Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Brandon Wright]

Brandon,

Yep, EQ can't fix polar response..
What is the difference in SPL between the 4kHz on axis dip and the vertical off axis peaks?
How many degrees off axis is a "slight movement" ?

Art

4 or 5 db and as little as 5-10 degrees.

[Scott Holtzman]

What does "taking a few rotations out of the box" mean?  I have never heard that term.

Sent from my SM-T800 using Tapatalk

[Peter Morris]

4 or 5 db and as little as 5-10 degrees.

That’s the price you pay for what’s more or less a flat sided horn with a conical expansion.  I expect cutting holes in it for the entry of the multiple drivers does not help.

A conical horn has a different expansion rate depending on what part of the horn you consider.  Initially it expands very fast which is fine for high frequencies.  Toward the mouth its expansion rate is lower which good loading for the low frequency drivers.

This trick provides good loading for all the different drivers in a synergy horn, but the price is controlling HF directivity without creating lumps and bumps ... that’s why “normal” HF horns have curved sides.

Here is a comparison of two speakers which both use similar HF drivers.  The Danley is 90 degrees x 60 degrees.  The DIY is 90 degrees x 50 degrees.

[John L Nobile]

How are the settings working for you? I know they did not translate well for me into powersoft armonia. One thing I keep coming back to over and over again with these speakers is how to deal with the lobing , yes I said lobing, in the vertical at 4k. EQing for a flat response on axis makes for a ridiculous ice pick in the forehead response at any slight movement off axis in the vertical plane. This anomaly is born out in Danley's measurements and is discussed by Merlin in his article. I have been going back and fourth finding a happy medium in both the on and off axis response. Don't get me wrong it is a wonderful box and gets stupid loud (i've found that one SH96HO to four dbh218 is about the right ratio), but it does have this one particular wart.

On a brighter note, I have found that taking a few rotations out of the box really seems to bring the vocals up front and adds separation. Confirmation bias is a beautiful thing.

The one thing I notice with Danley speakers over any other speaker, is even with a somewhat ragged frequency response, I still hear things in recordings that I have never heard before. However, as has been discussed before, this is sometimes a good thing and a bad thing. I'm finding a lot of my usual source material unusable.

P.S. sorry for the crappy cell phone pic.

Can you explain what the graph is? I see a pic but no caption.

And what is taking a few rotations out of the box?

I've not noticed the 4k response in the vertical. I do notice how smooth the response is in the horizontal. I'll have to check the 4k vertical response in the next few days.

4 DBH218's for one SH96HO? I'd love to hear that. I'm doing great with 1 per side of each though I don't get stupidly loud.

[Brandon Wright]

Can you explain what the graph is? I see a pic but no caption.

And what is taking a few rotations out of the box?

I've not noticed the 4k response in the vertical. I do notice how smooth the response is in the horizontal. I'll have to check the 4k vertical response in the next few days.

4 DBH218's for one SH96HO? I'd love to hear that. I'm doing great with 1 per side of each though I don't get stupidly loud.

Sorry, the graph is a phase plot before and after some FIR correction. The Y axis is +- 720 degrees of phase rotation. This is the rotation I was referring to.

[Brandon Wright]

That’s the price you pay for what’s more or less a flat sided horn with a conical expansion. 

I suspect you are right. It certainly isn't as smooth as say the 4594he on hf950 combination that I have been playing with based on some unnamed fella's experiences.  But, audio is all about compromise I suppose.

Art, have you experienced this with your straight walled, conical synergy horn? 

[Peter Morris]

I suspect you are right. It certainly isn't as smooth as say the 4594he on hf950 combination that I have been playing with based on some unnamed fella's experiences.  But, audio is all about compromise I suppose.

Art, have you experienced this with your straight walled, conical synergy horn?

Here is a quick model in Hornresp that compares a conical with an exponential horn - same length and mouth size

[Ivan Beaver]

Not to be "nitpicky", but unless the scales are the same (both the color difference and the distances in the plot), it is hard to make accurate comparisons.

Here are the scales from an ad that a manufacturer used to "prove" that their product was behaving better than others.

I have removed the offending name.

They showed the same coverage map-BUT THEY CHANGED SCALES!!!!!!!!!!

I'm sorry-but you lose ALL credibility when you do that.

I have also included a standard EASE SPL scale for reference.

I see the same sort of "dual lobe" shape in your coverage map-and just wonder if the scales were equal would the results appear to be different.

Of course, everything is a trade off.

So let's assume the scales are equal, then what I see is that the SH96HO has a much greater "out of pattern" rejection-meaning that less energy is being put onto the walls/ceiling etc.

This is NOT meant to be a product to product comparison, but rather just pointing out that the SCALES do matter, and it is easy to make the results look different with a different scale.

Just like smoothing in a loudspeaker response graph-or changing the vertical scale etc.

People have to be careful what they are looking at.

[Peter Morris]

Not to be "nitpicky", but unless the scales are the same (both the color difference and the distances in the plot), it is hard to make accurate comparisons.

Here are the scales from an ad that a manufacturer used to "prove" that their product was behaving better than others.

I have removed the offending name.

They showed the same coverage map-BUT THEY CHANGED SCALES!!!!!!!!!!

I'm sorry-but you lose ALL credibility when you do that.

I have also included a standard EASE SPL scale for reference.

I see the same sort of "dual lobe" shape in your coverage map-and just wonder if the scales were equal would the results appear to be different.

Of course, everything is a trade off.

So let's assume the scales are equal, then what I see is that the SH96HO has a much greater "out of pattern" rejection-meaning that less energy is being put onto the walls/ceiling etc.

This is NOT meant to be a product to product comparison, but rather just pointing out that the SCALES do matter, and it is easy to make the results look different with a different scale.

Just like smoothing in a loudspeaker response graph-or changing the vertical scale etc.

People have to be careful what they are looking at.

I tried to align the scales best I could, they are close but not perfect.

The SH96 is from your software the other is an ease plot.

Given the huge difference in the mouth size of the horns I would expect the SH96 have better cut off outside the pattern.   

Even though the plots are not exact they do seem to agree with what’s happening in the real world. 

Interestingly enough the rough SIM model also agrees as does the off axis frequency response plots that I have seen on your software and measurements I have made.

… as you said, "everything is a trade-off".

[Art Welter]

 

Art, have you experienced this with your straight walled, conical synergy horn?

Brandon,
My 85 x 35 SynTripP design does not seem to suffer much from off axis peaks being louder than on axis response.
Worst case in the vertical is at 5 degree off axis being 1.5 dB louder than on axis at 1590, 5500 and 8000 Hz.
Worst case in the horizontal is at 5 degree off axis being 3 dB louder than on axis at 2500 Hz, and 1 dB louder at 5.5 kHz.
Although I didn't chart more than 5 degree increments, while slowly tuning the test turntable I never have seen (or heard) anything "jump out" that would make me think any frequencies or dispersion angle is any worse than shown.

The response of Peter's "DIY" chart shows lobing of about 1 dB, as well as a pattern inconsistent with a stated 50 degree vertical (perhaps it is showing the 90 degree horizontal pattern?) pattern.

Although the 5 degree spacing on my chart won't reveal all the warts, there is no ambiguity regarding levels. Colored charts without a dB scale for those colors don't tell us the depth of any problems.

That said, an on axis dip at 4kHz is the worst possible frequency, as average ears have their maximum sensitivity there- and noise induced hearing loss also is centered at the same, so deaf sound engineers would not be aware of the potential problem.

Ouch.

Art

[ProSoundWeb Community]