The DSP Rolling & How-To Thread

[WoodyLuvr]

@Strangelove424 This is turning out to be a real kick-butt thread! Again, many thanks for taking the time putting this one together; very much appreciated and surely by all. Well done good sir!

TDR VOS SlickEQ
Your work around (aka "wonky solution") has solved my problem; thank you!

Case's Meier Crossfeed for Foobar
Would be nice to have some more insight added to this section regarding the how/why of setting the slider with some recommended settings.

 

[WoodyLuvr]

This is now my current DSP setup:

 

[ironmine]

while a bit extreme this comment confirmed what I thought, you can achieve similar results with a computer than with miniDSP. It save me a few hundred euros

For $800, one can buy a second-hand but decent DAC (or, probably, even a DAC/Head Amp combo) AND a laptop (as a source with DRC).

 

[gregorio]

while a bit extreme this comment confirmed what I thought, you can achieve similar results with a computer than with miniDSP.

No you can't, although it depends on what you mean by "similar". Room correction is a bit of a misleading term because what you end up with is not a "correct" room, it's a somewhat improved response but not a correct response. We can improve a room's response with EQ but EQ only addresses the total energy at a range of frequencies, not the other main factor of time. Resonance (room or any other type of resonance), doesn't just affect the level of the signal but also the duration, a resonant frequency isn't just louder, it's lasts longer and is smeared in time. It's typical therefore when using just EQ for correction to over compensate, end up with peak levels somewhat lower than the target (although at the target over time). The Dirac processing inside the miniDSP attempts to correct the smearing in conjunction with EQ correction, ending up with much an overall better impulse response, IE. Tighter bass, less phase induced error, etc. It's still not a correct room but the results are typically pretty good and far better than you can achieve with a computer.and an EQ plugin or two. Whether it's worth $800 to you personally is a different question though.

G

 

[sonitus mirus]

I'm getting close, but I just can't seem to grasp how to use the VST host to play music through MathAudio Room EQ.

Because I use a DAC connected to either USB or optical, I have to use some virtual cable solution to channel the sound through any software.

I use the VB-Audio Hi-Fi cable input as the default playback device with a sample rate of 44100 and the recording device as VB-Audio Hi-Fi cable output.

With this setup I can stream my music with a browser or via a desktop player.

I was able start a 64-bit VST host, use the Umik-1 microphone as the input, chain MathAudio EQ as a VST plugin, copy the mic's calibration.txt into the MathAudio EQ folder, and I was able to take room measurements with the frequency sweep and saved a .snr file.

Where I get lost is how to play music back in step 11 of this guide.

http://mathaudio.com/room-eq.htm

In the VST host, I can select a chain that seems to detect music being played, I can hear the music over the speakers, but the MathAudio Room EQ plugin does not seem to do anything to the sound when I adjust the sliders or switch back and forth from Room EQ and Bypass.

I've tried various input and output setting in the VST host. I've had echos, silence, and what appears to be a little static that can be heard if I adjust the volume on my DAC's preamp, but nothing I change on the MathAudio plugin seems to do anything.

 

[Strangelove424]

@Strangelove424 This is turning out to be a real kick-butt thread! Again, many thanks for taking the time putting this one together; very much appreciated and surely by all. Well done good sir!

TDR VOS SlickEQ
Your work around (aka "wonky solution") has solved my problem; thank you!

Case's Meier Crossfeed for Foobar
Would be nice to have some more insight added to this section regarding the how/why of setting the slider with some recommended settings.

Thank you, Woody. I’m excited to see where the thread goes! Everyone’s contributions and discussion has already uncovered a lot of progress. This DSP stuff is a lot of fun, very practical, and budget friendly. It's a great way to work out our hankering for experimentation, improvement, and dabbling with technology.

Glad I could help with the presets problem!

There’s some theory behind the Meier crossover linked in the description, but I’m not sure if that’s what you’re looking for. In regards to guidelines for use, it’s hard to say because everyone has their own preference and style of music. At 100, you are hearing full mono essentially, but in my experience the crossfeed squashes the sense of space as low as 15. Others might prefer something higher. There’s no way to recommend a universal setting.

I feel like I missed a crossfeed plugin somewhere. Did somebody mention they were using one other than Meier or BS2B?

 

[Strangelove424]

No you can't, although it depends on what you mean by "similar". Room correction is a bit of a misleading term because what you end up with is not a "correct" room, it's a somewhat improved response but not a correct response. We can improve a room's response with EQ but EQ only addresses the total energy at a range of frequencies, not the other main factor of time. Resonance (room or any other type of resonance), doesn't just affect the level of the signal but also the duration, a resonant frequency isn't just louder, it's lasts longer and is smeared in time. It's typical therefore when using just EQ for correction to over compensate, end up with peak levels somewhat lower than the target (although at the target over time). The Dirac processing inside the miniDSP attempts to correct the smearing in conjunction with EQ correction, ending up with much an overall better impulse response, IE. Tighter bass, less phase induced error, etc. It's still not a correct room but the results are typically pretty good and far better than you can achieve with a computer.and an EQ plugin or two. Whether it's worth $800 to you personally is a different question though.

G

I think with the use of a convolver plugin, impulse response correction can be imported from one of the free room correction programs. However, putting aside the features or comparative cost, there are quite a few people in the hobby who aren’t extremely technical, and got into audio (and most of their hobbies) long before the advent of computers. Some of the older folk I’ve met simply wouldn’t be up to the task, and that’s not to say anything of their intelligence, because these guys used to build DIY amps. They’re just from a different time. Room correction DSP, if it were a custom software process only, would effectively leave some very devoted people out. For that reason, and despite the steep prices, I think standalone DSPs like this certainly have a worthy market if they perform competitively. Nor do they have to be too steep. The Behringer DEQ2496 is $250, though I’m not sure it does impulse correction.

On a side question, does anyone know if REW and Room Math EQ do multi channel?

 

[jgazal]

(...) room correction programs. However, putting aside the features or comparative cost, (...)

While you are in this topic of room enhancement programs, particularly in the bass band, would you please add that the Realiser, more than just be priced to match head tracking feature, also allows virtual room enhancement:

The group of five faders on the bottom right are a new feature to the Realiser. These sliders allow the listener to alter the reverberation characteristics of the virtual room in the frequency domain, much like a 5 band graphic equalizer except it is decay time that is altered rather than gain. This feature has the ability to dramatically improve the quality of any mediocre sound room particularly in the low frequency regions.
http://smyth-research.com/downloads/additional_KS_info.pdf

And that it allows to cut/fade the reverberation decay with window/taper values:

WINDOW -- The PRIR measurement captures the room acoustic over time, including reverberation. The maximum measurement length is 850 ms (the default), but this can be reduced. As the length becomes shorter, the sound becomes drier (less reverberant), and at 5 ms thesignalislargelyanechoic.
One reason to reduce the reverberation length is that, as the reverberation dies down, any remaining measured sound is noise in the room. This noise is avoided if the length is reduced to the time from the impulse to the time when the reverberation tail drops below the ambient noise. Another reason would be for analysis of recorded material so that the reverberation of the measured room does not mask small details or errors. Very short reverb times such as 5 ms may sound unnatural.
To adjust the reverberation length, press MENU, the button for the speaker in question (L in this example) and move the cursor to L vSPKR SETUP2. The screen will say:
> WINDOW: 850 ms
REL VOL: 0 dB
DELAY: 0 ms
PROXIMITY: OFF
The cursor should be on WINDOW, but if not, use the ^ and v keys to move it there. Use the < and>keystochoosethereverberationlength;thechoicesare 850,400,200,100,50,25, 10 and 5 ms. Press OK to set, or CANCEL.
The screen indicates that the WINDOW value has been adjusted by displaying a W:
1 john doe
studio a
A 20:01 10-OCT-08
.W. . . .
The WINDOW value applies only to the current preset. If you wish to retain this setting, save or re-save the preset.
The WINDOW control is somewhat similar to the taper settings available when a PRIR is loaded into a preset. (See Other Controls for a discussion of taper.) The differences are: (1) the taper is just that, a linear taper providing a fade to zero, whereas the WINDOW control truncates the reverberation tail for an abrupt cutoff; and (2) the taper must be set upon loading the PRIR into the preset, whereas the WINDOW control can be varied easily after the fact. One might use WINDOW to get a quick idea of the reverberation time present in a PRIR, and to judge whether to reduce it; then use the taper for real listening since a taper will sound much more natural than a truncation.
http://www.smyth-research.com/downloads/A8manual.pdf

It would be great if you could also insert a link to the “Smyth Research Realiser A16” thread here in head-fi and, if you don’t mind (maybe off-topic?), to the thread “A layman multimedia guide to Immersive Sound for the technically minded (Immersive Audio and Holophony)”.

I know it is not a cheap program, but you may want to add the Bacch4Mac Pro software that not only cancels crosstalk with speakers by measuring a PRIR and using the Mac camera for head tracking, but also virtualize speakers with headphones. Since it measures a PRIR it also deals with “detrimental effects of spatial comb filtering and non-idealities of the listening room, the loudspeakers and the playback chain”.

 

[Strangelove424]

I included your Smyth Realiser and Immersive Audio links in the op. Feel free to link to this thread from yours.

Not sure what to make of "virtual room enhancement". Doesn't Smyth Realizer simulate any environment one wishes, why would you bring all your real life acoustic problems into a virtual environment? I am thoroughly confused.

How expensive is Bacch4Mac? If the price is insanely high, and nobody has subjective experience with it, perhaps I'll include a note stating such. I'd like to keep this a reliable source of (budget friendly) information from people who have direct experience.

 

[jgazal]

I included your Smyth Realiser and Immersive Audio links in the op. Feel free to link to this thread from yours.

Not sure what to make of "virtual room enhancement". Doesn't Smyth Realizer simulate any environment one wishes, why would you bring all your real life acoustic problems into a virtual environment? I am thoroughly confused.

How expensive is Bacch4Mac? If the price is insanely high, and nobody has subjective experience with it, perhaps I'll include a note stating such. I'd like to keep this a reliable source of (budget friendly) information from people who have direct experience.

Thank you.

I used the expression virtual room enhancement, because @gregorio said room correction is not precise:

(...) Room correction is a bit of a misleading term because what you end up with is not a "correct" room, it's a somewhat improved response but not a correct response. We can improve a room's response with EQ but EQ only addresses the total energy at a range of frequencies, not the other main factor of time. Resonance (room or any other type of resonance), doesn't just affect the level of the signal but also the duration, a resonant frequency isn't just louder, it's lasts longer and is smeared in time. It's typical therefore when using just EQ for correction to over compensate, end up with peak levels somewhat lower than the target (although at the target over time). (...)
G

The Smyth Realiser cannot synthesize reverberation according to user set room geometry. The Realiser emulates real rooms with their pros and cons. The equalization in the time domain helps to tame bass overhigh that a PRIR from a real room would emulate otherwise.

On the other hand, the Bacch4Mac pro allow binaural synthesis. Professor Choueiri claims that it takes into consideration room geometries and typical room materials set by the user.

Unfortunately, the Bacch4Mac is in the $7,000.00 range in its flagship package...

 

[jgazal]

Feel free to link to this thread from yours.

F) The DSP Rolling & How-To Thread (excellent thread about DSP started and edited by @Strangelove424)

 

[Strangelove424]

Thanks, Jgazal. I'm still not sure I understand the idea of virtual room enhancment/correction when the room itself is a product of Smyth algorithms, and has no inherent error. But I think the ability to simulate environments is very important so I added a sentence that says: The Realiser can mimic any listening environment the user wishes by altering tone, timing, and reverb. I think that's accurate and also covers what you want to express, yes?

$7,000 yikes! That's definitely expensive. I don't want to leave out options for people, so will include the BACCH, but I will make a note of the cost and relative lack of user experience and subjective impressions due to price point in the description.

 

[jgazal]

Thanks, Jgazal. I'm still not sure I understand the idea of virtual room enhancment/correction when the room itself is a product of Smyth algorithms, and has no inherent error. But I think the ability to simulate environments is very important so I added a sentence that says: The Realiser can mimic any listening environment the user wishes by altering tone, timing, and reverb. I think that's accurate and also covers what you want to express, yes?

Yes.

The PRIR might have inherent error depending on the acoustics of the room measured.

$7,000 yikes! That's definitely expensive. I don't want to leave out options for people, so will include the BACCH, but I will make a note of the cost and relative lack of user experience and subjective impressions due to price point in the description.

BACCH4Mac Audiophile Edition 3D Audio Playback System Read - audiostream.com
BACCH4Mac 3D Audio Playback System Follow-Up - audiostream.com

 

[Strangelove424]

Yes.

The PRIR might have inherent error depending on the acoustics of the room measured.


BACCH4Mac Audiophile Edition 3D Audio Playback System Read - audiostream.com
BACCH4Mac 3D Audio Playback System Follow-Up - audiostream.com

Do all the PRIRs they use contain the natural acoustic errors (or characteristics lets call them) of the environment they measured? And if I understand you correctly, you are saying that the Realiser can correct for the problems of each PRIR in its virtual representation of it, or serve it up raw without correction? That's quite fascinating.

BTW, How common is it to find PRIRs for the Realiser? Could I go download a PRIR right now for famous venues and studios?

Yes.

The PRIR might have inherent error depending on the acoustics of the room measured.


BACCH4Mac Audiophile Edition 3D Audio Playback System Read - audiostream.com
BACCH4Mac 3D Audio Playback System Follow-Up - audiostream.com

I see an entry price at $5,000.

 

[jgazal]

Do all the PRIRs they use contain the natural acoustic errors (or characteristics lets call them) of the environment they measured? And if I understand you correctly, you are saying that the Realiser can correct for the problems of each PRIR in its virtual representation of it, or serve it up raw without correction? That's quite fascinating.

That is also what I understood. Quite fascinating indeed.

BTW, How common is it to find PRIRs for the Realiser? Like, could I go download a PRIR right now for famous venues and studios?

The personal room impulse response contains HRTF and RIR components.

While you can use binaural room impulse response made by others in stellar rooms, you will have the HRTF component mismatched.

There will be an Exchange site where people can load their PRIRs to donate or sell them.

Such Exchange site will also allow personalization of BRIR’s, but:

(...) Stephen Smyth says the first tier in performance is a PRIR in room with good acoustics and a second tier would be BRIR in a room with good acoustics, personalized in the exchange site, using a PRIR made in any room you have available.

I see an entry price at $5,000.

Post #68 has a link to the Bacch4Mac webpage where you can see the price and features of each package and the optional upgrades for all packages.