Lets imagine 2/3way speaker with 25-30mm soft dome tweeter. Most of soft domes does not perform very well 10kHz+ offaxis. We can control directivity with the waveguide, but 10 kHz+ we dont have enough energy offaxis. There are some harddome tweeters with good behavior, but they are expensive.
My idea is - what if I will add two small 16-19mm tweeters (cheap ones) aiming to the sides (or rear/up/down...) and cross them at 10-12 kHz to make directivity in that region as wide as possible (add energy to reflections)? If done well, it will have little influance on the direct sound.
How it will sound like? Is it beneficial?
Can we localize direction 10kHz+? What if those extra tweeters will not be nearby main tweeter?
Absorption in typical room is usually high at those frequencies - will we hear the difference?
My idea is - what if I will add two small 16-19mm tweeters (cheap ones) aiming to the sides (or rear/up/down...) and cross them at 10-12 kHz to make directivity in that region as wide as possible (add energy to reflections)? If done well, it will have little influance on the direct sound.
How it will sound like? Is it beneficial?
Can we localize direction 10kHz+? What if those extra tweeters will not be nearby main tweeter?
Absorption in typical room is usually high at those frequencies - will we hear the difference?
I'd encourage you to design and perform such experiments in general, for one or more specific auditory requirements such as dispersion, sense of air, etc. I have thought about it but not done it with tweeters. Bipoles and some dipoles do it full-frequency. Up-firing "omni" fullrange drivers with deflectors or waveguides or just too-peaky highs (Harmon Kardon Citataion 360 using Lowther PM6). They all add or redistribute sound energy including HF and ultrasonic, their effects highly dependent on room boundaries/treatment/etc. i.e. complex.
Looking back (to Covid Era), my dozens of diy experiments have all dealt with HF in some way, usually unorthodox; the Full Range Forum Photo Gallery has ~20 of them spread over the last ten pages (200 msgs). Recently, I "solved" a long-standing 15" near-field (dubbed Axia) by bouncing a tweeter off of the convex dustcap, thus achieving wide and near-uniform HF dispersion, in addition to time and phase alignment i.e. virtual point-source by reflection. You might want to try this trick, possibly with distributed tweeters.
Anyway this may lead to a new technique of HF management and audio science, if people with more experience/knowledge and DSP/measurement tools were to take it up, eventually. (I'm gathering a variety of candidate drivers to continue diy experiments, and thinking about experimental set-up.)
https://www.diyaudio.com/community/...ials-with-ragged-response.408887/post-7783316
Videoclips given in my Audio-lensing thread.
https://www.diyaudio.com/community/...on-theory-and-experiments.417695/post-7794302
Looking back (to Covid Era), my dozens of diy experiments have all dealt with HF in some way, usually unorthodox; the Full Range Forum Photo Gallery has ~20 of them spread over the last ten pages (200 msgs). Recently, I "solved" a long-standing 15" near-field (dubbed Axia) by bouncing a tweeter off of the convex dustcap, thus achieving wide and near-uniform HF dispersion, in addition to time and phase alignment i.e. virtual point-source by reflection. You might want to try this trick, possibly with distributed tweeters.
Anyway this may lead to a new technique of HF management and audio science, if people with more experience/knowledge and DSP/measurement tools were to take it up, eventually. (I'm gathering a variety of candidate drivers to continue diy experiments, and thinking about experimental set-up.)
https://www.diyaudio.com/community/...ials-with-ragged-response.408887/post-7783316
Videoclips given in my Audio-lensing thread.
https://www.diyaudio.com/community/...on-theory-and-experiments.417695/post-7794302
Last edited:
So in the end I did not test setup above, but I have another idea I would like to test. One of the reasons, why I wanted to extend top octave was to get bigger soundstage by adding echo sound field at highest frequencies and have better soundstage.
New idea:
To make this work, it is needed to have reasonable delay between direct sound and echoes. Lets say, that one of the reasons, why it is beneficial to have speakers like 2m from rear wall is (aside of sbir elimination), that you have reflection from the rear wall, which is delayd like 15-25ms and this gives our brain idea about space size and creates sound stage depth. This is missing with onwall/inwall speakers. So my idea is, if we can recreate it.
How?
Each speaker will have extra midtweeter section (like 1000Hz+) with high directivity. This secondary speaker will be placed 2m from the front wall, aimed to front wall. Thanks to its directivity it will not play in listener direction and only create difuze sound field via reflection from the front wall with desired delay. This secondary speaker will be 6-10db attenuated and with slight decrease above 10-15kHz (to fit typical room rt60 behavior). Secondary speakers will be paralel to mains.
What I need?
Reasonable price horn (CD?) with very narrow angles (40x40°?) working to 1kHz. Any tips (available in EU)?
What do you thing? Does it worth the effort? Should be beneficial in all onwall/inwall and cardioid designs.
New idea:
To make this work, it is needed to have reasonable delay between direct sound and echoes. Lets say, that one of the reasons, why it is beneficial to have speakers like 2m from rear wall is (aside of sbir elimination), that you have reflection from the rear wall, which is delayd like 15-25ms and this gives our brain idea about space size and creates sound stage depth. This is missing with onwall/inwall speakers. So my idea is, if we can recreate it.
How?
Each speaker will have extra midtweeter section (like 1000Hz+) with high directivity. This secondary speaker will be placed 2m from the front wall, aimed to front wall. Thanks to its directivity it will not play in listener direction and only create difuze sound field via reflection from the front wall with desired delay. This secondary speaker will be 6-10db attenuated and with slight decrease above 10-15kHz (to fit typical room rt60 behavior). Secondary speakers will be paralel to mains.
What I need?
Reasonable price horn (CD?) with very narrow angles (40x40°?) working to 1kHz. Any tips (available in EU)?
What do you thing? Does it worth the effort? Should be beneficial in all onwall/inwall and cardioid designs.
You can easily experiment with multiple tweeters using the Visaton Boxsim software. It allows you to put drivers on any side of the cabinet and allows for adding delays and active crossovers for quick modeling. I made a small cube and put drivers on the sides and front. To avoid comb filtering effects the tweeters need to be practically co-located.
The tiny dome by Peerless is a good candidate. These can be mounted back to back.
https://www.parts-express.com/Peerless-OC25SC65-04-1-Textile-Dome-Tweeter-264-1018?quantity=1
I tested two of these Dayton tweeters side by side at 60 to 90 degree angle and had some pretty good results.
The tiny dome by Peerless is a good candidate. These can be mounted back to back.
https://www.parts-express.com/Peerless-OC25SC65-04-1-Textile-Dome-Tweeter-264-1018?quantity=1
I tested two of these Dayton tweeters side by side at 60 to 90 degree angle and had some pretty good results.
a single extra tweeter on backside and aimed 45deg upwards might be enough. Same highpass as front tweeter. But speaker distance to walls and wall's reflectiveness varies. And listening impression is here the most important criterion, so experiments will tell. Simulations are not worth doing IMO. Measurement of csd (decay) at spot can show differencies, also spl graph far from spot "kitchen"
Be careful with net impedance! Without dsp, setting gain/spl right might be difficult...
Be careful with net impedance! Without dsp, setting gain/spl right might be difficult...
I had pretty good sounding results with these guys, though a DSP delay worked best. I wouldn't worry too much about needing high directivity speakers for the ambience if you use DSP delay, the idea here isn't to recreate the waveform of a recording (which makes no real sense, anyway, as any real room will completely prevent that possibility anyway, even without adding ambience tweeters).
https://www.diyaudio.com/community/threads/ambience-tweeters-using-small-bmr-drivers.326055/
https://www.diyaudio.com/community/threads/ambience-tweeters-using-small-bmr-drivers.326055/
It will be part of home theater setup, so extra dsp and amp is complicated, but I really like your results 😉 your setting is surprisingly close to my idea. DSP way solves possible issues /negative effect of direct sound from ambient speakers to lisener. On the other hand - which solution will create better difuse sound field?
um, kind of worth thinking about whether that rolloff isn't taken into account in the music mastering process. I mean, yes, tweeters perform flatter on axis but is that even how most speakers are intended to be listened to? And how much energy is really in that last half octave anyway? It's generally not what makes a speaker sound better or not, it's the dispersion everywhere else that seems to make the most importance.
Of course, always fun to experiment with a variety of designs. 🙂
Of course, always fun to experiment with a variety of designs. 🙂