Hmm, in retrospect I can see how it could be interpreted this way, though my point/question was asking how compressed were your recordings due to this abnormally low 80 dB/SQ requirement which would include real world transients of up to at least 30+ dB [clipped!] in some symphonic recordings.
IOW what's krivium's crest factor?
A really valid question these days as so many modern recordings I listen to both OTA and on the net are apparently compressed to virtually flat for whatever reason, or so it seems based on how little power is required to play louder than most folks can/will tolerate in a HIFI/HT app.
As for the rest, not being able to keep up in much detail ATM can only reiterate that the lower the frequency the wider its BW and as the desired range is shifted upward the requisite HF corner will in theory increase to well above the 22 kHz? upper recording limit based on a 16*22,000 = 352,000 reference, so virtually all recordings will theoretically be clipped from the get-go, though no clue how this is handled fro an SQ POV.
So again; if we use Dolby's 105 dB peaks for the mains, CC, surrounds and 115 dB peak/ for < ~120 Hz, all at an 85 dB avg. and the listening position [Lp] and its attendant ~ 6 dB/octave roll-off/doubling of distance, how much dynamic headroom does your studio system require because TTBOMK this is your SQ Xmax limited reference goal nowadays.
Once again if you look at pink noise you'll see that these 30db peaks are only in reference to where the average is....The average spl is never going to be 115db at 300hz-20khz.
80db is just a good level to work at. A propose 76-77db, by some, to be the best level to work at. I have my own theories as we all do of course...theres a bit of art to it.
This is to show the headroom of the AXi+Horn, potentially. 300xo on the far right graph. More than enough. I think the Axi has 2mm peak to peak xmax??
Is there any evidence that an isobaric midrange performs more like an open baffle mid...blocking out cab resonances...I've been reading on the internets again....
80db is just a good level to work at. A propose 76-77db, by some, to be the best level to work at. I have my own theories as we all do of course...theres a bit of art to it.
This is to show the headroom of the AXi+Horn, potentially. 300xo on the far right graph. More than enough. I think the Axi has 2mm peak to peak xmax??
Is there any evidence that an isobaric midrange performs more like an open baffle mid...blocking out cab resonances...I've been reading on the internets again....
Attachments
Last edited:
....and as I keep pointing out, the easiest/cheapest way to do it is to make a bigger baffle!
Actually, best to put them in wall or at least a big enough false wall baffle.
This allows a typical 500 - 800 Hz XO with only an 8" woofer with decent Xmax or 12" at most to match up with a 120 Hz XO, which with the Axidriver might allow a 1st order XO.
OK, then good to go, just forgot to refresh to see later postings.
I don't see how as the drivers are in a sealed or vented cab and if on an OB, then the drivers need to be sealed and it's my understanding that much of its isobaric properties are lost unless it's internally baffled to keep its internal volume [Vb] as small as practical.
That said, I've seen some done clam shell mounted with/without a baffle with a simple suspended phase plug to reduce it.
No, the Axi2050 has about 0.8mm Xmax, 1.6mm peak to peak at the voice coil location near the center of the ring radiator. Since the outer perimeters of the annular ring diaphragm are fixed in place, the average is probably little more than a standard 4" diaphragm with 0.5 mm Xmax.
The Axi2050 used on a horn that loads well down to 300 Hz should be able to produce 115 dB at 1meter without the diaphragm hammering the phase plug, but a single 15 could easily have 10dB more headroom available at that frequency.
I still think using your 15" drivers on a multiple entrance horn with the Axi2050 is a far better concept for your intended near/far use than a separate front-loaded 15" cabinet- you can be on axis to the MEH regardless of the distance you are listening at.
Chris A certainly has found that to be true.
Art
An insightful and practical solution to the available drivers and intended use, it'll never catch on
So 0.8 mm is confirmed Vs this response's stated 0.5 mm?: New Celestion "AxiPeriodic Driver"
An insightful and practical solution to the available drivers and intended use, it'll never catch on
Yea. I can't imagine why a large, powerful, accurate, point source would work - snarc
How about Dick Burwen's approach for an alternative?
20,000 Watt Home Hi-Fi System
Horns in the corners...
Mine i know: i'm between 83 and 85 dbspl at listening point and i allowed 20db for the few records i have / projects i worked on which have a 'wide' dynamic range.
But most of what i listen to is well under 12db crest factor ( even some of my vinyl!).
I thought i was already well on the darkside but after listening to some of the reference Camplo gave i realize i like music which 'breathe' a bit regarding dynamic... and i'm now a dinosaur wrt what kids like!
The Axi2050 Xmech, contact of the center of the diaphragm with the phase plug is 1mm. 0.8mm would leave a bit of headroom before the crash
As Dr Jack Oclee-Brown mentioned at about 55:38 in the video describing the Axi2050:
PART 2: Wideband Compression Driver Design - Dr Jack Oclee-Brown - YouTube
“We’ve got a 1mm gap between the front of the diaphragm and the phase plug at it’s widest point which comes down to almost nothing where it’s clamped”.
He also mentions the displacement of the annular ring AXi2050 is about the equivalent to that of a 4” dome.
The AXi2050 has less low frequency (below 600 Hz) second harmonic distortion from mechanical limiting of the suspension the (presumably BMS) dual diaphragm driver he compared it to has.
As he also mentioned, “it’s not so good at high frequencies..”
The AXi2050 lack of mechanical limiting should also allow the diaphragm to hammer against the phase plug with less voltage applied, but also with less harmonic distortion before the mechanical limit is reached.
There is no "R" in Vituix
I think you are describing a gain block which uses the triangle symbol, this is similar but not the same as an operational amplifier.
You can use the gain block to change the gain at the point where it is inserted and also delay. The A is the change in gain in dB's positive or negative.
During this project I have made happy and sad mistakes. My latest happy mistake was thinking that a ka=2 had a more narrow beam width based off of the graph used in another study on directivity that used different scaling (ka=2 looks like about -4db at 90 degrees?). So the new revelation is just in time because I may have a sad mistake regarding how I can actually manipulate the signal via eq and gain vs what I thought I was about to do.
It appears the trouble area created by the slot isn't playing nicely. Its not so much my ability to create a flat response with filters....I've done so in several configurations, its just that in my ignorance I adjusted gain as if it was freely available, based on the ideas that the spl levels I aimed for were achieved pre filter. EQ filter gain is going to be limited to what, +12db? I'd probably want to leave unity gain alone here. The input or output gain might be enough but I have used numbers like 20 and higher for the +db of the amps in VituixCad. I'm pretty sure that 20 is probably getting towards the cap of gain I have safely available? If I focus on the slot loaded woofers, I have 625-2000 watts rms to use in 16ohms.
More so 625 since I rather be able to use stereo mode instead of bridged. So how much gain should I limit myself while trying to manipulate the low response into something that can useable.
Attachments
Last edited:
ok, when we look at using two drivers, we end up with a repaired FR yet GD is there. "If you "fixed the FR with filters," then you also fixed the GD." - In this situation have we not "fixed" FR? The way I read the above perspectives is that unless its something like diffraction, I can take a filter, bring FR to 0 transfer function and GD should be 0 to compliment. It appears that If I use two group delayed responses to accumulate to the proper FR (crossband) the area is still delayed. Which that makes sense.
"For discrete 3D modes we do generally need "more than EQ" to solve the problems - like multiple subs at LFs with EQ." - So room modes, or boundary related modes, which would include modes related to a vented enclosure and even the boundary related resonances of sealed enclosures. Eq is not going to simply fix GD from these 3d "mechanical" filters.
If the red line is an unprocessed measurement of your slot loaded sealed subwoofers, I’d argue that can’t be fixed with any reasonable amount of filtering and requires a redesign of the enclosure.
Why are you taking ka=2 as your upper cutoff? Do you want an omnidirectional response? If so, why the interest in a horn?
Also remember that a is the equivalent radius of a circular piston having the same surface area. You need to convert the area of the mouth or port from a rectangle if the drivers are not directly radiating.
The plots you’re comparing are likely different because they’re based on different analytical constructs - either a round, rigid piston in infinite baffle, duct (with or without flange and/or baffle), rectangular piston, etc.
If you want to see what EQ can 'fix' just normalise your polar map prediction to the axial response.
Why are you taking ka=2 as your upper cutoff? Do you want an omnidirectional response? If so, why the interest in a horn?
Also remember that a is the equivalent radius of a circular piston having the same surface area. You need to convert the area of the mouth or port from a rectangle if the drivers are not directly radiating.
The plots you’re comparing are likely different because they’re based on different analytical constructs - either a round, rigid piston in infinite baffle, duct (with or without flange and/or baffle), rectangular piston, etc.
If you want to see what EQ can 'fix' just normalise your polar map prediction to the axial response.