Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: MathAudio Room EQ for Foobar2000 (Read 74292 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Re: MathAudio Room EQ for Foobar2000

Reply #200
Also I can't use it with Cubase. If I use mathaudio as a system equalizer, it works if I play an audio file, but not with Cubase. if i load it as cubase vst plugin, it beeps even though i put the key code in the registry.
Hi Enri-audio, the registration button is located in the "About" dialog. If you have any other questions concerning the VST version of the plug-in, please contact us directly at https://mathaudio.com/contact.htm 

Re: MathAudio Room EQ for Foobar2000

Reply #201
News
New version released ( v2.7.7 ).
v2.7.7 uses newer versions of Foobar2000 SDK and Windows SDK.
Windows XP and Windows Vista are no longer supported.

Re: MathAudio Room EQ for Foobar2000

Reply #202
Hello @k2k,
thanks for the news!

However, I am not so expert on the matter of SDK and I would like to know what's the improvement achieved?
I mean, I am running Foobar on Windows 10, and does this v 2.7.7 run better? Less memory usage? Less CPU load?

If I am happy with the v 2.7.6, why I should upgrade it?
Thanks and regards, Andrea

Re: MathAudio Room EQ for Foobar2000

Reply #203
Hello @k2k,
thanks for the news!

However, I am not so expert on the matter of SDK and I would like to know what's the improvement achieved?
I mean, I am running Foobar on Windows 10, and does this v 2.7.7 run better? Less memory usage? Less CPU load?

If I am happy with the v 2.7.6, why I should upgrade it?
Thanks and regards, Andrea
Hello Andrea,
v2.7.7 is compiled with a newer compiler which knows and uses the advantages of the latest types of CPU. Most probably v2.7.7 puts less load on the CPU if you use a last-generation CPU chip, though we did not make such measurements. We updated the plug-in to make it more compatible with future releases of Windows and Foobar2000. We wish to be sure that the plug-in will not stop working after the next Windows update. Our previous ten-years-old compiler was optimized for Windows XP and nobody guarantees its compatibility with future updates of Windows 10. That is why we decided to move to a newer technology.
The new version of Room EQ doesn't bring any changes in the sound, so there are no urgent reasons to upgrade.

Re: MathAudio Room EQ for Foobar2000

Reply #204
Hi, any idea why the measurement sweep signal sounds crackling if Windows default format set on 48 kHz and no crackling on 44.1 kHz using Lexicon Alpha? (I want to show it with a screen recording, but in that no crackling.)

Re: MathAudio Room EQ for Foobar2000

Reply #205
Hi, any idea why the measurement sweep signal sounds crackling if Windows default format set on 48 kHz and no crackling on 44.1 kHz using Lexicon Alpha? (I want to show it with a screen recording, but in that no crackling.)

Nothing matters, either (now default) WASAPI shared or exclusive output used, as soon as the Room Measurement radio button selected the system sounds distorted (with 48 kHz default formats) or noisy even without playback (with 16 bit default formats). The only usable setting is 24 bit and 44.1 kHz.
Standard USB audio device drivers used by Windows 10 for Lexicon Alpha.

Re: MathAudio Room EQ for Foobar2000

Reply #206
Just out of curiosity (because it's a obsolete HW) I tested (v2.7.7) with my workplace PC's VT1705 VIA codec and even the Start Measurement button not working (foobar v1.6.1 default - WASAPI shared - output).

Re: MathAudio Room EQ for Foobar2000

Reply #207
Hi, any idea why the measurement sweep signal sounds crackling if Windows default format set on 48 kHz and no crackling on 44.1 kHz using Lexicon Alpha? (I want to show it with a screen recording, but in that no crackling.)
I don't know the reason of this problem, but the accuracy of measurement doesn't depend on the sampling rate. You can safely use 44.1 kHz for measurement.

Re: MathAudio Room EQ for Foobar2000

Reply #208
Just out of curiosity (because it's a obsolete HW) I tested (v2.7.7) with my workplace PC's VT1705 VIA codec and even the Start Measurement button not working (foobar v1.6.1 default - WASAPI shared - output).
Click the "In/Out" button and select the necessary audio output. Make sure that nothing takes exclusive control of the audio device's driver (including Foobar2000).

Re: MathAudio Room EQ for Foobar2000

Reply #209
I don't know the reason of this problem, but the accuracy of measurement doesn't depend on the sampling rate. You can safely use 44.1 kHz for measurement.

Yes, fortunately accuracy not affected, and at least one default format still works. The whole thing just occurred in the first place because with this device the Windows default is 24/48. It's only a suspicion that one reason might be the rework for the new foobar (beside driver or HW reasons), but without previous version I cannot test this.

Re: MathAudio Room EQ for Foobar2000

Reply #210
Yes, fortunately accuracy not affected, and at least one default format still works. The whole thing just occurred in the first place because with this device the Windows default is 24/48. It's only a suspicion that one reason might be the rework for the new foobar (beside driver or HW reasons), but without previous version I cannot test this.
Most probably this is a driver problem. If you set your audio card to the default sampling rate of 48 kHz, the audio player should be able to switch it to 44.1 kHz when necessary. If the driver doesn't allow switching the sampling rate, it may result in the "crackling". Try to find and install the latest driver for your audio interface. If the driver is not designed specifically for your version of Windows, it is better to set it to the default frequency of 44.1 kHz because 44.1 kHz is the sampling rate of majority of audio files in existence. You can check your audio driver like follows. Set the audio driver to 48 kHz, remove Room EQ from Foobar2000, switch Foobar2000 to a conventional audio driver (not WASAPI) and try to play a conventional audio file with the sampling frequency of 44.1 kHz. If you hear the "crackling", most probably this is a driver problem.

Re: MathAudio Room EQ for Foobar2000

Reply #211
I've just found out that Foobar2000 v1.6.1 uses a new default WASAPI output, so you cannot use it to test DirectSoud drivers. In any case I think that the problem is caused by the audio driver. A lot of audio cards have problems with automatic sample rate switching. This is not a large problem because you can manually set your card to 44.1 kHz for the measurement.

Re: MathAudio Room EQ for Foobar2000

Reply #212
I've just found out that Foobar2000 v1.6.1 uses a new default WASAPI output, so you cannot use it to test DirectSoud drivers. In any case I think that the problem is caused by the audio driver. A lot of audio cards have problems with automatic sample rate switching. This is not a large problem because you can manually set your card to 44.1 kHz for the measurement.
Yes that's what I wanted to emphasize. But I wrongly misinterpreted in your latest change log that foobar's change to WASAPI shared output is one reason for the modernization of the latest version. And that's why I maintained the possibility that Room EQ device access through foobar might be one reason. The driver is the latest, but as I found out only the USB device node using manufacturer components. There is no problem when I play 44.1/48 kHz sources with WASAPI exclusive which demands sample rate change from the device. Anyway, thanks a lot. If I want, I will do DirectSound tests with previous foobar version, and next week I hope I can start the measurements and corrections with your free plugin which I very much appreciate.

Re: MathAudio Room EQ for Foobar2000

Reply #213
foobar2000 1.6.2 fixed the problem most likely with this: "Added workarounds for wrong audio mix format reported by certain buggy soundcard drivers. (beta 3, beta 4)". Room EQ does an awesome job with cleaning my most compromised 100-500 Hz range.

Re: MathAudio Room EQ for Foobar2000

Reply #214
MathAudio Room EQ of Main Speakers with Sub

All, what is the best approach for using MathAudio Room EQ with main speakers and a sub? I am using Foobar2000 as the music player (outputting via ASIO or other direct means) and use two DACS: Meridian Explorer and, lately, Khardas Tone board. Also, measuring with a miniDSP UMIK-1. I run the main speakers full on, then via pre, feed the sub which I can pick up where the main speakers roll off. These are Nautilus 804s with overall frequency response is from 45Hz-20kHz +/- 2dB (the crossovers are 350Hz from bass to midrange and 4kHz from midrange to tweeter). The sub is a Def-Tech Super Cube II with frequency response 14-200 Hz and continuously variable low-pass 40-150 Hz crossover. Will set both sub and speaker volumes with pink noise to around 75db (or comfortably loud setting).

This is what I am thinking:

  • equalize for main speakers and room
  • equalize sub and room probably starting with low-pass setting as dictated by the results, but, figuring somewhere around 50 hz
  • equalize mains, subs, and room together
  • iterate results around crossover frequency

Does this make sense or is there a better approach to this method and sequence?

If I need to, I can rewire the main speakers to have them be filtered by the sub (introduces the need to to buy or cut my current nice speaker cables into two pieces each; would prefer not to do that).

Re: MathAudio Room EQ for Foobar2000

Reply #215
  • equalize for main speakers and room
  • equalize sub and room probably starting with low-pass setting as dictated by the results, but, figuring somewhere around 50 hz
  • equalize mains, subs, and room together
  • iterate results around crossover frequency

Does this make sense or is there a better approach to this method and sequence?

If I need to, I can rewire the main speakers to have them be filtered by the sub (introduces the need to to buy or cut my current nice speaker cables into two pieces each; would prefer not to do that).
The third approach looks preferable. By the way, the frequency response of cheap cables will be corrected by Room EQ, so you don't need to buy too expensive cables.

Re: MathAudio Room EQ for Foobar2000

Reply #216
K2k, I was afraid someone would say that.

Is the right approach to go speakers first, then sub, then both? Or just go speakers, then speakers and sub?

PS my UMIK-1 just came in today.

Re: MathAudio Room EQ for Foobar2000

Reply #217
Is the right approach to go speakers first, then sub, then both? Or just go speakers, then speakers and sub?
There is not much sense in measuring your sub separately. Measure your sub and your speakers simultaneously. Try to adjust the crossover of your sub to make the total frequency response of your audio system as flat as possible. After that move down the vertical slider of Room EQ as it is described at https://mathaudio.com/room-eq.htm

Re: MathAudio Room EQ for Foobar2000

Reply #218
So, I rewired the speaker cables, now using the sub's high-pass cross over. Using pink noise, I level matched the speakers to 75 db, then just the sub. Then brought them together and lowered overall level down to 75 db. I then did some measurements (added the calibration curve as indicated) and iterated the sub high pass filter from about 65 to around ~40 hz. I then brought the sub volume up slightly and fiddled with the crossover a bit. Got it as somewhat flat (okay hilly) as I could. Save that. Then used the neutral curve and brought down the overall level about 8-10 dbs. Saved settings. I did play around with flattening the curve more for a flatter (or neutral curve) response, and I could not really detect any muffled or muddy sound. So, not sure where I will end up with the overall attenuation setting using the neutral curve. Tomorrow I will do some fine tuning and multiple measurements around the sitting position. I am seriously going to have to electrically tape the sub volume and crossover knobs plus the slight change to the balance on the pre so they don't get mistakenly changed. So, a good start, May redo some of the steps and levels and measurements, too. A lot of variables to optimize. So far, I think I heard some improvement in sound, but, I've not done critical listening yet. 

Thank you to K2K for some great work, for making this available to the Foobar2k community, and for your helpful directions and responses (you care about your work and the attentive community). I've also been using the headphone version with my Philips X2s (hope the curve for the X2s is the same as the X2HRs...should be close, anyway). Again, great work.

As many times as I read the instructions, not until I tried it with headphoned and then room eq (measurements w/new calibrated microphone) was I able to see that  it was very intuitive, quite easy, and maddening fun. 

Re: MathAudio Room EQ for Foobar2000

Reply #219
Well, interesting day:

* Blew the tweeter on my right B&W Nautilus 804. Not sure when, but, started seeing drop off on right channel high-frequency  and knew something was wrong. Ear to the tweeter, no sound. Uuuuggh. This happened to me a long time ago, replaced the tweeter, but, now they are tough to find. B&W support indicates low-stock (they say contact them). I emailed them to see if I can buy one (or two). Lesson learned. Don't calibrate at too high a level (be careful of any pops, etc...although I had none). Well, for this exercise, I do have the calibration curves from the night before and earlier in the morning. So the first attached image file is the curves for the N804 calibration (pre-tweeter blow) with the DefTech SuperCube II. Mainly addressed the peaks around 30hz and higher one at 60-70hz.

* Okay, moved my N804s to the rear and pulled up my KEF LS50s. Very nice speakers and I can see fairly linear response; they  just don't provide the fuller sound of the N804s (even with the DefTech Sub). In terms of imaging also think the N804 s provide a better sound stage (maybe its the fuller sound). So, the second attached image file is curves for the Kef LS50 calibration. So, I brought down the overall measurement level as don't need to blow one of these UniQ drivers. I did numerous measurements around the sitting position: nine left mike and nine right mike (not sure it matters if it is one microphone, but the extra measurements don't hurt). These came out pretty good. I again see peak (room) resonances of (slight) 30hz and a larger one around 60hz I I did not see these as pronounced with the N804s, but, they were there. I had iterated the sub volume and filter to bring up the 30hz. So, now brought down the Neutral Room EQ curve to compensate for those two pronounced peaks. Outcome looks good, Sounds good.

Question: for both of these calibrations, are these approximately the proper levels for the equalization? Should they be broght a bit lower? I could go back and further change the sub crossover and volumes and remeasure. 

* Next, I attached a picture of my Foobar2000 setup. Been evolving it from scratch for 10s of years. Tabs for playlists, biography on left. Album cover below that. Main playlist window with song information including codec, bit encoding, frequency, bitrate, dynamic range average for album and specific for song, etc (the measurement meter is included in the setup). Below that, tabs with spectrum analyzer, equalizer, video (for youtube playlists & links), lyrics, selection specifics, & console. To the right, dynamic range plot of the song selection. Below that are tabs for vu meters, coverflow, and Shpeck visualizer.  Song play info also displayed in the status bar at bottom (including HDCD, anyone?).

* Foobar2000 andMathAudio are a lot of processing for an older lap laptop (updated to latest Win10 and patches). I can see that the added MathAudio algorithm, even in normal resolution mode, tends to slow and delay a few things during song start-stops, or selections within Foobar, plus moving cursor around the windows desktop. But, these are manageable.

* Lastly, I added a picture of the entire setup. This is a pretty wide open room (moved ping pond table off to far side), drop ceiling, hard walls, carpeted floor. Speakers about 4-5 feet off the back wall, everything centered to the sofa chair and parallel with the back walls.

Re: MathAudio Room EQ for Foobar2000

Reply #220
Question: for both of these calibrations, are these approximately the proper levels for the equalization? Should they be broght a bit lower? I could go back and further change the sub crossover and volumes and remeasure. 
I'd prefer a lower position of the vertical slider. Try different positions of the vertical slider and use the "Bypass" radio button and the "Bypass signal volume" slider to compare the corrected sound with the original sound as it is described in item 10 at https://mathaudio.com/room-eq.htm  Listen to the processed sound at least 30 minutes, then switch to the bypass mode: this method helps to clearly hear the contrast between processed and original sounds.

It looks like your first pair of speakers has problems with HF crossover (left speaker, 2000 Hz) and LF crossover (right speaker, 200 Hz).

Set your listening volume to measure your speakers.

Re: MathAudio Room EQ for Foobar2000

Reply #221
Thanks again. I read in audiophonids.fr the following: 'Positioning this slider too high can cause too little improvement in frequency response, and too low a position can lead to excessive excursions of your speaker membranes as well as nonlinear distortion.' So, a little confused here. You recommend bringing the MathAudio curve to a lower level (to the valleys across the spectrum). So, does this level create unnatural excursions, distortions, and phase issues? Or is a moderate attenuation a better more natural approach? I hear the difference in sound with my unprocessed setup coming across a bit more punchy while the neutral equalized a bit more, well, neutral. I would like to have the best imaging while not introducing distortion.

Re: MathAudio Room EQ for Foobar2000

Reply #222
Thanks again. I read in audiophonids.fr the following: 'Positioning this slider too high can cause too little improvement in frequency response, and too low a position can lead to excessive excursions of your speaker membranes as well as nonlinear distortion.' So, a little confused here. You recommend bringing the MathAudio curve to a lower level (to the valleys across the spectrum). So, does this level create unnatural excursions, distortions, and phase issues? Or is a moderate attenuation a better more natural approach? I hear the difference in sound with my unprocessed setup coming across a bit more punchy while the neutral equalized a bit more, well, neutral. I would like to have the best imaging while not introducing distortion.
The current setting of the vertical slider (-6 dB) allows your speakers (the first pair) work from 30 Hz. You can make them work from 20 Hz if you set the vertical slider to -30 dB, however, the large resultant excursion of the sub speaker diafragm can produce a substantial amount of upper harmonics. It is always better to not overload your speakers to avoid a muddy sound. I'd prefer the slider position of -12...-13 dB (in your case). If you usually use your speakers on a very small fraction of their maximum power, -15..-17 dB can sound even better. Ideally, you may want to try a few positions of the vertical slider and to find a position which correspond to a most natural and transparent sound.

 

Re: MathAudio Room EQ for Foobar2000

Reply #223
I can see that the added MathAudio algorithm, even in normal resolution mode, tends to slow and delay a few things during song start-stops, or selections within Foobar, plus moving cursor around the windows desktop.
When you will finish creating  EQ curve, you can create impulse for convolver and use convolver instead of equalizer on playback. It will be much easier for CPU.

Re: MathAudio Room EQ for Foobar2000

Reply #224
Well, now listening to the 2nd pair (LS50s) of speakers until/if I can get the tweeter on the N804s replaced.

Did not mean to indicate I would try to compensate to 20hz. That's  the limiting capability of the sub so I am fine with that.

On the 2nd pair (Ls50s), I was thinking of how much attentuation to apply across the rest of the resonse.

On the last suggestion of creating an impulse curve for a convolver. I will have to learn about that (new to me). I've seen links to convolver plugins, but, any link you can point me to will be appreciated. Thank you.

 
SimplePortal 1.0.0 RC1 © 2008-2021