REW questions

[Guest BadEnglish]

Honestly I donno. Mine are Usher MD2.

 

I didn’t like straight fire, coz the soundstage was wide but a tad fuzzy. I toed in a little to lock in the center image better, and with it came along more precise imaging.

 

 

Sent from my iPhone using Tapatalk

 

I think they are.

How far are they apart ?

Distance from back and side wall ?

 

Position of speakers has lot of impact on soundstage and tone.

 

I hope you are not trying to solve the problem never existed.

[wechnivag]

 

 

Just sharing my analysis... I was wondering about those peaks @ 8.2m, 9.0m (my room is only 4m * 4m), those peaks must be due to complex tangential and oblique modes.

 

0.55m = L speaker, front wall

0.8m = R speaker, side wall

1.5m = L speaker, ? ?

2.1m = L+R speaker, floor bounce

4.6m = L+R speaker, rear wall bounce

5.1m = ? ?

8.2m = ? ?

9.0m = ? ?

 

 

The closest 2 reflections have a big impact on the sound. The very short delay from main impulse can cause apparent image shift, and the large magnitude, in the case of the 0.8m is almost as large as the main impulse..

 

Try to isolate exactly where these are coming from. Like Bryan says, use your moab or a thick blanket, pillows etc to isolate the specific reflections. . Measure and confirm the fix, and most importantly listen to hear how it is different from before.

 

With small room, there is only so much reflection you can eliminate.. Use diffusion to scatter the sound in both space and time, so you get a spread of energy vs a single concentrated pulse of specular reflection.

 

Cheers

 

Sent from my X9009 using Tapatalk

 

 

[Boxerfan88]

I think they are.

How far are they apart ?

Distance from back and side wall ?

 

Position of speakers has lot of impact on soundstage and tone.

 

I hope you are not trying to solve the problem never existed.

 

L-R about 2m apart.

L-R about 0.9m from side walls.

About 0.55m (wall to speaker face) from front wall.

Toe-in about 10*.

Due to the speaker size and weight, there is very little room to move them; what's available is toe & small moves to front wall (+-10cm).

 

The main problem I am trying to solve is the 44Hz mode with my stereo setup.

This is now more or less controlled using the MathAudio EQ plugin tweaks in foobar2000.

 

With the above more or less done, now I am in learning mode with REWS to expand my knowledge.

With all the sharing on this thread, I'm just experimenting to see if further improvement can be made with some of the suggestions made here (eg. moving MOAB around targeting the problematic 1st reflection, etc...)

[Boxerfan88]

 

The closest 2 reflections have a big impact on the sound. The very short delay from main impulse can cause apparent image shift, and the large magnitude, in the case of the 0.8m is almost as large as the main impulse..

 

Try to isolate exactly where these are coming from. Like Bryan says, use your moab or a thick blanket, pillows etc to isolate the specific reflections. . Measure and confirm the fix, and most importantly listen to hear how it is different from before.

 

With small room, there is only so much reflection you can eliminate.. Use diffusion to scatter the sound in both space and time, so you get a spread of energy vs a single concentrated pulse of specular reflection.

 

Cheers

 

Sent from my X9009 using Tapatalk

 

Thank you for your suggestions!

[Guest BadEnglish]

L-R about 2m apart.

L-R about 0.9m from side walls.

About 0.55m (wall to speaker face) from front wall.

Toe-in about 10*.

Due to the speaker size and weight, there is very little room to move them; what's available is toe & small moves to front wall (+-10cm).

 

The main problem I am trying to solve is the 44Hz mode with my stereo setup.

This is now more or less controlled using the MathAudio EQ plugin tweaks in foobar2000.

 

With the above more or less done, now I am in learning mode with REWS to expand my knowledge.

With all the sharing on this thread, I'm just experimenting to see if further improvement can be made with some of the suggestions made here (eg. moving MOAB around targeting the problematic 1st reflection, etc...)

 

Oh my, you are in very tight situation.

44Hz is about the musical note F1, so just curious what kind of music do you listen on stereo ? House ?

With a small room and a pair of big floor standing speakers , just watch out a little bit on mid and high frequency : music is balanced in frequency.

I have encountered D2 before in my friend's place very difficult to tame in a small room.

 

[ronildoq]

L-R about 2m apart.

L-R about 0.9m from side walls.

About 0.55m (wall to speaker face) from front wall.

Toe-in about 10*.

Due to the speaker size and weight, there is very little room to move them; what's available is toe & small moves to front wall (+-10cm).

 

The main problem I am trying to solve is the 44Hz mode with my stereo setup.

This is now more or less controlled using the MathAudio EQ plugin tweaks in foobar2000.

 

With the above more or less done, now I am in learning mode with REWS to expand my knowledge.

With all the sharing on this thread, I'm just experimenting to see if further improvement can be made with some of the suggestions made here (eg. moving MOAB around targeting the problematic 1st reflection, etc...)

 

There are two things u need to look at, one is amplitude response aka the frequency response of the speakers. This information doesnt tell you what happen in the time domain, that is why you need to look at the energy time curve. The other is the Spectogram or waterfall graph, that will tell you everything, amplitude, freq and the effects in time domain, especially for low frequencies down to 10hz. The tools are there to help you,

 

1. Identify the problems in your room

2. Efficiently help you solve problems in your room once you have identified it

3. Lets you understand your room and gear better

4. Save you literally hours/days/years of time

 

Great job and effort there my friend, you have completed 1 above. By the end of this, you will understand not only your room and gear, but your appreciation for sound, music, quality and understanding of physics. No pain no gain!

 

On the 44hz dip you have there, there are 2-3 ways you can solve it.

 

1. Move your listening position

2. Use a Helmholtz resonator specifically to solve this 44hz

3. Sub placement, (if u using subs 2.1 stereo)

 

of course option 1 is the best, its free!

[Jag]

IMO, every room will have dips. Moving subs/listening position helps alot, but you might have new dips/peaks.

 

Don't be too clinical and expect to have zero peaks/dips. That's just an academic ideal.  If the dips are 2-3hz wide, ignore that and enjoy the music.

 

Play a sweep tone in REW. If sine tones at frequencies at those peaks/dips don't sound grossly louder/softer to your ears, you can ignore them.

 

 

[Guest BadEnglish]

IMO, every room will have dips. Moving subs/listening position helps alot, but you might have new dips/peaks.

 

Don't be too clinical and expect to have zero peaks/dips. That's just an academic ideal.  If the dips are 2-3hz wide, ignore that and enjoy the music.

 

Play a sweep tone in REW. If sine tones at frequencies at those peaks/dips don't sound grossly louder/softer to your ears, you can ignore them.

 

+1

[wechnivag]

Nice illustration that explains what happens with a bare wall reflection, vs absorption and diffusion.

 

Absorption based treatment focus the attention on direct sound from the speakers only, yielding a more precise, pinpoint imaging.

 

A direct reflection from a bare wall, if it is very close to the direct sound, and very high level, can cause comb filtering dips and peaks. Depending on frequency and severity, this will always show up on measurements by a microphone, but not necessarily audible due to how our ears are able to differentiate between the direct sound and the delayed reflection sound (within fusion zone) , which is integrated with the direct sound to provide timbre and tonality. It is this mechanism that is exploited by diffusers.

 

The lower peak of a diffused reflection, as well as the spread out time domain really complements the hearing mechanism and provides a lot of timbre, richness of tone and body to the direct sound.

 

 

 

 

 

Sent from my X9009 using Tapatalk

 

[Boxerfan88]

Thanks for sharing, very informative.

 

 

Sent from my iPhone using Tapatalk

[ronildoq]

Nice illustration that explains what happens with a bare wall reflection, vs absorption and diffusion.

 

Absorption based treatment focus the attention on direct sound from the speakers only, yielding a more precise, pinpoint imaging.

 

A direct reflection from a bare wall, if it is very close to the direct sound, and very high level, can cause comb filtering dips and peaks. Depending on frequency and severity, this will always show up on measurements by a microphone, but not necessarily audible due to how our ears are able to differentiate between the direct sound and the delayed reflection sound (within fusion zone) , which is integrated with the direct sound to provide timbre and tonality. It is this mechanism that is exploited by diffusers.

 

The lower peak of a diffused reflection, as well as the spread out time domain really complements the hearing mechanism and provides a lot of timbre, richness of tone and body to the direct sound.

 

 

 

 

 

Sent from my X9009 using Tapatalk

 

as long as within the 1st 40ms, beyond that it we hear it as echo..... Nice Sharing Gavin! "Depth" is exactly the word to describe it, i noticed....

 

Comb filtering is the worst, that is the last thing we want... Keep the listening area as simple as possible, the lesser the items and things in the surrounding bubble, the better it is....

[Boxerfan88]

Today I started to measure the HT multi-channel system via HDMI, and I hit a very very weird problem. Intermittently when doing "Check Levels" or Measurements, I hear random click/pop/stutter; and if I continue, ASIO4ALL will hang along with REW, and the only way out is to reboot PC. Switch back to JAVA mode, all is good - but can only measure L-R.

 

After some serious troubleshooting, I think I found the fix (recorded here for the benefit of all):

1. Open elevated CMD window

2. bcdedit /set disabledynamictick yes

3. Reboot PC

 

So far so good, managed to complete the measurement of all 5.1 speakers without ASIO4ALL hanging.

 

ps: it wasn't PC capability, it's quite a powerful box (i7-6800K + 32GB RAM)  -  Stupid Win10!!

[Jag]

Nice sharing, boxerfan.

[Boxerfan88]

Received my UMIK1 shipment...happie.  :)

 

Was re-measuring the HT system, and I got some questions on "system delay". It is supposed to show the time delay for the various speakers (referenced against the acoustic timing reference signal). I used the center channel as the acoustic timing reference for the measurement, and I notice the "system delay" jumps around quite a bit. I kind of expected the "system delay" for CC to be close to zero, and consistently close to zero; but it is not the case. Any ideas why it is so?

 

CC-Center Channel

 

FL-Front Left Channel

[Jag]

I would not suggest to use the CC as your timing reference Channel. Use one of the other surrounds as the timing reference.

[ronildoq]

Received my UMIK1 shipment...happie.  :)

 

Was re-measuring the HT system, and I got some questions on "system delay". It is supposed to show the time delay for the various speakers (referenced against the acoustic timing reference signal). I used the center channel as the acoustic timing reference for the measurement, and I notice the "system delay" jumps around quite a bit. I kind of expected the "system delay" for CC to be close to zero, and consistently close to zero; but it is not the case. Any ideas why it is so?

 

CC-Center Channel

 

FL-Front Left Channel

 

All speaker settings need to be set to 0 or same distance in your AVR before measuring.

 

Pick the speaker that is nearest in distance from where you put the MIC (MLP) as your reference speaker for ATR, so you wont have negative delays..

 

Make sure speakers are in Full Range when taking measurement.

 

Make sure the mic is exactly at the same spot without moving, even a slight movement affects the impulse response and time

[Boxerfan88]

I would not suggest to use the CC as your timing reference Channel. Use one of the other surrounds as the timing reference.

 

Will try measuring with surround as timing reference, and post back results.

Any reason why not to use the CC as timing reference channel?

[Boxerfan88]

All speaker settings need to be set to 0 or same distance in your AVR before measuring.

 

Should I not run YPAO first?

What's the reason to reset to blank before using REW?

Thanks for sharing.

 

Pick the speaker that is nearest in distance from where you put the MIC (MLP) as your reference speaker for ATR, so you wont have negative delays..

 

I see. My surrounds are nearer to MLP than CC. This kind of explains Jag's comment about using Surround instead of CC for the acoustic timing reference channel.

 

Make sure speakers are in Full Range when taking measurement.

 

It is full range for FL, CC, FR; but not SL,SR.

I see where you're going with this...pure speaker performance measured without bass management.

I will try.

 

Make sure the mic is exactly at the same spot without moving, even a slight movement affects the impulse response and time

 

Yep, mic is exactly same spot when doing all the measurements.

Mic is at ear level, and facing ceiling.

I am using the 90degree correction file.

 

Thank you for the pointers, much appreciated.

 

[ronildoq]

Should I not run YPAO first?

What's the reason to reset to blank before using REW?

Thanks for sharing.

 

I see. My surrounds are nearer to MLP than CC. This kind of explains Jag's comment about using Surround instead of CC for the acoustic timing reference channel.

 

It is full range for FL, CC, FR; but not SL,SR.

I see where you're going with this...pure speaker performance measured without bass management.

I will try.

 

Yep, mic is exactly same spot when doing all the measurements.

Mic is at ear level, and facing ceiling.

I am using the 90degree correction file.

 

Thank you for the pointers, much appreciated.

 

Depends on what you are trying to achieve. Are you measuring delays to check if YPAO sets the distance correctly post EQ?

 

If it is, then yes set all speakers to 0 distance settings (take a pic before that) , use the nearest speaker to MLP as the one for acoustic timing reference. Then measure all other speakers (except for subwoofer) and determine the delays/distance. See if YPAO sets them correctly . If the differences in delays Is marginal , ie <0.1ms, trust  the room correction software.

 

In order for the above to be effective, the mic must be placed at the same exact location , run a one point YPAO eq. Then compare with one without any eq 0 distance,see if YPAO sets it correctly.

 

Audyssey, accueq, Dirac, all does it quite accurately

 

The only difference between the above is how the software integrates mains with subs, that is where Dirac takes the lead....

 

A lot of this is to do with the subs being EQ’ed as 1.

 

When the subs are already phased matched with a good response, this then frees up resources for the EQ programs to work out the best delay combinations for mains with subs, this is how I approach EQ

[Boxerfan88]

Depends on what you are trying to achieve. Are you measuring delays to check if YPAO sets the distance correctly post EQ?

 

Yes, exactly my objective. Thank you for your sharing. So much to learn...

[Boxerfan88]

Playing around with REW today. I thought I'd compare UMIK microphone facing ceiling vs. facing front.

Similar (shape) yet different (peaks) -- interesting how 90 degrees can differ.

Question: for HT why measure facing ceiling when our ears face front?

 

[Jag]

Facing ceiling allows for all 7 ear level speakers to be measured with being affected by mic directionality.

 

If you use front facing, all speakers you xcept L&R speaker’s arrive at the mic’s transducer at different angles. I.e, the rear surrs will be affected by the mic’s base before arriving at the little transducer within the mic it self.

 

 

[Boxerfan88]

Facing ceiling allows for all 7 ear level speakers to be measured with being affected by mic directionality.

 

If you use front facing, all speakers you xcept L&R speaker’s arrive at the mic’s transducer at different angles. I.e, the rear surrs will be affected by the mic’s base before arriving at the little transducer within the mic it self.

 

Thanks for the explanation. Much appreciated.

[Boxerfan88]

Reading up on acoustics and playing around with REW, some interesting findings that I thought I'd share. It all started when I started to notice the intermittent/very slight honkiness of male vocals on my stereo setup. After much research and reading up, I kind of guess it may be due to SBIR. Popping over to an online SBIR calculator http://tripp.com.au/sbir.htm, filling up my room dimensions and speaker parameters, it says I'd have a peak caused by front wall of around 450Hz & 1050Hz.

 

 

Taking the REW measurement, smoothing it to 1/1, the 450Hz hump is pretty obvious on both speakers. I decided to use use 1/1 smoothing so as to try to keep to 1 "wide-Q" filter. Used REQ to compute the equalization for both speakers, it resulted in these:

 

 

 

After programming the PEQ in foobar2000, then tested with a few male vocal songs that I know had some honkiness (Josh Groban), the result was very good - his voice became "clearer" and slightly better clarity on instrument's mid bass; overall a nice improvement. Resultant measured response is slightly flatter at the lower mid range:

 

 

 

The "noise floor" was measured using RTA. With windows & door closed, aircond at lowest fan speed, I thought it is quite good at <40dB.

 

Conclusion: using EQ to "voice" the speaker+room combination can produce a very good outcome.

 

Comments?

[Jag]

Once you start EQ-ong the mains, you basically can voice the speaker in a way that was not intended but the manufacturer. That leads to the possibility of buying ANY reasonably built speaker and EQ it to sound different from its original voice.

 

IMO, no harm doing EQ above the Schroeder frequency as long as it’s an improvement to your ears. Besides, many auto-EQ do that anyway.