I think the concept should be more or less scalable. Make an exact copy at half the size and you should have similar directivity at double the frequency. If that is true, you should be able to use a 6.5" driver up to about 2 khz.
That's what I'm doing too: a passive crossover and a miniDSP for corrections.
All in all I do think the extended directivity has some merit in terms of sound quality, but I also think Earl is more qualified to say anything about that - although I fear he may be a bit biased. However, I do remember him saying he would like to make a speaker even bigger than the Summa to extend directivity to about 500 hz, but that the impracticality of increased size withheld him. Below the Schroeder frequency it probably doesn't make much of a difference.
Did you say this wrong?
Its not getting "omni", it gets more directional, but no longer cardiod and has a constantly changing (increasing) directivity that the enclosure design has no effect on.
If it were possible to increase the directivty to ever lower frequencies WITHOUT OTHER COMPROMISES then sure, do it! WHy not? It's the compromises that I object to. Even adding a mini-DSP would push my system costs up substantially. Does this give the customer "good value"?
I may be biased, but I did not come by my solutions without looking into all the options. For me, there are no audible problems in the LFs (below 500 Hz) that I find objectionable so I am not out looking for a "solution". I am certainly not one who looks upon the ideal as a goal. I have found practical to be a much better guide to the best overall system performance. Fix the real problems, not the "perceived" ones (those in our minds, not our hearing).
Another good way is to use the free software "holmimpulse". It allows you to keep the input/output stream of your sound card active (see pic of setting). This keeps the latency of the sound card fixed.
By enabling the setting "time zero locked (time alignment)" (see pic of setting), you can see the actual delay between the drivers by the time difference between the peaks of their impulse responses.
Tip: first measure the woofer, and then press the button "use last detected time zero", then remeasure the woofer. The peak of the woofer impulse will now be perfectly at "0 ms". You can now align the tweeter by trying to get the impulse peak also at zero.
Attachments
I think the bad rep of over stuffing with damping materials come form the fact that many damping materials move along with the vibrating air effectily mass loading the cone (and affecting the sound quality). When you use a stiff open material (like open cell reticulated foam) this vibration effect is completely gone, and you can stuff a speaker as full as you want. Unfortunately open cell reticulated foam is still quite expensive for "big" speakers...
This is indeed a good choice for giving the customer good value.
But, many of us DIYers are always looking for further ways to improve our system (otherwise we would not have a hobby anymore...), and many of us also have some sort of active crossover in place (because this allows us to get a good result faster, as designing a good passive crossover is not easy and very time consuming...)
This means that the additional cost is very low, it is fun to do, and because we made a change and see a good measurement (like the polar keyser showed), in our head it will sound way better also J
It was mentioned earlier that I should look at doing this in my speakers. I was commenting on why that is not likely to happen. DIYers do all sorts of things that don't make any sense in a commercial venture. I know this is a DIY site but I was asked to come here and comment, otherwise, I don't come to DIY too much anymore because thats not what I do. I'm getting away from DIY because its not a good business. DIYers are always comparing what you sell to what they can do with tools that are already paid for and their time is free. Thats simply not the real world in business.
I fully understand about the DIY community not being ideal for business.
You would have probably had more commercial succes by making your speakers more flashy and 5 times as expensive while paying for reviews in every hifi magazine around the world.
I feel very sorry that so many attacks are made to good contributors like yourself. Some just do not respect the time and effort that is spend here trying to explain what we do not understand.
I, and I am sure many others, hope that you keep posting here, because we will not learn it from reading the hifi magazines...
Do you know if anyone (by measurments) have demonstrated the LPF properties of various density open cell reticulated foam - or have suggestions to get the same dampening properties that Keyser has achieved?
Hi Wim,
Good hints. I have used the "time zero locked" feature but was not aware of the input/output stream issue.
I tend to find an offset that makes the tweeter phase flat in its midband and then lock that for woofer and mid measurements (assuming the tweeter is nearest). Note that the second pair of phase curves that the OP showed still looked like they had a lot more delay than minimum phase...?
David S.
Try glasswool. It's one of the best damping materials you'll find, and it's cheap to boot!
You may find this worth your while: http://www.bobgolds.com/AbsorptionCoefficients.htm
You may find this worth your while: http://www.bobgolds.com/AbsorptionCoefficients.htm
The main problem is to get the higher densities in small quantities. From 10 to 30 PPI is commonly used in filters for ponds. This can be get in all sorts of shapes and quite inexpensively. I used 30 PPI in my speakers to fill the center, and used basotect to the sides. Pictures can be seen here: http://www.diyaudio.com/forums/multi-way/153199-enclosure-stuffing-26.html#post2323640
If you read to the entire thread posted above (and leave out the BS), some very usefull insigts are shown on the working of a damping material.
It was Earl who suggested using open cell foam, and after reading that thread, it maked very good sense to me.
David,
The sound of the woofer arives earlier than that of the tweeter, while it is usually the other way around. So there is some excess phase visible for the tweeter. Moreover, the graph shows the individual phase curves for both drivers with the crossover. The highpass and lowpass on the tweeter and woofer respectively cause even more phase rotation and make things look messy.
The sound of the woofer arives earlier than that of the tweeter, while it is usually the other way around. So there is some excess phase visible for the tweeter. Moreover, the graph shows the individual phase curves for both drivers with the crossover. The highpass and lowpass on the tweeter and woofer respectively cause even more phase rotation and make things look messy.
In my previous speaker project I just glasswool. It is indeed also very good, but a pain to work with when you want to experiment alot (once in, it should stay in). As explained, after reading the thread about stuffing (see previous post), I was convinced that open cell foam was even better (although probably not woth the money cost/performance wise).
But as Earl correctly stated, some DIYers (read: me) are always looking for the ideal, and thats why I had to try open cell foam. I anyone knows a good source for 60 to 80 PPI open cell foam (affordable), I am willing to by some and do some measurements on it and post them here.
Last edited:
Probably getting a little off topic but it isn't clear what you are attempting to optimize while stuffing? I can think of two parameters that matter to me. One would be to track f3 (or fs) while varying stuffing density. Lowest f3 wins.
Secondly, optimum stuffing can be defined as that which most reduces internal standing waves. I've often drilled a hole in the corner of a cabinet to insert a 1/2" mic and measure the standing waves while trying different stuffing materials. this is very revealing.
When people start talking about flow density and such without referring to practical tests of performance, I'm not sure where they are coming from. What am I missing?
David S.
Secondly, optimum stuffing can be defined as that which most reduces internal standing waves. I've often drilled a hole in the corner of a cabinet to insert a 1/2" mic and measure the standing waves while trying different stuffing materials. this is very revealing.
When people start talking about flow density and such without referring to practical tests of performance, I'm not sure where they are coming from. What am I missing?
David S.
My goal was maximum damping of standing wave, without any additional loading on the cone. I had bad experiences before with overstuffing using baileys long hair cheepwool, and was looking for the optimum solution.
I confess I never did any measurents inside of the enclousure with different types of damping materials, I merely followed advice that looked convincing.
Did you post those measurements somewhere, I would like to see these results.
Yes, of course, a fairly deep tweeter horn. It still seems like you have more phase roatations than strictly necessary, although it isn't a big deal and you seem to be able to see through the rotations.
In very general terms every driver is a bandpass and can be thought of as having little or no delay in the center of its bandpass when the airpath delay to the acoustic center is accounted for. That being the case you can find a time offset that makes the phase curve in the center of the bandpass of the nearest driver (the woofer in your case) approximately flat. Then there will be phase advance at low frequencies and phase retard for high frequencies. All other sections will have more delay (more downward rotations) when the same transit time is subtracted out.
David S.
Then I can only tell you how I did it:
- Calculate the minimum phase on the basis the entire impuls of the woofer.
- Set the delay for phase estimation so that it leads to a measured phase response that best matches the calculated minimum phase.
- Leave the cursor and marker at the same place and use the same delay to get the phase for the tweeter.
- Calculate the minimum phase on the basis the entire impuls of the woofer.
- Set the delay for phase estimation so that it leads to a measured phase response that best matches the calculated minimum phase.
- Leave the cursor and marker at the same place and use the same delay to get the phase for the tweeter.
of course! I should proof-read what I write more often.
Doing whats practical is good engineering.
I seem to remember one study, where they tested how different bandwidth reflections can pull/smear auditory images. Oddly it was found the 800Hz and lower reflections having greater influence than reflections around 2-4kHz - if I remember this correctly. Does this study seem familiar to anyone? Would like to check it again, but can't find it no matter how I search.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.