I was watching this earlier: Galion TS Voyager
I'm not normally one for LS3/5A type speakers - especially at the prices asked these days - but I thought this was interesting.
Not least - a fully labelled XSim with the drivers and crossover components.
The designer is clearly no idiot, and identifies why he's crossed rather high.
But that must result in a directivity mismatch and like many I suspect I'd have gone for an LR at 2.5-3k.
Any thoughts? Is it contributing to the sound signature described?
I'd have guessed one of the pricier Monacor drivers, but refreshing that its so inexpensive at Dayton, and so few crossover parts. And interesting that he didn't go for a 19mm tweeter.
I'm not normally one for LS3/5A type speakers - especially at the prices asked these days - but I thought this was interesting.
Not least - a fully labelled XSim with the drivers and crossover components.
The designer is clearly no idiot, and identifies why he's crossed rather high.
But that must result in a directivity mismatch and like many I suspect I'd have gone for an LR at 2.5-3k.
Any thoughts? Is it contributing to the sound signature described?
I'd have guessed one of the pricier Monacor drivers, but refreshing that its so inexpensive at Dayton, and so few crossover parts. And interesting that he didn't go for a 19mm tweeter.
For this little 2-Ways speaker, I went for a Fc = 4.5kHz transistion frequency :
My attempts with Fc = 3kHz and 3.5kHz were not satisfactory aurally speaking, despite the fact that the Monacor DT-280 tweeter has a Fs at 1100Hz,while the woofer Monacor SPH-170 reaches 5kHz off-axis (quite close to Fc = 4.5kHz). I listen off-axis at 15°, and everything is fine.
T
My attempts with Fc = 3kHz and 3.5kHz were not satisfactory aurally speaking, despite the fact that the Monacor DT-280 tweeter has a Fs at 1100Hz,while the woofer Monacor SPH-170 reaches 5kHz off-axis (quite close to Fc = 4.5kHz). I listen off-axis at 15°, and everything is fine.
T
How so, off axis?
The radiating area is declared at 137cm^2 here: https://www.scan-speak.dk/datasheet/pdf/18w-4424g00.pdf
Which is the same as this ScanSpeak 18cm midwoofer: https://www.scan-speak.dk/datasheet/pdf/18w-4424g00.pdf
That is certainly not effective off-axis at 5kHz.
You are assuming that it is a major factor. Do you assert that it would be better if the crossover is lower - that the directivity mismatch trumps the phase behaviour through the upper midrange'? I'm inclined to think it needs a 10-12cm driver 3-way to get a good balance on paper - but I've heard 2-ways that clearly can't do it and yet are very pleasant. Most of them are crossing higher than I would think iwould be sensible if the beaming is as important as some might say, and on another thread there is a graph suggesting that in-box the off axis is worse than expected and even a 180cm driver should cross at little more than 1.2kHz.
There‘s enough proof directivity is a major factor. It is pretty logical too, once you start thinking about sound power vs sound pressure wrt listening in normal domestic rooms.
The crossover shows more than 5dB level difference within a 1000Hz range. On axis. No clue what this speaker does off axis. But I’m pretty sure the sound power per frequency plot shows an impressive bump at 4kHz, the same frequency the on axis level peaks.
[Edit) Looked at the baffle. Diffraction sure poses a problem. He possibly tried to avoid it by picking a high crossover frequency. But you can’t get this really right, actually.
The crossover shows more than 5dB level difference within a 1000Hz range. On axis. No clue what this speaker does off axis. But I’m pretty sure the sound power per frequency plot shows an impressive bump at 4kHz, the same frequency the on axis level peaks.
[Edit) Looked at the baffle. Diffraction sure poses a problem. He possibly tried to avoid it by picking a high crossover frequency. But you can’t get this really right, actually.
Last edited:
Are you not assuming now?
In any case, most people that argue this point do it the other way around, that the tweeter should be the one to cover as wide a band as possible.
Another point is the Harsch type crossover which John Kreskovsky IIRC studied in great detail a couple of decades ago, and stopped using it. It is known for having messy vertical lobing behaviour.
Well, that has been my assumption and I've bought tweeters with low resonance and flat responses and traded directivity and power response at high frequencies on the grounds that I'm nearly 60 and figured I can't hear much over mid teens anyway. But I think its worth keeping an open mind. When I sweep with REW I'm always reminded that what I think of as a high pitch is still upper midrange.
Diffraction is interesting. The total sound energy is largely still there, and we get used to room acoustics - so its unlike a driver that measures flat to 5k on-axis but is beaming. Interesting that retro big boxes are somewhat in fashion.
Also interesting that we are told that roundover etc on the internals of baffles and at the cabinet edges makes a difference, when Olsen's 'rounded edges' were very large radius compared to a baffle. If I recall, Dickason suggests that small radius edge treatments don't make much impact.
I have some new old stock Seas coax drivers - is it, for example, worth flush mounting them given that bass is unaffected and the treble is already bouncing around in a horn? I wonder whether I have been suckered into 'Always flush mount drivers to avoid nasty reflections'. It certainly complicates construction for beginner woodworkers.
Some of the most classic speaker designs in the history of audio would exhibit horrible directivity plots.
This is indeed a speaker that's nothing special these days.....but not for the reason you're implying.
Dave.
For a nice piece of furniture that probably sounds Ok and is just $800 retail - that seems quite good to me, to be fair.
Interesting that it is suggested for nearfield, too. I have a pair of powered Behringers on my desk (admittedly somewhat behind my monitors) and they are uncouth - I'd quite like something 'better', even if I have to have a crazy stand system over the screens.
I think these are by one of our locals, XRK971, his thread for them is here: https://www.diyaudio.com/community/threads/vanguard-speaker.388184/
I don't know his thoughts about matching directivity. Maybe he'll stop in here, or you can ask him on his thread for the product...
I don't know his thoughts about matching directivity. Maybe he'll stop in here, or you can ask him on his thread for the product...
Some time ago I tried to combine a regular dome tweeter (no WG) with a 13 cm midwoofer. The goal was to obtain linear SPL and Power Response characteristics. Attempts with a crossover frequency at 2-3kHz were unsuccessful. I was achieving either linear SPL either linear PR, but never both at the same time.
I was somewhat surprised to discover that there is a way to achieve my goal but i had to raise the crossover frequency to 5kHz. The reason turned out to be that it's easier to match drivers directivity at a high frequency than at low.
Following is resultant directivity graph (red) for a low crossover frequency (c.a. 2.5kHz). Midwoofer directivity is blue and the tweeter's is yellow.
Now similar graph for the same drivers but the crossover frequency is high (c.a. 5kHz):
Assume that system SPL is approximately constant. SPL, PR and directivity are described by the formula
PR=SPL-directivity
which implicates that if SPL and PR have to be linear, then directivity must also be linear. Looking at the graphs above, for 2.5 kHz it won't be possible, but for 5 kHz it was achieved to some extent.
Following is the Power Response characteristic achieved (green curve) which was was close to linear. The red curve is an estimation of the PR characteristic if the crossover frequency was 2.5kHz.
Coming from a prosound background in general and Altec in particular, all aspects of polar response was a 'big deal' with the majority of their tech notes devoted to it over the decades, so you'll have to decide how much it matters in your app, keeping in mind that the room dominates to a great extent, so hard to critique others designs, especially ones designed for studio apps.
That said, the pioneers designed/chose simple 1st, 2nd order XOs based on matching driver sizes exponentially, so limited to the driver's theoretical pistonic BW, ergo no TL and/or breakup modes BW XO points to 'muddy up' its polar matching.
I'm sure it's a good speaker. The problem is there's about 6000 different speakers just like it. 🙂
That's what makes it nothing special.
Dave.
I didn't read your link yet (I will), but I'm strongly of the belief that the crossover frequency is a huge decision in any speaker. Really important. One of several key considerations in a design. If it's wrong, it'll remain a flaw until addressed, regardless of however many other good decisions were made. It ties directly into the tweeter parameters, the crossover type, the range it's used in, and significantly impacts phase coherency. When you get into a mid/tweeter crossover situation (woofer and mid too, but that distance is a lot more generous), the distance between their acoustic centers is pretty major decision. At least within the practices I know how to do, they need to be crossed over at a frequency no greater than the length of the distance between the two acoustic centers. The crossover frequency also dictates the slope required, and whole lot more.
IMO it's an art to identify how well materials work together. While measurements are certainly very possible and helpful, but it's a trial and error exercise for most people. Potentially very important, but with some latitude for some margin of error.
IMO it's an art to identify how well materials work together. While measurements are certainly very possible and helpful, but it's a trial and error exercise for most people. Potentially very important, but with some latitude for some margin of error.
Last edited:
It's hard to peddle yet another (amongst thousands) 5+1" bookshelf two-way, so obviously designer has to construct some far-fetched mythology around a regular squirrel coffin.
Looking into the video, they show regular LS3/5A frequency response as a sort of a vague reference (there are dozens subtly different kinds of 3/5A), which is not that linear, but in general is OK.
Next is TS Voyager Bamboo, which I suppose is production version.
I guess 8 dB broad hump centered on 1 kHz fits their definition of "extremely musical". Politely speaking, this speaker is voiced for a certain 'refined' taste, but some folks at ASR would be compelled to call it garbage. There is also the prototype, which seems to be somewhat closer to LS3/5A, including 4 khz notch/bump.
Not sure if this half-hearted attempt of a speaker warrants furter discussion. However, I second @XMechanik in that high crossover point can, with care and deliberation, lead to a good SP/ER and avoid edge diffraction. Doubt such care was taken in the device in question.
Looking into the video, they show regular LS3/5A frequency response as a sort of a vague reference (there are dozens subtly different kinds of 3/5A), which is not that linear, but in general is OK.
Next is TS Voyager Bamboo, which I suppose is production version.
I guess 8 dB broad hump centered on 1 kHz fits their definition of "extremely musical". Politely speaking, this speaker is voiced for a certain 'refined' taste, but some folks at ASR would be compelled to call it garbage. There is also the prototype, which seems to be somewhat closer to LS3/5A, including 4 khz notch/bump.
Not sure if this half-hearted attempt of a speaker warrants furter discussion. However, I second @XMechanik in that high crossover point can, with care and deliberation, lead to a good SP/ER and avoid edge diffraction. Doubt such care was taken in the device in question.
I had a 3 way with SBA Satori drivers. An MR16P 6.5" midrange and TW29D soft dome 29mm tweeter. I subjectively preferred Fc at 2300Hz over a lower 1800Hz. I removed any bias via automated random DSP switching between 10 samples identifying both version correctly 9 times
That is probably because the baffle itself behaves like a waveguide too. Tiny narrow baffles push the transition to half space (think we better could talk about a lobe with considerably smaller exit angle than 180 deg) up in frequency. I guess (but haven’t checked this) this is even more so with small radiating surfaces like tweeters. The directivity for a given tweeter can thus rise faster than that of the midwoofer. Alas you end up with cone misbehavior at 5k, so compromises, compromises…
I pulled my video monitor "somewhat" out in front of my flanking Tannoy 6.5 PBM speakers and the sound was instantly degraded, some might say a "night and day" difference, but with my old deaf ears not quite that much 😉
Putting most any speaker somewhat behind video monitors will make them sound "uncouth", as reflections cause comb filtering and the diffraction makes a wreck of polar patterns, phase and frequency response.
If you want something to sound "better", start by aligning them with the video monitors face, HF driver centered at ear level.
Cheers,
Art