I would like to use the miniDSP 2x4 HD for room correction for my speakers. I don't exactly want to spend $1000 for a Dirac system. I'm wondering is there a guide on how to do room correction with FIR filters to achieve something similar to Dirac's room correction?
Thanks
Thanks
That's an easy question, the answers are all around you, depending on what you're willing to do.
Search for RePhase or DRC-FIR as it will not only bring the main threads about the software but probably some users actually using the software and sharing enough info to get started.
Prepare to do a lot of research on how to correct and/or what to correct for. DSP isn't an answer for every problem. I personally use DRC-FIR, but try and use it more as speaker correction, my specific speakers do need EQ to be able to do something useful at all.
Some useful links (lot's to read and learn):
http://www.diyaudio.com/forums/multi-way/221434-rephase-loudspeaker-phase-linearization-eq-fir-filtering-tool.html
http://www.diyaudio.com/forums/full-range/275730-convolution-based-alternative-electrical-loudspeaker-correction-networks.html
More reading material, from a paid package:
http://www.diyaudio.com/forums/full-range/284916-room-correction-speaker-correction-what-can-we-do-dsp-power-now-availabl.html
Useful tutorial of yet another software package (but the theory applies to (for instance) the free DRC-FIR):
Computer Audiophile - Acourate Digital Room and Loudspeaker Correction Software Walkthrough
and
Computer Audiophile - Advanced Acourate Digital XO Time Alignment Driver Linearization Walkthrough
First get an idea what the DSP part would need to do for you. After that you can pick the one you'd need.
My own thread filled with DRC-FIR mumbo jumbo (but probably way to big to wade trough):
http://www.diyaudio.com/forums/full-range/242171-making-two-towers-25-driver-full-range-line-array.html
Lot's more examples on here, both on going and finished projects.
Search for RePhase or DRC-FIR as it will not only bring the main threads about the software but probably some users actually using the software and sharing enough info to get started.
Prepare to do a lot of research on how to correct and/or what to correct for. DSP isn't an answer for every problem. I personally use DRC-FIR, but try and use it more as speaker correction, my specific speakers do need EQ to be able to do something useful at all.
Some useful links (lot's to read and learn):
http://www.diyaudio.com/forums/multi-way/221434-rephase-loudspeaker-phase-linearization-eq-fir-filtering-tool.html
http://www.diyaudio.com/forums/full-range/275730-convolution-based-alternative-electrical-loudspeaker-correction-networks.html
More reading material, from a paid package:
http://www.diyaudio.com/forums/full-range/284916-room-correction-speaker-correction-what-can-we-do-dsp-power-now-availabl.html
Useful tutorial of yet another software package (but the theory applies to (for instance) the free DRC-FIR):
Computer Audiophile - Acourate Digital Room and Loudspeaker Correction Software Walkthrough
and
Computer Audiophile - Advanced Acourate Digital XO Time Alignment Driver Linearization Walkthrough
First get an idea what the DSP part would need to do for you. After that you can pick the one you'd need.
My own thread filled with DRC-FIR mumbo jumbo (but probably way to big to wade trough):
http://www.diyaudio.com/forums/full-range/242171-making-two-towers-25-driver-full-range-line-array.html
Lot's more examples on here, both on going and finished projects.
Adding one more to Ronald's excellent list above, if you are considering using a PC and JRiver:
Room Correction Using Freeware and JRiver
For sure, lots more examples abound.
Have fun!
Room Correction Using Freeware and JRiver
For sure, lots more examples abound.
Have fun!
That's exactly what I did using multiple 2x4HDs in a 4way
After year of learning rephase I got a Dirac box and my own filters I made perform almost as good. Almost meaning pretty close. I'm not expert level at rephase I would say I'm intermediate to novice and could still make pretty decent filters.
I got the Dirac just for added fun but honestly didn't really need it .
Rephase is great and very accurate as well. What you see really is what you get.
After year of learning rephase I got a Dirac box and my own filters I made perform almost as good. Almost meaning pretty close. I'm not expert level at rephase I would say I'm intermediate to novice and could still make pretty decent filters.
I got the Dirac just for added fun but honestly didn't really need it .
Rephase is great and very accurate as well. What you see really is what you get.
How do you use multiple 2x4 HD, expecially with a digital input?
Please explain. What keeps the hardware units sync'd up without wandering?
Thanks
Powered optical splitter 1 in 3 out (15$)
Newest Digital Audio Splitter SPDIF / Toslink 1 In to 3 Out Powered Amplifier | eBay
The non powered ones work but powered will sync.
Have separate configurations for each driver set and toggling between platforms isn't THAT
Bad , I've looked and this combo is one of the only ways to accomplish it.
Separate minishqrks might be better, if you want to run a different dac but the akm in the minis is very decent. Or if you wanted to use 48k for more filter taps the mini shark might be the better choice , and they sell curry man DAC that pretty much plugs right in Thru a i2s cable ,
Bad , I've looked and this combo is one of the only ways to accomplish it.
Separate minishqrks might be better, if you want to run a different dac but the akm in the minis is very decent. Or if you wanted to use 48k for more filter taps the mini shark might be the better choice , and they sell curry man DAC that pretty much plugs right in Thru a i2s cable ,
you can use REW to measure the system and then use DRC to create filters.
IMO they sound way better than dirac filters and I wouldnt use filters because you have no control over the strength of the correction filters. the only thing you control with dirac is the frequency response, which IMO is not good enough.
if you want to spend just a little bit more and have a far more powerful system I would recommend using a dedicated pc with linux for your room correction. you will have more i/o, more taps, and more options.
I was using 2x 2x4 hds and I switched to a pc. best move ever.
I use rephase and drc for crossover and room correction
Did you just say you can use rew (auto eq) to make fir filters?
If so how? Is there a write up?
I read on rephase thread v1.2 does something with rew but I couldn't figure out what everyone was talking about.
Would be very curious to try.
Dirac does more than just frequency response that at least I understand it.
I'm not 100% but I think it makes a inverse Mag filter in IIR and a global phase offset in fir. Than allows the user to shape the magnitude any way you want and however you make it it shapes phase to that curve. It takes nine measurements but each measurement has 3-4 sweeps. So it's more like 27-36 sweeps which is decent averages. Not sure what you mean about strength of correction filters, the user can adjust the boost or cut of the filter based on the average displayed. But no can't tell it not to boost unless you put the target under lowest null. It does a fairly decent job, at least in the live version(not the trial)
you cant use rew to make fir filters just yet 🙂 . you just use it to make a measurement . then DRC will take that measurement, and make a convolution filter to correct it
rew auto eq now spits out to rephase so you can just import the rew file into rephase and the equalization is already done for you just generate the filter. it used to be a time consuming and painful process before this 🙂
you can shape the magnitude response of dirac anyway you want(hint- the beautiful after prediction of dirac is nowhere close to what it actually is when measured in REW)
BUT, you cannot adjust the time correction. we have no idea how many ms dirac corrects at any frequency and we cant change that so you are stuck with what dirac thinks is best. and I dont like their correction because I think its too strong.
with DRC there are more than 150 settings you can change(about 10 main ones) that alter how the sound is adjusted in the time domain.
Then you're probably looking at wrong data. It has been shown that the average response (the pretty curve in Dirac Live) is exactly the same when measured in REW or similar. There's data showing this over at AVS.
Very interesting stuff that I will be exploring. I have a pretty terrible sounding room, and I think correction in the time domain is going to be the key in getting better sound in my room.
Can you explain a little what made a PC better than 2x 2x4 HD's? I understand that PC's have a lot more processing power, so more taps, but shouldn't the 2x4 HD's have more than enough taps for midrange and treble correction?
Thanks
Can you explain a little what made a PC better than 2x 2x4 HD's? I understand that PC's have a lot more processing power, so more taps, but shouldn't the 2x4 HD's have more than enough taps for midrange and treble correction?
Thanks
That's not correct, Dirac FIR/IIR correct for phase (linear and minimum) as well as impulse response. In Theory, Dirac should be about as good as any other software, hardware solution. However, that may not be reality.😉
Interesting stuff. So I need to explore drc ... I've never used that one.
As far as the Dirac , if I make measurements and select a target and optimize to that target I get about 1db variations in REw using spacial averaging from about 250hz on up , so in my experience Dirac is pretty much on point.
Under 250hz there's about 3db (or less) variations. However your correct in that you can't adjust the time settings, it just moves everything to be in phase from the mic positions and if you adjust the mag response it moves phase with it. So no matter how you move amplitudes or shape the magnitude phase will follow (like a normal minimum phase system should). It doesn't take any thinking to use and works decently well. I have had better results with tape measure setting with delays than run Dirac rather than let it do the time adjustments for me. It can do time changes it just does a way better job if there close to begin with.
please post this data. when i did a multipoint measurement even the average was not anywhere close to how pretty dirac was showing as the after in their graph.
imo frequency correction will give you the best improvement.
room correction products can improve the phase a little bit but the consumer ones dont do a great job fixing the impulse. dirac did a poor job fixing impulse response, and also inverted the phase of the speaker.
pc is also easier to apply filters. you can also cascade filters so you can do crossover and at the same time drc.
the number of taps on the 2x4 hd is only enough to do room correction above about 500hz, where it is least needed.
but how much control do you have on how dirac corrects for in the time domain? none.
can you share a graph of this?
thanks
Attachments
^^^
Absolutely can.
Give me 24hrs I'll have to bust out all the equipment tomorrow and take some measurements🙂
Btw it's in car measurements don't know how much that makes a difference plus I have my own correction in rephase on 3-2x4HDs to give Dirac a helping hand but I can do AB w w/o Dirac 🙂
Absolutely can.
Give me 24hrs I'll have to bust out all the equipment tomorrow and take some measurements🙂
Btw it's in car measurements don't know how much that makes a difference plus I have my own correction in rephase on 3-2x4HDs to give Dirac a helping hand but I can do AB w w/o Dirac 🙂
Well okay.
If I get board I won't be able to resist tho 😉
So we'll see I guess.
But to your point, I did try Dirac without any pre tuning (high efficiency w horns) on speakers with dsps and Dirac left a lot to be desired, but my understanding is it's intended to correct a loudspeaker that's already pretty well tuned and just add to it. In that instance, it's pretty good, only had it two weeks now and it made a big improvement. I would bet, on a set of conventional drivers (bookshelfs computer speakers etc) it would do pretty well by itself.
sure, if its not too much to ask? 🙂
thanks for posting. mine were no where near that close , and the predicted impulse shown in DIRAC was grossly inaccurate compared to the actual one(shown in my post above).
Attachments
- Status
- Not open for further replies.