Just for the record, there was a reason why cabinets like this used to be a thing, but are now essentially obsolete. That reason is amplifier power. A horn is efficient for power, but inefficient in total volume.
I can understand it as a science experiement or a passion project. Or maybe you want to achieve THX reference level with a tube amp? But it is sooo big! I mean, you could divide that cabinet into like four 18" subwoofers!
The physics of how a horn works are fascinating, but, well, you've been warned!
I can understand it as a science experiement or a passion project. Or maybe you want to achieve THX reference level with a tube amp? But it is sooo big! I mean, you could divide that cabinet into like four 18" subwoofers!
The physics of how a horn works are fascinating, but, well, you've been warned!
So the spiky frequency response and ringing of the FLH is OK, but a delay equivalent to a 1/128th note at 180BPM is...not for you
Last edited:
So the spiky frequency response and ringing of the FLH is OK, but a delay equivalent to a 1/128th note at 180BPM is...not for you
To support your point, according to Earl Geddes, the point at which we can begin to detect a delay at sub 100hz frequencies is 50ms. And he is rumored to know a thing or two about audio.
We can detect very small delays at higher frequencies, but not low frequencies.
Lol I never said that FR was good, but I know for certain, unlike some, a 10ms is 100% intolerable, on my mouse nor my audio monitoring system, and for the skeptics, its always been detectable by myself, without a doubt, and I recently tried a 10ms on subwoofer, to remind myself of what it sounds like, and it was a big nope 👎 I can't expect everyone to be able to match the level of ear training that I've had on timing as a drummer with 20+ years of experience. I hate saying that cause it sounds like bragging but how else do you explain it?View attachment 1301368
So the spiky frequency response and ringing of the FLH is OK, but a delay equivalent to a 1/128th note at 180BPM is...not for you
Those are generalizations just like the equal loudness contour.
I've had the privilege to talk to Dr. Geddes on a daily basis for the last 4-5 years, and yes he is a technical power house. I also just recently went over group delay, in my thread. I set up an all pass filter as well did experiments within my DAW. 50ms was detectable at down to 20hz for myself. Its a matter of testing what to listen for but frankly 50ms is a huge delay whether time to peak energy or literal delay and if you can't detect it you just have bad ears. Testing with tones and headphones makes the effect more vulnerable. Detecting in the grey areas with music is likely impossible, never the less A delay, which is signal delay, not group delay, is detectable, with music, at 10ms, for myself, crossing at 200hz.
@weltersys I recently have grasped the concept of ringing vs FR and I keep forgetting... thanks for the reminder
Last edited:
I tried to the test of delay, but honestly I don't remember what I did before.... and now I am wondering did I even have a low pass on when I did it... I have been experimenting for a while so its not like I am just pulling these ideas out of my butt, but, I do make mistakes and confuse things time to time so lets do what you all should do... Stop taking perception advice at face value and do the test yourself. Before we jump into it, lets remember that whats detectable with test tones, isn't always detectable with music. This was the case with the group delay issues of my horn. Where to draw the line is up to you, as in, what's acceptable for your expectations of accuracy. I could argue that several different FR's wouldn't be detectable with music unless the areas of speculation are pointed to some how or direct compare and contrast are presented but non the less, one may seek to achieve a certain level of smoothness in FR for the sake of accuracy, so please don't shoot me.
I made this (attachment) for you guys to test yourself, every 8 beats I add 10ms until I hit 50ms.... There is a linear phase, steep, crossing at 20hz that is -6db at 200hz... I realize this isn't a perfect crossing but this is what I got. What I want you to keep in mind is that there are musical situations where the kick might be slightly delay for whatever reason, natural drumming error, or the instrument, like electronic 808's might inherently have its lower spectrum material delayed happenstantially. Likening to GD peaks of a horns raw response, which might be just below perceptibility, a steep cross over only causes the group delay to worsen. With the same set up, a signal delay in your subwoofer will push something that might of been just below the realm of perception, into, perception, regardless of why. I have an ear for timing, so I'm probably anal about the situation but to each its own, no love lost.
It is less noticeable as testing in the past, which to me, says, the results can be source dependent, but non the less I can tell the blur in transient quality at 10ms... by 30ms, anyone should be to tell...
Comment if youd like
I made this (attachment) for you guys to test yourself, every 8 beats I add 10ms until I hit 50ms.... There is a linear phase, steep, crossing at 20hz that is -6db at 200hz... I realize this isn't a perfect crossing but this is what I got. What I want you to keep in mind is that there are musical situations where the kick might be slightly delay for whatever reason, natural drumming error, or the instrument, like electronic 808's might inherently have its lower spectrum material delayed happenstantially. Likening to GD peaks of a horns raw response, which might be just below perceptibility, a steep cross over only causes the group delay to worsen. With the same set up, a signal delay in your subwoofer will push something that might of been just below the realm of perception, into, perception, regardless of why. I have an ear for timing, so I'm probably anal about the situation but to each its own, no love lost.
It is less noticeable as testing in the past, which to me, says, the results can be source dependent, but non the less I can tell the blur in transient quality at 10ms... by 30ms, anyone should be to tell...
Comment if youd like
Then maybe you want to ask him about his comments in this video? The context is low frequency subwoofer optimization, and here he says both 50ms and "somewhere between 30-50ms" is where you can sense a delay is present.
(sorry, had wrong link first...)
I like the way Floyd Toole debunks the idea of "golden ears". People can only maintain objectivity about sound quality until they can see the speaker or can identify it in some way. When they can identifiy it in some way, their preconceived biases sway their perception.
This conversation is about the power of suggestion. The suggestion is that delay = bad. And because delay = bad, anyone who takes pride in their listening ability must hear this delay and know it is an unacceptable bad problem. It's like The Emperor's New Clothes; only a fool can't see them!
If you haven't noticed, I'm not very susceptible to suggestion. So if you want to convince me that delays are bad, please explain to me how my brain is going to differentiate between 1) the direct signal of a delayed subwoofer, 2) the direct signal of another subwoofer with no delay, and 3) the multiple out of phase reflections from the same 2 subwoofers. Small room acoustics are a mess, and what matters is peak energy time, not whether different drivers in different locations are actuating at the exact same moment.
To be honest I'm not interested in debating the topic... if it sounds good to you, then its good enough. How can you miss what you cannot detect? You'd argue that a person whos spent decades in sound design and live music has the same listening ability as Joe Schmo. To me, from a psychological standpoint, that is illogical. I honestly believe you can't hear it. I have enough tools to experiment with and decide for myself.
Did you listen to the test file I created? When were you able to discern a difference?
"30-50ms is where you can sense a delay is present."
Vs
"by 30ms, anyone should be to tell."
What a coincidence that I came to a similar conclusion. Just because I say I can perceive a difference at 10ms... why are you at odds? I once read the smallest line during a physical, imagine the doctor saying "you can't do that, studies say you can't" 🤣 I remember once in highschool the teacher had the CRT TV on with a blank screen but he didn't know it was still on and it was time to take a test and the whole room was quiet and all I could hear was "eeeeeeeeee"... I asked him to turn it off because it was distracting and he told me it was off and to get to work kind of scolding me... after begging him again later he finally reached up and pressed the button and turned it off he looked at me like you could hear that? No one else in class could hear it btw.... you'd just be a peer saying "I don't hear it thus you can't"
After a while you get tired of people telling you what you can and can't do... I could probably live with a 10ms delay on a subwoofer, the difference between 0ms delay is negligible but Im not going to pretend that with an A and B comparison I can't perceive a difference, just to make you feel better lol
Did you listen to the test file I created? When were you able to discern a difference?
"30-50ms is where you can sense a delay is present."
Vs
"by 30ms, anyone should be to tell."
What a coincidence that I came to a similar conclusion. Just because I say I can perceive a difference at 10ms... why are you at odds? I once read the smallest line during a physical, imagine the doctor saying "you can't do that, studies say you can't" 🤣 I remember once in highschool the teacher had the CRT TV on with a blank screen but he didn't know it was still on and it was time to take a test and the whole room was quiet and all I could hear was "eeeeeeeeee"... I asked him to turn it off because it was distracting and he told me it was off and to get to work kind of scolding me... after begging him again later he finally reached up and pressed the button and turned it off he looked at me like you could hear that? No one else in class could hear it btw.... you'd just be a peer saying "I don't hear it thus you can't"
After a while you get tired of people telling you what you can and can't do... I could probably live with a 10ms delay on a subwoofer, the difference between 0ms delay is negligible but Im not going to pretend that with an A and B comparison I can't perceive a difference, just to make you feel better lol
Last edited:
I feel the same way about information, people can only remain objective until they can identify the source lol, and once they do, their preconceived biases sway their perception.
I'm using a similar sized box for 2 18's. $ 18's in a slightly bigger box would tickle my fancy but I have different views on the why factor. Efficiency is a matter of radiation mass size. 4 18's in a similar sized box as the horn will be as electrically efficient, if not more.... but cost efficiency per output, not so much.
Camplo, have you done with any work with tactile? Let's say you have your feet up on the cab or leaning back on it, what is the implication for that with delay? Does it stand out more? I understand what you say about the drumming, I am like-minded about complex and speedy things I arrange with FL Studio sequencer and jam with it with fast electric bass playing dancing around the kick
I'm not sure. When you are touching the cab you are already really close to the source. I would think that you would be able to be more critical of the source, more than anything. As a whole, being really close to the source allows you to drown out reverb. Most of my listening is in headphones or with 1 meter so I am accustomed to direct sound. All I am saying about being a drummer whos into loudspeaker design is that I've spent a large amount of hours, years, studying the distance between transients... that is literally what drumming is all about. Why should someone who hasn't put in that work, have ears trained to hear what I am listening for. At the the end of day we don't have to spend all discussing it because really, what you can't hear, can't hurt you lol.... I am not listening for amplifiers and speaker cables. If you have to try really hard to hear it, is it really worth addressing is the truth. I think I fall into the aim for technically correct, regardless, group. That is until it because inconvenient for my design wishes lol! Then I am looking to figure how far I can move away from technical perfect without serious detriment. Transient issues? I am automatically searching for flaw, that's what drummers do.
Your HF FLH ( High Frequency Front Loaded Horn) has ~2ms time of flight delay, to align with a LF FLH with a 10ms time of flight would require delaying the HF by ~8ms.
If not time aligned, the frequency response may suffer depending on the constructive and destructive phase relationship between the two. The Haas (precedence) effect also tends to make the leading wavefront seem to sound louder, requiring more LF SPL to be perceived as loud, adding to LF room reverberation "mud". Simple and cheap to correct the problem with DSP.
Your attachment with additional delay to the LF (low frequency) every 8 beats, adding 10ms up to 50ms demonstrates a timing lag between low and high frequencies.
At the extreme of 50ms, most should begin to perceive the latency, it's longer than the slap-back echo popularized by Sun Studios
.If clicks were used alone, perception of latency, rather than a single fused sound can drop to as little as 1-5ms.
An acoustic drum set puts a time of flight acoustic delay of ~4ms between the kick drum's acoustic output and your ear, and each drum or cymbal may vary in distance from that by a few ms.
Unless you are playing along, any delay is imperceptible-no one can tell the difference between something recorded/reproduced 2ms, 2 minutes, two days, or two years ago.
Electric guitar and bass players learn to play with typical delays between them and their speakers of 4ms to 25ms.
Orchestral performers often are separated by greater distances, but still manage to play in time with each other.
That said, studio musicians that want the most precise and instantaneous monitoring have preferred to use an analog audio path with headphones, the time lag between playing and hearing can be dropped to a fraction of a millisecond.
Art
Exactly. There would be a grey area between clicks and music, 1-5ms to 30ms. Ones ability to dissect music into separate spectral events would play a huge factor on perception is all. Then it turns into not if you can, but how can you. For example, If you gave me the snare drum sample I made, seperated, into unlabelled segmented files, of 0ms, 10ms, 20ms, 30ms, 40ms , and 50ms, I am very confident that I could reordered them from lowest to highest latency, repeatedly. In a listening session of multiple identical systems playing music, at each interval, it will be a struggle between 0ms and 10ms, and possibly 20ms, but as long as I can A/B them, I could probably make the right choice to some degree. Am I going to walk up to a stereo and be like, theres a 10ms delay on your subwoofer, unlikely. I prefer to aim for transient perfection, out of technicality. Technically speaking this is direable.Your attachment with additional delay to the LF (low frequency) every 8 beats, adding 10ms up to 50ms demonstrates a timing lag between low and high frequencies.
At the extreme of 50ms, most should begin to perceive the latency, it's longer than the slap-back echo popularized by Sun Studios
If clicks were used alone, perception of latency, rather than a single fused sound can drop to as little as 1-5ms.
Thank you for the detailed response, it is enlightening
"Transient perfection" has nothing to do with the length of the horn, and everything to do with time/phase alignment.