• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Share your in-room measurements?

cab1024

Member
Joined
Jun 2, 2023
Messages
7
Likes
2
With Audyssey, I suggest using Dynamic EQ. As another member said, the response looks "bright," as in low bass relative to mids and highs. Dynamic EQ will elevate the bass region (and do a few other things). That should address the brightness / weak-ish bass sound.

I assume this is Audyssey "Reference" EQ, not "Flat."

The "offset" setting (15, 10, 5, 0) for each input like for music, TV, movies, should be based on how loud you normally listen. The amount of "dynamic EQ boost" is the same for each, but each assumes a different target volume level and add more boost the farther you go from that volume level (eg., every 5dB or notches down, it adds X dB more boost on a curve). Movies are usually mixed with 0 on the dial as the target, so if you listen to movies and TV at around -20, you get much more boost with a "0" offset setting (it senses 20dB low) versus if you set it to "15" (it senses you are only missing the target volume setting by 5dB). For music inputs like streaming or a CD player, your normal volume setting might be -30 to -35, so an offset setting of "0" will add a ton of boost, and a setting of "15" might sound much more natural.

On my Marantz receiver, I use an offset of 15 for my music inputs, and 0 / 5 for TV/movie inputs, adjusting to taste with different material as needed.

If you do REW readings with Dynamic EQ turned on, you will get different graphs depending on your input level and what the volume dial is set to. For example if you have offset set to 15 and the volume dial is at -15 in order to get enough volume to register on the mic / REW, you will not see any boosts in the response. If you set DEQ offset to 0 and have the volume dial at -15, you will see the boost you get at 15 below the offset number (-30 dial on 15 offset, -25 dial on 10 offset, -20 dial on 5 offset, -15 dial on 0 offset). Conversely, if you set the dial to 0 to get enough volume to do REW readings, and the DEQ offset is at 15, you will see a reduction in the bass because it thinks you are 15dB past max volume level and it is protecting your speakers.

You can see the boost graph here, and that there is some slight boosting to frequencies over 8K as well.
Thanks for your reply. Yes, I do use Dynamic EQ. I suspect that the benefits of that dynamic adjustment do not show up in the seconds-long test tone. I would say my sound is anything but lacking in bass response. That said I've taken your response and @PeterNL's as well and adjusted the Curve Editor to make the downwards sloping bass-heavy curve. That has further increased the bass response, though I swear the results do not show up in the graphs below.

This morning's results look decent but very similar despite all the different speaker and EQ settiigs:
Day 2-morning all.jpg


Now, this evening so far, I have added the bass heavy slope to the Curve Editor -- and the bass response is 40 dB lower than the rest of the frequency range, across the board in all EQs and speaker settings (with and without the subwoofer:
Day 2-evening 1 all.jpg


So I'm completely confused by the above graph now. I really appreciated your explanation of the Reference Offset, though it still left me somewhat confused, given what I hear. It sounds like the setting that will produce the most low end would be the 0 Offset, which is what I kind of understood from the manual, but your explanation is more thorough (even if I can't definitely say I understand the effect). At any rate, changing to 10 and 15 seemed to produce less bass when listening to Taylor Swift - Lavender Haze at 50 on the dial, which has become my primary bass testing song of late. At any rate, getting frustrated with the graph above and the bizarrely quiet bass response it shows, I've been listening to actual music with my "Audyssey Slope2 without subwoofer at volume50" settings and there is insane good bass, and balanced with crisp, defined vocals and instruments front and center in the mix, while enveloped in deep wide bass. Even though bass didn't seem overly weak before, it's definitely not now. I'll keep listening to my stereo-testing playlist, then put the new copy of Dave Brubeck that arrived on vinyl today and see if analog jazz comes across just as lovely.
*p1 speaker setting has no Subwoofer; p2 has a subwoofer; v60 means volume set at 60, which is pretty loud for streamed music; Slope2 is the second bass heavy curve int eh Curve Editor
 

ExPerfectionist

Active Member
Joined
Apr 28, 2023
Messages
127
Likes
138
Location
San Francisco Bay Area
Thanks for your reply. Yes, I do use Dynamic EQ. I suspect that the benefits of that dynamic adjustment do not show up in the seconds-long test tone. I would say my sound is anything but lacking in bass response. That said I've taken your response and @PeterNL's as well and adjusted the Curve Editor to make the downwards sloping bass-heavy curve. That has further increased the bass response, though I swear the results do not show up in the graphs below.

This morning's results look decent but very similar despite all the different speaker and EQ settiigs:
View attachment 321014

Now, this evening so far, I have added the bass heavy slope to the Curve Editor -- and the bass response is 40 dB lower than the rest of the frequency range, across the board in all EQs and speaker settings (with and without the subwoofer:
View attachment 321015

So I'm completely confused by the above graph now. I really appreciated your explanation of the Reference Offset, though it still left me somewhat confused, given what I hear. It sounds like the setting that will produce the most low end would be the 0 Offset, which is what I kind of understood from the manual, but your explanation is more thorough (even if I can't definitely say I understand the effect). At any rate, changing to 10 and 15 seemed to produce less bass when listening to Taylor Swift - Lavender Haze at 50 on the dial, which has become my primary bass testing song of late. At any rate, getting frustrated with the graph above and the bizarrely quiet bass response it shows, I've been listening to actual music with my "Audyssey Slope2 without subwoofer at volume50" settings and there is insane good bass, and balanced with crisp, defined vocals and instruments front and center in the mix, while enveloped in deep wide bass. Even though bass didn't seem overly weak before, it's definitely not now. I'll keep listening to my stereo-testing playlist, then put the new copy of Dave Brubeck that arrived on vinyl today and see if analog jazz comes across just as lovely.
*p1 speaker setting has no Subwoofer; p2 has a subwoofer; v60 means volume set at 60, which is pretty loud for streamed music; Slope2 is the second bass heavy curve int eh Curve Editor

My suggestion is to start with measuring the response with DEQ turned OFF. The DEQ settings will alter the signal and the readings (combination of offset setting and dial volume level). With the no-boost readings, then you can know in your head that DEQ will boost bass below 200Hz or so (and a little above 8K, plus levels of surround speakers) depending on the offset setting. For music you may prefer the 15 offset, and something in the 0~10 range for movies/TV.

I'm not sure about your graphs, and there's a lot of lines to decipher. If you're listening with a sub, just look at the readings with the sub integrated (although with/without can be useful to compare for setting crossover frequency).
 

ExPerfectionist

Active Member
Joined
Apr 28, 2023
Messages
127
Likes
138
Location
San Francisco Bay Area
Thanks for your reply. Yes, I do use Dynamic EQ. I suspect that the benefits of that dynamic adjustment do not show up in the seconds-long test tone. I would say my sound is anything but lacking in bass response. That said I've taken your response and @PeterNL's as well and adjusted the Curve Editor to make the downwards sloping bass-heavy curve. That has further increased the bass response, though I swear the results do not show up in the graphs below.

This morning's results look decent but very similar despite all the different speaker and EQ settiigs:
View attachment 321014

Now, this evening so far, I have added the bass heavy slope to the Curve Editor -- and the bass response is 40 dB lower than the rest of the frequency range, across the board in all EQs and speaker settings (with and without the subwoofer:
View attachment 321015

So I'm completely confused by the above graph now. I really appreciated your explanation of the Reference Offset, though it still left me somewhat confused, given what I hear. It sounds like the setting that will produce the most low end would be the 0 Offset, which is what I kind of understood from the manual, but your explanation is more thorough (even if I can't definitely say I understand the effect). At any rate, changing to 10 and 15 seemed to produce less bass when listening to Taylor Swift - Lavender Haze at 50 on the dial, which has become my primary bass testing song of late. At any rate, getting frustrated with the graph above and the bizarrely quiet bass response it shows, I've been listening to actual music with my "Audyssey Slope2 without subwoofer at volume50" settings and there is insane good bass, and balanced with crisp, defined vocals and instruments front and center in the mix, while enveloped in deep wide bass. Even though bass didn't seem overly weak before, it's definitely not now. I'll keep listening to my stereo-testing playlist, then put the new copy of Dave Brubeck that arrived on vinyl today and see if analog jazz comes across just as lovely.
*p1 speaker setting has no Subwoofer; p2 has a subwoofer; v60 means volume set at 60, which is pretty loud for streamed music; Slope2 is the second bass heavy curve int eh Curve Editor

This may or may not help, but basically the different offset levels produce increasingly more bass (and slight boost in highs, and levels of surround channels).

*This isn't accurate -- I couldn't find the actual #s for how much boost is added -- this is just to show the concept of the boost based on offset compared to volume dial setting. Note that these are "0 = max/reference" on the dial, and turning UP goes towards 0, turning DOWN goes bigger negative numbers

Offset // Volume (Dial) // Bass boost added
15 >> -30/25/20/15/10/5/0 >> +8/+5/+3/0/-3/-5/-8
10 >> -30/25/20/15/10/5/0 >> +10/+8/+5/+3/0/-3/-5
5 >> -30/25/20/15/10/5/0 >> +12/+10/+8/+5/+3/0/-3
0 >> -30/25/20/15/10/5/0 >> +15/+12/+10/+8/+5/+3/0

Music is mastered louder, so your volume level is usually lower (eg., -35 on the dial), and movies/TV are mastered with more dynamic range, so you might listen at -20 on the dial for comfortable dialog levels, and you have loud and quiet SFX based on that volume setting (Dynamic Range Compression / Audyssey Dynamic Volume alter the dynamic range of the quiet~loud sounds).
 

cab1024

Member
Joined
Jun 2, 2023
Messages
7
Likes
2
This may or may not help, but basically the different offset levels produce increasingly more bass (and slight boost in highs, and levels of surround channels).

*This isn't accurate -- I couldn't find the actual #s for how much boost is added -- this is just to show the concept of the boost based on offset compared to volume dial setting. Note that these are "0 = max/reference" on the dial, and turning UP goes towards 0, turning DOWN goes bigger negative numbers

Offset // Volume (Dial) // Bass boost added
15 >> -30/25/20/15/10/5/0 >> +8/+5/+3/0/-3/-5/-8
10 >> -30/25/20/15/10/5/0 >> +10/+8/+5/+3/0/-3/-5
5 >> -30/25/20/15/10/5/0 >> +12/+10/+8/+5/+3/0/-3
0 >> -30/25/20/15/10/5/0 >> +15/+12/+10/+8/+5/+3/0

Music is mastered louder, so your volume level is usually lower (eg., -35 on the dial), and movies/TV are mastered with more dynamic range, so you might listen at -20 on the dial for comfortable dialog levels, and you have loud and quiet SFX based on that volume setting (Dynamic Range Compression / Audyssey Dynamic Volume alter the dynamic range of the quiet~loud sounds).
Thank you for this. I understand now. I tend to listen with Offset at 0, but I see it will give me the option of toning it down if too much in situations when it may sound muddy. Though, with the new curve via the Curve Editor and bass-aplenty, it sounded great at 0 with all the music I played.

Next time I take REW readings I'll set DEQ off and see if that normalizes my readings.
 

garyrc

Active Member
Joined
Nov 4, 2021
Messages
112
Likes
119
I see it will give me the option of toning it down if too much in situations when it may sound muddy.

I have noticed a very slight muddiness when using DEQ, compared to the rough equivalent bass boost using the bass tone control. With the tone control, I notice no muddiness at all. I don't have my data from 2011 at hand, or I would share it with you.
 

Chromatischism

Major Contributor
Forum Donor
Joined
Jun 5, 2020
Messages
4,843
Likes
3,771
Setting the reference offset to -15dB is ideal for my listening volumes. I generally listen to music at 55 to 65 (on the 0-98 scale), depending on the type of music and the mastering. Most surround and atmos mixes are quite dynamic and can be listened to at 65 without fatigue or annoyance. At this volume, DEQ with a -15dB offset does nothing, as it should! But at lower volumes, it is still not for me. I prefer having it off.

My next goal is to purchase the MultEQ X pc app and try to get the best EQ I can for my room. That counts more than using DEQ at this stage. I read everywhere on this website that frequencies above 250/500 should be left alone, but in my case it does sound better if I go higher, for the surrounds at least to 1KHz and for the fronts and centre to 6KHz. It might be theoretically incorrect to do that, but it's all better balanced. My speakers are well designed, I think the problem is the room :)
It really, really depends on the speakers you're working with. Speakers with a more laid back presentation will benefit more from full-range Audyssey, especially for movies. With neutral speakers, you'll want to cut them off lower.
 
Last edited:

Albert Dagger

Member
Forum Donor
Joined
Jul 29, 2023
Messages
50
Likes
32
Location
Melbourne, Australia
It really, really depends on the speakers you're working with. Speakers with a more laid back presentation will benefit more from full-range Audyssey, especially for movies. With brighter speakers, you'll want to cut them off lower.
I've purchased a new Marantz AV 10 and a licence for MultEQ-X.

I've corrected frequencies up to 400Hz for front, centre and surround speakers and 300Hz for in-ceiling speakers. Above these frequencies, I've added parametric EQ to adjust the response of the front and centre speaker (in the range 2K-6K) and the in-ceiling speakers (again, 2K-6K). So, these parametric EQ corrections are similar to using tone controls, while the main room modes are corrected by Audyssey. The result is very pleasing.

You can see my settings in this screenshot. Since I took this screenshot, I decided to increase a bit more the 2K-6K range of the in-ceiling speakers. The tonality is very consistent now.

I've also adjusted manually the trims for all speakers. The automatic values were slightly off. I lowered the rears by 0.5/1dB and increased the centre by 0.5dB. I used pink noise that only focuses on the 500Hz-2KHz region for this manual adjustment. Now all my demo tracks with a lot of panning play as they should.

Marantz AV10 MultEQ-X corrections 2023 10 21.jpg


Here you can see the target curves applied to each channel (see the curves at the bottom).

1698301431675.png


Just for curiosity, these are the results of the Audyssey measurements. I used a calibrated mic for these, not the standard mic coming with the AV processor:

Marantz AV10 MultEQ-X Measurements 2023 10 21.jpg


I've also double-checked the results with REW. Here you can see the frequency response with 1/6 smoothing:

1698301368931.png



I've also used an online calculation to visualise the room modes of my living room. I've drawn an accurate plan of the space. See for example this screenshot:

1698301629722.png


The nasty room modes are all between 50Hz and 150Hz. Before applying Audyssey, there was a big peak at 50Hz, and the dips at 60Hz and 90Hz have not been fully fixed.

Overall, MultEQ-X is a great PC app. Incredibly easy to use and I am already happy with these results. I could probably spend more time and come up with a better fix for the low frequencies, but this is already "good enough". I'm also happy that I've got a better understanding of my room and its behaviour.
 

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
968
Likes
397
Location
Yerevan "Sydney Born"
Small heavily treated room with my DIY design speaker, horn CD with a 15 inch and dual subs. Someone simulated the CD with the horn to properly directivity match it with the woofer. I used Audiolense XO to tie it all in. The large photo measurement is left and right(Red) while and second measurement attached below is left(blue) both with 1/6 smoothing.

1698318935873.png
 

Attachments

  • Left Main.jpg
    Left Main.jpg
    113.7 KB · Views: 41
Last edited:

PeterNL

Member
Joined
Jan 22, 2019
Messages
91
Likes
79
Location
Netherlands
Small heavily treated room with my DIY design speaker, horn CD with a 15 inch and dual subs. Someone simulated the CD with the horn to properly directivity match it with the woofer. I used Audiolense XO to tie it all in. The large photo measurement is left and right(Red) while and second measurement attached below is left(blue) both with 1/6 smoothing.

View attachment 321316
Could You provide graph with 50dB scale and 20Hz-20kHz ?
 

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
968
Likes
397
Location
Yerevan "Sydney Born"
Could You provide graph with 50dB scale and 20Hz-20kHz ?
Is the scale on the top right hand side with the plus and minus? And do I chose DBu DBw or DBv? And how do I chose te 50db?

Sorry , to many questions I am sure its simple enough....
 

PeterNL

Member
Joined
Jan 22, 2019
Messages
91
Likes
79
Location
Netherlands
Is the scale on the top right hand side with the plus and minus? And do I chose DBu DBw or DBv? And how do I chose te 50db?

Sorry , to many questions I am sure its simple enough....
Use "limits" button, low limit for example 50dB, high 100dB, frequency range 20Hz-20kHz or 15Hz (in your case) to 20kHz.
Use SPL.
 

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
968
Likes
397
Location
Yerevan "Sydney Born"
I think I got it right, I did see some parameters at the start of the thread such as 1/6 smoothing, either these limits are common knowledge or I just missed, I suppose this is a better representation right? Red both left right, and blue just left.

1698321870538.png

1698321917918.png
 

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
968
Likes
397
Location
Yerevan "Sydney Born"
I tried investigating the dip in the left/right measurement around 4k, its not there in the seperate left or right measurements and apparently it is phase differences possibly from the room unlikely it is audible but not much I can do. Unless anyone has any ideas...
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,115
Likes
2,337
Location
Canada
I tried investigating the dip in the left/right measurement around 4k, its not there in the seperate left or right measurements and apparently it is phase differences possibly from the room unlikely it is audible but not much I can do. Unless anyone has any ideas...

Try to apply frequency dependent windowing (say, FDW 6 cycles) and/or phase "cross correlation align" prior vector averaging.
 

Trdat

Addicted to Fun and Learning
Forum Donor
Joined
Sep 6, 2019
Messages
968
Likes
397
Location
Yerevan "Sydney Born"
Try to apply frequency dependent windowing (say, FDW 6 cycles) and/or phase "cross correlation align" prior vector averaging.
Is the point of this to give me an idea of the audibility of that dip? And if so do I add the seperate left and right measurements in the vector average?
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,115
Likes
2,337
Location
Canada
Is the point of this to give me an idea of the audibility of that dip? And if so do I add the seperate left and right measurements in the vector average?

Assuming it is the room that's causing the skew, frequency dependent windowing will window out some of the room effect. If the microphone is not exactly centered, cross corr align will adjust the impulse centering between the selected measurements based on the first one. If the phase response of the speakers happen to not match quasi(anechoically) I would attempt to correct it.
 

DJBonoBobo

Major Contributor
Forum Donor
Joined
Jan 21, 2020
Messages
1,396
Likes
2,942
Location
any germ
Isn't it perfectly normal to have cancellations/disturbances in treble when measuring both channels together? I thought this was the reason why everyone says one should measure one channel at once?
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,115
Likes
2,337
Location
Canada
Isn't it perfectly normal to have cancellations/disturbances in treble when measuring both channels together? I thought this was the reason why everyone says one should measure one channel at once?

There is a very remote chance that one of the speakers' windowed phase response deviates far enough from what's expected. I may be wrong, but Trdat's speaker appear DIY as well. I would at least try to investigate rather than assume the two speakers are identical...
 

garyrc

Active Member
Joined
Nov 4, 2021
Messages
112
Likes
119
I don't know, but if your speakers cross over to the tweeters between 3K and 7K could just one tweeter be wired with the wrong polarity? Probably just the right one? As I said, I'm stumped, but might it be worth reversing the wires to the right tweeter only, just to see?

As far as audibility is concerned, I'd bet that, at that frequency, the dip would be audible on some music.
 

TurtlePaul

Addicted to Fun and Learning
Joined
May 19, 2020
Messages
632
Likes
1,032
Location
New York
Here we go. I have been tinkering with the MiniDSP 2x4 HD to integrate the switch to Revel F206 speakers. I think this is done now. This is measured at main listening position, but please note that I tuned to an average response which does not just target MLP (I averaged several seats and other positions in the room). Sub is rolled off 24 dB Linkowitz-Riley at 60 hz, inverted and has a 19 ms delay to align phase slopes. The highs are cut 12 dB Butterworth which combines with their natural roll-off at 60 hz. I added an additional cut via a peaking filter at 25 hz to the mains to remove some bass cancelation below 30 hz and cut below the port tuning. A total of eight filters are used with the highest at 323 hz. Max boost is 5.7 dB at 182 hz as the uncorrected response has a mild mid-bass suckout to fill and three room modes to cut.

Frequency Response.jpg
Spectrogram.jpg
 
Last edited:
Top Bottom