So you are indeed undercharging given that the increase in cost per speaker is $150, however your charge is only $200 per speaker.
A $50 profit makes you a saint to your customers, but certainly underpriced from a business standpoint
.A thumbs up for you to have more orders so you can recuperate from quantity sales
.Anand.
If the Summas can be offered at about $5000 a pair, that would probably forever stop those who crave for active pro monitors such as the Klein-Hummel o410 or even the o500c and the new Genelec 8620a! One can always power the Summas with any existing amps or receivers found in the house and not having to worry about compatibility or power output, and most people have at least one sub, just get 2 or 3 additional inexpensive subs such as the Polk PSW 10 and we'll have a system that will be equal or superior to most of the mega-buck audiophile speakers out there! There are already a few enlightened ones out there who have replaced their prestigious high-priced audiophile models with the Abbeys! I'm starting my savings today
jheyau: Curious that I just read your comment. I was literally just eyeing up the Genelec 1037, 8260a or an Abbey....
Now, I confess I have never really heard any horns, so I don't know what I am getting into. I also have a fairly small listening room (4m x 8m ish) which is an odd shape and needs to have reasonable WAF factor.
I have spoken to Earl when he first started producing his waveguide speakers and I was very interested back then and it was only the size of the Suma which put me off, however, the Abbey looks workable (sizewise)
So the question is really how you would compare the 3way Genelecs with the Abbey? The published data is reasonably hard to compare, but as I see it the Genelec figures look pretty ruler flat over a wider range and further towards the 20Khz mark. They also appear to decline slower off axis and appear to offer extremely well controlled dispersion.
In contrast the Abbey appears to have a narrower flat on axis range. The low end doesn't worry me too much because I will be trying to integrate with my IB sub, but the high end seems to roll off a little faster than desired? The off-axis seems to decline much faster (which I think is what I expect from a horn?) and seems very well controlled as we go off-axis. It's hard to figure out the graph key to the off-axis plots on the gedlee site though?
Any comments on the relative audio performance of the Abbey vs the Genelec? Especially interested if Earl wanted to offer his 2p?
Thanks
Now, I confess I have never really heard any horns, so I don't know what I am getting into. I also have a fairly small listening room (4m x 8m ish) which is an odd shape and needs to have reasonable WAF factor.
I have spoken to Earl when he first started producing his waveguide speakers and I was very interested back then and it was only the size of the Suma which put me off, however, the Abbey looks workable (sizewise)
So the question is really how you would compare the 3way Genelecs with the Abbey? The published data is reasonably hard to compare, but as I see it the Genelec figures look pretty ruler flat over a wider range and further towards the 20Khz mark. They also appear to decline slower off axis and appear to offer extremely well controlled dispersion.
In contrast the Abbey appears to have a narrower flat on axis range. The low end doesn't worry me too much because I will be trying to integrate with my IB sub, but the high end seems to roll off a little faster than desired? The off-axis seems to decline much faster (which I think is what I expect from a horn?) and seems very well controlled as we go off-axis. It's hard to figure out the graph key to the off-axis plots on the gedlee site though?
Any comments on the relative audio performance of the Abbey vs the Genelec? Especially interested if Earl wanted to offer his 2p?
Thanks
The faster fall off off-axis is highly desirable IMO, it's what I believe makes the Abbey the better choice. I can't speak to other companies products (I don;t know the details on whats inside), but Genelec has a good name and their stuff is very good, I agree, but I do not think that the price ranges for the same basic systems (take woofer size for example) are comparable.
I would also say that there are "horns" and then there are my waveguides. The two things are entirely different - thats why mine have a patent.
It is a shame about "graphs" because everybody does them differently. That makes comparisons even more difficult. I would say that it would be wise for you to learn to read these graphs so that you are comfortable with the comparisons, because until you can do that, this kind of data, which is the most relavent, is not going to suite its purpose for you. I tried to post information on my site about what the data means and how to read it. Thats not always easy to do.
To me the narrower directivity is a key element in the comparisons that you are making, but you seem to take the position that the faster fall off is a bad thing, To me its a key design requirement - especially for a small room.
I'm not sure I know what youre having trouble with. Read the paper on directivity - it explains polar maps pretty well I think.
err, well actually I have a suspicion that a faster falloff could be helpful in a small room, that's why I'm investigating waveguide solutions rather than openbaffle designs (say). I believe the Genelec solution is also technically a waveguide? However, I lack the experience of having heard a range of stuff in various sized rooms to be doing this from more than an intuitive guesswork position
Regards your graphs there isn't obviously a key on the right hand graph so it's not immediately obvious which line relates to what number of degrees of axis? Are they 5 degree measurements?
I will re-read the papers you referenced, but unless you updated them recently I think I have read them a few times over the last 5+ years? It's possible that you are mis-reading my post though...?
Oh, you didn't comment on the HF rolloff from the Abbey? Seems a fairly steep rolloff from a little over 10Khz, but the graph is kind of hard to see up there? How does this bear out audibly?
Any happy Abbey (new and improved) owners hang out here?
Regards your graphs there isn't obviously a key on the right hand graph so it's not immediately obvious which line relates to what number of degrees of axis? Are they 5 degree measurements?
I will re-read the papers you referenced, but unless you updated them recently I think I have read them a few times over the last 5+ years? It's possible that you are mis-reading my post though...?
Oh, you didn't comment on the HF rolloff from the Abbey? Seems a fairly steep rolloff from a little over 10Khz, but the graph is kind of hard to see up there? How does this bear out audibly?
Any happy Abbey (new and improved) owners hang out here?
"Technically" the Genelec thing is a waveguide, but it's not a waveguide in the same sense as I use the term. Its far too wide to work in the manner in which I intend my to work.
The directivity paper is only a few months old.
The curves are every 7.5 degrees, from 0 - 90.
Yes, the HF rolloff. I used to not even test above 10 kHz because I give this region such little importance. There is signal to about 16 kHz, but then ,yes it drop significantly above that. Is this important. IMO not in the least. Its "specsmanship" that drives the belief that > 10 kHz has any relavence. There is also a LP anti-aliasing filter out there somewhere as well. This is also a factor.
As far as happy owners go, I update the reviews on my site regularly so you can read from several owners.
In the past I compared Gedlee vs Genelec measurements. The Gedlee measurements look better to me. Firstly there is the issue of directivity as mentioned. Earl's recent paper on directivity shows a very clever way around the problem of the sound stage collapsing into the nearest speaker. The higher directivity yet with a flat shelf at a lower level for the Geddes speakers has a clever synergy here. The Genelec doesn't have enough directivity control for this to work. So it could be that the 2 Geddes speakers could replace 3 Genelecs in a HT system and for many situations where a centre speaker is a pest, that's a huge plus.
Some of the Genelec speakers have some problems at 30 degrees and beyond. The 1039 has big dips around 500 and 3k for 30 degrees and beyond. The Geddes speakers are better behaved IMHO. In comparing, also keep in mind that you are looking at measurements out to twice the angle - 90 degrees. So you need to disgregard half of the Geddes data in making a direct comparison.
Overall, comparing the Abbey to say the Genelec 1037, the Abbey looks better behaved. The response at different angles are smooth and there is consistency. The main anomaly is directly on axis and has a simple solution while the Genelec isn't as consistent off axis. If you were using DSP you could Eq the Abbey very flat and the result would work at all angles pretty well with a smooth and gradual transition. While the Abbey rolls off sooner on the top end, the response is much better behaved while the Genelec has some issues. You can't do anything to it with eq as it changes with angle, suggesting problems needing an acoustic solution.
If the top end roll off were an issue then you could experiment with a super tweeter. You could do it actively with any tweeter, or passively with just about any efficient tweeter. I'm not sure it's a great idea, but it's something you could try if that's important to you.
The Abbey uses a CD where the Genelec uses a dome (cheaper). It's a vented 3 way with a 5" efficient mid similar to what you find in many commercial cinemas. It makes sense to use a simpler approach where it works. To me this looks like an expensive complex solution that works well vs a simpler solution that is probably much cheaper yet with most likely superior performance. If common sense were the rule rather than the exception, then I think we'd see a LOT more of these kits around and Earl would have more orders than he could possibly put together without help.
Some of the Genelec speakers have some problems at 30 degrees and beyond. The 1039 has big dips around 500 and 3k for 30 degrees and beyond. The Geddes speakers are better behaved IMHO. In comparing, also keep in mind that you are looking at measurements out to twice the angle - 90 degrees. So you need to disgregard half of the Geddes data in making a direct comparison.
Overall, comparing the Abbey to say the Genelec 1037, the Abbey looks better behaved. The response at different angles are smooth and there is consistency. The main anomaly is directly on axis and has a simple solution while the Genelec isn't as consistent off axis. If you were using DSP you could Eq the Abbey very flat and the result would work at all angles pretty well with a smooth and gradual transition. While the Abbey rolls off sooner on the top end, the response is much better behaved while the Genelec has some issues. You can't do anything to it with eq as it changes with angle, suggesting problems needing an acoustic solution.
If the top end roll off were an issue then you could experiment with a super tweeter. You could do it actively with any tweeter, or passively with just about any efficient tweeter. I'm not sure it's a great idea, but it's something you could try if that's important to you.
The Abbey uses a CD where the Genelec uses a dome (cheaper). It's a vented 3 way with a 5" efficient mid similar to what you find in many commercial cinemas. It makes sense to use a simpler approach where it works. To me this looks like an expensive complex solution that works well vs a simpler solution that is probably much cheaper yet with most likely superior performance. If common sense were the rule rather than the exception, then I think we'd see a LOT more of these kits around and Earl would have more orders than he could possibly put together without help.
paulspencer: Thanks for your comments, very interesting
Yes, I haven't previously owned any high sensitivity speakers, eg horns or wave guided devices (lets disregard the design specifics). However, I have very positively experimented very positively with digital EQ and once gone down that route I find it seems to test drivers and amps more, so I have found larger gains by adding extra drivers or adding extra/better amps, etc. All I guess pretty much as you expect, EQ reduces headroom, etc
Anyway, I think I currently want to experiment with both controlled directivity and high sensitivity drivers. I have various horn designs that I can't quite bring myself to build. I very much buy the theory behind Earl's designs and of course the pro audio world has been building "similar" based designs for years (I say "similar" in the sense that their design criteria is controlled directivity, how they achieve it may be different). It seems like the right direction for the kind of listening room they need to work in.
The Genelec 1037 costs something like $4,500/speaker and the 1038s are something like 25% more. I think it's comparable to the Abbey/Suma in the sense that you are buying similarly sized drivers, but we can instead compare with the 1032s to use a two way design for I think around $2,000. The Genelec's include amplifiers per channel, but on the flip side this is "pro-audio", ie mass market and so the prices are usually much more keen for a given design (versus say the home audio world). So overall yes the Abbey/Suma are quite a bit lower in price and we should judge everything proportionally.
Thanks
Yes, I haven't previously owned any high sensitivity speakers, eg horns or wave guided devices (lets disregard the design specifics). However, I have very positively experimented very positively with digital EQ and once gone down that route I find it seems to test drivers and amps more, so I have found larger gains by adding extra drivers or adding extra/better amps, etc. All I guess pretty much as you expect, EQ reduces headroom, etc
Anyway, I think I currently want to experiment with both controlled directivity and high sensitivity drivers. I have various horn designs that I can't quite bring myself to build. I very much buy the theory behind Earl's designs and of course the pro audio world has been building "similar" based designs for years (I say "similar" in the sense that their design criteria is controlled directivity, how they achieve it may be different). It seems like the right direction for the kind of listening room they need to work in.
The Genelec 1037 costs something like $4,500/speaker and the 1038s are something like 25% more. I think it's comparable to the Abbey/Suma in the sense that you are buying similarly sized drivers, but we can instead compare with the 1032s to use a two way design for I think around $2,000. The Genelec's include amplifiers per channel, but on the flip side this is "pro-audio", ie mass market and so the prices are usually much more keen for a given design (versus say the home audio world). So overall yes the Abbey/Suma are quite a bit lower in price and we should judge everything proportionally.
Thanks
Thats really my point. For the prices that Genelec gets I could do a lot more as well. But the point that I often make is that the last doubling of the price gets you only a few % of improvement. I'ts much like a musical instrument where the really good ones are ten times as expensive as the "good" ones but to most people the sound quality difference is minimal. And just like speakers, if the product is not used properly that audible difference is lost altogether. If you do not build your room around the speakers, you will not notice these improvements.
Dr. Geddes, I know I've talked about this in different ways numerous times, but what do you think about the use of active filters with your speakers. Specifically, what about using FIR filters to create minimum phase at the crossover. So not aggressively eqing the speakers above the transition point, not utilizing complex active filters, just a more simple FIR filter that "Fixes" the phase. There aren't a lot of affordable devices currently able to do this, but I don't think that is due to a tangible limitations, but rather the newness of it's use in audio. I believe that the miniDSP I talked about before is capable of this, just not with the software they currently offer (the chip can handle it).
Alternatively, what about offering a version of the speaker that is active using something like the new Hypex AS2.100 Digital, which would add about 800 dollars to the price per pair to any of your speakers. Maybe silly on the smaller cheaper models, but on the Abbey and Summa, maybe a worthwhile option. That unit uses IIR filters for the crossover and can do one FIR filter to fix the phase around the crossover.
I think that the 400 dollar price per speaker for going active isn't outrageous (even if it was 600 to cover some profit and additional work for you) given that we are talking about it including the amplifier, and offering a potentially better system than you would get with passive crossovers and no FIR filter for the phase. I know that, compared with your pioneer receiver option, it's more money than the entire receiver probably cost, but for many of us, the EQ and amplifier benefit is inline or cheaper than what we already are using. You could even add to your room optimization service a custom filter program that relies on a set of measurements taken in the room, and then optimizing the response over a broader range.
Alternatively, what about offering a version of the speaker that is active using something like the new Hypex AS2.100 Digital, which would add about 800 dollars to the price per pair to any of your speakers. Maybe silly on the smaller cheaper models, but on the Abbey and Summa, maybe a worthwhile option. That unit uses IIR filters for the crossover and can do one FIR filter to fix the phase around the crossover.
I think that the 400 dollar price per speaker for going active isn't outrageous (even if it was 600 to cover some profit and additional work for you) given that we are talking about it including the amplifier, and offering a potentially better system than you would get with passive crossovers and no FIR filter for the phase. I know that, compared with your pioneer receiver option, it's more money than the entire receiver probably cost, but for many of us, the EQ and amplifier benefit is inline or cheaper than what we already are using. You could even add to your room optimization service a custom filter program that relies on a set of measurements taken in the room, and then optimizing the response over a broader range.
I'm a big fan of this kind of thing also. Bar the cost of implementation it completely frees the designer to draw any crossover they like without needing to worry about the size of components or funny impedances building up around crossover points. Also all the components become theoretically perfect with perfect tolerances. Finally of course you have the option to experiment with very steep and non causal filters (or a blend of each)
The disadvantage is the increase in the number of DACs and amplifiers... However, it seems to be easier to come across reasonably priced amps and dacs which are approximately linear that it does speakers...
I have seen some folks claim that you can get enough horsepower out of a little Alix board to do 2-4 channels of complex FIR filtering. These are just over $100 each (need to figure out how to add your DAC). Alternatively there are a bunch of inexpensive Atom boards which have infinitely more horsepower for probably only twice the price. I should also imagine the onboard HD audio chipsets are not even the limiting factor of most people's system (yep, I realise that comment is going to go down like a lead zeplin...)
Good luck
Ed W
The disadvantage is the increase in the number of DACs and amplifiers... However, it seems to be easier to come across reasonably priced amps and dacs which are approximately linear that it does speakers...
I have seen some folks claim that you can get enough horsepower out of a little Alix board to do 2-4 channels of complex FIR filtering. These are just over $100 each (need to figure out how to add your DAC). Alternatively there are a bunch of inexpensive Atom boards which have infinitely more horsepower for probably only twice the price. I should also imagine the onboard HD audio chipsets are not even the limiting factor of most people's system (yep, I realise that comment is going to go down like a lead zeplin...)
Good luck
Ed W
The minidsp I mentioned has the power, and they are planning on releasing an FIR plugin in the near future. This would be the cheapest and easiest option to implement what I suggested. It would not include the necassary amps for an active setup, but could be used simply to correct the impulse response for linear phase. This would actually be far more simple. You would take a good clean impulse response measurement, create an FIR filter that is the inverse of that, and load that into the software. The hardest part is the clean impulse response (I think Dr. Geddes could get this best), and having a good method of creating that inverse filter.
The same device could also be utilized with IIR crossover filters, eq, and then the FIR for phase correction. This seems to be a commonly preferred method to avoid the problems inherent in FIR filters such as ringing and latency. I think this would be what I would do if I was converting Dr. Geddes design to fully active designs.
I've mentioned many times before, and still believe that a passive crossover with weird phase or impedance issues or major inefficiencies are due not to problems with passive crossovers, but poor crossover design. Speakers are not perfect linear phase flat impedance devices, and many of the problems people show as proof of the deficiencies in passive crossovers are really due to either a design problem, or speaker driver/enclosure problem. Dr. Geddes unusual crossover design is actually very good, and uses techniques a few other creative designers have used, which ensures far less of these impedance or phase anomalies.
I've posted the system impedance of my Abbey's before, and it's really quite good, better than most all speakers I've seen. The key, I believe, is that he isn't using the traditional topologies to get the slope order he was looking for. This method has also ensured that the phase remains far more linear than other designs have. I mean, if we define linear phase as the phase transitioning in an even and smooth fashion, instead of abruptly at the crossover point, then his designs show unusually good phase response for a passive multiway speaker.
The last area of contention between passive and active crossovers is inefficiency. I think Dr. Geddes has agreed with me before that with properly designed crossovers, passive crossovers are not as inefficient as they are presented to be, and that the implication in actual every day use is very minor. There is no denying that the parts get hot, that power is being converted to wasteful heat, but in terms of added power compression, loss of efficiency, etc. it's pretty minor. Certainly spending 1000's of dollars to fix it would be a huge waste. But it is waste, and if the cost was the same as or similar to the cost of a passive crossover, it's not such a bad idea in my opinion, with some advantages.
In terms of active crossover limitations, the conversion to and from digital incorporates issues that are only now being overcome. Basically, headroom with the signal the DSP operates on means that, if not optimized for the system, you either have horrible distortion from not enough headroom, or a signal that is mostly so low in level that the bit rate from the ADC into the DSP is too low for proper resolution, thus causing it's own distortion. One of the great fixes to this, besides the much improved digital gain controls being developed, is the much higher internal bit rates that the DSP's can operate at. The final issue isn't able to be resolved easily, and that is that volume is still controlled at a preamp before the DSP, and thus some bit loss is inherent at lower volumes. A fix is a volume control after the DSP, but this isn't feasible for the average person in an integrated fashion.
Interesting comments. I seem to recall Earl has said he's tried active but didn't find a worthwhile advantage.
One of the aspects of active I find appealing is the extra freedom and headroom regarding amplifiers. Even with a dome tweeter, very little power is needed. With a CD it's even less. Good news for tube enthusiasts. Class A becomes more feasible. Even for the midbass, this becomes possible. Sure, there are 100w tube amps that will drive an efficient speaker but with active 40w + 5w goes a long way. Or some might use a pair of chip amps that otherwise might not be up to the task.
Another advantage is making changes and tweaking. Earl mentioned a crossover upgrade - no need to buy and solder with digital active.
One of the aspects of active I find appealing is the extra freedom and headroom regarding amplifiers. Even with a dome tweeter, very little power is needed. With a CD it's even less. Good news for tube enthusiasts. Class A becomes more feasible. Even for the midbass, this becomes possible. Sure, there are 100w tube amps that will drive an efficient speaker but with active 40w + 5w goes a long way. Or some might use a pair of chip amps that otherwise might not be up to the task.
Another advantage is making changes and tweaking. Earl mentioned a crossover upgrade - no need to buy and solder with digital active.
Note also that if you are into the whole digital EQ thing then clever maths such as implemented by the DRC or Acourate software packages (or Matlab if you are a smart cookie) can significantly correct the imperfections of the analogue crossover. Effectively such problems are largely min phase and hence can be corrected after the fact just as easily as before the fact (linear + linear = linear)
This is often overlooked...
This is often overlooked...
I've done my speakers actively several times - best case they sound the same, worst case they just don't work right and you end up scraping the whole idea. I just don;t see any advantages, only excessive cost increases and headaches. In virtually every case that they were implimented actively they were converted back to passive.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.