A how to for a PC XO.

[Thunau]

Hi Jan

I've finally gotten around to checking out the Allocator/Arbitrator.

Things are looking pretty good so far.

I'll post a thorough writeup in after I've had a lot more exposure.

A couple of things missing that immediately struck me as being obviously needed:

- Ability to setup seperate XO schemes for the left and right channels. This is important however trivial it may sound. Imaging in particular really takes on a pin sharp quality once you have very closely matched pairs of speakers. I should imagine this feature will be fairly easy to implement.

From my own and other's experience it hardly ever helps to optimize crossovers separatly for left and right channels. It could be advantageous to use an EQ before each channel and correct the small differences due to room or possibly loose tolerances.
The workaround for the real tweekers would be to start two instances of the Allocator and leave one side of each not assigned to any input or output. This would net very similar CPU load as a single stereo instance. Than you could import different frd's into each side and tweak to perfection.
I'll see if going additional two layers of GUI is easy to do, but I really don't think there would be many users actually taking advantage of this feature.

- Dynamic skins; The program chops information off the display and those controls are inaccessible with any less than 1024x768. This isn't much good for me as my main display is a widescreen projector with a resolution of 1280x720 and my secondary display is a small 7.5" TFT with a resolution of 800x600. This may not be a problem for the masses but its something thats annoying for us that have less tradition displays. I always really like the waves stuff because they were compact but had all the options laid out nicely.

Yes, I had that request before from someone with a remote desktop running on a small display. I guess scrolling would be easy to implement. Expanding skin- probably much harder as all the graphics you see are static png's. The whole GUI would have to be rewritten to accomodate dynamic sizing.

The following isn't essential but would greatly increase the appeal of the program:

- FIR filters implemented. If this is done I'd like to see the ability to customise the coefficients and window functions, as well as the more generic types. The design package LEAP is excellent in this regard but it doesn't have a crossover emulator to allow you to 'listen' to these. Again this is a big ask and I'd totally understand if you turn around and say 'ain't gonna happen buddy'. Didn't you mention a possible FIR version of the software in our emails?

There will be a separate piece of software called Frequency Affirmer (had to put the FIR somewhere in the name ) It will let you do very steep slopes with no phase shift in the transfer function. It will also have a form of DRC preceeding the crossover filters.

BTW do you own LEAP? I could use a transfer function or two generated by it.

- DRC integrated into the whole package so as to prevent cluttered plugins syndrome (perhaps you could get in touch with Dennis and see about using his freely available source and packaging it with your own program, all you'd then need to do is implement a convolver). This can be achieved through using Console with the addition of Voxengo Pristine Space or a similar convolver such as Waves IR1.

See above

- The ability to measure drivers and generate .frd profiles within the program without relying on 3rd party apps to do the job, Without the .frd files, I've found that Allocator/Arbitrator is very limited in scope otherwise. Its full potential is only realised when your using actual driver data measured from the final enclosures. I know this is NOT a trivial thing to implement and perhaps relying on 3rd party apps is essential to maintain a reasonable cost for the app but I'd strongly suggest writing a more thorough manual just for this subject as you'd certainly be aiding ease of use for those newer to this. For those who've been messing with this stuff for a while now and have programs like ETF and SC, this isn't a problem.

I enquired with a couple of guys who wrote a measurement utility about getting a custom code but had no luck so far. My developer is really good with the DSP part, but has no experience with the measurement side of it. I didn't want to impose a huge learning task on him. But who knows, Maybe I can get it to the point where the measurements will be fully automated in the Allocator itself.

Without frd's the Allocator and Allocator light are still better than your run-of-the-mill digital crossovers. You can recall patches for the standard LR and Butterworth curves just like the DXC or others AND you can tweak q's and spread the cut-off points of the biquads to arrive at totally custom curves. You can't do that with the hardware boxes.
The Phase Arbitrator algorithm is not available anywhere.

Finally I don't mean to sound negative. All of these were just things that I noticed within the first 20 minutes of playing around with the program. Overall I think it looks promising. I'll be comparing it to the DEQX in terms of sound quality alone. I've been extremely pleased with the DEQX over the last 6 months so it will be interesting to see if I can approach and hopefully exceed what I've done with that.

I'd also love to compare it the Waves/Voxengo/DRC setup I had going very early on this year - I still swear it was better sounding than the DEQX but alas. I some 'issues' seemed to creep into the implementation. These still seem to be at large so I doubt any meaningful comparison can be drawn.

I'll get back to you with more thoughts as I get into testing more deeply.

Cheers
Ant

Thanks for your input. It's all noted.

 

[ShinOBIWAN]

From my own and other's experience it hardly ever helps to optimize crossovers separatly for left and right channels. It could be advantageous to use an EQ before each channel and correct the small differences due to room or possibly loose tolerances.
The workaround for the real tweekers would be to start two instances of the Allocator and leave one side of each not assigned to any input or output. This would net very similar CPU load as a single stereo instance. Than you could import different frd's into each side and tweak to perfection.
I'll see if going additional two layers of GUI is easy to do, but I really don't think there would be many users actually taking advantage of this feature.

I was looking more at using the program as EQ for the stuff under 200hz primarily. I'm sure you'll agree that there's often rather large discrepancies between loudspeakers placed in room at and below this frequency. Again this isn't a problem if you use DRC as it uses seperate correction filters for the L&R channels.

If you got the ear of the developer, you might as well bend it

Yes, I had that request before from someone with a remote desktop running on a small display. I guess scrolling would be easy to implement. Expanding skin- probably much harder as all the graphics you see are static png's. The whole GUI would have to be rewritten to accomodate dynamic sizing.

If you get a chance to implement it, then the panning thing would be perfect. There's only a small portion chopped off (the EQ section on 800x600 display). I do however understand that 99% of users have displays capable of a vertical res of 768 and higher.

[qyite]There will be a separate piece of software called Frequency Affirmer (had to put the FIR somewhere in the name ) It will let you do very steep slopes with no phase shift in the transfer function. It will also have a form of DRC preceeding the crossover filters.[/quote]

Now that sounds like something I'd be even more interested in. Any idea's on a release date for that one. I'd definitely shell out good money if its comparable to the waves/voxengo setup.

BTW do you own LEAP? I could use a transfer function or two generated by it.

Yes, verion 5. I'll email you to discuss.

I enquired with a couple of guys who wrote a measurement utility about getting a custom code but had no luck so far. My developer is really good with the DSP part, but has no experience with the measurement side of it. I didn't want to impose a huge learning task on him. But who knows, Maybe I can get it to the point where the measurements will be fully automated in the Allocator itself.

I think that feature would really complete the package. Its not absolutely essential but it would be nice to have a stripped down measurement facility that excelled at just this one task. The 3rd party apps don't do that quite so seemlessly,

Without frd's the Allocator and Allocator light are still better than your run-of-the-mill digital crossovers. You can recall patches for the standard LR and Butterworth curves just like the DXC or others AND you can tweak q's and spread the cut-off points of the biquads to arrive at totally custom curves. You can't do that with the hardware boxes.
The Phase Arbitrator algorithm is not available anywhere.

Sorry Thuneau, I can seem like a right negative bas**** sometimes, when my actually intent was praise.

Totally agree. Its certainly got configurablility down well. I wasn't drawing conclusions that the program has limited value without the actual driver response but rather the end results are of limited to a game of guessing. Good results can still be got by ear but despite all that you read, it doesn't get you an accurate speaker and that's what I'm in this for. So to me the .frd are essential and elevate the program and end results into a whole other league.

 

[Thunau]

Sorry Thuneau, I can seem like a right negative bas**** sometimes, when my actually intent was praise.

Totally agree. Its certainly got configurablility down well. I wasn't drawing conclusions that the program has limited value without the actual driver response but rather the end results are of limited to a game of guessing. Good results can still be got by ear but despite all that you read, it doesn't get you an accurate speaker and that's what I'm in this for. So to me the .frd are essential and elevate the program and end results into a whole other league.

No offense taken. But, how do you achieve good results with DCX? You can only use your ears or a mic with an RTA and tweak the overall response. The same procedure applies to the Allocator if you don't have frd's.

BTW, I bought an used Ashly XR4001 4-way crossover and a DBX graphic EQ for my basement speakers (my wife likes to listen to her old LP records when doing laundry). They are 3-way Dynaudio and Morel drivers (all acquired used) in old yard sale cabinets. I fired up SMAART and tweaked the analog knobs for a while. I now have a set of killer speakers that are almost too good for the basement.
So yes, tweaking preset curves can lead to good results. But, when I fire up my other 3-way speakers with carefully optimized custom crossovers (based on measurements), the jump in quality
is definitely noticable.
Anyone can get good results without measuring and using frd's. The real power users will make the extra effort to measure and be rewarded with superior speakers.

 

[DoomPixie]

Did anyone ever try this with the m-audio delta 1010 in the end? just asking as i have a spare.. Use 2 of them in my Studio PC and one of them went down but seeing as i needed the PC the next day as i had someone comeing in to use the studio i went and ordered a new one and then ofcourse i got the replacement back from RMA and now i have 3.... Great cards.. they have 10 analog in and 10 analog out as standard.. couldnt fault them even if i wanted too..lol..
Owen
EDIT: forgot to mention.. i have used 16 In's and 4 Outs simultaneously useing these cards without any problems.. not sure if anyone else useing the studio has tried useing more than that but they would have told me if they had any problems im sure..lol.

 

[Thunau]

Yes, M-audio cards have been tested with the Allocator and Allocator Lite. No real known problems. If all your other ASIO apps work with it, you should have no trouble using it as a crossover.

 

[DoomPixie]

I guess i'll give it a try and see how i get on then..
Thanks,
Owen

 

[capslock]

From my own and other's experience it hardly ever helps to optimize crossovers separatly for left and right channels. It could be advantageous to use an EQ before each channel and correct the small differences due to room or possibly loose tolerances.
The workaround for the real tweekers would be to start two instances of the Allocator and leave one side of each not assigned to any input or output. This would net very similar CPU load as a single stereo instance. Than you could import different frd's into each side and tweak to perfection.
I'll see if going additional two layers of GUI is easy to do, but I really don't think there would be many users actually taking advantage of this feature.

Well, from my wife's favorite listening position, the distances to the two mains are pretty different, so having a different delay for each side would be great. Also, the first order reflections are quite different, so some eq might be helpful.

If two instances of Allocator full work, this might be a solution, albeit a cumbersome one.


However, there seems to be a problem with the ASIO drivers of my Delta 410 in that I see outputs only in pairs, i.e. when I assign outputs, I see entries like out 1/2, out 3/4, out 5/6, out 7/8, and there are two of each. I will have to see if they are actually 1 and 2, in spite of having exactly the same name.

 

[Thunau]

Well, from my wife's favorite listening position, the distances to the two mains are pretty different, so having a different delay for each side would be great. Also, the first order reflections are quite different, so some eq might be helpful.

If two instances of Allocator full work, this might be a solution, albeit a cumbersome one.


However, there seems to be a problem with the ASIO drivers of my Delta 410 in that I see outputs only in pairs, i.e. when I assign outputs, I see entries like out 1/2, out 3/4, out 5/6, out 7/8, and there are two of each. I will have to see if they are actually 1 and 2, in spite of having exactly the same name.

ASIO outputs are mono by definition. The labeling might be confusing. Each "duplicate" entry is actually left and right side of the pair.

 

[maoumaou]

Hi guys

Congratulations for the excellent job you are doing here

I have an Audigy 2 + kx drivers which I used in the past for my first attempt on a simple 2-way PC XO. I was left with an open mouth at that time with the results on my low-end b&w speakers. The passive xo they have just seemed crap. I soon abandoned the project cause I wanted to use all the outputs for a 5.1 system.

Now after starting reading this thread I begun imagining a pc with two audigy 2 cards linked via spdif, one for playing up to 5.1 channels and the other for using the front two channels of the first one to do a 2-way XO with CONSOLE.

Since I don’t have the second audigy card yet I started playing with CurveEQ. Just for the record, for those who can’t afford expensive cards and since so nicely pointed out before the sound card isn’t the weakest link I used my audigy 2 with ASIO in the CONSOLE and it worked fine. I liked the result. I did get an improvement in the low frequency response of my room. Also I used it for my center channel as well and now the L-C-R speakers are correctly matched. But that was the easy part. Now I have some questions.

1) Correct me if I’m wrong. To use the Pristine Space VST (for time domain corrections right?) you capture the IR of both L+R using for example an MLS from a measuring software (in my case Speaker Workshop), saving it as a wave file with 16-bit 48kHz resolution and importing it in the VST plug-in. Then you invert the IR and that’s it, right?
2) Isn’t the actual IR wrong since the output of the system is already delayed by a few msec due to latency?
3) Even in Real-time priority, the sound skips some times (i.e. minimizing-maximizing apps), so it is inevitable that the XO PC must be purely dedicated for that job, right? The CPU power is not an issue (3200 Athlon).
4) Can anyone tell me what the worst latency you came across in CONSOLE was? Theatertek has a lip sync range of +-200msec. Is that enough?

That’s all for now. More ??????? will follow in the future. I hope I’m not getting u tired but I have no one else to ask.

 

[peufeu]

Sorry, felt lazy reading the 50 pages

Has anyone mentioned that BruteFIR+Jack on Linux will do active crossover with very little CPU use ; and it's opensource and costs 0 € ?

 

[ewildgoose]

Frequently...

..Please post some details of your own experiences and any config scripts that you use here:

http://www.duffroomcorrection.com

It would be nice to see some of the waves/console, etc stuff documented on there as well to avoid others having to wade through the 50 pages of stuff to try and find the latest conclusion... Any takers?

Ed

 

[RR]

Sorry, felt lazy reading the 50 pages

Has anyone mentioned that BruteFIR+Jack on Linux will do active crossover with very little CPU use ; and it's opensource and costs 0 € ?

Yes, but it has become diluted. Absolutely, it's the best combination to use all the power of your CPU, and use if desired weird 64K coeff. FIR filters, with two instances of series global precision equalization, and use only about 20% CPU time in any decent modern machine.

Aside for tight control of per driver delay, 64 bits processing, dither to 24 bits if needed. Quite refined, I'd say.

I've done a killer machine (and silent, btw) for about 500 €, including soundcard (M-audio Delta 1010-LT).

Cheers,

Roberto

 

[ewildgoose]

I've done a killer machine (and silent, btw) for about 500 €, including soundcard (M-audio Delta 1010-LT).

That's quite a keen price if it includes case and it's quite quiet

Can we see a writeup with pictures added to the wiki sometime soon...? ...we all learn by reading about how others have progressed, so please do take some time to document what you have achieved so that we can all take a step forwards

Good luck

Ed W

 

[RR]

Hello Ed:

This is the second time that I say to you "I will do", and I feel quite ashamed, but... I will do.

Since I have the bill around, let´s see:

Producto : Celeron D 326 (2533 MHz)
Cantidad : 1
Precio unidad : EUR 54,-
Importe total : EUR 54,-

---------------------------------
Producto : 775I65GV (Intel i865GV)
Cantidad : 1
Precio unidad : EUR 46,-
Importe total : EUR 46,-

---------------------------------
Producto : Big Typhoon (4000+)
Cantidad : 1
Precio unidad : EUR 36,-
Importe total : EUR 36,-

---------------------------------
Producto : CF Carddrive (CF)
Cantidad : 1
Precio unidad : EUR 44,-
Importe total : EUR 44,-

---------------------------------
Producto : CompactFlash Card (256 MB)
Cantidad : 1
Precio unidad : EUR 14,-
Importe total : EUR 14,-

---------------------------------
Producto : DIMM 256 MB DDR-400 (Para PC)
Cantidad : 1
Precio unidad : EUR 25,-
Importe total : EUR 25,-


Total: 219 plus VAT.

The interesting parts are the compact flash acting as a hard disk (no noise) and the Big Thyphoon, that is surprisingly silent from the listening position (I don't notice it, but it sounds in close-up). Look for something with 12 cm. fan.

I had the case already, an AOpen H450B. It has a power supply with 12 cm. fan also. No aditional fans are connected. 64 €, for a total of 283.

Add 190 € for the soundcard, VAT and shipping, and you are more or less in the 500 € point.

The machine acts only as a convolver, taking the digital out of a external CD player, although using analog sources is not a problem.

Cheers,

Roberto

 

[RR]

And a picture. See the Marantz AV-4600 receiver, handy for digital crossovers (analog 8 channels volume control, and 7 power amps). ~300 €. A power amp for bass channels can be a useful adition, possible using the preamp outs.

On the back you can see a DIY 6 channel line cable, made with microphone cable (borrowed from the national TV), connectors are nice WBT clones from a local shop (2 €/piece).

An externally hosted image should be here but it was not working when we last tested it.

 

[Ninfendo]

DRC in Windows Vista

Hello, I just read this thread on AVS Forum, seems like Vista will have DRC built in, all you will need is a soundcard with Vista compatible drivers. Does anyone have any opinions on this?

http://www.avsforum.com/avs-vb/showthread.php?t=713073

Also, Audiotrak is releasing its latest prodigy card the 'Prodigy 7.1 HiFi'. It will sell for under GBP 100. It uses the following hardware: VIA Envy24HT (ICE 1724) audio controller and the Wolfson WM8776 / 8766. How does this compare to the 7.1LT?

http://www.audiotrak.co.uk/home_theatre/prodigy71hifi/prodigy71hifi.shtml

 

[ewildgoose]

Hmm, wait until it's released. Microsoft are experts at selling you something that isn't even going to be released for ages yet...

Bear in mind their history for releasing a product which *nearly* solves the problem (bar the numerous bugs and deficiencies that you spend more time working around than if you have just written your own from scratch), but not quite solving the problem and you may see that happen again here

I will place bets that the end result is a hideously complicated "bass management system on steroids" which isn't quite a programmable filter system, yet is *nearly* good enough that people waste a ton of their lives trying to get by with it

Remember that room correction is all about the filter generation process. Applying the filters to the audio (ie the audio framework) is technically the easy bit... There is nothing in the article to suggest that this product does more than a bit of a parametric tweak to the freq response to even it out a bit - surely great for desktop speakers, but not really "hifi"

Ed

 

[ShinOBIWAN]

Hmm, wait until it's released. Microsoft are experts at selling you something that isn't even going to be released for ages yet...

Bear in mind their history for releasing a product which *nearly* solves the problem (bar the numerous bugs and deficiencies that you spend more time working around than if you have just written your own from scratch), but not quite solving the problem and you may see that happen again here

I will place bets that the end result is a hideously complicated "bass management system on steroids" which isn't quite a programmable filter system, yet is *nearly* good enough that people waste a ton of their lives trying to get by with it

Remember that room correction is all about the filter generation process. Applying the filters to the audio (ie the audio framework) is technically the easy bit... There is nothing in the article to suggest that this product does more than a bit of a parametric tweak to the freq response to even it out a bit - surely great for desktop speakers, but not really "hifi"

Ed

Its nice to see this sort of thing creeping into the mainstream though. With Vista, DRC(whatever form that may take) will be in millions of homes worldwide. If anything it will only raise awareness and maybe even pave the way for more competant and professional 3rd party applications.

I also doubt that this will be a suitable option in its first release but judging by MS's previous performance with video playback(which rivals scalers costing thousands) it may someday become a realistic solution. One things for sure DRC is here to stay and is now popping up in receivers, standalone hardware/software and is a feature that consumers are starting to demand, evolution and market forces will see this become a standard feature in much the same way as Dolby Digital is now.

After playing around with other OS's I can conclude that Windows is the only option for a well rounded PC. Hardware drivers on Linux are a problem and gaming/video support is well below what I consider acceptable.

Oh and Kmixer is dead in Vista Thank god for that.

 

[ShinOBIWAN]

Actually I just read this:

No you can not. If you bypass the mixer, you bypass the effects system where these things hook in.

Hopefully, you don't have to bypass the new mixer in Vista . It works far, far better than kmixer in XP.

This was a reply regarding drivers that bypass the mixer in Vista, apparently you lose the DRC functionality should this happen.

We all know that any self respecting pro audio card *will* bypass that mixer and use a mixer that is tied into the hardware DSP of the soundcard eg. RME, EMU, Lynx, Audiotrak and M-Audio.

This fact says to me that 80% of us here won't be considering this as anything but superflous in much the same way as those 3D spatialisers. We'll see anyway...

 

[Ninfendo]

I think you are correct in that more people will get into DRC thanks to Microsoft. But those serious about it will soon discover that there are better ways then the MS way.

Btw, Thuneau I'm am very impressed with your work! I'm using the Frequency Allocator Light version as I use my System for a lot of TV and movies. If you were to include DRC into the Allocator what latencies would we be looking at?