I thought it didn't as well - and IMHO it shouldn't. Which is why I find the difference in target level between channels quite surprising.
Room Mode + Speaker Anechoic Measurement Correction
- Thread starter lempogi
- Start date
slartibartfast
Major Contributor
- Joined
- Apr 18, 2024
- Messages
- 4,538
It appears to choose different target levels depending on the chosen frequency range which it also shouldn't do. It should analyse the full frequency range to choose the target level regardless of the chosen range.
That is my view as well, I absolutely agree.
lempogi
Member
Thanks, I'm taking notes of what I'm learning here.
I don't know really. Can you take a look at the pictures I attached? The only difference of the left speaker is the wall in the left is shorter than the wall in the right (left wall is 72cm while right wall is 2.69m). Woofer center of left speaker is 93cm from the left wall, which is the same distance compared to the right speaker to the right wall. Woofer center to the floor is 82cm while woofer center to the ceilling is 1.62m.
103-85 inches = 18 inches = 45.72cm x (1ms/34.3cm) = 1.33ms = ~2ms
SB-1000 Pro = 6ms
Total = 8ms
Is that correct? I rounded it to 2 instead of 1?
It's a lot better. Best sound I heard from my setup. Super happy!And most importantly - how do you like the sound after correction?
But I learned from you guys that still it has issues, and even a RoomFit error/bug?
Attachments
I'm really glad to hear that!It's a lot better. Best sound I heard from my setup. Super happy!
Please beware that a lot of us here are audio enthusiasts, so things we sometimes get hung up on may not be *that* important to achieve really good sound reproduction. I.e. even if you solved the "issues" we indicated so far you may find there's not that big of an improvement to the sound you are already getting. But the graphs would look nicer, and it definitely wouldn't hurt.
So I'd only advise to continue looking into further improvements if you're curious to learn more about how all this works. If not, I'd suggest to just sit back and enjoy some music!
Entering these values into this online SBIR calculator I get:
As you can see, with your current placement, the sidewall reflection will cause a null at ~92Hz and floor reflection will cause a null at ~105Hz.
This is all relatively close to the crossover frequency (80Hz), so I suspect a combination of these 3 factors is what is causing the relatively wide null between 80-100Hz in your left channel.
If you could place your loudspeakers >110cm away from side walls and >110cm above the floor the related SBIR nulls would go below the crossover frequency and effectively disappear. You can try these values in the calculator I linked above and see for yourself.
However, looking at your pictures that kind of side-wall distance may not be practical in your room.
So alternatively you could try to set the subwoofer crossover higher (say 90Hz) and push the speakers just a couple of inches further away from the side walls (e.g. 98cm instead of current 93cm) - just to get the SBIR dip below the new crossover point (90Hz).
After this I'd remeasure before trying to adjust speaker height at all, and re-evaluate. But the same principles apply to floor reflection as well, and you can use the same SBIR calculator to see how changing the relative distances/height would affect the nulls/dips in measured response.
The sub/speaker delay optimization would be the second topic to tackle (once above is done). That brings us to:
These are very reasonable assumptions, yes!
However, there may be some tolerance, so once placement and crossover frequency is optimized I'd try to test a few ms lower and higher delay values as well, and see which one results in highest level at the crossover frequency.
Hope this is helpful!
Last edited:
lempogi
Member
No worries, I would like to try all the possible practical ways to maximize the capabilities of my audio setup. There will be compromises for sure, but it will be ok to me.I'm really glad to hear that!
Please beware that a lot of us here are audio enthusiasts, so things we sometimes get hung up on may not be *that* important to achieve really good sound reproduction. I.e. even if you solved the "issues" we indicated so far you may find there's not that big of an improvement to the sound you are already getting. But the graphs would look nicer, and it definitely wouldn't hurt.
So I'd only advise to continue looking into further improvements if you're curious to learn more about how all this works. If not, I'd suggest to just sit back and enjoy some music!
Noted. I'll try to adjust the crossover and move the left and right speakers nearer to the tv stand.
I'm curious with this one. I learned somewhere that the best height for speakers is, the tweeter should be at the same level with listener's ears. So I measured it and bought a 28 inch speaker stand for that. So that means it's a compromise regarding SBIR measurement then? I think I'll stick with my speaker stands because my wife may get angry if I'll buy another one.
Excited to do this next after we conclude the first issue. Thank you so much! Please don't get tired of helping us here, it's really invaluable!These are very reasonable assumptions, yes!
However, there may be some tolerance, so once placement and crossover frequency is optimized I'd try to test a few ms lower and higher delay values as well, and see which one results in highest level at the crossover frequency.
Hope this is helpful!
Your understanding it absolutely correct!
The tweeter should ideally be pointing right at your ears; but if your seat is low, and the speaker is pointing straight forward, then this might not be ideal for SBIR - as you correctly concluded.
Note however that there's some leeway with this because of two main reasons:
A) Loudspeaker directivity. Your loudspeaker produces a pretty similar response within about +/-5° of its tweeter axis vertically (just eyeballing it from this graph):
(source)
That means that if you sit 85" away from such a loudspeaker you can be sitting up to 7,5" higher or lower than the tweeter and still basically get close to ideal on-axis direct sound at the listening position. It is still ideal to be in the middle of this range to allow for some vertical movement, though - but it is not true that only one single speaker height results in good sound.
B) You could angle the loudspeaker downwards from the stand to point it to your ears, even if the tweeter is a bit above your head.
If it's any consolation I agree with your wife that this is probably not good justification to go and buy new speaker stands.I think I'll stick with my speaker stands because my wife may get angry if I'll buy another one.
It will be very interesting to see what you are able to achieve, if you're willing to share the results as you progress through the iterations.
No problem, TBH it makes me happy when I'm able to help!
lempogi
Member
Thanks for your very good explanation. But yeah, I think I'll just accept that. If only I bought a 36 inch stands so that the floor null will be at 84Hz based from the calculator. But it is what it is. And I also don't like my speakers to be tilted for the sake of the looks aspect, sorry.Your understanding it absolutely correct!
The tweeter should ideally be pointing right at your ears; but if your seat is low, and the speaker is pointing straight forward, then this might not be ideal for SBIR - as you correctly concluded.
Note however that there's some leeway with this because of two main reasons:
A) Loudspeaker directivity. Your loudspeaker produces a pretty similar response within about +/-5° of its tweeter axis vertically (just eyeballing it from this graph):
(source)
That means that if you sit 85" away from such a loudspeaker you can be sitting up to 7,5" higher or lower than the tweeter and still basically get close to ideal on-axis direct sound at the listening position. It is still ideal to be in the middle of this range to allow for some vertical movement, though - but it is not true that only one single speaker height results in good sound.
B) You could angle the loudspeaker downwards from the stand to point it to your ears, even if the tweeter is a bit above your head.
If it's any consolation I agree with your wife that this is probably not good justification to go and buy new speaker stands.
But what if I'll also try a 110Hz crossover to overcome the floor null? Is it very high already?
For sure I'll share it all here. Thank you!It will be very interesting to see what you are able to achieve, if you're willing to share the results as you progress through the iterations.
No problem, TBH it makes me happy when I'm able to help!
By the way, the one error you guys found out about RoomFit correction. Will it still be better to use individual correction compared to the stereo correction?
You can try it, sure, though In general I'd try to keep it lower than that.
110Hz is significantly higher than the most typical values (<85Hz), so you might start to localize sound coming from the sub in the corner, or have a feeling the stereo image is being 'pulled' to the left.
If you don't experience any of these effects then 110Hz would be perfectly OK!
Absolutely nothing wrong with experimenting! I believe it is the best way to learn.
I'd still suggest to use the individual correction, yes, at least for now. When you make modifications the imbalance might be reduced - we'll see as you go forward. But I would definitely suggest to report this issue through the WiiM Home App "Feedback" function.
lempogi
Member
Ok I'll try it also just for curiosity.You can try it, sure, though In general I'd try to keep it lower than that.
110Hz is significantly higher than the most typical values (<85Hz), so you might start to localize sound coming from the sub in the corner, or have a feeling the stereo image is being 'pulled' to the left.
If you don't experience any of these effects then 110Hz would be perfectly OK!
Absolutely nothing wrong with experimenting! I believe it is the best way to learn.
I'll do that later after my work. Thanks!
So, again taking a opportunity of these discussion I did some testing:
My listening (for this setup) room is my bedroom, sqarish, about 12 m2, there is not much space because of master bed, and my TV is located on the closet which is not centered. I put speakers by TV, one on stand and one on closet next to TV. They (speakers) are around 15cm from rear wall (that's way I decided to buy front ported), left one is about 80 cm from side wall, right one is around 1,3m from side wall. Sub is located between left speaker and side wall (circa 15 cm from the corner).
My RoomFit settings as for today test:
BK Curve 20-400hz
Cut-Only Mode
Positive Gain 3,5 db, Negative -12db, Q 12db
Individual L/R
Mic Dyton IMM 6C, Calibration file enabled
Single Sweep
Sub cross: 70Hz
Phase on sub and in Wiim App: 180
Speaker vs sub latency : 3ms (calculated by Wiim, sub in the same line as speakers)
Sub by pass mode and Mains output bass : off (although I got the feeling that with this second option sound is fuller)
Results from today:
and here is control sweep in HouseCurve
I can't get rid of dip around 150-170 no matter of setting I use. I tried to play with subwoofer cross, phase (witch brings some impact to overall result) but this dip is higher than crossover so I don't know if it matters. Another dip is around 70 so maybe it is sub cross related.
I can't rearrange speakers much, I can put them very close to wall, or toe them in but rather can't replace it
The problem that I try to fight from a long time is (and I don't know how to explain it properly): when there is bass line (most scenarios) some notes are played with full deep bass but when melody pattern goes, others are thin, like it can be heard that they are with less volume and body. There are songs with like 3 notes bass pattern and I can hear audible differences between for example second and third one. When I hear the same song on my AirPods I can hear then as melody goes all bass line notes are played with the same volume/body/bass impact so to say. And I don't know what to do frankly speaking. Is is the gear, the room itself, gear placement/quality or acoutitcs.
In terms of harshness reduction, Elac's Spinorama eq seetings work similar to 1,6 kHz HS filter that I experimented with. In some scenarios it works, but for some it affects bass/mids/trebles balance and bass and mids are too much presented, especially if RoomFit is more bass oriented.
Here are 5 cents from my side , hope someone can assess my results and help get better ones, and resolve my problems.
best regards
K
My listening (for this setup) room is my bedroom, sqarish, about 12 m2, there is not much space because of master bed, and my TV is located on the closet which is not centered. I put speakers by TV, one on stand and one on closet next to TV. They (speakers) are around 15cm from rear wall (that's way I decided to buy front ported), left one is about 80 cm from side wall, right one is around 1,3m from side wall. Sub is located between left speaker and side wall (circa 15 cm from the corner).
My RoomFit settings as for today test:
BK Curve 20-400hz
Cut-Only Mode
Positive Gain 3,5 db, Negative -12db, Q 12db
Individual L/R
Mic Dyton IMM 6C, Calibration file enabled
Single Sweep
Sub cross: 70Hz
Phase on sub and in Wiim App: 180
Speaker vs sub latency : 3ms (calculated by Wiim, sub in the same line as speakers)
Sub by pass mode and Mains output bass : off (although I got the feeling that with this second option sound is fuller)
Results from today:
and here is control sweep in HouseCurve
I can't get rid of dip around 150-170 no matter of setting I use. I tried to play with subwoofer cross, phase (witch brings some impact to overall result) but this dip is higher than crossover so I don't know if it matters. Another dip is around 70 so maybe it is sub cross related.
I can't rearrange speakers much, I can put them very close to wall, or toe them in but rather can't replace it
The problem that I try to fight from a long time is (and I don't know how to explain it properly): when there is bass line (most scenarios) some notes are played with full deep bass but when melody pattern goes, others are thin, like it can be heard that they are with less volume and body. There are songs with like 3 notes bass pattern and I can hear audible differences between for example second and third one. When I hear the same song on my AirPods I can hear then as melody goes all bass line notes are played with the same volume/body/bass impact so to say. And I don't know what to do frankly speaking. Is is the gear, the room itself, gear placement/quality or acoutitcs.
In terms of harshness reduction, Elac's Spinorama eq seetings work similar to 1,6 kHz HS filter that I experimented with. In some scenarios it works, but for some it affects bass/mids/trebles balance and bass and mids are too much presented, especially if RoomFit is more bass oriented.
Here are 5 cents from my side , hope someone can assess my results and help get better ones, and resolve my problems.
best regards
K
Last edited:
slartibartfast
Major Contributor
- Joined
- Apr 18, 2024
- Messages
- 4,538
You could try increasing the allowed positive gain. At the moment you have a bigger dip around 60Hz in the corrected curves than in the original measurement.
That makes no sense.
Having both, the sub and the WiiM invert the polarity get's you back to non-inverted polarity. It's the same as having both set to 0°. I would strongly encourage you to set the "phase" to 0° on the sub but keep it at 180° on the WiiM. This big dip appears to be pretty much exactly at the crossover frequency of 70 Hz. This is the typical sign of the sub and the mains not being "in phase".
If the dip turns into a (slight) peak at 70 Hz after setting the sub's phase to 0° it's time to further adjust the delay. In 1 ms steps.
Having "Mains Output Bass" enabled (all other settings remaining the same) should certainly sound fuller, but not necessarily better.
The smaller your speakers, the more they will benefit from not having to output bass. It's also usually easier to get a good integration between mains and subs with both settings disabled.Better don't change two parameters at a time.
In particular, do not disable No-Boost Mode when allowing high values of max gain.
Similar threads
