If I understand you correctly, you do not disconnect the internal passive crossover?
Actually, I find it much better to disconnect the internal crossover before using digital XO and this is not difficult to do. Just need to unscrew the drivers and disconnect the connection between the circuit board to the drivers.
But then, doing this way will connect the drivers directly to the amp and if there is DC out of the AMP, you can theoretically damage the driver. However, I have not had any problem so far!
Actually, I find it much better to disconnect the internal crossover before using digital XO and this is not difficult to do. Just need to unscrew the drivers and disconnect the connection between the circuit board to the drivers.
But then, doing this way will connect the drivers directly to the amp and if there is DC out of the AMP, you can theoretically damage the driver. However, I have not had any problem so far!
PCXO
...yes i will disconnect the X-over of the speakers and plug the speakers directly to each amp.
My question ist how to play a music wave or compressed flac or whatever through "Console" to get a x-over "on the fly" and output the music (frequency-devided) to each separate channel.
Ralf
...yes i will disconnect the X-over of the speakers and plug the speakers directly to each amp.
My question ist how to play a music wave or compressed flac or whatever through "Console" to get a x-over "on the fly" and output the music (frequency-devided) to each separate channel.
Ralf
Re: PCXO
How you play music back through Console/VST plugin is fairly specific to your sound card.
Some audio chains I use with my RME HDSP9652 sound card.
1) Sony jukebox into sound card. Console set to the inputs receiving the sound from the jukebox. This works great. I do have to set my sound card to "slave" synch. And I also have to use input1 using SPDIF instead of ADAT in. I am also running through a M-Audio CO3 box to remove copybit flag. I dont remember is this was necessary but it seems that it's now necessary.
2) Play music with Winamp, Windows Mediaplayer, etc. out one physical connector on sound card and into another. I physically loop the sound out of the same sound card and back in. I have to set audio synch to "master".
I wish there was a way with this particular card to not have to physically loop the sound. Some other RME cards do support looping the sound without using a physical cable.
You could debate whether Winamp using ASIO drivers is more bit-perfect than Windows Media Player. In recent posts I've read that in some cases dependent on sound card, using WDM output is bit-perfect. I don't know what to believe on this issue anymore.
I get the feeling that all music avoid's windows kmixer when I use the RME card's driver.
ralf said:
How you play music back through Console/VST plugin is fairly specific to your sound card.
Some audio chains I use with my RME HDSP9652 sound card.
1) Sony jukebox into sound card. Console set to the inputs receiving the sound from the jukebox. This works great. I do have to set my sound card to "slave" synch. And I also have to use input1 using SPDIF instead of ADAT in. I am also running through a M-Audio CO3 box to remove copybit flag. I dont remember is this was necessary but it seems that it's now necessary.
2) Play music with Winamp, Windows Mediaplayer, etc. out one physical connector on sound card and into another. I physically loop the sound out of the same sound card and back in. I have to set audio synch to "master".
I wish there was a way with this particular card to not have to physically loop the sound. Some other RME cards do support looping the sound without using a physical cable.
You could debate whether Winamp using ASIO drivers is more bit-perfect than Windows Media Player. In recent posts I've read that in some cases dependent on sound card, using WDM output is bit-perfect. I don't know what to believe on this issue anymore.
I get the feeling that all music avoid's windows kmixer when I use the RME card's driver.
ackcheng said:
I am going to use active monitors, and I guess they have amplifiers and crossover on the same board or somehow integrated, so it will not be easy to separate them. Instead, I want to separate drivers into different speakers. My understanding is that digital processing will fully compensate for the crossover anyway -- is that right?
Ralf, you need a soundcard able of internal routing of sound from wav output to asio input. I have a Emu-0404 wich can do this, other soundcards are for.ex. RME, Audiotrak, ESI.
2 channel 3-way with 24bit/192kHz
Hi diyer,
Can RME HDSP 9632 do the PC OX for Output 2 channel 3-way with 24bit/192kHz?
I am looking for 24bit/192kHz DA output 2 channel.
Is there any sound card able to the same job?
Thank you.
Hi diyer,
Can RME HDSP 9632 do the PC OX for Output 2 channel 3-way with 24bit/192kHz?
I am looking for 24bit/192kHz DA output 2 channel.
Is there any sound card able to the same job?
Thank you.
i got it
thanks,
i solved the problem by using stream_boy, i put the music stream out of winamp with strem_boy and loaded it in Console with a second instance of stream_boy. so i can use more channels an more X-Fi`s
greez,
Ralf
thanks,
i solved the problem by using stream_boy, i put the music stream out of winamp with strem_boy and loaded it in Console with a second instance of stream_boy. so i can use more channels an more X-Fi`s
greez,
Ralf
Hi,
Started going through this thread the other day(on page 24...) thought i'd ask a a question.
Is it possible to use a pc-xo and play video at the same time? Most of the software players have lip-synch delay so any latency I don't think would be an issue...But how would DD/DTS be handled? I'm looking for just a 2 channel 3 way setup with stereo subs, and i'm still on the fence as to whether or not I want to go the pc route or a standard active crossover/Pre-Pro.
I'm pretty new to all this, so I apologize in advance if the answer is obvious 😕 but if someone could clarify how I would do this or point me in the right direction that would be great!
Thanks in advance!
Tristan
Started going through this thread the other day(on page 24...) thought i'd ask a a question.
Is it possible to use a pc-xo and play video at the same time? Most of the software players have lip-synch delay so any latency I don't think would be an issue...But how would DD/DTS be handled? I'm looking for just a 2 channel 3 way setup with stereo subs, and i'm still on the fence as to whether or not I want to go the pc route or a standard active crossover/Pre-Pro.
I'm pretty new to all this, so I apologize in advance if the answer is obvious 😕 but if someone could clarify how I would do this or point me in the right direction that would be great!
Thanks in advance!
Tristan
Whats the maximum number of channels a PC xo solution could handle?
I guess that depends on how many channels your soundcard will handle but how about whats the maximum the software will handle?
I'm interested to see if its possible to build three way speakers that will be used in a full HT setup. That means 3*7 for a total of 21 channels of crossover frequencies.
Is this even possible?
What kind of hardware and software will support such a system requirement?
Cheers,
Exipnso
I guess that depends on how many channels your soundcard will handle but how about whats the maximum the software will handle?
I'm interested to see if its possible to build three way speakers that will be used in a full HT setup. That means 3*7 for a total of 21 channels of crossover frequencies.
Is this even possible?
What kind of hardware and software will support such a system requirement?
Cheers,
Exipnso
exipnos said:
Frequency Allocator Light can be run in multiple instances within the Console.
As long as you have a professional 24-channel sound card- like RME HDSP with external converters or MOTU 24I/O- you could easily setup a fully active 3-way 7.1 system.
Any modern computer like Athlon 64 or Intel Core will not have any problems running 4 instances of Allocator Light.
You could run into buffering issues with 4 instances of full Frequency Allocator, but I don't recommend using it with HT setup anyway because of the high latency, which makes the sound lag behind video.
Allocator Light will work with negligible latency (5msec or less).
Is anyone using the "ffdshow" audio decoder to enhance their playback or apply convolution or FIR filters? Seriously, check this out, it's free! I'd used the video decoder in the past, but wasn't aware of the audio decoder. The only downside is there is virtually no documentation.
http://sourceforge.net/project/showfiles.php?group_id=173941
http://sourceforge.net/project/showfiles.php?group_id=173941
Windows Vista Audio
Has anyone managed to use Windows Vista for Filtering and/or DRC?
DRC seems to be built in, I don't know about Filtering though.
Has anyone managed to use Windows Vista for Filtering and/or DRC?
DRC seems to be built in, I don't know about Filtering though.
Hi all,
I think this is my first post about DRC in this forum.... I wanted to share with everyone my progress so far in case some of what I've found is helpful to anyone....
I've been following the DRC thing for many months now, and initially started messing around with my old SB Live card, knowing full well that it resamples, and probably wasn't worth my time.
So, I finally bought an Audiotrak Prodigy HiFi 7.1 card about two weeks ago, and I've got it set up on a separate PC with all 8 channels connected directly to a Denon AVC-A11SR amp on the EXT IN inputs.
I also got hold of some WM-61A mic capsules (a bit difficult to get hold of in the UK), and built a quick preamp with an OPA2604 opamp....
(I'm using the normal.drc file included with the DRC package and sampling at 44KHz btw.)
The first impulse I recorded with a modified WM-61A and the original preamp sounded pretty boring and a bit harsh with the "flat" target curve on DRC. So, I changed the target to BK.txt, and it was much warmer and solid sounding... Maybe a tad too warm if anything.
The main problem I'm having is working out which is the best method of recording the speaker / room impulses, and which method most accurately represents my room??
So, I spent a bit of time earlier comparing impulses recorded using the Cool Edit / Aurora method, and using rec_imp. Once I figured out how to use rec_imp (again), the resulting DRC filters sound VERY similar.
I think it's safe to say that either method is worth using, as they both seem to produce extremely similar sounding filters. In fact, I must explain that I don't have a proper loop back signal when using rec_imp, and I'm looping back a digital signal instead using Directwire on the Audiotrak driver. So, the slight difference in sound from the Cool Edit method is probably due to the direct loop back.
It's difficult to explain, so I've attached a screenshot instead.
With the Cool Edit method, no loop back signal is used, but with rec_imp, I've bodged it a bit to make it work. In the screenshot (attached), you can see that I've connected the left channel of the microphone input (line in / mono in this case) to WDM input 1. I then use rec_impDS to record the sweep via the DirectSound (WDM) interface....
I also have to loop back each of the 8 ASIO outputs back to WDM input 2, so that rec_imp receives a direct loop back signal (the reason for looping all 8 outputs is because the sweep is obviously only output on one channel at a time. ie. it doesn't "stay" in one place.).
But, here's where it gets much more complex.... Because I eventually want to record the log sweeps of each speaker for 6.1 / 7.1 surround, the WDM (DirectSound / Windows) output is redirected into the ASIO inputs, I then use Console to be able to route the signals to any "real" output channel that I want.
So, when I record a sweep of say the "front left" speaker using rec_imp, I simply connect ASIO input 1 (sweep output from rec_imp) directly to ASIO output 1 (real front left output to amp). I also connect ASIO input 1 to ASIO output 6 (subwoofer), as you can apparently record the main speaker signals with the sub still connected, and DRC takes care of the rest.
But then (wait for it!), I realized that I should have some crossover filters for the main speakers so they don't receive the really low frequencies, which could possibly damage them. So, under Console, I'm using the "LinearPhaseGraphicEQ 2" VST plugin on each output channel to work as crossovers at 80hz.
Once I've recorded the impulses for each speaker, I run them through DRC (using normal.drc / bk target), then I use Voxengo Pristine Space VST as the 8-channel convolver. (Thanks to ShinOBIWAN for the tips, and whoever else who has mentioned similar progs!)
Ok, so the questions are....
Does anyone here think the LPGEQ2 filters are good enough for using all the time (even while recording the sweeps through them)? The reason I chose this plugin is because it has very low latency, as I want to stream TV audio through the PC. I also had trouble finding anything better so far. Would the slight latency of the filters / ASIO affect the log sweep recording?
Do you definitely leave the sub connected while recording the sweep through each main speaker? (my sub is an M&K VX-4, which unfortunately is sitting in the rear-right corner of the room).
Is there a way of calculating whether or not the log sweep you've recorded has a decent SNR, and if the mic preamp gain / input level gains are correct? ie. "quality" of log sweep recordings?
Does anyone know if the Denon A11SR still applies the bass management (crossovers) when using the EXT IN inputs?
If you untick the "autorange result" box in Aurora (for convolving the sweep with the clipboard), is this then similar to the way rec_imp works? ie. NOT normalizing the impulse(s)? I only ask because I want DRC to correct the relative channel levels as well as everything else.
Oh, and btw, the seriously bad (hollow / harsh) results I was getting with DRC before were due to not having a preamp connected near the mic, so I was getting a huge "peak" of noise at 50Hz on the recordings (mains hum). This happened with both the SB Live card, and the Audiotrak Prodigy, so it's not just a sound card problem. Please make sure you find a decent mic preamp! (I'm still using two bits of cheap audio cable joined together, so that doesn't help either).
OzOnE.
P.S. Apologies for the huge post - I may have to trim it down a bit in future. 😉
I think this is my first post about DRC in this forum.... I wanted to share with everyone my progress so far in case some of what I've found is helpful to anyone....
I've been following the DRC thing for many months now, and initially started messing around with my old SB Live card, knowing full well that it resamples, and probably wasn't worth my time.
So, I finally bought an Audiotrak Prodigy HiFi 7.1 card about two weeks ago, and I've got it set up on a separate PC with all 8 channels connected directly to a Denon AVC-A11SR amp on the EXT IN inputs.
I also got hold of some WM-61A mic capsules (a bit difficult to get hold of in the UK), and built a quick preamp with an OPA2604 opamp....
(I'm using the normal.drc file included with the DRC package and sampling at 44KHz btw.)
The first impulse I recorded with a modified WM-61A and the original preamp sounded pretty boring and a bit harsh with the "flat" target curve on DRC. So, I changed the target to BK.txt, and it was much warmer and solid sounding... Maybe a tad too warm if anything.
The main problem I'm having is working out which is the best method of recording the speaker / room impulses, and which method most accurately represents my room??
So, I spent a bit of time earlier comparing impulses recorded using the Cool Edit / Aurora method, and using rec_imp. Once I figured out how to use rec_imp (again), the resulting DRC filters sound VERY similar.
I think it's safe to say that either method is worth using, as they both seem to produce extremely similar sounding filters. In fact, I must explain that I don't have a proper loop back signal when using rec_imp, and I'm looping back a digital signal instead using Directwire on the Audiotrak driver. So, the slight difference in sound from the Cool Edit method is probably due to the direct loop back.
It's difficult to explain, so I've attached a screenshot instead.
With the Cool Edit method, no loop back signal is used, but with rec_imp, I've bodged it a bit to make it work. In the screenshot (attached), you can see that I've connected the left channel of the microphone input (line in / mono in this case) to WDM input 1. I then use rec_impDS to record the sweep via the DirectSound (WDM) interface....
I also have to loop back each of the 8 ASIO outputs back to WDM input 2, so that rec_imp receives a direct loop back signal (the reason for looping all 8 outputs is because the sweep is obviously only output on one channel at a time. ie. it doesn't "stay" in one place.).
But, here's where it gets much more complex.... Because I eventually want to record the log sweeps of each speaker for 6.1 / 7.1 surround, the WDM (DirectSound / Windows) output is redirected into the ASIO inputs, I then use Console to be able to route the signals to any "real" output channel that I want.
So, when I record a sweep of say the "front left" speaker using rec_imp, I simply connect ASIO input 1 (sweep output from rec_imp) directly to ASIO output 1 (real front left output to amp). I also connect ASIO input 1 to ASIO output 6 (subwoofer), as you can apparently record the main speaker signals with the sub still connected, and DRC takes care of the rest.
But then (wait for it!), I realized that I should have some crossover filters for the main speakers so they don't receive the really low frequencies, which could possibly damage them. So, under Console, I'm using the "LinearPhaseGraphicEQ 2" VST plugin on each output channel to work as crossovers at 80hz.
Once I've recorded the impulses for each speaker, I run them through DRC (using normal.drc / bk target), then I use Voxengo Pristine Space VST as the 8-channel convolver. (Thanks to ShinOBIWAN for the tips, and whoever else who has mentioned similar progs!)
Ok, so the questions are....
Does anyone here think the LPGEQ2 filters are good enough for using all the time (even while recording the sweeps through them)? The reason I chose this plugin is because it has very low latency, as I want to stream TV audio through the PC. I also had trouble finding anything better so far. Would the slight latency of the filters / ASIO affect the log sweep recording?
Do you definitely leave the sub connected while recording the sweep through each main speaker? (my sub is an M&K VX-4, which unfortunately is sitting in the rear-right corner of the room).
Is there a way of calculating whether or not the log sweep you've recorded has a decent SNR, and if the mic preamp gain / input level gains are correct? ie. "quality" of log sweep recordings?
Does anyone know if the Denon A11SR still applies the bass management (crossovers) when using the EXT IN inputs?
If you untick the "autorange result" box in Aurora (for convolving the sweep with the clipboard), is this then similar to the way rec_imp works? ie. NOT normalizing the impulse(s)? I only ask because I want DRC to correct the relative channel levels as well as everything else.
Oh, and btw, the seriously bad (hollow / harsh) results I was getting with DRC before were due to not having a preamp connected near the mic, so I was getting a huge "peak" of noise at 50Hz on the recordings (mains hum). This happened with both the SB Live card, and the Audiotrak Prodigy, so it's not just a sound card problem. Please make sure you find a decent mic preamp! (I'm still using two bits of cheap audio cable joined together, so that doesn't help either).
OzOnE.
P.S. Apologies for the huge post - I may have to trim it down a bit in future. 😉
Attachments
Ok, an update for your reading pleasure (or boredom!).....
I just tried measuring the impulses again but using a modified Panasonic WM-61A, and an unmodified WM-61A to see what the difference was. I also tried the modded mic with a 70% input gain and 100% input gain (on the Audiotrak mixer)....
The input gain appeared to make no real difference, so I'll keep it at around 90% from now on. I just wanted to be sure it wasn't clipping at 100%.
The biggest difference was when I used the modded mic capsule instead of the unmodded capsule. The modded mic was what I originally used when I got the nice warm sounding DRC filter (BK curve), it is perhaps slightly lacking in detail / treble, but pretty damn good sounding anyway.
When I tried the unmodded capsule again, it sounded noticably worse.... It's not amazingly bad, but is harsher and hollower sounding. This is a similar sound to what I got before when using a cheap unmodded / unbranded mic.
So, it seems best to always use modded Panasonic capsules (if you're using them), but what causes this difference in the apparent response?
And this brings me to the biggest problem with DRC as far as I can see.... Without owning big buck calibrated sound analysing equipment, having standardised preamp circuits and calibrated mics etc. - how do we know when we've got everything set up correctly?, and how can we tell when DRC is working properly (ie. what is it "supposed" to sound like?)
So far, the biggest benefits of DRC seem to be the very linear and smooth sounding bass response, the lack of harshness I used to have before DRC, the increased "3D" sound of the out-of-phase parts of music, and the dramatic improvement of the sound stage.
The vocal (ie. mono) parts of songs now appear directly between the two front speakers (it used to sound quite "separated"). You would swear that the center speaker is turned on, but it's not!
So my only real gripes are that I seem to get different results depending on how I connect things (virtually or otherwise), which mic I use, which preamp circuit I use, and the fact that I can never be sure if what I'm hearing is correct?
OzOnE.
I just tried measuring the impulses again but using a modified Panasonic WM-61A, and an unmodified WM-61A to see what the difference was. I also tried the modded mic with a 70% input gain and 100% input gain (on the Audiotrak mixer)....
The input gain appeared to make no real difference, so I'll keep it at around 90% from now on. I just wanted to be sure it wasn't clipping at 100%.
The biggest difference was when I used the modded mic capsule instead of the unmodded capsule. The modded mic was what I originally used when I got the nice warm sounding DRC filter (BK curve), it is perhaps slightly lacking in detail / treble, but pretty damn good sounding anyway.
When I tried the unmodded capsule again, it sounded noticably worse.... It's not amazingly bad, but is harsher and hollower sounding. This is a similar sound to what I got before when using a cheap unmodded / unbranded mic.
So, it seems best to always use modded Panasonic capsules (if you're using them), but what causes this difference in the apparent response?
And this brings me to the biggest problem with DRC as far as I can see.... Without owning big buck calibrated sound analysing equipment, having standardised preamp circuits and calibrated mics etc. - how do we know when we've got everything set up correctly?, and how can we tell when DRC is working properly (ie. what is it "supposed" to sound like?)
So far, the biggest benefits of DRC seem to be the very linear and smooth sounding bass response, the lack of harshness I used to have before DRC, the increased "3D" sound of the out-of-phase parts of music, and the dramatic improvement of the sound stage.
The vocal (ie. mono) parts of songs now appear directly between the two front speakers (it used to sound quite "separated"). You would swear that the center speaker is turned on, but it's not!
So my only real gripes are that I seem to get different results depending on how I connect things (virtually or otherwise), which mic I use, which preamp circuit I use, and the fact that I can never be sure if what I'm hearing is correct?
OzOnE.
There are a number of programs you can use to measure the frequency response post DRC. That is the easiest/best way to sanity check your DRC. I use qloud.
The hardest part of the whole process is getting a good impulse response measurement. It is particularly important not to have any clipping in the recording. The frequency response measurement post-DRC will help spot this problem. It is also pretty obvious on the recording itself if you display it using Ardour or the like.
The hardest part of the whole process is getting a good impulse response measurement. It is particularly important not to have any clipping in the recording. The frequency response measurement post-DRC will help spot this problem. It is also pretty obvious on the recording itself if you display it using Ardour or the like.
Hi, houstinian,
Thanks for the tip. I've been looking for something to view the frequency response of the impulses, but I can't seem to find something which is accurate?
I really need something which runs under Windoze, since I have the DRC stuff in place already.
I've tried things like Dirac, Cool Edit, Voxengo Span etc., but I'm not I can't seem to find something reliable which works with the 32-bit filters / speaker impulses? - Any suggestions?
I did some more messing with DRC yesterday, and found that I can't seem to make a DRC filter which sounds as good as the very first one I made when I got the Prodigy 7.1.... This measurement was done by simply connecting the front outputs of the card directly to the EXT IN inputs on the amp and measuring each speaker in turn. The right channel output of the sound card was left connected to the subwoofer pre-input on the amp.
The only difference now is that I now have all 8 outputs from the sound card connected to the 8 pre-ins on the amp, and to measure each speaker, I'm connecting each channels virtually using Console / ASIO ?
Would there be any obvious reason why doing the log sweep measurements while sending the WDM output through ASIO / Console would affect the results?
I don't think it has to do with the 'fake' loop back I'm using with the Prodigy drivers, as the results are very similar between using the Cool Edit method (no loop back) or rec_impDS (digital loop back)?
The original DRC filter I made sounds much warmer and nicer than anything since. The only thing that has really changed is directing the sound via Console during log sweep recording? The mic / preamp is the same, the mic position is the same, the volume and all levels are roughly the same (within a few dB's)?
I suppose I'd better try a more direct measurement tomorrow without going through Console. The original idea was to use rec_impASIO to measure each speaker automatically (using a batch file), since rec_impDS only 'sees' two channels. But how do I handle the loop back signal with rec_imp? Can it be left out altogether?
I was also getting audible clicks and glitches when using rec_impASIO (buffers might need increasing) - is it generally preferrable to use ASIO instead of DS for recording the log sweeps?
Oh, and has anyone else found major differences in the frequency response between unmodded and modded Panasonic capsules?
OzOnE.
P.S. Again, sorry for going on a bit, but DRC throws up so many questions!
Thanks for the tip. I've been looking for something to view the frequency response of the impulses, but I can't seem to find something which is accurate?
I really need something which runs under Windoze, since I have the DRC stuff in place already.
I've tried things like Dirac, Cool Edit, Voxengo Span etc., but I'm not I can't seem to find something reliable which works with the 32-bit filters / speaker impulses? - Any suggestions?
I did some more messing with DRC yesterday, and found that I can't seem to make a DRC filter which sounds as good as the very first one I made when I got the Prodigy 7.1.... This measurement was done by simply connecting the front outputs of the card directly to the EXT IN inputs on the amp and measuring each speaker in turn. The right channel output of the sound card was left connected to the subwoofer pre-input on the amp.
The only difference now is that I now have all 8 outputs from the sound card connected to the 8 pre-ins on the amp, and to measure each speaker, I'm connecting each channels virtually using Console / ASIO ?
Would there be any obvious reason why doing the log sweep measurements while sending the WDM output through ASIO / Console would affect the results?
I don't think it has to do with the 'fake' loop back I'm using with the Prodigy drivers, as the results are very similar between using the Cool Edit method (no loop back) or rec_impDS (digital loop back)?
The original DRC filter I made sounds much warmer and nicer than anything since. The only thing that has really changed is directing the sound via Console during log sweep recording? The mic / preamp is the same, the mic position is the same, the volume and all levels are roughly the same (within a few dB's)?
I suppose I'd better try a more direct measurement tomorrow without going through Console. The original idea was to use rec_impASIO to measure each speaker automatically (using a batch file), since rec_impDS only 'sees' two channels. But how do I handle the loop back signal with rec_imp? Can it be left out altogether?
I was also getting audible clicks and glitches when using rec_impASIO (buffers might need increasing) - is it generally preferrable to use ASIO instead of DS for recording the log sweeps?
Oh, and has anyone else found major differences in the frequency response between unmodded and modded Panasonic capsules?
OzOnE.
P.S. Again, sorry for going on a bit, but DRC throws up so many questions!
OzOnE_2k3 said:
Hi,
Cool Edit works fine with raw floating point audio, maybe you are not using the correct settings... ?
Anyhow, if you don't mind exporting your raw pcm to integer 32 bits wavs (with cool edit, for instance, or better yet scripting with sox), you may love Praxis as I do. http://www.libinst.com/
It is gratis soft, provided that you don't want to measure with it, but you have your measured impulses yet. It can import integer wavs of any bitlength, and the postproccessing capabilities are excelent.
Cheers,
Hi RR / Uli,
I've been using Cool Edit and Aurora fine for measuring the speaker impulses and saving to 32-bit 16.8 format for DRC, it's just that every DRC filter I make now sounds terrible...
Praxis looks handy for 32-bit stuff (I'll give it a go), and Acourate looks great, but is a bit out of my price range at this point tbh (trial has run out). I might consider buying Acourate in the future, but only when I can confirm that I'm measuring everything properly and getting consistant results with DRC.
It's a mystery to me as to why I can't seem to make another decent DRC filter? I'm using the exact same normal.drc file, 44k sampling (no resampling), BK target curve etc. Same preamp / mic / levels?
I tried measuring again using ACXO Player, but also going through ASIO / Console so I could apply the 80Hz crossover filters. I've also tried it without any filters at all (going directly out of WDM), but I got similar results to the rec_imp / Cool Edit method?
The FFT graphs on ACXO Player seem to confirm that there are no serious problems with the microphone or anything... ie. when I placed the mic about 10cm from the speaker, it gave a fairly "flat" response. The "Verification" test on ACXO Player gave extremely similar results to the test convolution graph, so something must be working?
Ok... I tried it again just now, but going directly through WDM using the Cool Edit / Aurora method, but this time playing via S/PDIF rather than analog cables. The resulting filter sounds a bit better, but is pretty much the same as using the analog inputs? (again, the sub was left working during the sweep and the Denon would have handled the crossovers.)
Maybe I should mention at this point that I'm using M&K LCR-55 speakers for the front three, Mordaunt Short MSB-20 for all four rears, and an M&K VX-4 sub. (I'm not too worried about doing DRC on the surround speakers as yet until I can get a consistantly good stereo filter.)
The LCR-55's are only rated at 100Hz - 20KHz according to the manual, so they require some form of crossover (digital or otherwise). I still use the THX standard 80Hz crossover though, as the LCR-55's seem to produce 80Hz easily.
All the latest filters sound fairly similar to the "best" filter for frequency response, but the phase doesn't quite sound right? With the best filter, the vocals in music are solidly placed between the two front speakers, but with any filter I've tried since, the imaging isn't as good?
I have noticed that sometimes the impulses appear to be inverted in Cool Edit after convolving with the clipboard (Aurora)? Is it common to have this happen, or could it be a problem with the sound card drivers or mic? Is there a simple test for confirming the mic / preamp's phase polarity?
Could it be that the position of the sub is affecting the sweep recording? (sub is in rear-right corner of the room?) How should the crossovers be handled when using the analog inputs on the amp? (ie. which VST filters are considered to be good for doing the crossovers while measuring the sweeps for DRC?)
Please help, as I'm close to giving up with DRC as just using the "Enhancer" plugin with Winamp! 😉
Your help / suggestions are always greatly appreciated.
OzOnE.
I've been using Cool Edit and Aurora fine for measuring the speaker impulses and saving to 32-bit 16.8 format for DRC, it's just that every DRC filter I make now sounds terrible...
Praxis looks handy for 32-bit stuff (I'll give it a go), and Acourate looks great, but is a bit out of my price range at this point tbh (trial has run out). I might consider buying Acourate in the future, but only when I can confirm that I'm measuring everything properly and getting consistant results with DRC.
It's a mystery to me as to why I can't seem to make another decent DRC filter? I'm using the exact same normal.drc file, 44k sampling (no resampling), BK target curve etc. Same preamp / mic / levels?
I tried measuring again using ACXO Player, but also going through ASIO / Console so I could apply the 80Hz crossover filters. I've also tried it without any filters at all (going directly out of WDM), but I got similar results to the rec_imp / Cool Edit method?
The FFT graphs on ACXO Player seem to confirm that there are no serious problems with the microphone or anything... ie. when I placed the mic about 10cm from the speaker, it gave a fairly "flat" response. The "Verification" test on ACXO Player gave extremely similar results to the test convolution graph, so something must be working?
Ok... I tried it again just now, but going directly through WDM using the Cool Edit / Aurora method, but this time playing via S/PDIF rather than analog cables. The resulting filter sounds a bit better, but is pretty much the same as using the analog inputs? (again, the sub was left working during the sweep and the Denon would have handled the crossovers.)
Maybe I should mention at this point that I'm using M&K LCR-55 speakers for the front three, Mordaunt Short MSB-20 for all four rears, and an M&K VX-4 sub. (I'm not too worried about doing DRC on the surround speakers as yet until I can get a consistantly good stereo filter.)
The LCR-55's are only rated at 100Hz - 20KHz according to the manual, so they require some form of crossover (digital or otherwise). I still use the THX standard 80Hz crossover though, as the LCR-55's seem to produce 80Hz easily.
All the latest filters sound fairly similar to the "best" filter for frequency response, but the phase doesn't quite sound right? With the best filter, the vocals in music are solidly placed between the two front speakers, but with any filter I've tried since, the imaging isn't as good?
I have noticed that sometimes the impulses appear to be inverted in Cool Edit after convolving with the clipboard (Aurora)? Is it common to have this happen, or could it be a problem with the sound card drivers or mic? Is there a simple test for confirming the mic / preamp's phase polarity?
Could it be that the position of the sub is affecting the sweep recording? (sub is in rear-right corner of the room?) How should the crossovers be handled when using the analog inputs on the amp? (ie. which VST filters are considered to be good for doing the crossovers while measuring the sweeps for DRC?)
Please help, as I'm close to giving up with DRC as just using the "Enhancer" plugin with Winamp! 😉
Your help / suggestions are always greatly appreciated.
OzOnE.
- Status
- Not open for further replies.