Hi Tom,
No doubt, garbage in, garbage out. Its one reason so many in DIY choose to design for slight high frequency reduction, so that garbage recordings (of which there are many) sound tolerable, at a small penalty to un-hyped recordings.
Also agree that the room must be factored. As I said earlier, its the real value prop of DIY to be able to tailor for your acoustics.
Regarding the baffle diffraction step, and its affect at differing listening distances, I once studied this for an average size 2-way loudspeaker. I found ~ 0.3dB max difference at 1m vs 2 m, and less than that at 2m vs 4 or 5m. I take all measures at 2m unless teh system is very large, and go from there.
Any design can and really should account for diffraction effect change with distance through diffraction simulations. Of these, I found the BDS (Baffle Diffraction Simulator) quite accurate, "TheEdge" less so, based on sims vs actual anechoic data I had available.
Dave
My DSS implementation conclusions...
As promised, I've been living with DSS shelving for sometimes.
I conclude that the DSS shelving filter is not realistic in my dipoles. They did not sound natural.
By chance I experimented with waveguide + compression driver combination like CS2. Upon this short experiment I found that the dome tweeter's polar response is the biggest offender. Good polars can be obtained by combining dipoles and waveguides (attached).
This convinced me even more that dome tweeters cannot provide satisfactionary polar response when used with dipoles. The DSS filter on my speakers could not fix it although it provides a different sound.
Let us know if you have experimented with DSS and share us your finding.
As promised, I've been living with DSS shelving for sometimes.
I conclude that the DSS shelving filter is not realistic in my dipoles. They did not sound natural.
By chance I experimented with waveguide + compression driver combination like CS2. Upon this short experiment I found that the dome tweeter's polar response is the biggest offender. Good polars can be obtained by combining dipoles and waveguides (attached).
This convinced me even more that dome tweeters cannot provide satisfactionary polar response when used with dipoles. The DSS filter on my speakers could not fix it although it provides a different sound.
Let us know if you have experimented with DSS and share us your finding.
Attachments
Looking back to recall what "DSS" is, I have not found it.
What are your angles in that plot, please?
And which waveguide?
CS2?
What are your angles in that plot, please?
And which waveguide?
CS2?
]
II is kind of a repeat of I with the improved test setup. Probbly not worth buying I. What was in III?
I'll look for the Moulton paper. Queen wrote about the fr of the reflections also. It is somewhat at odds with the Lipshitz and Vanderkooy paper. Of course the direction of the reflection is key. Reflections from the sides can be separated by spatial hearing. The floor and ceiling bounce, in line with the direct sound, can't, and are heard more as response changes.
David S.
Is this the Queen paper? This is what I took from it:
"The effect of Loudspeaker Radiation Patterns on Stereo Imaging and Clarity", Queen, JAES 1978
- for 0 to 3 ms, images are smeared following an intensity rule, phase of the reflection didn't appear to affect this.
- short term reflections at least 10 dB below the first incidence do not create image shift (Olive found the threshold was 7 dB)
- horizontal mounting of woofer and tweeter caused 15 degree image wander when side wall reflections were present. Main cause was strong side lobes due to woofer/tweeter interaction (lesson here: use steep xovers if mounting woofer and tweeter horizontally).
- very narrow baffles reduced image shift in real rooms
- unconscious head move aids low frequency localization ability
- multiple vertical sources reduce image ambiguity (there's one for the line-source lovers)
Bech III modeled the direct sound + 17 discrete reflections, and reverb field. He measured detection thresholds for discrete reflections, in the presence of other reflections and reverb, simulated as omni in a room of flat absorption coeff and also simmed a 2 way directionality with absorption coeff like a real room. Main findings:
- reflection above 2 kHz that determined the affect on localization of individual reflections
- only the first order floor reflection (and not other reflections) contributed separately to source localization. The other reflections contributed additively
Haas effect (precedence) did not apply:
- reflections of delay 1 to 15 ms and -10 to -20 re direct sound can influence the3 auditory detection of direction of the direct sound.
- noise allowed reflections to be detected at much more sensitive thresholds than with speech
- detection of changes in timbre and localization are based on different auditory mechanisms: so threshold for comb filtering being detected as a timbral change is different than threshold for reflection causing localization change
All in all what looks like a pretty useful paper.
Dave
Yep. Shelving down the tweeter to correct the power response issue screws up the direct sound.
Could SL have hit the "right" level though with the Orions? Or is this a non-possibility?
Have you tried it at one stage in developing the NaOs and their variants? I only know one other dipole lover who tried the DSS and he did not like it as well.
I would like to hear the NaNote to see if I like "Flat"... But you can count me as one who quite enjoys the effect of the DSS shelf.
My impression of implementing the DSS change is that the speakers physical location is not so apparent and they literally disappear on the stage. I've had people swear I must have a center channel playing. It also allows me to listen to female vocalists and in general to listen at higher levels without fatigue.
I doubt it. I have gone through this in detail before and don't care to discuss it again. It's nothing more that a treble control. If you like the effect great. If you think that is represents some type of innovation or break through that is another issue.
All my designs have some type or tweeter level adjustment of shelving boost/cut control. I don't presume to know what is best for your preference or room.
Whether or not you like flat with the Note will still depend on your room and source material. I would say that what might be optimum for speaker A in your room would not likely be optimum for speaker B. But if adjusting the treble works for you in your room with your speakers and source material then by all means use it. My position would be that most likely a shelf of this type is just masking the real source of the problem, be it the speaker, the room or the source material.
The simple observation that some people like it and others don't ought to tell us something.
Hi Gainphile,
That could possibly be me 😱
When I first experimented with the DSS, I was using my digital XO and I had the feeling the system/room could benefit from the DSS. It did not work for me upon the first trials. But after a while I indeed found a setup that provided improvement (but that was not anywhere close to the DSS transfer function). Then I converted from digital to analogue filters (without DSS) and the desire for the DSS was gone. The only things that changed were the shelving filters for the mid and woofers, which cannot be implemented/mapped 100% correctly with my digital X-over.
My listening spot is at the apex of an equilateral triangle because that provides the sweetest spot for me (so that is in accordance with the DSS application).
I plan to experiment further with the DSS (out of curiosity) in the analogue x-over. Everything is prepared I just did not get to it because of some other construction sites related to speakers.
cheers,
Oliver
That could possibly be me 😱
When I first experimented with the DSS, I was using my digital XO and I had the feeling the system/room could benefit from the DSS. It did not work for me upon the first trials. But after a while I indeed found a setup that provided improvement (but that was not anywhere close to the DSS transfer function). Then I converted from digital to analogue filters (without DSS) and the desire for the DSS was gone. The only things that changed were the shelving filters for the mid and woofers, which cannot be implemented/mapped 100% correctly with my digital X-over.
My listening spot is at the apex of an equilateral triangle because that provides the sweetest spot for me (so that is in accordance with the DSS application).
I plan to experiment further with the DSS (out of curiosity) in the analogue x-over. Everything is prepared I just did not get to it because of some other construction sites related to speakers.
cheers,
Oliver
Last edited:
I could not edit my previous post so let me correct something here. I miss read this. It is possible that what SL has done with the orion is create a shelf the brings the power response of the tweeter section into alignment with the mids. But since the polar response remains the same the ratio of direct to reflected sound at those affected frequencies will remain unaltered. The problem is that the reflected sound is a function of the room acoustics while the direct sound is not. Thus, by shelving the response down he may have brought the reflected sound field into alignment with the reflected sound field of the midrange which may be beneficial to the sound over all, but it is at a loss of detail as provide by the direct sound.
Here is a plot showing what I mean:
An externally hosted image should be here but it was not working when we last tested it.
Assuming flat on axis response, the red trace is the power you would expect for a dipole mid in free space with about a 1.5k crossover. Note it is 4.8dB below the 0dB reference of a monopole. The violet trace is the power contributed from front and rear dome tweeters similarly crossover at 1.5k hz. As you can see, the power rises almost to 0dB as the front and rear tweeters basically act independently in the front and rear 1/2 spaces. As the frequency rises the power rolls off due to tweeter directionality. Now, the orange trace is a shelving filter response, sort of like the DSS, and the pale blue trace is the power of the system at high frequency with the shelf in place. The on axis (direct) response, which was previously flat, now follows the read trace combined with the orange trace at high frequency.
You can argue about why this might sound better or worse, depending on room acoustics, position of the speakers from reflecting surfaces, etc, but aside from the hand waving, this is pretty much what the shelving filter does. In a highly reverberant room it should be obvious that the excess power radiated above the x-o could result in a significant spectral imbalance while in a deader room, or a room with greater absorption in the offending frequency range, such shelving could result is making the system sound dull.
So from my point of view the problem is power response and room acoustics. Now this leads to some interesting observations. Recall that originally the Orion did not have a rear tweeter. As a result, with flat axial response the power response issue was not as significant. But there is another issue with regard to the spectral balance of the direct sound and the reflected sound. With a dipole mid the majority of the reflections come from the wall behind the speaker. Without a rear tweeter this reflected sound is imbalanced. Reflections from side walls are predominately from the tweeter, again with imbalanced content. You might suspect that the imbalances between the rear and side reflections would equal out but there are also timing and directional issues. This was obvious to me very early on with my initial designs and I added a rear tweeter before the design was completed. I also recognized the power response problem and attempted to circumvent it by using a narrower baffle and a higher crossover point. These helped but do not totally eliminate the problem.
Now that brings us to how to address the problem. I think this is a lot like making a comparison between a Mustang, Camaro, a Firebird and a Cobra. If you will bear with me, the NaO II is the Mustang and the Orion (with rear tweeter) the Camaro. They are different, but both are very similar. Both have similar strong and weak points. Both suffer the power response problem though, perhaps to different extents. With the shelf the Orion becomes the Firebird. It is tuned differently but underneath it is the same platform. It has the same problems as the Camaro. It may be more refined, it may perform better, but it still has the same platform related problems. The other approach is to move past the performance limitations of the platform and address them head on with new platform. That is what I did with the Note. It moved from Mustang to Cobra. It is a different platform with different performance objectives. There may still be a lot of similarities, but underneath it is different.
Now I know I get a lot of flack when I am critical of SL. I accept that. And I don't mean to get all carried away here, but to me this flat is not correct issue is just a manifestation of the GM syndrome: Stay with the same platform and create the illusion that it is now some how different and better. Add a bell or a whistle. It doesn't mean that the the product is bad. But it is a reluctance to reinvest in something better. I fully understand that. The Orion is a business venture as far as I can see. I'm not into all that. I'm interested in trying to design a better speaker than what I did before. At best, I'm just trying to support a habit, not turn a profit.
Anyway, right or wrong, that is my position.
Part of being human is hearing higher frequencies as closer and lower frequencies as distant b/c HF is so easily absorbed by air. We've just adapted that way. That could explain the disappearing speaker relative to the old flat response.
Dan
Hello,
Certainly the flat on axis and flat off axis response cannot be correct because it does not reproduce what is intended. 99% of the recording monitoring speakers do not have flat off axis response. Many of them can have smooth off axis response but not certainly flat.
You will not be able to hear what the recording engineer was hearing if you use speaker with flat on and off axis responses. In this light a loudspeaker having flat responses is an invalid system component, to be avoided.
You must abandon all speakers that are so called constant directivity to be able to achieve the accuracy the recording engineer was driving for because their response will not be correct to reproduce the recording that were created using non flat speakers.
😀
- Elias
Certainly the flat on axis and flat off axis response cannot be correct because it does not reproduce what is intended. 99% of the recording monitoring speakers do not have flat off axis response. Many of them can have smooth off axis response but not certainly flat.
You will not be able to hear what the recording engineer was hearing if you use speaker with flat on and off axis responses. In this light a loudspeaker having flat responses is an invalid system component, to be avoided.
You must abandon all speakers that are so called constant directivity to be able to achieve the accuracy the recording engineer was driving for because their response will not be correct to reproduce the recording that were created using non flat speakers.
😀
- Elias
you must abandon everthing but the studio, studio speakers, engeneer's seating position and his ears.
Elias,
if your logic would be 'sound', you would need to have the same room acoustics at home as the recording engineer had in his studio too.
Logic works another way: The mastering engineer (that's the man who is really in charge) will see to it, that the mastered recording is working in common home environments - over a larger spectrum of acoustical conditions.
Haven't you read the Bech papers 'sanedesign' was talking about?
Rudolf
Elias is obviously right. Whatever the recording engineer is doing, he/she is certainly not trying to make absolute constant directivity speakers sound right to listeners.
I think it would be more useful to think of speaker/room systems in dual but separate terms: (1) capturing what the recording producers wanted to get to your ears and (2) adjusting to the circumstances of your room and perhaps individual taste*. These should be different design undertakings but are lumped together in judging a final result.
BTW, aspects of this thread read like the dreaded speaker cable thread with people making enormous sensitive subjective interpretations based on 1 or 2 dB of change. And some of the subjective judgments are either hasty impressions (comparisons to what you just heard or the sensory norm you are used to) or after some time getting used to the changes ("mother's burnt soup" comparison or "brain finally done learning the room").
*Elsewhere I have commented on the lack of proper psycho-acoustical "sound controls" which need to be available to users whatever Magic Foo Foo Dust the arrogant designer has expended on his/her speakers in their showroom to make them "perfect" and therefore not needing tweeter controls, etc.
I think it would be more useful to think of speaker/room systems in dual but separate terms: (1) capturing what the recording producers wanted to get to your ears and (2) adjusting to the circumstances of your room and perhaps individual taste*. These should be different design undertakings but are lumped together in judging a final result.
BTW, aspects of this thread read like the dreaded speaker cable thread with people making enormous sensitive subjective interpretations based on 1 or 2 dB of change. And some of the subjective judgments are either hasty impressions (comparisons to what you just heard or the sensory norm you are used to) or after some time getting used to the changes ("mother's burnt soup" comparison or "brain finally done learning the room").
*Elsewhere I have commented on the lack of proper psycho-acoustical "sound controls" which need to be available to users whatever Magic Foo Foo Dust the arrogant designer has expended on his/her speakers in their showroom to make them "perfect" and therefore not needing tweeter controls, etc.
Last edited:
Not sure this follows directly above comment, but I tend not to have much trouble with certain sound parameters (like tonal color) under the control of the producers AND which are the ones the speaker makers fool with. In a way, those parameters are not much more helpful than critiquing the colors in a painting or photo print... kind of a matter of taste.
Where you do have major issues are choices which speaker designers are entirely unable to address: how much ambience is on the recording, the strange perspective (or helpful sound) of doing very close-up recording of a flute, multi-micing pan-potting placements and stuff like that that can make a recording really screwy to listen too.
But tonal coloration, like I said, is a much lesser issue although it does permit convergence from speaker designers. Yet that is what speaker designers (must) fuss with and the other stuff have little control over... whatever their directivity index.
Where you do have major issues are choices which speaker designers are entirely unable to address: how much ambience is on the recording, the strange perspective (or helpful sound) of doing very close-up recording of a flute, multi-micing pan-potting placements and stuff like that that can make a recording really screwy to listen too.
But tonal coloration, like I said, is a much lesser issue although it does permit convergence from speaker designers. Yet that is what speaker designers (must) fuss with and the other stuff have little control over... whatever their directivity index.
- Status
- Not open for further replies.