Eric Weitzman said:
Clearly a True Believer. Fair enough.
Perhaps if you re-read my earlier post as to why I suggested dumping the RF filter my reasons for doing so would be clearer. However, you seem to have made certain assumptions about what I said, rather than reading the actual words. If you read the post, you will note that I did not say that it would reduce the number of opamps--it's the RF filter itself I object to.
And of course it's analog...jeez, did you think Linkwitz was suddenly going to stick a digital filter in there?
If an RF filter you must have, at least raise the frequency.
You seem also to have completely, entirely absolutely missed the point I was making about the LS3/5a, but that's okay. You seem, likewise, to have either missed or misunderstood the portion of my post where I referenced the Walsh/Ohm speaker. Perhaps I should hit that nail a little harder.
For those who are not familiar with the Walsh/Ohm (two different companies using the same concept), they use a dynamic driver mounted upside-down and face down on top of the enclosure. When you look at the speaker, you see the back of the driver only; the front fires down into the cabinet. You can do what you want with the front wave--infinite baffle, bass reflex, transmission line, etc. Given that the back of the speaker is exposed on top of the enclosure, a speaker of this sort is truly omnidirectional. By definition the off-axis radition is exactly the same as the on-axis radition. As such, it's a good conceptual model for considering Linkwitz's assertion that the off-axis radiation must be the same for good imaging.
Not surprisingly, the speakers have no hot spot. You have much more freedom as to where to sit. And from wherever you sit, you get an image. It's a spooky image, spaced out and diffuse, but it's huge. The problem is that it lacks anything like image depth and it's bloated and out of focus. Spectacular as an effect, but not real. Anyone who thinks that sort of image is 'real' needs to listen to more live music.
In his presentation, Linkwitz compares the image from the omni speaker to the dipole and says something to the effect that the dipole has 'a little more depth.' I can hardly contain my surprise. This, among other things, leads me to wonder about his statement that the speakers sound so similar.
Again, if you read what I wrote, you'll see that I agree that a speaker's off-axis radiation should be consistent with its on-axis radiation. My point was that it's important for other reasons, just not for imaging.
If anyone needs further clarification, take some speakers outside and listen. If you're free from walls to reflect sound, the off-axis radiation falls out of the equation. Granted, things sound very different, and it will take a lot more power to get a volume that you consider normal, but if you do so you'll find that speakers that image well still image well and speakers that don't...don't.
Grey
After decades of searching and building, I proudly admit to being a true believer. Someone once used the term Linkwitz fanboy/lemming and I wear that badge with pride! SL's Beethovens 15 years ago would have made me truly happy but there were too expensive. Having pursued other paths and effectively waiting for the Orions, I discovered that the Orions are a big step forward compared to the Beethovens because of driver advances and SL's constant quest, and they're substantially less expensive too.
Your reasons had nothing to do with the filter's function and everything to do with your preconceptions that they would audibly affect the response at 20kHz. I calculated the attenuation of this third order filter at 20kHz and it's a whopping 0.03db. Would be that the tweeters were as flat and consistent as that... Then you talk about cascading all these 1/10th db drops. May I ask precisely what are these other nightmarish drops at 20kHz in the design that populate your imagination? Left as an exercise for the fearful perhaps? And impact on the majority of us who can only hear to about 15kHz, three tones below 20kHz?
I didn't, but it appeared that you did...Since the thread evolved (after Nelson' post) into the (perceived) problem of too many opamps and a redesign to remove the opamps being beneficial, one would legitimately assume that you were addressing the topic. You certainly weren't addressing the OP's question... Your prior posts were also about the (perceived) problems with opamps. "Yes, the music gets through opamps, but it's been processed into McSound." and "I confess that just looking at the "Orion ASP Function block diagram" gives me a case of the heebie-jeebies. So many opamps--and chip opamps, at that--must sum to several hundred transistors...just in the crossover." Please accept my apology for assuming your criticism of the RFI filter was based on your previously expressed notions.
But you wrote this regarding the LS3/5A . This is what I responded to. May I not object to your premise instead of addressing your conclusion?
Repeat, "image like a bat out of hell". It's quite appropriate for me to argue with this claim with a counter claim. You can't reply to my criticism because you have no experience listening to the Orions reproduce recordings. But you can't just get away with deflecting my criticism.
But why not address your conclusion, the point I think you think I missed regarding the LS3/5A? You argued that the felt treatment hones the hf radiation pattern into a laser, but then concede the bass remains omni. So? Does this refute any of SL's observations about uniform polar response and bursts the bubble? No, in fact, this entirely supports what he says. The LS 3/5A doesn't come close to the realistic soundstage of the Orions, and it's polar response is the one part of their problem, this problem being discussed in SL's paper. I'll let pass (as you did) its lack of realistic dynamics and extension and, ultimately, SPL.
Apparently, Grey, you have a bone to pick with SL. Ignorant of the details of the ASP design and of the Orion's sound, you criticize based on nothing valid on its face so far. What's the history here? At least JohnK was gentlemanly enough to admit to his history with SL and then step aside.
Please go ahead and take the last word and end round two. Two rounds of this is more than enough so I will probably not react to your next post.
- Eric
I'm admittedly a "True Believer" and I've found Grey's posts disruptive in the context of this thread.
May I respectfully ask to remain on topic here?
Of course Grey is entitled to have any opinion about the Orion design, maybe it would be more appropriate to discuss about it in a separate thread?
Regards
Giorgio
P.S. there is no "waveform fidelity" that can be hurt by a 140KHz RFI filter if the signal is sampled by a 30KHz bandwidth limited microphone, let alone stored on a 44.1 KHz digital media.
May I respectfully ask to remain on topic here?
Of course Grey is entitled to have any opinion about the Orion design, maybe it would be more appropriate to discuss about it in a separate thread?
Regards
Giorgio
P.S. there is no "waveform fidelity" that can be hurt by a 140KHz RFI filter if the signal is sampled by a 30KHz bandwidth limited microphone, let alone stored on a 44.1 KHz digital media.
Grumpy_Git said:
No kidding.
We have a poster who doesn't believe filters can be done more efficiently...I guess he's never looked at an RIAA filter, wherein two resistors and two caps do the work of three separate first order filters, several posters who feel that chip opamps are wonderful...the more the merrier (somehow, adding more opamps undoes the damage--perhaps there's a null point when you reach five opamps...no?...add another five...we'll find that null somewhere), another poster who can't even be bothered to see which speaker I'm talking about, not to mention any number of people who only read three out of every ten words I wrote then jump to conclusions blindly...
Against religious fervor, reason strives in vain.
Grey
Zen Mod said:
What we see is a group of blind men who each only feels a small part of the elephant. Nobody has provided any certain proof of who's right, just opinions based on each person's understanding of that portion of the elephant that they've had contact with. Every thinking person can see that Grey's comment about religious fervor and reason cuts both ways... Anyways, the only person likely to be "right" in this discussion isn't even participating.
A crossover for a three-way dipole can be, and has been, done passively. SL did it in the worst environment (at speaker power/voltage levels) with the Audio Artistry Vivaldi speaker. The blind elephant watchers should seek to discover why the particular engineering choices in the Orion that were made were made.
This blind elephant watcher sees two main issues doing the Orion ASP mostly passively.
The first big issue is lumping filters together without buffering. It would take a really smart fellow to chain four or five filters together, with each frequency-dependent output impedance driving the next filter instead of fixed input/output impedances. The inspired designer would go from a cascaded, series design to a single multifunction network. This seems to me as complex as designing a series speaker crossover with five or six drivers.
The second big issue is gain. Since the Orions have dipole equalization from 1500Hz down to 20Hz, there would be 36db of insertion loss from 20Hz to 1500Hz. Throw in the varying driver sensitivities, floor gain for the woofers, and baffle step, there may be a little more or less loss. Either way, there'd be a serious noise problem in the highs if the input voltage is simply dumped using passive, attenuating equalization.
I was a member of the minimalist, no NFB, high efficiency club for a decade. In retrospect, I see the problem with the other clubs is not that the electronics suck a bit, but that the speakers suck a lot: flea-power amps require the use of speaker technologies that are more efficient but which, more importantly, are by their nature free from more types of distortion. My last horn/SET system, devoid of transistors and opamps, was replaced by the Orions. Such Apostasy! Given that this is the Pass Labs forum, I would add I've had an Aleph 3 in my system at one point and it sounded great. The Orions (with even crude amps) outperform every other speaker technology I've heard. I'm completely open to a passive solution if it solves more problems than it creates, but at this point I don't hear any problems with the ominous cavalcade of opamps. I could be wrong and would like to know -- old beliefs die hard. We should all try to grok the whole elephant.
- Eric
Eric Weitzman said:
Any person with a simulator, a knowledge of filter theory, a methodical approach and a willingness to spend the time can achieve this.
Cascaded filter interactions are not always significant enough to need buffers. Try different combinations and take it one step at a time.
If stuck, try freestyle engineering to a transfer function. It may sound daunting, but with any topology at your disposal (some good examples of which can be found at the Linkwitz site), making for example kinky curves in one step, this proposal's success should be largely a matter of time for the designer.
Eric Weitzman said:
One of the things I addressed in my NaO design the all active design approach. There are arguments for all active, but there is also the cost/perfromance issue, particualrly if high quality amplification is used. As such it seems reasonable to use a hybrid appaoach with a passive crossover between midrange and tweeter and design the system as a bi-amplified system, eliminating the need for a separate tweeter amplifier. This results in the potential for a fairly simple active circuit with only 4 active stages directly in line with the main panel, 3 if the adjustable bass boost is removed.
An externally hosted image should be here but it was not working when we last tested it.
This circuit can easily be addapted to the Orion woofer/mid eq and crossover. The all pass delays between the mid and woofer really aren't required at all. The resulting phase error between mid and woofer through the crososver has negilgable effect on the on axis response or on the polar response. And another issue which seems to have been missed is that any time delay applied to correct for physical offset with respect to the on axis response results in an increase in the phase error due to offset from the rear side. After all, this is a dipole and both front and rear response need to be considered. http://www.musicanddesign.com/Dipole-offset.html
Nelson Pass said:Something worth pointing out is that generally the objection to
large amounts of circuitry refers to circuits in series. You
can reduce this by splitting off the W/M/T early so that various
filters and delays are applied only to the narrower bands, not
to the signal as a whole.
Examination of the Phoenix circuit diagram on the Linkwitz site shows only a passive 140kHz low pass RFI filter and a 2.8 kHz notch filter before the signal is paralleled to the respective woofer, mid, and tweeter legs. On the Orion, the 2.8 kHz notch filter does not exist, so SL does just as NP suggests.......
In this case I was referring to my copy of the Orion block diagram
in which the signal which is fed to the tweeter filters first goes
through the subsonic HP, then the midbass shelf, then the 2 times
2 pole midrange hipasses and then 2 low-mid phase shifters.
Conceivably the tweeter circuit input could be taken at the input
to the system, bypassing those electronic blocks and simplifying
the signal path for the tweeters.
in which the signal which is fed to the tweeter filters first goes
through the subsonic HP, then the midbass shelf, then the 2 times
2 pole midrange hipasses and then 2 low-mid phase shifters.
Conceivably the tweeter circuit input could be taken at the input
to the system, bypassing those electronic blocks and simplifying
the signal path for the tweeters.
How to proceed?
I love the idea of alternative XOs for the Orion cabinets.
However, the Orion cabinet itself is proprietary. It would be rude to reverse engineer it from photos. So, even if we measure the drivers in the cabinet, and design from there, that's already rude.
Nelson, maybe you could visit SL (or vice versa) and discuss options? He seems to accept visitors, so I bet he'd be more than happy.
David
I love the idea of alternative XOs for the Orion cabinets.
However, the Orion cabinet itself is proprietary. It would be rude to reverse engineer it from photos. So, even if we measure the drivers in the cabinet, and design from there, that's already rude.
Nelson, maybe you could visit SL (or vice versa) and discuss options? He seems to accept visitors, so I bet he'd be more than happy.
David
My understanding from a post here is that someone talked to SL at the London AES show and he said he didn't care as long as it isn't called an Orion (or a Pluto I assume) Seriously, it is clearly not going to be identical to the Orion much as people can try, so it seems that it is OK to proceed..
SL's response to O'Brien ASP
At the recent London lecture, in consideration of Linkwitz Lab offering for sale Orion plans and finished electronics, so as to clear the way for our project, I asked Siegfried Linkwitz and told him about NP and others interested in improving on or simplifying the op-amps, and he said just don't call it 'Orion.' A few days ago on the Orion users' forum (where we took some flak) he clarified further and made some observations about the task.
David, who said anything about reverse engineering Orion cabinets from photos? Still, it would be kind of interesting for a summit meeting...
At the recent London lecture, in consideration of Linkwitz Lab offering for sale Orion plans and finished electronics, so as to clear the way for our project, I asked Siegfried Linkwitz and told him about NP and others interested in improving on or simplifying the op-amps, and he said just don't call it 'Orion.' A few days ago on the Orion users' forum (where we took some flak) he clarified further and made some observations about the task.
David, who said anything about reverse engineering Orion cabinets from photos? Still, it would be kind of interesting for a summit meeting...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.