to be quite powerfull as far as its processor goes. So do u need 2 cards, and if so, is that even possible?

Thanks

Jon King

If there's Xfi linux drivers available you can try the BruteFIR method. But it won't work with Console.

It will however do DRC with just music using players such as winamp and foobar2k.

 

[oehlrich]

The "Perfect" XO for High End Stereo Systems?!

Hello folks out there,
reading this forum I found a lot of discussion about audio filters, ringing, steepness, linearphase and all that stuff.

Some time ago I started a project building a real nice standalone (96kHz, digital in, analogue out) fanless BruteFir engine. The engine is up and running now and I currently operate an active tri-amping setup with XOs and DRC on it. The audible result is really phantastic!

During the last months I read through a bunch of research literature regarding linear phase filters, audibility, room correction, jitter theory and all that stuff. I decided to share what I have learned during this project with you hoping it would help you on your way to your optimal High End Stereo System.

Today I will start with XOs.

I have prepared an article for you with my research, thoughts, experiments and tools regarding the "Perfect" XO for High End Stereo Systems ( see: http://freerider.dyndns.org/anlage/MyDrc/Perfect-XO-for-High-End-Stereo-Systems.zip ). In this article I also reference some freeware tools for XO calculation and coefficient generation for BruteFir or other convolvers.

Hope this material would be useful for you.

Charly

 

[ShinOBIWAN]

Thanks Charly,

I'll take a look at that.

 

[voodooless]

I had the DCX before I tried all this, my advice would be to avoid. Average at best would be my first thought but personally I'd rather go with passive than that route. The DCX is cheaper than the other methods for a reason.

Sure it is, but it is also affordable. In my point of view a huge advantage.

If you think that FIR filters are disadvantaged compared to the other methods available, you need to look and hear again. Like I said in my previous post, check out the DEQX reviews from publications that generally prefer the old methods and you'll quickly realise that it can turn average systems into a heavy weight one and top flight system into something pretty special. This is my experience with the PC setup in the past and something that neither passive nor analogue active will ever be able to do.

I don't think FIR filtering is a disadvantage! Quite the opposite actually And I know the DEQX is a great device. But hey, check the price tag! I'm not going to spend 3500 euros for it (I would take the one with the pre-amp). That's more than my speakers and amps will be going to cost together, and there is no way I'm going to think it's worth it. Don't tell me that I have to cheap stuff, it's all top of the line. The Brutefir solution will cost me only a pre-amp and a soundcard (I already have a RME 9652 + ouput board to start with). Besides that it will only cost me a lot of time to get it to work the way I want. For this, I get a lot of experience back (something you most probably have also gained, a lot!).

If you already knew all that then why are you generalising filter characteristics and suggesting the wrong things?

Again, I did quite the opposite form my point of view... It might have sounded otherwise

@oehlrich: sound very interesting: I'll look into that!

 

[ShinOBIWAN]

Sure it is, but it is also affordable. In my point of view a huge advantage.

It is if your just starting out and want something to test the waters with. But I think you'd quickly out grow it or at least I did. I'm not sure but I think I only had it a couple of weeks - a record for me and I swap kit pretty often. Its not that its a bad product, just that there's much better stuff at a price of course. How much your willing to spend is probably relative to the rest of your kit.

I don't think FIR filtering is a disadvantage! Quite the opposite actually And I know the DEQX is a great device. But hey, check the price tag! I'm not going to spend 3500 euros for it (I would take the one with the pre-amp). That's more than my speakers and amps will be going to cost together, and there is no way I'm going to think it's worth it. Don't tell me that I have to cheap stuff, it's all top of the line. The Brutefir solution will cost me only a pre-amp and a soundcard (I already have a RME 9652 + ouput board to start with). Besides that it will only cost me a lot of time to get it to work the way I want. For this, I get a lot of experience back (something you most probably have also gained, a lot!).

I had the 9652 when I used two PC's for my setup. Its a decent card but the expansion boards use the inferior DAC's. I would have gone for the 9632 with the AO4-192S, also owned one of those too. Unless of course there was a specific reason why you went with the 9652, like using external DAC's. Eitherway you have what is needed for BruteFIR method although the Console way will have to be done using physical loopback via ADAT and wastage of the related IO's, the reason why is because RME recently disabled loopback on the drivers for this card since there were incompatibilities.

I'd also be inclined to start with BruteFIR any how, it might not offer the luxury of real time adjustments, absolute tweakability and ease of use but it could be better in the long run once you've sorted out your filter parameters. It also looks like it has lower latency.

 

[voodooless]

It is if your just starting out and want something to test the waters with. But I think you'd quickly out grow it or at least I did. I'm not sure but I think I only had it a couple of weeks - a record for me and I swap kit pretty often. Its not that its a bad product, just that there's much better stuff at a price of course. How much your willing to spend is probably relative to the rest of your kit.

Better stuff at the price? Like what?

I had the 9652 when I used two PC's for my setup. Its a decent card but the expansion boards use the inferior DAC's. I would have gone for the 9632 with the AO4-192S, also owned one of those too. Unless of course there was a specific reason why you went with the 9652, like using external DAC's. Eitherway you have what is needed for BruteFIR method although the Console way will have to be done using physical loopback via ADAT and wastage of the related IO's, the reason why is because RME recently disabled loopback on the drivers for this card since there were incompatibilities.

I took the 9652 + 8x DAC because I could get it cheap second hand. If I don't like Burutefir, nothing is lost, and I have a nice soundard in my PC. If I do like it, I can simple add better DAC's if I want. They are not that bad actually, at least to start with.

I'd also be inclined to start with BruteFIR any how, it might not offer the luxury of real time adjustments, absolute tweakability and ease of use but it could be better in the long run once you've sorted out your filter parameters. It also looks like it has lower latency.

Well, since I'm going to make a webinterface, I can add all those things, but I guess, it will take me some time. I might still go for a second hand DCX in the meantime to play arround with

 

[ShinOBIWAN]

Better stuff at the price? Like what?

Two cups and a peice of string?

Seriously though in the grand scheme of things the DCX is the merely average. Sorry.

 

[voodooless]

Seriously though in the grand scheme of things the DCX is the merely average. Sorry.

Might be, but for the price, I don't think you should complain. The DEQX costs more than 10 times more. And if the difference between avarage and very good is that much, I'll might as well stick to avarage for now. If I win a lottery, I can always buy a DEQX

I'll stick with the brutefir idea for now though. I first got to get the amps together

 

[seppstefano]

Re: The "Perfect" XO for High End Stereo Systems?!

Thank you charlie,
exactly what I needed !

Bis Bald,

Stefano

 

[dwk123]

If you read my post on the problem you would see that I have indeed tried quite a few plugin, including IIR... still think its the plugins

I haven't been following the details of this discussion all that closely, but am I understanding things correctly that you have been undertaking this effort without advanced measurement capability? ie sound easy, praxis etc? If you are getting distortion, I would imagine that carefully measuring input/output characteristics in the digital domain with such a package would be step 1. From the brief descriptions I've seen, I'd guess that you're getting digital clipping due to bandpass gain in your mid xover - if so, this would show up pretty clearly by running a full-scale signal through and capturing the results (this is the beauty of the ASIO/Console method - instead of sending the outputs to the physical outs on the soundcard, you can loop them back to Windows wav inputs for your measurement package)

I just want something that works.

I can relate to this sentiment. I have probably been playing around with this stuff as long or longer than anyone. My first DSP xover attempt was with a Motorola DSP kit almost 10 years ago. Gave up on that when I discovered BruteFir (early in its life, and it's been around for a while). (Temporarily?) shelved the PC route and went with a DCX for prototyping, and am now back to experimenting with some ASIO based approaches. In all that time, I've never had a 'stable' system that just lets me listen to music.

The event that finally gave me a stable system? Winning a pair of speakers in a raffle. I'm now running commercial speakers in basic 2-channel mode off my modified Panny receiver, and listening to more music than at any point in my DSP oddysey.

Doesn't mean I'm giving up on it though

 

[ShinOBIWAN]

I haven't been following the details of this discussion all that closely, but am I understanding things correctly that you have been undertaking this effort without advanced measurement capability? ie sound easy, praxis etc? If you are getting distortion, I would imagine that carefully measuring input/output characteristics in the digital domain with such a package would be step 1. From the brief descriptions I've seen, I'd guess that you're getting digital clipping due to bandpass gain in your mid xover - if so, this would show up pretty clearly by running a full-scale signal through and capturing the results (this is the beauty of the ASIO/Console method - instead of sending the outputs to the physical outs on the soundcard, you can loop them back to Windows wav inputs for your measurement package)

I have tons of loudspeaker softwares from LEAP5 through to Speaker workshop.

On the measurement side I have:

ETF 5
SIA Smaartlive 5
Dirac
TrueRTA
Audio Tester
Sample Champion
WinMLS Pro

Some of these are more for room and acoustic measurement but a couple such as ETF and WinMLS would be very suitable. Any how, no I haven't done any measurements at all. I'm pretty lazy these days and don't expect I can be bothered at the moment. Besides, I'm solving a big problem by moving away from the PC setup to a standalone hardware box.

I can relate to this sentiment. I have probably been playing around with this stuff as long or longer than anyone. My first DSP xover attempt was with a Motorola DSP kit almost 10 years ago. Gave up on that when I discovered BruteFir (early in its life, and it's been around for a while). (Temporarily?) shelved the PC route and went with a DCX for prototyping, and am now back to experimenting with some ASIO based approaches. In all that time, I've never had a 'stable' system that just lets me listen to music.

The event that finally gave me a stable system? Winning a pair of speakers in a raffle. I'm now running commercial speakers in basic 2-channel mode off my modified Panny receiver, and listening to more music than at any point in my DSP oddysey.

Doesn't mean I'm giving up on it though

Now all these rings very true with me. I miss just being able to play something and take it for granted that it works perfectly everytime. The PC way doesn't 100% do that even without the problems I've recently encountered.

My speakers are also a point of contention, after a year of construction or re-construction will they ever be finished? Sure they work but look ugly as hell in raw MDF and sealer everywhere. My system at the minute is kinda like the house that jack built - rough.

I'm getting out of the swapping around game and going for something that simply works and yes you do have to pay for the privledge and that's the reason why the DEQX is expensive. But at this point its actually looking like a great deal to me.

 

[voodooless]

yes you do have to pay for the privledge and that's the reason why the DEQX is expensive. But at this point its actually looking like a great deal to me.

Well, I can immagine that. Ans I also agree that you'll need to be able to just listen to the music! But I guess, that is in inherent problem of any high configurable system, you just keep tweaking it. Maybee, buying some commercial speaker and amp is a solution to that, but then you'll keep swapping out useless interlinks, powercords, and speakercables. Then I still prefer something that actually does make a difference

We shouldn't be to hard on ourselves

 

[ShinOBIWAN]

Well, I can immagine that. Ans I also agree that you'll need to be able to just listen to the music! But I guess, that is in inherent problem of any high configurable system, you just keep tweaking it. Maybee, buying some commercial speaker and amp is a solution to that, but then you'll keep swapping out useless interlinks, powercords, and speakercables. Then I still prefer something that actually does make a difference

We shouldn't be to hard on ourselves

Agreed and I've been down the commercial route and felt unfulfilled. It costs an arm and a leg for the really good stuff and there's far too much trimming that doesn't actually add anything to the sound.

I had to laugh the other day when I saw a pair of Aurum Cantus speakers on a British forum that somebody had taken pictures of. It wasn't the speakers as such (beautiful BTW) but the fact that they came with a pair of cotton gloves so you didn't have to touch that lovely gloss finish. I mean come on, are these some sort of sacred artifact or a pair of speakers?

Don't get me started on cables either.

 

[wigginjs]

Re: The "Perfect" XO for High End Stereo Systems?!

Charly:

Read your paper, very interesting. Anymore details you could provide about your particular implementation (hardware and software) and also if you wouldn't mind showing examples of the files that you generated and how you implement them. I generated a few filters using the spreadsheet, but when I graph them using Excel they seem to simply oscillate back and forth between ~1 and ~-1. Is this correct? I'd like to try to implement these, as well as DRC filters, using BruteFIR when my dedicated hardware arrives.


Thanks again!!!

Hello folks out there,
reading this forum I found a lot of discussion about audio filters, ringing, steepness, linearphase and all that stuff.

Some time ago I started a project building a real nice standalone (96kHz, digital in, analogue out) fanless BruteFir engine. The engine is up and running now and I currently operate an active tri-amping setup with XOs and DRC on it. The audible result is really phantastic!

During the last months I read through a bunch of research literature regarding linear phase filters, audibility, room correction, jitter theory and all that stuff. I decided to share what I have learned during this project with you hoping it would help you on your way to your optimal High End Stereo System.

Today I will start with XOs.

I have prepared an article for you with my research, thoughts, experiments and tools regarding the "Perfect" XO for High End Stereo Systems ( see: http://freerider.dyndns.org/anlage/MyDrc/Perfect-XO-for-High-End-Stereo-Systems.zip ). In this article I also reference some freeware tools for XO calculation and coefficient generation for BruteFir or other convolvers.

Hope this material would be useful for you.

Charly

 

[Vil]

there is pretty interesting Linear Phase EQ :

http://refinedaudiometrics.com/products-plpareq.shtml

they say " Phase Linear Operation is achieved by processing your sound in both the forward-time and reverse-time directions through classic filters - all in realtime. This completely removes the phase warping caused by IIR filtering "

so thats not FIR but IIR filter with linear phase . I tested it , phase is really linear , and there is no frequency ringing up to -140dB .

 

[Vil]

picture

 

[jtavan]

I believe this is how the Weiss EQ1-LP equalizer performs its linear phase filtering. I'd wondered if that was a technique proprietary to them, or if someone else had done something similar.

 

[Vil]

>>>>I believe this is how the Weiss EQ1-LP equalizer performs its linear phase filtering

yes , thats true . more info is here :

http://www.weiss.ch/eq1/images/eq1-lp_update.PDF

 

[voodooless]

they say " Phase Linear Operation is achieved by processing your sound in both the forward-time and reverse-time directions through classic filters - all in realtime. This completely removes the phase warping caused by IIR filtering "

so thats not FIR but IIR filter with linear phase .

Who sais it's fot FIR filteren. Since it uses both future and past samples, I bet this is simply FIR filtering, but they don't want to tell you