Recommended PC based DSP hardware and software?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello and happy new year everyone! Long time reader, first time poster. Over a decade of reading, love the forum.

I am in the planning stages of what I think is considered a simple 2.0 setup for my office/gaming/computer room. But it may actually be a 2.2 setup, I'm not sure.

I am looking to have 2 smallish to medium sized towers, on each side of my desk, so I can relax and and listen to music/videos and such without having my headphones on.

For practice purposes I would like to make each speaker box a 4 stage unit, with an active crossover setup. Tweeter, highs, mid, low, sub. I realize that is completely unnecessary, but I would like to try and do it, again, for practice and a challenge. I am hoping to get some practice so I can successfully upgrade our current projector/movie room's audio setup past the auto-configuration in the receiver. :p

My question to this part of the forum is, what hardware and software will I need to do this? I have been doing some research but have a hard time finding the correct terms to get good results. My main computer is a Windows PC, but I have extra computers available if it is recommended I made a dedicated Windows or Linux box specifically for the speakers.

What I have found so far is:
Hardware-
USB Sound card : https://www.amazon.com/StarTech-com-7-1-USB-Sound-Card/dp/B002LM0U2S
I am concerned that is will be low quality?
Maybe a PCIE sound card if that will help? Like this one? : https://www.amazon.com/Creative-Blaster-Audigy-Performance-Headphone/dp/B00EO6X7PG/

Amplifier: I have been looking for an 8 channel amplifier, but have only been able to find this 12: Dayton Audio MA1240a Multi-Zone 12 Channel Amplifier

Cables: 4 of these 3.5 MM to RCA https://www.amazon.com/AmazonBasics-3-5mm-2-Male-Adapter-Stereo/dp/B01D5H8JW0

Microphone : miniDSP UMIK-1 Omni-directional USB Measurement Calibrated Microphone

Software: REW and Peace equalizer? Is that really all that is needed?

Other than the speaker wire and the speakers themselves, is this all I need hardware wise to complete the project? Do you have any other recommendations?

Thank you.
 
Last edited:
I think a good sound card is essential for a pc based setup.
Asus makes a good one if your budget allows it: Essence STX II 7.1 | Sound Cards | ASUS Global
They make cheaper cards too.
Creative has been around for a long time, I guess their cards are fine too. I would stay away from cheap usb cards.

One important factor is volume control, if you're going to use the pc's volume control I suggest you use a sound card with the lowest amount of noise.
 
Shampless self-plug: You might want to try my crossover software for linux/pulseaudio:
Pulseaudio Crossover Rack :: T5! DIY Audio Software & Hardware .
There's also a thread here: Pulseaudio Crossover Rack - multi-way crossover design & implementation with linux

As a recommendation for hardware I'd go for an x86 based mini PC (Intel NUC or similar, old Laptop etc) instead of one of the many arm based solutions (Raspberry Pi etc).

Currently I'm using an ASUS Xonar U7 7.1 soundcard, sounds pretty decent. The recommended Essence STX II 7.1 should also work with linux.
 
The best desktop sound I've enjoyed was from a pair of simple 100 mm Audax full-range drivers (in sealed boxes, for an F3 somewhere below 100 Hz). (It was always intended to become a 2.1 setup, but I never got around to finishing the active crossover, and then the foam surrounds rotted...)

I suspect that when multi-way speakers are used up close, any movement of your head means a significant change in path differences between the drivers, which will affect stuff like imaging.
 
Amplifier: I have been looking for an 8 channel amplifier, but have only been able to find this 12: Dayton Audio MA1240a Multi-Zone 12 Channel Amplifier

A basic surround receiver has a lot of amplifier channels. So if you can find a surround receiver with more loudspeaker channels than you need, that is suitable as well. (Sub-channels are pre-amplifier only, and often have a bandwidth limitation.)

If the surround receiver has audio input over HDMI, you could skip a dedicated soundcard because you don't need the analog outputs.

Also, after a certain period, the on-board soundcards have adequate specs (linear frequency response, high sample rates and bit rates, low noise). If you have a PC with decent on-board multi-channel audio, you might not need the dedicated soundcard. It depends on the quality of the overall PC design. I have various laptops with very similar Intel codecs, some of them are dead quiet and sound nice over the analog output, some are hissing and humming and unusuable.
 
Generally PC soundcards (internal, external) run higher risk of ground loops (hissing, humming). IME whenever my internal soundcard produced humming (PC operation sounds: increased CPU load, mouse moving, screen refreshing), the culprit were ground loops. After their elimination (typically using balanced inputs or class II amplifier) the sound even from internal soundcards was surprisingly decent.
 
Ground currents on Pi3 board between pins of the GPIO connector (i.e. the ground loop there) are negligible, compared to the potentially huge currents flowing somewhere along the path between onboard audio output ground and the ATX PSU 0V terminal (the PC audio ground loop).
 
The past two years I've had a behringer 204hd, minidsp 2x4hd, and Asus xonar u7mkii. U7mkii for the win. The behringer was way too noisy on my horn system. The minidsp was loads better but still slight slight hiss from 1m or less. Ended up wanting to go 3 way so tried the u7mkii on a lark and damn if it isn't dead absolute quiet with my ear inside the mouth of the horns. I'm sure the bigger minidsp solutions are great but a xonar u7 is only $100 for eight dead silent channels.

I'm using equalizer apo half the time and pulse xoverrack the other half. Still deciding what I prefer. They are both great and both are free and open source too.

I also tried brutefir and it's too glitchy and fiddly with three different sound cards when used as an alsa plug but does a better job under jack. Prefer the xorack though. Tfive is the man for Linux active xo as far as I am concerned. But I still can't part with eapo I like the fact it is "script based" and can be easily configured using a gui I designed for myself where I can set tilt on the fly and stuff like that. I think I can get there with tfives xorack too just need to quit being lazy and finish up a python UI I've been working on here and there.
 
Last edited:
It's hard to beat the Xonar U7 in the performance/buck category IMHO. I got mine used in mint condition for 50€ on ebay. The only thing that you have to keep in mind is that it punps out quite a hefty thump on turnon, if you have large filter caps or DC coupled inputs on your amps always wait a little after turning on the soundcard before turning on the amp. I can confirm that it's dead silent during normal use though.
 
Shampless self-plug: You might want to try my crossover software for linux/pulseaudio:
Pulseaudio Crossover Rack :: T5! DIY Audio Software & Hardware .
There's also a thread here: Pulseaudio Crossover Rack - multi-way crossover design & implementation with linux

As a recommendation for hardware I'd go for an x86 based mini PC (Intel NUC or similar, old Laptop etc) instead of one of the many arm based solutions (Raspberry Pi etc).

Currently I'm using an ASUS Xonar U7 7.1 soundcard, sounds pretty decent. The recommended Essence STX II 7.1 should also work with linux.


Hey, this looks just the ticket for me. I want to incorporate a sub into my stereo system to take a bit of low end duty off my magnepans.
I already have a Ubuntu based laptop as my main audio source going into my DAC.
I was contemplating going down the route of getting a MiniDSP but it looks like i could do this with your software, I have 2 questions
1: does it allow for other DSP filters/effects on top of the cross over (Ideally id like room eq, time correction and the ability to sum the low passed portion of the crossover to mono for my sub)
2: Will it recognise more than one interface, my current DAC only has on stereo output so I'd need to add another one to handle the analogue output for the sub.

Thanks
 
Pulseaudio does handle multiple soundcards well, though I would not recommend it. Better get a nice multi-channel soundcard so that all channels are perfectly in sync. You mentioned time correction which only makes sense if there is ZERO time dependant skew between channels. By that I mean a delay between channels, that is changing over time... In any case you can always test it with a second soundcard an report back, cheap and easy :)



As for the other questions: You can do as many EQs/filters/delays as you like. You can also sum the output for the sub to mono.
 
Pulseaudio does handle multiple soundcards well, though I would not recommend it. Better get a nice multi-channel soundcard so that all channels are perfectly in sync. You mentioned time correction which only makes sense if there is ZERO time dependant skew between channels. By that I mean a delay between channels, that is changing over time... In any case you can always test it with a second soundcard an report back, cheap and easy :)



As for the other questions: You can do as many EQs/filters/delays as you like. You can also sum the output for the sub to mono.

Thanks for your reply.
sorry, perhaps i didnt use the correct terminology with regards to time. I was referring to time alignment. I've read a lot of people who use miniDSP use time alignment to compensate for delay (usually of the sub) so the sound from the speakers and the sub reach the listening postion at the same time.

Either way, sounds like this does everything I need and saved me £250 from buying a miniDSP that does the same job.
 
Last edited:
Well, if the sub lags behind then you add a delay to the tweeter. By doing that you can effectively time-align them.

This will not work properly though, if the two(four) outputs for the sub/tweeter do not have a fixed delay between them at any time. I.e. if they live on different sound cards / DACs that are driven by different clocks.

In recording studios this problem is solved by using some kind of clock synchronization aka "Word Clock" distribution. This is pretty uncommon to see in consumer devices. So my advice for you would be to get a decent multi-channel sound card.
 
Well, if the sub lags behind then you add a delay to the tweeter. By doing that you can effectively time-align them.

This will not work properly though, if the two(four) outputs for the sub/tweeter do not have a fixed delay between them at any time. I.e. if they live on different sound cards / DACs that are driven by different clocks.

In recording studios this problem is solved by using some kind of clock synchronization aka "Word Clock" distribution. This is pretty uncommon to see in consumer devices. So my advice for you would be to get a decent multi-channel sound card.

It just so happens I produce music and have an interface that does just this. :) So I can test the full solution with my studio interface before considering getting another DAC for a more permanent set up.

Either way though, your software looks fantastic. I cant wait to finish work and get tinkering this evening
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.