There's no processing going on in the soundcard, just AD and DA conversion, and software-based DSP wont change that. A few cards also have additional functions like digital volume control, digital mixer or even sample rate conversion, but no soundcards I know have any advanced DSP functions.
You nailed it - DRC is best applied below the Schröder frequency, usually ~200 Hz.
You nailed it - DRC is best applied below the Schröder frequency, usually ~200 Hz.
Why 128 SNR? In a real-life situation, you wont get more than 90 dB SNR anyhow, limited by the background noise in your room and the max SPL of your speakers.
Everything has tradeoffs. You will loose resolution with an analog attenuator as well, as the signal will in the end disappear into the noisefloor....
Everything has tradeoffs. You will loose resolution with an analog attenuator as well, as the signal will in the end disappear into the noisefloor....
A commercial DAC thats not accurate (most) , sinks into nonlinearity very fast adds various harmonics and stuff . Surely not analogue tape alike nonlinearity or noise.
With CD's the dither is not straight line white noise but shaped, it reaches even -70, -80 around 20khz, but below that, in the critical freq's it s actually below -120 dB!
When they add the tracks together in the mixer, those tracks are still 32bit, and right until the final dithering theres information there, especially if it was attenuated during mixing. So for at least 15 years now all plain, professionally mastered cd's can contain useful information up to 20th LSB. I'd better listen to those come out in a linear way.
With CD's the dither is not straight line white noise but shaped, it reaches even -70, -80 around 20khz, but below that, in the critical freq's it s actually below -120 dB!
When they add the tracks together in the mixer, those tracks are still 32bit, and right until the final dithering theres information there, especially if it was attenuated during mixing. So for at least 15 years now all plain, professionally mastered cd's can contain useful information up to 20th LSB. I'd better listen to those come out in a linear way.
Last edited:
Maybe there is information on our CD's at -120 dB, but how do you expect to hear it? Even if you had a totally quiet room you would have to play it back at 120 dB or more peak level. In a typical home situation, the background noise is at least 30 dB, masking everything below -80 dB anyhow.....
Yes, a good reason to ensure you have properly set analog levels when doing digital volume control. You should approach 0 dBFS at your preferred listening level. I have an 8-channel dac with adjustable analog levels. I have set those so the digital levels are as close to 0 dBFS as possible.
Filter for 120Hz crossover
Hi,
I am currently in the process of tuning the filter for my orion-lookalikes:
(note: I pull the speakers away from the wall when I listen to music)
2xSeas 27tdfc
1xSeas W15CH001
1xSeas L22RN4X/P
2xPeerless SWR269
I am searching for good vst filters(preferably free ones). I found this page:
Christian’s Blog EQs & Filters
Has anybody tried these filters? Is there a chance that this can give comparable results as the more expensive filters? I have played around with them in The VST Plugin Analyser from the same guy:
Christian’s Blog Measurement
Seems quite nice, doesn't it? I am going to give them a try in my rig this evening. It would be interesting to see how more expensive filters perform in the VST Plugin Analyser. You can even export the filter response to matlab for further analysis.
My current filter uses linear-phase filters for the tweeter-midrange and the midrange-midbass crossovers. The midbass-bass crossover is currently a simple 4th order minimum-phase filter. I would like to use linear-phase crossovers for the midbass-bass filter as well, but the plugins I use today doesn't allow steep enough filters in the bass-region. I have read some comments that you do not want to use such a filter in the bass-region, as this will cause noticeable pre-ringing. Is this true? Do you guys use linear-phase filters for the lower frequency XO-filters?
Some words about my set-up:
Crossover-pc: MSI 785GM-E65 + Underclocked AMD 235e (2gHz)
Soundcard: EMU-1820M
Amplifiers: Primare SPA21 + a couple of power-amps (I am looking for a multi-way class D amp)
VST Host: Console VST (Yes, I did pay for a license. Nice software)
regards,
Øyvin Eikeland
Hi,
I am currently in the process of tuning the filter for my orion-lookalikes:
An externally hosted image should be here but it was not working when we last tested it.
(note: I pull the speakers away from the wall when I listen to music)
2xSeas 27tdfc
1xSeas W15CH001
1xSeas L22RN4X/P
2xPeerless SWR269
I am searching for good vst filters(preferably free ones). I found this page:
Christian’s Blog EQs & Filters
Has anybody tried these filters? Is there a chance that this can give comparable results as the more expensive filters? I have played around with them in The VST Plugin Analyser from the same guy:
Christian’s Blog Measurement
Seems quite nice, doesn't it? I am going to give them a try in my rig this evening. It would be interesting to see how more expensive filters perform in the VST Plugin Analyser. You can even export the filter response to matlab for further analysis.
My current filter uses linear-phase filters for the tweeter-midrange and the midrange-midbass crossovers. The midbass-bass crossover is currently a simple 4th order minimum-phase filter. I would like to use linear-phase crossovers for the midbass-bass filter as well, but the plugins I use today doesn't allow steep enough filters in the bass-region. I have read some comments that you do not want to use such a filter in the bass-region, as this will cause noticeable pre-ringing. Is this true? Do you guys use linear-phase filters for the lower frequency XO-filters?
Some words about my set-up:
Crossover-pc: MSI 785GM-E65 + Underclocked AMD 235e (2gHz)
Soundcard: EMU-1820M
Amplifiers: Primare SPA21 + a couple of power-amps (I am looking for a multi-way class D amp)
VST Host: Console VST (Yes, I did pay for a license. Nice software)
regards,
Øyvin Eikeland
I use Linear-Phase filters at all XO points in my 4-way system. The problem with Linear-Phase filters seems to be that high accuracy and high Q at low frequencies gives you very very long latency, because the filters need large "blocks" of data, several wavelenghts at the desired frequency.
I have not yet found any free or shareware Linear-Phase plugins worth considering... so I paid 1000$ for the wonderful sounding PLParEQ. A much better way to spend 1000$ than to totally waste it on worthless cables..... that excuse worked for me!
Nice "clones" by the way.. what are the XO points?
I have not yet found any free or shareware Linear-Phase plugins worth considering... so I paid 1000$ for the wonderful sounding PLParEQ. A much better way to spend 1000$ than to totally waste it on worthless cables..... that excuse worked for me!
Nice "clones" by the way.. what are the XO points?
Thanks,
I have concentrated on getting the bass right first. The midrange and tweeter are just level-matched with a pretty random XO-point.
Bass-midbass: 120Hz (Linear-phase 4th order)
Midbass-Mid: 344Hz (Steep(Q=6) Linear-phase)
Mid-Tweeter: 2497Hz (Medium steep(Q=4) Linear phase)
I had plans to extend the midbass to about 500Hz, but I get a resonance that I do not quite understand at 550Hz. I also had a resonance at 373Hz that was cured by mounting the 8" by the magnet. The 550Hz resonance remains however. The sidewall is about 12cm deep next to the middle of the driver. This should give a resonance at 708Hz I believe. Could it be this one? I hope that the midrange is able to handle the lowish XO-point. What do you think?
It would be great if you or some of the other experienced guys on this forum would give the linear-phase crossovers from Christian W Budde a test and give us your verdict. Or maybe not... I will not be able to use them if I know that they are noe up to par
best regards,
Øyvin
Know your plugins
Hi,
I have been fooling around with Christian W. Buddes VST-analyser for a couple of hours:
Christian’s Blog Measurement
It is a really great tool to get some insight into what is going on inside those impeccable vst-plugins.
Here you have a little horror-show of what waves linear-phase EQ might look like:
My first attempt at a lowpass-filter. It didn't sound any better than it looks.
This is my next XO-point. I actually had this playing for a couple of days. Ugh!
Finally I have included a worst case for the 2.5k low pass. I am glad I never used this one.
It wouldn't be fair to post this without including some more favourable screenshots as well:
I was not able to get the 120Hz crossover to look nice. Here I have added a 258Hz filter on top to hide some of the uglyness.
The 344Hz filter is shaping up pretty good with Q=3 and Method set to low-ripple.
If you back off to Q=4, the 2.5k low pass is squeaky-clean.
At the moment I am using the dual Linkwitz-Riley filters from Christian W Budde for all my cross-overs:
Christian’s Blog EQs & Filters
Sure, I do not get to say that I have linear-phase filters, but it sounds better than what I had before with Waves. I might try waves again now that I have a way to see what is really going on inside.
Any suggestions for other/good linear-phase filters?
regards,
Øyvin
Hi,
I have been fooling around with Christian W. Buddes VST-analyser for a couple of hours:
Christian’s Blog Measurement
It is a really great tool to get some insight into what is going on inside those impeccable vst-plugins.
Here you have a little horror-show of what waves linear-phase EQ might look like:
An externally hosted image should be here but it was not working when we last tested it.
My first attempt at a lowpass-filter. It didn't sound any better than it looks.
An externally hosted image should be here but it was not working when we last tested it.
This is my next XO-point. I actually had this playing for a couple of days. Ugh!
An externally hosted image should be here but it was not working when we last tested it.
Finally I have included a worst case for the 2.5k low pass. I am glad I never used this one.
It wouldn't be fair to post this without including some more favourable screenshots as well:
An externally hosted image should be here but it was not working when we last tested it.
I was not able to get the 120Hz crossover to look nice. Here I have added a 258Hz filter on top to hide some of the uglyness.
An externally hosted image should be here but it was not working when we last tested it.
The 344Hz filter is shaping up pretty good with Q=3 and Method set to low-ripple.
An externally hosted image should be here but it was not working when we last tested it.
If you back off to Q=4, the 2.5k low pass is squeaky-clean.
At the moment I am using the dual Linkwitz-Riley filters from Christian W Budde for all my cross-overs:
Christian’s Blog EQs & Filters
Sure, I do not get to say that I have linear-phase filters, but it sounds better than what I had before with Waves. I might try waves again now that I have a way to see what is really going on inside.
Any suggestions for other/good linear-phase filters?
regards,
Øyvin
ReaFir filter
At the moment I am using the dual Linkwitz-Riley filters from Christian W Budde for all my cross-overs:
Christian’s Blog EQs & Filters
Sure, I do not get to say that I have linear-phase filters, but it sounds better than what I had before with Waves. I might try waves again now that I have a way to see what is really going on inside.
Any suggestions for other/good linear-phase filters?
regards,
Øyvin[/QUOTE]
Hi,
try Reafir standalone Filter.
this makes a very steep filter, its phase linear and you can adjust the frequency response, doing some equalizing.
At the moment I am using the dual Linkwitz-Riley filters from Christian W Budde for all my cross-overs:
Christian’s Blog EQs & Filters
Sure, I do not get to say that I have linear-phase filters, but it sounds better than what I had before with Waves. I might try waves again now that I have a way to see what is really going on inside.
Any suggestions for other/good linear-phase filters?
regards,
Øyvin[/QUOTE]
Hi,
try Reafir standalone Filter.
this makes a very steep filter, its phase linear and you can adjust the frequency response, doing some equalizing.
Attachments
I would like to see tested Algoritmix PrParlEQ and Ozone. I use the latter and sounds way better than waves.
Hi,
it seems like you have some experience with the VST plugin-analyser. Have you tried the export function? I have not tried this myself, but it should work for what you are trying to do I think. Unless you really WANT to learn Octave that is
regards,
Øyvin
My experience with generating simple brick wall linear phase crossovers in Matlab is similar to oyvine's. I got poor stopband attentuation with Matlab's fir filter generator when run in convolverVST. The tweeter high pass only took the signal down about 24 db at 20 Hz, not very desirable. The same filter when tested in Matlab showed 100-120 db of attentuation so it may be roundoff error in convolverVST? I dont' really know, I ended up purchasing Thuneau's Frequency Allocator a couple of years ago and have been happy with it ever since. At the very least, you should be careful when working with brick wall filters.
Dan
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.