No basically i was giving a lamen's way of testing for off axis response without having to use multiple recordings and everything it entails to do off axis measurements of a speaker. I think this test may actually be better than having to record on and off axis responses then converting that into a waterfall or a gedlee chart because it uses perception. When you use perception and skip the measurement then any variable is accounted for in the overall transfer function of your brain and ears. Things should sound "natural" with pink noise as strange as that sounds or mystical. If you get too much weirdness when you move your head then your directivity is not controlled enough.
I'm not an expert but I've been told that all tweeters eventually beam and become directional. So we can not construct a true omni that is 20Hz-20kHz.
But my hunch is that actually omni can be more efficient and maybe somewhat more accurate to the recording or what the 1:1 recording potential should be. What i think may happen is that the higher end will be weak for the engineer if the tweeter is not omni or being used efficiently in a room - which will lead to them boosting slightly too much in the 7kHz range.
Peter Walker basically said his speakers sounded dreadful because they were more accurate and i think that you may get a slight bit of this effect as speakers become closer to being truly omni. Because of your system getting slightly more accurate than the recording engineer's.
But my hunch is that actually omni can be more efficient and maybe somewhat more accurate to the recording or what the 1:1 recording potential should be. What i think may happen is that the higher end will be weak for the engineer if the tweeter is not omni or being used efficiently in a room - which will lead to them boosting slightly too much in the 7kHz range.
Peter Walker basically said his speakers sounded dreadful because they were more accurate and i think that you may get a slight bit of this effect as speakers become closer to being truly omni. Because of your system getting slightly more accurate than the recording engineer's.
I use a coaxial waveguides for the tweeter so the sound comes out in a 360 pattern and it was measured the FR is constant 360deg. Seriously 40Hz-20Khz
http://www.diyaudio.com/forums/multi-way/303941-omnidirectional-work-progress-5.html#post5019727
The APL_TDA tool looks at the wave front a little different. That's why it's interesting to me to see it. It uses frequency dependent slices of the measured sound in an effort to separate direct sound (the first wave front) from reflections.
I've been following your thread from the start
. One of the reasons for me to ask you about this APL_TDA test.The measurements can tell us a lot, once we figure out how to interpret them. That has been my main interest in all of my own testing. Learning how we perceive the sound at the listening spot.
Interesting to read you spotted the 1.8 KHz "problem" Toole mentioned and noticed it was a vague "problem" at best. That's what the 'Phantom Centre' thread was primarily about. And reason for me to say it usually gets filled in by reflections in most rooms. Except when you absorb it all like I did. That's why I use the "generated" reflections (ambient channels) to fill them back in.
The Griesinger paper I linked to has a plot showing the differences of 30 degree stereo compared to 150 degree stereo. The 30 degree measurement shows the 1.8 KHz dip, just like Toole's own measurements. (missing the actual frequencies in the presented graph) It also shows the 150 degree dip being elsewhere, hence the success to fill the 1.8 KHz dip in from behind.
It also might be the primary reason for Grasso to attack Stereo although I'm sure he prefers the wider sweet spot and positional independence from an actual centre too. What I would love from Grasso though is to let us make up our own mind what we want to listen to. Basically we should be able to agree to disagree. We all have our own path we choose to follow. But we shouldn't expect everyone to see it in the same light as we do.
If you learn how it all works in our head a fix isn't that hard to find for 2 speaker Stereo I.M.H.O. All the more reason to see plots of things that work.
It does matter when reflections hit the listener. This does not mean we can't hear trough them as that's what our brain is very good at. There are perceptional differences. The very early reflections eat away the imaging ability. Later ones are way less detrimental. Still there is a difference even in the later arrivals (~6/7 ms and more)
If you remove even more early reflections you stand a bigger chance of hearing/picking up more of the room queues that are embedded in the recording itself (reason for me to clean up the first 20 ms). They are no longer drowned by the room response/reflections. (more chance of a "you are there" type of experience.
This is why studio's go for that semi-anechoic room. Never really anechoic as that would be an unpleasant environment to work in.
Even though there are way more than one 'ideal' Studio design floating around, most will feature a reflection free zone.
If you like the sound you get at the listening position, record a song right there. Play it back on headphones and listen to hear more of what the room adds. Even more daring: play that recording back trough your speakers and you'll hear double the room effects.
A warning though, it will be tough to not hear "the room" after a test like that! Been there, done that!
Last edited:
Any link to those speakers? And the room they were in? That's a pretty bold claim to make. One I personally do not buy without substantial evidence.
If you "get" how we listen and what/when we hear and know how the brain interprets it, you can make every piece of music more entertaining in that process. This may include deliberately deviating from a flat frequency response and things like that.
I've used measurements and psycho acoustic principles to measure my way to better sound. Following the works of many of the Guru's and keeping what they said about it in mind. Guess what: I may not always agree with mixing decisions but the music doesn't sound awful all of a sudden. On the contrary.
I see claims like this in audiophile circles as well. I don't buy it without substantial proof. What better way too boost your own ego
."Yeah the listening experience sucks, the music sounds terrible. It isn't my speakers though, really... it's just those professional mixing and mastering engineers that don't use my speakers, they just can't hear it properly."
I'm convinced if you keep in mind how we listen, you can improve the listening experience of most of the music out there. Not by looking at speakers alone, but by keeping in mind the room and speakers. I refuse not to be able to listen to my music. It's the sole reason for making the speakers in the first place! I love the music I've acquired over the years. My speakers measure pretty good though. It hasn't become a problem.
Last edited:
I was talking about the inventor of the electrostatic speaker. He went from speakers that went to like 15kHz tops to a speaker that could do 20kHz so back then you could hear all the 15kHz-20kHz 'scratch" on the record that the engineer couldn't hear.
What i am talking about now is much more subtle. By making small improvements in the high end you get this same effect just not very noticeable.
In terms of center channel comb filtering effect in your earlier thread i think this can be overcome by adding side channels to stereophony which actually effects the front and fills in the center a bit.
I use a 4 speaker equidistant layout so it essentially blinds me from the room. Where everything localizes the same as a stereo panner to headphones but my panner works 360 degrees.
I feel when you use a 60 degree front stereo set up that generally everything outside of the front 60 degrees is an error no matter what you do.
Personally i think unless you are getting some sort of a bad sound from the room that making it totally dead or anechoic is a bad idea. You only reduce the efficiency of your amplifiers. I think most room treatment is just changing the sound in an incomplete design that is error filled already. Anything outside of the front 60 is an error by nature.
What i am talking about now is much more subtle. By making small improvements in the high end you get this same effect just not very noticeable.
In terms of center channel comb filtering effect in your earlier thread i think this can be overcome by adding side channels to stereophony which actually effects the front and fills in the center a bit.
I use a 4 speaker equidistant layout so it essentially blinds me from the room. Where everything localizes the same as a stereo panner to headphones but my panner works 360 degrees.
I feel when you use a 60 degree front stereo set up that generally everything outside of the front 60 degrees is an error no matter what you do.
Personally i think unless you are getting some sort of a bad sound from the room that making it totally dead or anechoic is a bad idea. You only reduce the efficiency of your amplifiers. I think most room treatment is just changing the sound in an incomplete design that is error filled already. Anything outside of the front 60 is an error by nature.
Last edited:
Did you read that thread or this one? That's essentially what I do (even though I name the other speaker pair ambient speakers). They are not equidistant in my setup, as I cater to more than one listener. But the idea is the same, yet a little different.
Small improvements without evidence are no improvement for me... sorry. I don't do guessing. An improvement is an improvement if I can measure it. I know that's not up everyone's alley but there's no real magic in reproduction for me. The real magic is in the recoded works. I see it as my job to present it well enough.
What you talked about here was a measurable difference. That I can accept. The older recordings can be fun if you have output down low too! It can get scary if they didn't high pass the microphones. I love listening to that kind of thing.
I care about that octave below 40 Hz. It has given me a lot of pleasure to have that. I tried with and without and prefer to listen with that first octave.
Small improvements without evidence are no improvement for me... sorry. I don't do guessing. An improvement is an improvement if I can measure it. I know that's not up everyone's alley but there's no real magic in reproduction for me. The real magic is in the recoded works. I see it as my job to present it well enough.
What you talked about here was a measurable difference. That I can accept. The older recordings can be fun if you have output down low too! It can get scary if they didn't high pass the microphones
. I love listening to that kind of thing.I care about that octave below 40 Hz. It has given me a lot of pleasure to have that. I tried with and without and prefer to listen with that first octave.
Last edited:
Generally mics pic up a lot more reflections than your brain with the wrong matrix in place. Maybe a binaural head would be ok i am not sure. But monitoring on headphones usually reveals more ambiance than using just 2 speakers. Using 4 speakers with stereo can bring the ambiance up to par with the headphones.
The problem with what you propose is that you give no 1:1 layout for stereo sound and mics in terms of a playback matrix. Your illusion you create will most likely be flawed and not a very good illusion making the reflections come out of the mics and speakers much more than they do in a room with ears and reflections.
I think what i am catching is in the mastering phase. It's not always clear if they just do it on purpose to hype up the high end or if the high end is just a tad weak in the past due to weak dispersion of their tweeters. I think as you make the tweeter more omni it perceptually fills in the high end more and more maybe giving the illusion of a louder response in sensitive areas like 7kHz. 7kHz is strange because if you look at a perceptual curve at that area noise goes down 3dB or so while tone goes up 3dB or so. So an engineer might bring up the cymbals which are all noise to get them flat around 0dB while at the same time boosting the tone an extra 3dB. I think this may be the problem i hear sometimes. Taming high end some minuscule amount like 1.5dB actually goes a long way sometimes even though it should be pretty much impercievable.
No, you just missed the point I was trying to make. This test only shows you what our brain is capable of. Nothing more, nothing less.
If you do use binaural mics, your brain could get to work to do what it does at the listening spot. Making it as pleasing on headphones as it is in real life. Then you still don't know what your room adds.
But that wasn't the point. The wave front at your listening spot out in a room just isn't the same as the one embedded in the recording. It isn't a test that is supposed to make a good illusion possible, that wouldn't teach you the sound of the room.
Even with your 4 speakers you could do it. Just play back the recording over those 4 speakers. Double anything that's happening out in the room.
Be warned though. A lot easier to hear than it is to lose that impression
.
Last edited:
Faults like these could very well come form the cross talk comb pattern that was present in the setup used for mixing/mastering.
Of course we can pick up things like this, especially when you're trained to listen. Some people like it like that though. People often like the un-EQed sound of a full range speaker that would drive me nuts.
1.5 dB being impercievable? Not that far fetched for me that we can hear such a difference. I've been amazed about what I can hear after cleaning up the early response.
That's the whole point.
Now play it back on the same speakers. This only shows what our brain filters out. We do it every day wherever we go. I make it a sport to listen for rooms wherever I go. You can train yourself to hear the room.
You might not want to know but I do. Because I build speakers I want to know how we listen. Don't make any more out of it. Do it in a studio and you'll hear less basketball court, I promise. It's also the reason why that's the better place to judge sound. Not necessarily the most enjoyable place.
One might prefer an actual room, not a studio.
Last edited:
I guess i was missing your point. We are actually both on the same page. For some reason i was thinking that you were saying that the sounds the mic picks up are reasons to treat the room.
I think our brain uses all sorts of error corrections to make sounds more intelligible. Any of the error correction that we can come up with in math and technology might be already in place in the human brain due to possibly billions of years of trial and error with perception. So something like a pre loaded feed forward error correction or predictive error correction could be happening with the brain that our reproduction may be lacking.
What i guess i was saying is that if you start to approach accuracy with your matrix stereophony that these incidental reflections may start to disappear again because you are now getting closer to an accurate illusion or projection of natural sound.
I once helped someone translate a wave field synthesis thesis on this board from German. I think one of the things he said that was missing from recordings vs wave field synthesis is copies of the first sound. That if you want to add realism then add copies. I think adding these copies or reflections in the right way we may start to blank out the virtual recorded room in the same way we blank out a normal room.
I think our brain uses all sorts of error corrections to make sounds more intelligible. Any of the error correction that we can come up with in math and technology might be already in place in the human brain due to possibly billions of years of trial and error with perception. So something like a pre loaded feed forward error correction or predictive error correction could be happening with the brain that our reproduction may be lacking.
What i guess i was saying is that if you start to approach accuracy with your matrix stereophony that these incidental reflections may start to disappear again because you are now getting closer to an accurate illusion or projection of natural sound.
I once helped someone translate a wave field synthesis thesis on this board from German. I think one of the things he said that was missing from recordings vs wave field synthesis is copies of the first sound. That if you want to add realism then add copies. I think adding these copies or reflections in the right way we may start to blank out the virtual recorded room in the same way we blank out a normal room.
Last edited:
Any link to those speakers? And the room they were in? That's a pretty bold claim to make. One I personally do not buy without substantial evidence.
If you "get" how we listen and what/when we hear and know how the brain interprets it, you can make every piece of music more entertaining in that process. This may include deliberately deviating from a flat frequency response and things like that.
I've used measurements and psycho acoustic principles to measure my way to better sound. Following the works of many of the Guru's and keeping what they said about it in mind. Guess what: I may not always agree with mixing decisions but the music doesn't sound awful all of a sudden. On the contrary.
I see claims like this in audiophile circles as well. I don't buy it without substantial proof. What better way too boost your own ego
"Yeah the listening experience sucks, the music sounds terrible. It isn't my speakers though, really... it's just those professional mixing and mastering engineers that don't use my speakers, they just can't hear it properly."
I'm convinced if you keep in mind how we listen, you can improve the listening experience of most of the music out there. Not by looking at speakers alone, but by keeping in mind the room and speakers. I refuse not to be able to listen to my music. It's the sole reason for making the speakers in the first place! I love the music I've acquired over the years. My speakers measure pretty good though. It hasn't become a problem.
One experience I have had is similar to this in terms of what I heard but not the reasons for it.
After I completed the LX521 I started to listen to my test tracks and noticed something weird with the track 'Strong' by London Grammar. I kept hearing this bit of crackly distortion that sounded terrible at particular points in the track. I had never noticed it before when listening to any other system including a pretty good pair of AKG headphones.
I was convinced at first that it was my amplifiers but when I knew where to listen for it in the track I could hear it on the headphones but it didn't sound quite as objectionable. I analysed the track with Audacity and it was actually clipped at those points.
I had read something similar on Linkwitz site that he had noticed when building the Orion with a rear tweeter. It seems that a full range Dipole right up to the treble regions can in some cases reveal flaws in the recording that are not apparent on other systems.
Now I doubt this is the case for any audiphool type of explanation so I don't think it invalidates your point, but in certain circumstances the same sort of thing can happen.
I can't say whether the full range dipole is being more accurate or more real but I now find it hard to listen to that track anymore!
I agree with a lot of Wesayso’s opinions. If one manages to lower the ampitude of early reflections to inaudible and later (15-30 ms or so) get some strong reflections back from the room, one will hear more of what was acually recorded. Direct sound, followed by 0-30 ms of silence /inaudible reflections and then audible reflections, fools your brain of you having a much larger room than you actually have. A more spacious sound which is nice, while at the same time more revealing of what was recorded with no sound smearing reflections. I would spend most effort on getting the first 5 ms reflection free. (For some recordings, this may actually narrow the soundstage to appear closer to or within speaker width, not widen it to go inside the next door room. A matter of prefence I would say. A wider sound stage may sound more preferable but be less accurate. Like comparing a pleasant /acceptable listening room to a more accurate sounding control room.)
https://www.gearslutz.com/board/stu...diffusor-vertical-horizontal.html#post8295482 is a link to another forum on this subject. The mentioned ISD gap stands for Initial Signal Delay, the time in ms between direct sound and first audible room reflection. There are another 4 links in that one, which are also well worth reading. In no particaular order, I would pay more attention to posts from JhBrandt, Jens Eklund, SAC (mas), Boggy, Gullfo, Localhost127 and Avare.
In my previuos setup I managed to get -20 dB and lower for early reflections and then some strong lateral reflections back around 28 ms and a smooth decay afterwards. Quite an ”ear opener”!
https://www.gearslutz.com/board/stu...diffusor-vertical-horizontal.html#post8295482 is a link to another forum on this subject. The mentioned ISD gap stands for Initial Signal Delay, the time in ms between direct sound and first audible room reflection. There are another 4 links in that one, which are also well worth reading. In no particaular order, I would pay more attention to posts from JhBrandt, Jens Eklund, SAC (mas), Boggy, Gullfo, Localhost127 and Avare.
In my previuos setup I managed to get -20 dB and lower for early reflections and then some strong lateral reflections back around 28 ms and a smooth decay afterwards. Quite an ”ear opener”!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.