Do you see you have it set between 4 and 10K but your EQ is at 11?
This time I equalized the driver's ON axis response flat using REW auto EQ and tried to see the plot.
On axis response equalized flat from 800 Hz up (sort of)
Corresponding spectrogram
Some weirdness above 2kHz and 6-7kHz and above is now even more prominently visible.. 🤔
Attachments
But I had switched off the graphical equalizer when I was playing around with the Deesser.
The graphical equalizer was used initially when I tried to EQ down the 11kHz-ish frequency response hump in yetserday's spectrum from the song.
I will try the desser again as krivium suggested once more and report back.. 🙂
They all correspond perfectly with the little bumps as seen in the FR plot. That should indicate where they come from.
But I don't expect this to be at a level that is troublesome. Is this a close measurement?
If I create even the slightest ripple in a straight FR curve:
It starts to show up in the Burst Decay view:
But seeing how small this ripple is, it is nothing compared to what a room does for perception at the listening spot...
(or the cross talk that our ears perceive for that matter!)
Last edited:
@wesayso: This measurement is a 4ms gated response of the compression driver attached to the ST260KVAR horn, measured at 1m away from the horn mouth.
Also there is not a lot of sibilance in anyway that is unbearable or anything... I was mostly curious even about the small amount of sibilance that I heard..
I probably need to hear more sibilant tracks like the ones Krivium posted above and then decide how much resources I need to spend on remobing this sibilance effect, if it is even possible. 🙂
The other thing is I didn't notice any such sibilance in my 3way speaker compared to the current 2 way. This is why I became the most curious 😀
But your explanations about cross talk and room effects tells me that it could be a potential reason.
I thought if it is something driver related and easy to get rid of, I should do that first.. 😀
Hence all this driver-related enquiries..
Also there is not a lot of sibilance in anyway that is unbearable or anything... I was mostly curious even about the small amount of sibilance that I heard..
I probably need to hear more sibilant tracks like the ones Krivium posted above and then decide how much resources I need to spend on remobing this sibilance effect, if it is even possible. 🙂
The other thing is I didn't notice any such sibilance in my 3way speaker compared to the current 2 way. This is why I became the most curious 😀
But your explanations about cross talk and room effects tells me that it could be a potential reason.
I thought if it is something driver related and easy to get rid of, I should do that first.. 😀
Hence all this driver-related enquiries..
It could make sense due to having that horn on top, which will be quite different in how it reacts to a room than a dome tweeter would. Plus a dome tweeter could have varying diffraction problems that don't show up similarly in this neat horn-shape you have.
If you have a somewhat "standard" stereo setup, being speakers are about 3m from each other and between 2.5 to 3m towards listening position,
try a few simple EQ tweaks and see if you like it for the phantom center voices:
Create a PEQ of -3 dB at about 3.7 KHz, q is dependent on the equalizer, but it should span from 2.2 KHz to 5.5 KHz. If that works favourably for the center channel, do another PEQ at 7.2 KHz, -2.5 db cut spanning from 5.5 KHz to wherever that Q brings it on top. May need a small PEQ to make sure 5.5 KHz still is at the normal levels. Listen to that and focus on voices that are dead center. Do you like what you hear?
See this example:
Don't judge the side panned voices with these tweaks... this isn't anywhere near the total solution, just a quick check to see if you experience excess energy at typical crosstalk spots. The solution I use, uses linear phase EQ tweaks and is using corrections around 700 Hz as well and has way different settings for the sides.
Depending on exact speaker location and distance (as well as head size) these tweaks might have to move up or down in frequency. You can also vary the amount of the cuts. Don't feel bad if it doesn't do anything good, it is easy enough to try.
If you have a somewhat "standard" stereo setup, being speakers are about 3m from each other and between 2.5 to 3m towards listening position,
try a few simple EQ tweaks and see if you like it for the phantom center voices:
Create a PEQ of -3 dB at about 3.7 KHz, q is dependent on the equalizer, but it should span from 2.2 KHz to 5.5 KHz. If that works favourably for the center channel, do another PEQ at 7.2 KHz, -2.5 db cut spanning from 5.5 KHz to wherever that Q brings it on top. May need a small PEQ to make sure 5.5 KHz still is at the normal levels. Listen to that and focus on voices that are dead center. Do you like what you hear?
See this example:
Don't judge the side panned voices with these tweaks... this isn't anywhere near the total solution, just a quick check to see if you experience excess energy at typical crosstalk spots. The solution I use, uses linear phase EQ tweaks and is using corrections around 700 Hz as well and has way different settings for the sides.
Depending on exact speaker location and distance (as well as head size) these tweaks might have to move up or down in frequency. You can also vary the amount of the cuts. Don't feel bad if it doesn't do anything good, it is easy enough to try.
I tried @wesayso's EQ tweaks.. 🙂
Here are MLP measurements with and without the 2 different EQs (The EQ tweaks when applied together with the Room EQ is hard to distinguish in the plots below)
There is an L channel measurement, R channel measurement, L+R channel measurement, L+R measurement with the 3.3kHz dip as shown in wesayso's plot above and the L+R measurement with the 3.3kHz dip & 7.2kHz dips.
I tried to listen to a few songs, including the 'Everything but the girl-missing' song that @krivium linked in an above post.
Songs were played from Tidal and were lossless files.
I tried hard to hear if there is any change in the voices that appeared to come from the midpoint between the speakers. However, the differences would have been very minute... I hardly noticed any change. One thing I felt is playing lossless from Tidal is a good improvement w.r.t playing MP3s I have. Slightly less sibilance (even though there is not much sibilance to start with). Applying wesayso's EQ appeared to minutely increase the depth of the sound stage I think. 🤔
I honestly think that I could be imagining things at the moment since the differences appear to be very less. 😀
I kept on switching ON and switching OFF the EQ while playing the track. But only the above observations were what I could make out (if at all they are correct) 😀
Regards
Vineeth
Here are MLP measurements with and without the 2 different EQs (The EQ tweaks when applied together with the Room EQ is hard to distinguish in the plots below)
There is an L channel measurement, R channel measurement, L+R channel measurement, L+R measurement with the 3.3kHz dip as shown in wesayso's plot above and the L+R measurement with the 3.3kHz dip & 7.2kHz dips.
I tried to listen to a few songs, including the 'Everything but the girl-missing' song that @krivium linked in an above post.
Songs were played from Tidal and were lossless files.
I tried hard to hear if there is any change in the voices that appeared to come from the midpoint between the speakers. However, the differences would have been very minute... I hardly noticed any change. One thing I felt is playing lossless from Tidal is a good improvement w.r.t playing MP3s I have. Slightly less sibilance (even though there is not much sibilance to start with). Applying wesayso's EQ appeared to minutely increase the depth of the sound stage I think. 🤔
I honestly think that I could be imagining things at the moment since the differences appear to be very less. 😀
I kept on switching ON and switching OFF the EQ while playing the track. But only the above observations were what I could make out (if at all they are correct) 😀
Regards
Vineeth
When differences are very small, and both options sound very good anyway, it can take me a long time to decide which one is preferable. I have sometimes evaluated two different DSP filter options over a period of weeks before I could definitively say which one is preferred.
Thank you for sharing your progress !
j.
Thank you for sharing your progress !
j.
Thank you @hifijim.
Yes. I agree extended listening sessions are the only way to understand if I can make out some differences.
I am going to listen to these changes over the nest week or more and see if I have a better understanding of how/if the EQ changes the sound identifiably... 🙂
Yes. I agree extended listening sessions are the only way to understand if I can make out some differences.
I am going to listen to these changes over the nest week or more and see if I have a better understanding of how/if the EQ changes the sound identifiably... 🙂
Vineeth try to find EBTG Todd Terry's remix of 'missing' it's the worst about sibillance ( club oriented, with the house beat and heavy compression on voices).
I will try to find a selection of tracks which are troublesome wrt sibillance.
I will try to find a selection of tracks which are troublesome wrt sibillance.
All have sibillance and could be problematic on loudspeakers with issues but all sound great in their own way.
Dusty Springfield's is interesting: too old to be overprocessed so most likely her voice. But it sound GREAT and it seems louder than the most recent tracks... no overcompression used.
Selena Gomez, well typical USA stuff of the period. Not really my thing but sound nice and very well produced. That said it almost overload at times: loudness war. No depth because of it.
Petra Mogani is live take. A bit in same spirit as Dusty Springfield as it's not overprocessed on voice.
Teardrop. Elizabeth Frazer voice. Great test track for bass too. Production is top notch, lots of movements in the mix. Lack of depth in my view but with this kind of overall eq it's not surprising. Check how they dealt with the S. Clever mixer ( GENIUS mixer. All Massive Attack album are gems to me, but i'm an hardcore fan).
Sinead O'Connor: lot of sibilance in the intro up to 1min. Once the beat is in here voice production change, she sings louder and the reverbs used make them disapear. Great Prince's composition ( remains me of Purple Rain, the strings arrangement, Prince's typical). And i love her voice... i decided not to include Massive Attack's tracks with her but with 20 more years her voice is still nice and fresh. Impressing. And some tracks were sadly premonitory ( 'A prayer for England').
Last edited:
There's something to spot in the graph you posted,
I see the Left and Right separately, but if you measure both combined, they don't sum +6 dB over the entire range as expected, but already show a sizable dip at ~ 3.7 KHz. Did you record this at listening distance?
In general, it does somewhat explain why our ears have this much trouble hearing those little tweaks I suggested. Something tells me that the mic was slightly off center (as both our ears are as well, to the speakers) making that comb pattern with a dip at 3.7 KHz and a bump at 9 KHz. It does show that when both speakers are playing, the position of the ears matters. Luckily our head provides some shading from these type of effects and it won't be perceived as severe as moving a mic slightly off center.
The tweak I suggested should be subtle when heard in a room at normal listening distances. If one does the complete mid-side (with linear phase EQ) it gets a little less subtle what it does for imaging etc. But if you don't hear a difference in sibilant sources, than I guess the material itself is the culprit.
I see the Left and Right separately, but if you measure both combined, they don't sum +6 dB over the entire range as expected, but already show a sizable dip at ~ 3.7 KHz. Did you record this at listening distance?
In general, it does somewhat explain why our ears have this much trouble hearing those little tweaks I suggested. Something tells me that the mic was slightly off center (as both our ears are as well, to the speakers) making that comb pattern with a dip at 3.7 KHz and a bump at 9 KHz. It does show that when both speakers are playing, the position of the ears matters. Luckily our head provides some shading from these type of effects and it won't be perceived as severe as moving a mic slightly off center.
The tweak I suggested should be subtle when heard in a room at normal listening distances. If one does the complete mid-side (with linear phase EQ) it gets a little less subtle what it does for imaging etc. But if you don't hear a difference in sibilant sources, than I guess the material itself is the culprit.
I have measured this phenomenon, and the first time I found it, I thought something was very wrong.
Black curve is equidistant from both speakers. Red is moved to the left about 1 inch. Blue is moved to the left about 3 inches.
Black curve is equidistant from both speakers. Red is moved to the left about 1 inch. Blue is moved to the left about 3 inches.
As i find unfair to have only female singer and english speaking here is another round of 'sibilant' tracks, by men singers and mostly in french language ( not all are french singers: Arno and Jacques Brel are from Belgium):
All this guys are almost the best singers francophony had ever produced, some true legends. It span from 50's to 00's.
All have sibillance in these tracks, some more than others ( Gainsbourg...). I've never really been bothered about it, but i think it is mainly because of lyrics ( as i understand them and all are great poetry it's easy to be taken into the meaning rather than sound) and i'm used to the language.
It never occured to me french is so prone to sibillance but it seems to be the case...
For Brel and Bashung they used Neumann u67 tube microphones. Those are 'dark' rendering microphones... Gainsbourg was u47 tubes iirc. Bit more metallic sounding ( Sinatra and Beatles favourites) but not a Sony C800g or Akg C-12: would have been worst!. I don't know for Arno and Christophe.
All this guys are almost the best singers francophony had ever produced, some true legends. It span from 50's to 00's.
All have sibillance in these tracks, some more than others ( Gainsbourg...). I've never really been bothered about it, but i think it is mainly because of lyrics ( as i understand them and all are great poetry it's easy to be taken into the meaning rather than sound) and i'm used to the language.
It never occured to me french is so prone to sibillance but it seems to be the case...
For Brel and Bashung they used Neumann u67 tube microphones. Those are 'dark' rendering microphones... Gainsbourg was u47 tubes iirc. Bit more metallic sounding ( Sinatra and Beatles favourites) but not a Sony C800g or Akg C-12: would have been worst!. I don't know for Arno and Christophe.
Last edited:
@wesayso: Thanks.. 🙂 Yes, I think the mic was a little off centre. 😀
I thought, anyway I slightly move around while listening so why should we look only at measurement right at the midpoint.
But, I guess, at least for getting good L+R measurements, it is an important criterion.. I had taken some pain to do this while I cooked up the current room EQ but this time I was lazy.. 😀
@krivium: Dont know how to thank you enough 🙂
All these tracks you are pointing out are new to me, and some are quite a departure from the usual kind of music I hear. 😀
In addition to speaker building, I am learning more about what is the kind of music that the world listens to now and this is more fun.. 🙂
I thought, anyway I slightly move around while listening so why should we look only at measurement right at the midpoint.
But, I guess, at least for getting good L+R measurements, it is an important criterion.. I had taken some pain to do this while I cooked up the current room EQ but this time I was lazy.. 😀
@krivium: Dont know how to thank you enough 🙂
All these tracks you are pointing out are new to me, and some are quite a departure from the usual kind of music I hear. 😀
In addition to speaker building, I am learning more about what is the kind of music that the world listens to now and this is more fun.. 🙂
No problem,. as it nicely demonstrates what can potentially happen at ear positions. Only in the exact sweet spot, not with a microphone, but with the nose 😉, so both ears are equidistant to the speakers do these effects get the worst they can be. Because at that exact point, both ears get the exact same comb pattern ast the ear. Move slightly left or right, and both ears have slightly differing comb patterns to judge, and the overall tonal balance as perceived might even be considered as preferable compared to the exact sweet spot.
The M/S EQ (these tweaks were only a small part of it) I use softens that up, while not being perceived as worse upon a head movement.
In total, it mainly addresses the tonal balance difference as it can be perceived between phantom voices and side panned voices. It also frees up the imaging, actually making it a bit more 3D like.
One could do something very similar with an anti-cross talk signal. The difference is that this technique is much more place dependent. It works excellent in the exact sweetspot, but move slightly left or right and the phantom sound moves with you, it actually moves quite fast! So that's more of a "head in vice" solution, or one would need a head placement sensor or camera and change patterns according to position (such options do exist!).
A third option was discussed in the Fixing the Phantom Center thread. It consists of small "virtual" reflections which pattern differs from each other for the left and right channel. The theory (and origin) of that idea is discussed in the paper in the second post of that thread. (we also tried a phase-only version of that)
My own solution puts all above mentioned pieces together, in one way or another. With a bit of freedom taken on how to get there. It works marvelous for movies without needing an actual center channel and also works more than well for music listening. But to have it work like intended, the amount of early reflections at the listening position must be quite low.
How is the imaging you're getting with these speakers? Instead of only listening to sibilant sources, it might also be fun to learn of a few tracks that are quite fun as far as imaging goes.
I'll kick off a list of possible tracks to listen to (for imaging purposes):
Infected Mushroom, Converting Vegetarians - Avratz
My first introduction to this band, and in fact, this music style. There's more tracks on this album that are fun, but it can be a bit much at times.
Their follow up album with this same theme, Converting Vegetarians II actually has more accessible tracks, like:
Also fun imaging wise:
Malia & Boris Blank - Smouldering Ashes.
Overall a fun album to listen to for the different effects that Boris Blank (famous as one of the two piece band: Yellow) creates.
And Hotel California from the album Hell Freezes Over by The Eagles is a favourite in Audiophile circles for a reason:
I actually use these tracks (and many many more) to judge different facets of imaging. One more honorable mention:
The Doors - The Ghost Song (An American Prayer)
A song that was created after Jim Morrison had passed. They took one of his poems and put the music around it, it works wonderfully and very "3D".
The music is wrapped around that center vocal, if you have good imaging, this is quite beautifully done.
There are many more, I'm sure others can suggest a few too. As said, I use all of these as tools and come to love some of these tracks in the process.
My son claims Hotel California is the best song ever (he's 16) and became a huge Pink Floyd fan due to their entertaining imaging with Q-sound tricks and well recorded songs in general. He's also deeply into Infected Mushroom. It's not music, its musical art... he claims.
The M/S EQ (these tweaks were only a small part of it) I use softens that up, while not being perceived as worse upon a head movement.
In total, it mainly addresses the tonal balance difference as it can be perceived between phantom voices and side panned voices. It also frees up the imaging, actually making it a bit more 3D like.
One could do something very similar with an anti-cross talk signal. The difference is that this technique is much more place dependent. It works excellent in the exact sweetspot, but move slightly left or right and the phantom sound moves with you, it actually moves quite fast! So that's more of a "head in vice" solution, or one would need a head placement sensor or camera and change patterns according to position (such options do exist!).
A third option was discussed in the Fixing the Phantom Center thread. It consists of small "virtual" reflections which pattern differs from each other for the left and right channel. The theory (and origin) of that idea is discussed in the paper in the second post of that thread. (we also tried a phase-only version of that)
My own solution puts all above mentioned pieces together, in one way or another. With a bit of freedom taken on how to get there. It works marvelous for movies without needing an actual center channel and also works more than well for music listening. But to have it work like intended, the amount of early reflections at the listening position must be quite low.
How is the imaging you're getting with these speakers? Instead of only listening to sibilant sources, it might also be fun to learn of a few tracks that are quite fun as far as imaging goes.
I'll kick off a list of possible tracks to listen to (for imaging purposes):
Infected Mushroom, Converting Vegetarians - Avratz
My first introduction to this band, and in fact, this music style. There's more tracks on this album that are fun, but it can be a bit much at times.
Their follow up album with this same theme, Converting Vegetarians II actually has more accessible tracks, like:
Also fun imaging wise:
Malia & Boris Blank - Smouldering Ashes.
Overall a fun album to listen to for the different effects that Boris Blank (famous as one of the two piece band: Yellow) creates.
And Hotel California from the album Hell Freezes Over by The Eagles is a favourite in Audiophile circles for a reason:
I actually use these tracks (and many many more) to judge different facets of imaging. One more honorable mention:
The Doors - The Ghost Song (An American Prayer)
A song that was created after Jim Morrison had passed. They took one of his poems and put the music around it, it works wonderfully and very "3D".
The music is wrapped around that center vocal, if you have good imaging, this is quite beautifully done.
There are many more, I'm sure others can suggest a few too. As said, I use all of these as tools and come to love some of these tracks in the process.
My son claims Hotel California is the best song ever (he's 16) and became a huge Pink Floyd fan due to their entertaining imaging with Q-sound tricks and well recorded songs in general. He's also deeply into Infected Mushroom. It's not music, its musical art... he claims.
Oh yes Vineeth,
We are all 'trapped' into our cultural habits, style we listen too,... for this world wide web can be a nice antidote and source of wonder.
There is a section of forum called 'music' in which you can make discovery. Worth spending a bit of time in it imho, there is even some 'games' ( '15 songs album collection') in which i discovered bands, artists, style,...
Anyway, if you want to discover things from this side of our world we should open a thread and exchange on this rather than pollute your 'technical' thread: my own knowledge of your culture is more or less 'locked' on the 'classical/traditional' music of India ( From Ocora label, a division of Radio France dedicated to foreign culture documenting by Country/ region, kind of musicology approach)...
I've got poor knowledge of the 'pop culture' of your country ( Asia in general) and i'm curious about it. And i'm sure i'm not alone in this case in here, after all we are at first music lovers...
Nice selection of tracks Wesayso.
I don't get the Eagles though. I agree it is nicely engineered but... i don't know. I'm not thrilled by it.
Infected Mushroom is very nice.
For imaging i've got some interesting things too but i fear the cultural gap to be too much (i'm well into weirdos electronic things mainly sometimes a bit dark...).
Anyway:
A Serge modular synthesizer, Tb-303,Tr-808,Tr-909 some Lexicon reverbs, too much free time (and acid taken):
Electronic dub produced by mastering engineer:
All Massive Attack albums. Mezzanine, 100TH Window and Heligoland are gems about production and imaging.
https://youtube.com/playlist?list=PL-sQwttqzmN1YVqnGb9ChiRTffGyyJcA1
We are all 'trapped' into our cultural habits, style we listen too,... for this world wide web can be a nice antidote and source of wonder.
There is a section of forum called 'music' in which you can make discovery. Worth spending a bit of time in it imho, there is even some 'games' ( '15 songs album collection') in which i discovered bands, artists, style,...
Anyway, if you want to discover things from this side of our world we should open a thread and exchange on this rather than pollute your 'technical' thread: my own knowledge of your culture is more or less 'locked' on the 'classical/traditional' music of India ( From Ocora label, a division of Radio France dedicated to foreign culture documenting by Country/ region, kind of musicology approach)...
I've got poor knowledge of the 'pop culture' of your country ( Asia in general) and i'm curious about it. And i'm sure i'm not alone in this case in here, after all we are at first music lovers...
Nice selection of tracks Wesayso.
I don't get the Eagles though. I agree it is nicely engineered but... i don't know. I'm not thrilled by it.
Infected Mushroom is very nice.
For imaging i've got some interesting things too but i fear the cultural gap to be too much (i'm well into weirdos electronic things mainly sometimes a bit dark...).
Anyway:
A Serge modular synthesizer, Tb-303,Tr-808,Tr-909 some Lexicon reverbs, too much free time (and acid taken):
Electronic dub produced by mastering engineer:
All Massive Attack albums. Mezzanine, 100TH Window and Heligoland are gems about production and imaging.
https://youtube.com/playlist?list=PL-sQwttqzmN1YVqnGb9ChiRTffGyyJcA1
Last edited:
This is awesome @krivium and @wesayso 😀
Thanks a lot for for sharing these tracks. I will listen to every single one of them.. 🙂
@wesasyo: Imaging is really nice on this system from MLP. I hear the voices coming from dead centre, right in the middle of the space between the speakers in many songs.. But If I move significantly to the left or right, something like more than 30-40 cm, that effect of vocals arriving from the centre seems to collapse. It all sound very nice w.r.t tonality though.
The imaging and the dynamics is one of the reason I like this 2 way system better and keep going back to it again and again. Upon comparing with my 3 way speaker, this 2 way sounds more fun and exciting while the 3 way speaker sounds sort of more polite and the centre imaging seems to be not as well defined as this one. This 2 way also goes scary loud cleanly without any change in tonality. It could all be due to a number of reasons like the impact of room interaction, diffraction, cross-over integration, and all other things I don't know yet.. 😀
@krivium: Really nice idea about starting a new thread for sharing more about music.. 🙂 I will start a new thread for discussing about tracks for evaluation of speakers as well as sharing interesting music soon.. 😀
@fluid: I saw yesterday that the Yuichi A290 inspired mk3b2 horn files were shared for public use.. And I liked it a lot just from the looks of it. I am curious about what is the motivation behind the design of such a horn. Also what size of throat is it designed for? 2inch or 1.4inch.
Is the motivation behind the design more low frequency directivity control in the horizontal plane, better loading of the compression driver, possible better match with a 15inch driver, or some other things..?
Could you please share some details about these aspects..?
Thanks a lot for for sharing these tracks. I will listen to every single one of them.. 🙂
@wesasyo: Imaging is really nice on this system from MLP. I hear the voices coming from dead centre, right in the middle of the space between the speakers in many songs.. But If I move significantly to the left or right, something like more than 30-40 cm, that effect of vocals arriving from the centre seems to collapse. It all sound very nice w.r.t tonality though.
The imaging and the dynamics is one of the reason I like this 2 way system better and keep going back to it again and again. Upon comparing with my 3 way speaker, this 2 way sounds more fun and exciting while the 3 way speaker sounds sort of more polite and the centre imaging seems to be not as well defined as this one. This 2 way also goes scary loud cleanly without any change in tonality. It could all be due to a number of reasons like the impact of room interaction, diffraction, cross-over integration, and all other things I don't know yet.. 😀
@krivium: Really nice idea about starting a new thread for sharing more about music.. 🙂 I will start a new thread for discussing about tracks for evaluation of speakers as well as sharing interesting music soon.. 😀
@fluid: I saw yesterday that the Yuichi A290 inspired mk3b2 horn files were shared for public use.. And I liked it a lot just from the looks of it. I am curious about what is the motivation behind the design of such a horn. Also what size of throat is it designed for? 2inch or 1.4inch.
Is the motivation behind the design more low frequency directivity control in the horizontal plane, better loading of the compression driver, possible better match with a 15inch driver, or some other things..?
Could you please share some details about these aspects..?