What is '3 blob soundstage' ?
Aug 4, 2014 at 2:54 PM Post #2 of 12
Music sounds like 3 blobs-- sound coming at you from extreme left, sound coming at you from extreme, and sound coming at you from the middle.  Lack of an overall soundscape that pans from left to right.
 
I personally have never heard that particular phenomenon in headphones.
 
Aug 4, 2014 at 3:21 PM Post #3 of 12
  Music sounds like 3 blobs-- sound coming at you from extreme left, sound coming at you from extreme, and sound coming at you from the middle.  Lack of an overall soundscape that pans from left to right.
 
I personally have never heard that particular phenomenon in headphones.


i never heard it too. is this just an exaggeration?
 
Aug 5, 2014 at 1:45 AM Post #5 of 12
I don't know about "3 blob," but I use the term "(Poseidon's) Trident." Strong Left, Strong Center, Strong Right, with notes between L-C and C-R pushed to back and softer too. It happens when you have your headphone's drivers smack over your ear canals. Happens with some IEMs but I'm using one that sounds closer to a headphone even without Crossfeed. On headphones, it is most easily apparent on Grados, particularly the Prestige series; it can be remedied by moving the  earcup slightly forward so the drivers aren't directly over the ear canals, but of course there might be other side effects depending on the fit (ex. if you lose proper seal around your ears, you lose bass). 
 
I use trident because the sound is usually piercing and feels like what would be at your face if you manage to piss off this guy:

 
 
For those who don't hear this, it's either that the way you wear your headphones already have the drivers off the top of the ear canals, or you might not be aware of imaging, which is always the last thing people notice considering the impact of tonality and dynamics first. I didn't realize how wrong my car's audio system was until I heard a system with a time alignment processor, and the voice was smack in the middle of the dashboard (mine was offset three inches to the left), the drums rolled from left to right with precise locations that will make a cat look (mine rolled from extreme left to extreme right); and the bass drum despite the subwoofer in the trunk was imaged on the windshield. The equivalent of the trident shape in car audio is called the "rainbow stage," where sitting on the right and the unequal pathlength and on/off axis angles of the driver's side and passenger's side midrange can pull the soundstage one way, typically the driver's (the primary listener) side, and using balance control as opposed to actual time alignment results in very forward, big sound on the passenger side (since it's louder) while the sound on the driver's side is softer and pushed a bit towards the rear of the soundstage (windshield).
 
This is the sort of image that inspired the term rainbow, not one that you view from the center:

 
Aug 5, 2014 at 9:10 AM Post #6 of 12
Yeah it's a wild exaggeration.  Almost all headphones have the same soundstage placement.  It's only the openness and perceveived largeness of the soundstage that can differ a bit in my experience.
 
I've never heard a supposed underpowered planar exhibit this either.
 
Aug 5, 2014 at 12:12 PM Post #7 of 12
I forgot that I made a quickie job on Paint to illustrate that "Trident" soundstage issue, so assuming that's the same as "3 Blob," here's a few of them for comparison. Circles are percussion (save for the head), rectangles are (electric) guitars (acoustic guitars take a narrower, more precise position on L, C, or R, given they are more point-source than the wide dispersion of guitar amp speakers), triangle is the main vocalist. Note how the actual "size" of the soundstage does not actually change, contrary to many observations of "huge" soundstage that sound like headphones already got around the limitations posed by the physical position of the drivers (ie, on the head), which personally applies only to the K1000 (which doesn't really count).
 
1. Grado SR225, SR80 (worn the "usual" way with the on-ear pads' rear on the ear lobes)
- Larger circles = louder, more "upfront" drums; smaller circles = some drum hits are softer and imaged behind the other drums
- Drums on the far right and far left are at times imaged slightly forward of the vocals, owing to how loud they are relative to the other instruments
- Guitars also louder than vocals and appear to be coming from a wider swath along the front of the soundstage
- Worn a bit forward of the ears, the larger circles can be represented as slightly smaller, and pushed back (to the rear of the soundstage) a little bit.

 
2. Sennheiser HD600 (worn with the rear of the earpads having a slight forward push on the earlobes, for further non-alignment of the drivers to the ear canal)
- More even circle sizes - extreme right and left percussion sounds not too loud, not too forward; generally same volume range as other hits; along with bass drum, slightly behind main vocals
- Guitars along the same plain as the vocals, though sometimes the latter may seem more forward (usually with powerful female vocals)

 
 
 
3. Sennheiser HD800 (Just for comparison)
- Again, actual soundstage size didn't really change, particularly the width
- Greater perception of sense of depth (not so much actual depth) since notes aren't as forward anymore; however, if for example a test track for depth and width are used (voice moving around the recording mics, as on EMMA and Chesky CDs), it really won't get that much deeper. Main vocals, the best reference point for the relative position of other instruments, doesn't move outside the head for example, but for example the guitars aren't on the same point along the Z-Axis anymore.
- Again, total/overall soundstage size isn't affected, as shown by how wide off to the left or right the sound from a guitar amp can seem to come from, just that there's more space between each instrument. It does however give the illusion of standing farther back from the stage (not really as far as the vocals are concerned) as the drums are in a more precise point behind the vocals, and on a narrower X-axis coverage, but location tends to be more precise despite the "narrower" image (simply wider isn't always the correct one, as on the HD600 above).

 
Aug 5, 2014 at 1:06 PM Post #9 of 12
  wow excellent explanation, i experienced the same way you describe the HD800.
but... what about sounds coming from the rear?

 
It's physically impossible because the drivers are ahead of your ears - not even a good stereo speaker system can do that. However, if you play a stereo recording of a guy talking while walking around a mic on a 5.1 system, it wouldn't go around either since the recording was in stereo. If you output with the usual 4ch or 5ch (where the stereo output just gets sent everywhere, without actual processing like virtual surround on gaming headphones) settings on A/V receivers, you'll get it all messed up and everywhere because the sound was never meant to come out of more than two channels. Ditto with using the EMMA disc with that track on a car with the rear speakers still operational (serious set-ups disable those unless they have an A/V processor that can do concert DVDs, like James Hetfield's Pioneer-MTX truck).
 
The most that those walking around a mic test tracks will do is an egg-shaped walkaround path, where the depth towards the front of the listener (technically the rear of the soundstage) is deeper than around the rear of the listener's head (technically the front of the soundstage). Note that the orientation of front and rear is such as you and the performers are supposed to be facing each other, not as in surround sound music where some might be programmed to put the listener in the middle of the band (although a lot of concert DVDs I've listened to only have the crowd cheers and some of the live speakers, or just residual sound from them as they recorded the crowd, on the surrounds).
 
That forward bias is also helped by the fact that, since the band is in front and there's that trident/3 blob issue, the headphone drivers like speakers are moved forward of the ear canals and possibly at a toe-in angle as well, which again is preferable for stereo anyway since the band is supposed to be in front. If you for example take speakers and put them smack to your left and right, it will cause the same issues as with headphones, albeit a "larger" soundstage (more like an area where the sound is splattered around). Just try listening in a car and note how each instrument images in the cabin, with and without the rear speakers. Then listen to a desktop speaker system with active monitors - basically as the image is laid on  the desk at eye level is the goal that car audio competitors are trying to replicate on the dashboard, and that the vocals should be dead center despite sitting off center (unless you have a McLaren F1, or an Alpine show car). If you use the balance control to make the far side louder, here's what the "rainbow" soundstage is like:
 
1. Regular car audio set-up, even with aftermarket gear, but without proper time alignment processing, from driver's perspective.
- nearer sounds are louder, seem larger
- vocals biased to off-center

 
2. Using balance control to bias volume to passenger side (ROYGBIV drawn in, very roughly, to illustrate why it's called a rainbow)
- vocals can be centered
- Sounds more on driver side speakers are softer, seem pushed back
- Penalty: sounds more on passenger side speakers are louder, seem pushed forward (essentially reversed)
- can be more severe depending on speakers' dispersion angle and toe-in of the installation, the on-axis response and how much more on-axis to the driver (see next point) the passenger side tweeter (and midrange) is
- can actually be more severe with better aftermarket equipment:
1) the goal is also to align the sound along the Y-axis (height), so there is a tendency to use tweeters that go lower or have a midrange speaker also
2) having more frequencies playing from the dashboard height also pulls more of the image towards the nearer speaker, so you angle it away from the driver and towards the center (and up) to raise and center the vocals
3) this results in the passenger side midrange or tweeter (capable of midrange freqs) being more on-axis with the driver, pulling the image to that side, which can only be made worse by using the balance control instead of actual time alignment (note that correcting time alignment for the driver does not correct it for the passenger, and may make it worse)
 

------
 
So basically what I'm getting at is that every sound system is limited by the physical position of the drivers. In headphones, the soundstage won't really get outside of the head, although putting the drivers off-axis (and introducing toe-in mounting) along with more transparent drivers can result in more space between each instrument, and also place them in a more realistic position relative to each other, minimizing the stronger L-C-R image with recessed spaces between L-C and C-R. In cars, even if you correct it with time alignment, that only works for one seat, but hey it's your own car, and if you're going to spend money on its audio system, might as well make the most out of it.
 
Even on the standard stereo systems you will still be limited by room size (provided you've already controlled reflections and other acoustic issues) in the sense that a 16ft x 20ft room will not be able to replicate the physical width of an orchestra performance in an opera house where the stage alone is probably three of that entire room side by side, since they can't really image the sound where the speakers' dispersion angle cannot throw any sound - less toe-in means wider stage but no depth, too much toe in means minimal width; either way the sound will not be projected smack behind the speakers, save for non-conventional speaker designs (like open panel speakers or omnidirectional dynamic drivers, but even then, it won't be all that holographic, for now anyway).
 
Aug 5, 2014 at 2:47 PM Post #10 of 12
   
It's physically impossible because the drivers are ahead of your ears - not even a good stereo speaker system can do that. However, if you play a stereo recording of a guy talking while walking around a mic on a 5.1 system, it wouldn't go around either since the recording was in stereo. If you output with the usual 4ch or 5ch (where the stereo output just gets sent everywhere, without actual processing like virtual surround on gaming headphones) settings on A/V receivers, you'll get it all messed up and everywhere because the sound was never meant to come out of more than two channels. Ditto with using the EMMA disc with that track on a car with the rear speakers still operational (serious set-ups disable those unless they have an A/V processor that can do concert DVDs, like James Hetfield's Pioneer-MTX truck).
 
The most that those walking around a mic test tracks will do is an egg-shaped walkaround path, where the depth towards the front of the listener (technically the rear of the soundstage) is deeper than around the rear of the listener's head (technically the front of the soundstage). Note that the orientation of front and rear is such as you and the performers are supposed to be facing each other, not as in surround sound music where some might be programmed to put the listener in the middle of the band (although a lot of concert DVDs I've listened to only have the crowd cheers and some of the live speakers, or just residual sound from them as they recorded the crowd, on the surrounds).
 
That forward bias is also helped by the fact that, since the band is in front and there's that trident/3 blob issue, the headphone drivers like speakers are moved forward of the ear canals and possibly at a toe-in angle as well, which again is preferable for stereo anyway since the band is supposed to be in front. If you for example take speakers and put them smack to your left and right, it will cause the same issues as with headphones, albeit a "larger" soundstage (more like an area where the sound is splattered around). Just try listening in a car and note how each instrument images in the cabin, with and without the rear speakers. Then listen to a desktop speaker system with active monitors - basically as the image is laid on  the desk at eye level is the goal that car audio competitors are trying to replicate on the dashboard, and that the vocals should be dead center despite sitting off center (unless you have a McLaren F1, or an Alpine show car). If you use the balance control to make the far side louder, here's what the "rainbow" soundstage is like:
 
1. Regular car audio set-up, even with aftermarket gear, but without proper time alignment processing, from driver's perspective.
- nearer sounds are louder, seem larger
- vocals biased to off-center

 
2. Using balance control to bias volume to passenger side (ROYGBIV drawn in, very roughly, to illustrate why it's called a rainbow)
- vocals can be centered
- Sounds more on driver side speakers are softer, seem pushed back
- Penalty: sounds more on passenger side speakers are louder, seem pushed forward (essentially reversed)
- can be more severe depending on speakers' dispersion angle and toe-in of the installation, the on-axis response and how much more on-axis to the driver (see next point) the passenger side tweeter (and midrange) is
- can actually be more severe with better aftermarket equipment:
1) the goal is also to align the sound along the Y-axis (height), so there is a tendency to use tweeters that go lower or have a midrange speaker also
2) having more frequencies playing from the dashboard height also pulls more of the image towards the nearer speaker, so you angle it away from the driver and towards the center (and up) to raise and center the vocals
3) this results in the passenger side midrange or tweeter (capable of midrange freqs) being more on-axis with the driver, pulling the image to that side, which can only be made worse by using the balance control instead of actual time alignment (note that correcting time alignment for the driver does not correct it for the passenger, and may make it worse)
 

------
 
So basically what I'm getting at is that every sound system is limited by the physical position of the drivers. In headphones, the soundstage won't really get outside of the head, although putting the drivers off-axis (and introducing toe-in mounting) along with more transparent drivers can result in more space between each instrument, and also place them in a more realistic position relative to each other, minimizing the stronger L-C-R image with recessed spaces between L-C and C-R. In cars, even if you correct it with time alignment, that only works for one seat, but hey it's your own car, and if you're going to spend money on its audio system, might as well make the most out of it.
 
Even on the standard stereo systems you will still be limited by room size (provided you've already controlled reflections and other acoustic issues) in the sense that a 16ft x 20ft room will not be able to replicate the physical width of an orchestra performance in an opera house where the stage alone is probably three of that entire room side by side, since they can't really image the sound where the speakers' dispersion angle cannot throw any sound - less toe-in means wider stage but no depth, too much toe in means minimal width; either way the sound will not be projected smack behind the speakers, save for non-conventional speaker designs (like open panel speakers or omnidirectional dynamic drivers, but even then, it won't be all that holographic, for now anyway).


So is that why the HD800 is so big and shaped really odd? super nice explanation although im kinda confused
tongue.gif
(not good at math riddles).
still kinda weird that on certain song, there's sound that always come from the upper rear of my head, it doesn't even bleed to the center, front or L/R
 
Aug 5, 2014 at 11:22 PM Post #11 of 12
 
So is that why the HD800 is so big and shaped really odd? 

 
 
Yes, it allows for mimicking where properly placed speakers are relative to your ears. Here's another quickie diagram I made with Paint to illustrate that. Brown boxes are speakers with toe in (note: quickie job, not necessarily the proper toe in angle - only for illustrative purposes); red boxes are your regular headphone (move that a little bit forward and angled but not as much as the blue line for the HD600, pushed forward); blue line is very generally the HD800, AKG x7xx, Beyer T1/5P, etc. Notice which one looks more like where speakers should be. 
 
 

 
 
 
 
Specifically for the HD800, it's huge primarily because the drivers are huge - most hi-fi headphones use 40mm, the HD600 AFAIK uses 50mm (maybe the AKGs too), but the HD800 uses 70mm drivers - previously used only as bass headphones as more surface area generally can mean more bass (if you are designing it like a subwoofer). The  HD800 however has a lot of vents even on the front of the driver, which I suppose prevents excess bass resonance (which is in some ways is how closed, bass headphones make their bass). For the AKG x7xx series, they also have tiny vents (but overall area is much smaller), and the angle is achieved through angled pads that are thicker towards the rear, which means that with use (or "burn in," as other people call it), the sound can be affected as the pads wear out.
 
super nice explanation although im kinda confused 
tongue.gif
 (not good at math riddles).

 
You mean X, Y, Z? No equations to be solved here (although I did use Calculus for our philosophy classes, just to screw with Aristotle's fans)

X axis = Left to Right - guitar on left/right or center (depending on the recording; Metallica offsets rhythm to one side even if Hetfield is in the middle; on acoustic recordings if the vocalist is playing his/her guitar is in the center), vocalist in the center, drums pan left to right
 
Y axis = Top to bottom. This is only a concern in cars where the tweeter and/or midrange might be mounted higher, or all transducers mounted too low, or with nearfield set-ups with too low or too high speakers that aren't angled. In regular speaker set-ups, if your chair, height, relative speaker stand height, and even room size are all out of proportion (speakers' stands are to low or too high relative to your ears, which in some cases is minimized if you're sitting far enough away from them.
 
Z axis = front to rear of soundstage, ie vocalist in front, drums in the rear
 
 
still kinda weird that on certain song, there's sound that always come from the upper rear of my head, it doesn't even bleed to the center, front or L/R

 
It might have to do with the drivers' relative position on your head, or it might have been a resonance during the recording process. If it's a live recording, the mic recording the crowd might have actually been placed over the crowd, picking up more of the same notes from the stage or the extra monitors if it's a large venue (or the mics on stage picked up the sound from the additional monitors). In any case, I did not say that staging towards the rear was absolutely impossible, but that you can't expect a perfect circle around your head, hence the "egg shape" I discussed above.
 
Rough headphone imaging diagram overlaid on a chicken egg; blue circles illustrate the positions of a typical walk-around the mic/s test track in front of the listener's face, red circles illustrate the positions in the same track when they go around the head. See how its shape isn't symmetrical front to rear, like the egg having a rounder side and a longer side.
 

 

Users who are viewing this thread

Back
Top