[FullRangeMan]

Do you really want to know?

Yes I do, cause I have no chance to listen a LA like this. And this IDS25 is very well quoted in the manufacturer Site.

[Danny Richie]

When I went into the room to see these I was the only one in there (besides the guy running the room). So I handed him a familiar disc and gave them a listen.

Having no crossover they had no baffle step compensation. So from about 1000Hz and down they got a little lean. To compensate for the lack of bass the speakers were pushed right up to the wall.

They were positioned on the long wall and the rooms are 13 feet wide so I was about 10 to 11 feet away from the front of the speakers.

I also quickly noted that the top end was not there. I'd say everything from about 3kHz and up was really rolled off. All the air and space that went along with the piano that I was listening to was gone. Imaging was 2D and sound like it was all playing from the wall forward.

They remind me of the old saying about Bose. No highs, no lows.

To make matters worse, not only did they sound badly but they were very expensive.

[FullRangeMan]

Thankyou, as I live in Brazil any first hand info about a big LA is valuable to me.
Regards,  Gustavo

[Browntrout]

Thanks very much indeed for your advice Paul Hynes. 

[Paul Hynes]

Hello Danny,

I appreciate your concerns regarding measured comb effects when using drive units in an uncompensated vertical drive unit array and understand that your comments are based on your experiences as a loudspeaker designer and from personal experience with line design.

I can actually hear the comb effect close up to the B200 array, but the further away I listen, the less obvious it becomes and, at my usual listening position at around twelve feet away, it is not obvious at all. In fact, at the listening position, the musical program is very coherent and I value this parameter above all else.

In my opinion, more damage is done to the musical coherence, and the image dimension and stability, by wiring line drivers in series parallel. This essentially places a highly reactive element in series with each drive unit. The waveform corruption caused by this is not trivial and affects time as well as space.

Hello Browntrout,

My line source was not designed for commercial reasons. I was involved with loudspeaker design a lot in the 1980s and 1990s so have a general understanding of the main principals involved, although I would not consider myself an expert in this field. Measurement systems were pretty basic in those days and we did not have access to the test equipment available today, so we relied a lot on our ears and I still do this. Test equipment and simulation systems are useful tools but the final arbiter is how does it sound. I did learn that some parameters are important and some are not. Sometimes you can go against theory and get a good result if that parameter is not important.

Regards
Paul

[opnly bafld]

When I went into the room to see these I was the only one in there (besides the guy running the room). So I handed him a familiar disc and gave them a listen.

Having no crossover they had no baffle step compensation. So from about 1000Hz and down they got a little lean. To compensate for the lack of bass the speakers were pushed right up to the wall.

They were positioned on the long wall and the rooms are 13 feet wide so I was about 10 to 11 feet away from the front of the speakers.

I also quickly noted that the top end was not there. I'd say everything from about 3kHz and up was really rolled off. All the air and space that went along with the piano that I was listening to was gone. Imaging was 2D and sound like it was all playing from the wall forward.

They remind me of the old saying about Bose. No highs, no lows.

To make matters worse, not only did they sound badly but they were very expensive.

+1
I couldn't believe they were actually letting people in the room to listen to the things. 

Lin

[opnly bafld]

Has anyone implimented multiple Lowther full range drivers in an open baffle array? Say four or five drivers per side. With the idea being to generate good bass without losing speed and a boomy cabinet. Very interested to hear the results if someone has been brave enough.

Using a single fr driver with multiple ob bass only drivers will accomplish this quite well.

Lin

[Danny Richie]

I can actually hear the comb effect close up to the B200 array, but the further away I listen, the less obvious it becomes and, at my usual listening position at around twelve feet away, it is not obvious at all.

It should be very apparent as the peaks and dips will be very sever even twelve feet away.

For instance, let's say that you draw a line from your ear level to the center of the array and get 12' exactly. Then from your ear level to the top of the array you get 12' and 5".

The differential is 5".

A 5" wavelength is about 2.6kHz.

So you will have cancellation down to 2.6kHz and the effect will be seen (with a dip at that fundamental) carrying over in each direction covering that full octave.

The dip will be in the 15db range.

Likewise where they actually couple and are in phase then there will be at least a 6db peak.

And that is from 12 feet away.

Looking at the response of a single B200, it looks like it would be hard to listen to a single unit.



From only 100Hz to 8kHz it rises 15db. Wow, talk about your hot top ends. Then from there to 20kHz it drops 20db. Boy do those drivers need a corrective filter.

At lease running a group of them will elevate the bottom end (where they all couple) enough to even it out with the top end. At least it would be close if you averaged out the peaks and dips.
 

[Paul Hynes]

Danny and anyone else interested in open baffle line arrays,

I am not disputing that there is a comb effect, but stating that from my listening position, with the B200 line, it is not obvious.

Danny, your mathematical analysis is incomplete as it only considers two vertical points on the line so the cancellation at 2.6 KHz will not be in the 15 db range. Assuming one were to sit with ones ears level with the dead centre on the line, the output at the centre would display cancellation effects with the top and bottom drive units at, for example, your stated 2.6 KHz. This would leave a dip at 2.6 KHz as you say. However that would leave the equivalent of five drive units outputting their normal level at this frequency as no other pairs of drive units suffers the 5 inch distance discrepancy between the drive unit position and the listening position. The net drop in overall line array output level at 2.6 KHz would be around 6 db. As I do not listen with my ears aligned with the dead centre of the array this is actually still not representative of real life in my listening position.

I had heavy-duty brass spikes turned to fit the aluminium frame of the line array to provide three-point rigid coupling to the floor and these together with the seating height in the listening position place my ears slightly above centre of the third drive unit from the bottom of the array. Now only two points on the array show the 5 inch discrepancy instead of the three points if I were to raise my ears to dead centre of the array. Now the output from six of the drive units are giving full output at 2.6 KHz and the output level drop, from the whole line, at this frequency, will be closer to 3 db. This picture is repeated with different pairs of drive units showing different alignment distances, and therefore different frequencies, all along the line. The net effect is a level variation of around +/- 3 db at any given frequency, due to comb effects, along the line.

Regarding the B200, I think you are underestimating the performance based on what you see from the published response curve, which I suspect was recorded using a brand new drive unit. Before I bought drive units for the full line arrays, I tested a pair on open baffles for around a month of use, for approximately 5 hours each evening, to see how they behaved. Initially they did sound somewhat ragged at the top end and light in the bass/upper bass areas as you would expect from the published response curve. After a month the top end was much better as the presentation was noticeably smoother and the lower frequencies filled out more. A run-in period is essential before judgement. The most important factor during this period was the improvement in overall coherence. After the month was over I new instinctively that the drive unit was going to work well in an open baffle line source array, and this has proved to be the case. Using eight drive units has, as Danny stated, filled out the bottom/midrange giving a well-balanced frequency response through the operating bandwidth. The high end is no longer sounds ragged now that the drive units are well broken in. Whilst the HF is not as well extended as would be provided by adding a tweeter array, I am not prepared to sacrifice the coherency of the line by adding tweeters and having to use a crossover. The reproduction of instruments with lots of high frequency energy is very good. Cymbals sound like cymbals, not someone thrashing a dustbin lid. This is extended enough for me. The bass end goes very low with low distortion and, of course no cabinet effects. Eight 8 inch drive units can easily reproduce the weight of a bass transient from a large drum and I feel no need for a subwoofer.

There are now a lot of B200 drive unit users around the world, and, whilst we users would not claim that the B200 is the best drive unit on the planet, it is damn good at what it does. It is designed to work on an open baffle and it plays music well, once run in, irrespective of any theoretical issues with respect to the published data.

I must reiterate how important it is to wire line source drive units in parallel to get the best performance from the drive units. If you wire the drive units in series/parallel this will cause electrical interaction between the drive units that will impede their correct operation as a line. If anyone wants to experiment with parallel wiring I can provide circuit information for amplification that can drive the resultant low impedance load and I will be happy to help with any technical issues like matching the amplifier to your line. I posted information about the self-biased depletion mosfet output stages, that I use in my system, a while back in the Lab thread. I have recently resurrected my older design using diode biased lateral mosfets, as depletion power mosfets are relatively expensive and difficult to obtain through normal supply chains. When I have found time to finish the drawing I will publish it in the Lab for general reference.

Regards
Paul

[JohnR]

Paul - apologies if I missed it, but what is the size of baffle you used?

JohnR

[Paul Hynes]

Hello JohnR,

The baffles are 78.75 inches tall and 15.75 inches wide. The drive units are offset from the centre to avoid symetrical diffraction from the edges and the baffle edge is rounded to reduce diffraction effects. The drive units are also rebated to avoid diffraction from the edge of the drive unit baskets.

Regards
Paul

[Danny Richie]

Yes Paul, you will get some fill in from the other drivers in that range. Overall room response (room reflections) will fill in some areas as well. But as frequency increases and wavelengths are shorter you will still wind up with some large dips and peaks.

It will be a choppy ride, but if that ride is for you, then enjoy the ride.

I would have personally only allowed one driver to play full range and brought the others in at the lower ranges. This would give you a smoother response and sensitivity would have still been the same.

If you want to add a bit of upper end air without causing cancellation effects in the top couple of octaves, then you can try adding a rear facing tweeter.

[opnly bafld]

I would have personally only allowed one driver to play full range and brought the others in at the lower ranges. This would give you a smoother response and sensitivity would have still been the same.

I can attest to this working quite well as I have five B200s, one running full range and the other four being used for bass.
When I first tried them together with the four drivers playing up into the lower midrange everything sounded strange and the soundstage especially was very bad.

Lin

[Paul Hynes]

Danny,

At the risk of repeating myself I am not getting a choppy ride at the listening position with the B200 drive units. I really would be interested to know why you think I should be suffering large peaks and dips in the upper ranges.

How do you get a smoother response and maintain sensitivity in the upper frequencies using one drive unit running full range and the rest only running in the lower ranges? I?m getting 105 db for 1 watt at present and I would be reluctant to reduce the sensitivity back down to the sensitivity of 96 db for one drive unit.

This is an open baffle design. There is plenty of air from the rear reflection.  Fortunately the time delay from this rear reflection is long enough not to interfere with the initial wave-front from the front of the speakers and the image is not destroyed by it. It just sounds like hall ambience.

Opnlybafld,

Did you wire your B200s in series parallel or in parallel? If you wired them in series parallel I am not surprised they sounded strange and gave a poor soundstage. I can compare both wiring systems on my lines very easily by changing a few 4mm plug positions and the difference is not subtle. Series parallel wiring looses all coherence and this is what Danny thinks I should be hearing from my line source. This begs the question how can you appraise the performance of a line array under the conditions imposed by series parallel wiring. You may be attributing audible problems to the wrong cause.

Regards
Paul

[opnly bafld]

Opnlybafld,

Did you wire your B200s in series parallel or in parallel? If you wired them in series parallel I am not surprised they sounded strange and gave a poor soundstage. I can compare both wiring systems on my lines very easily by changing a few 4mm plug positions and the difference is not subtle. Series parallel wiring looses all coherence and this is what Danny thinks I should be hearing from my line source. This begs the question how can you appraise the performance of a line array under the conditions imposed by series parallel wiring. You may be attributing audible problems to the wrong cause.

Regards
Paul

Hi Paul,

I had them series/parallel for the easy 6 ohm load, but since I have an amp that can handle the 1.5 ohm load resulting from four in parallel I will try this.  The way I'm using them now works fine for me because they get loud enough to satisfy my listening habits.

Lin

[Browntrout]

A good discussion indeed. Thanks especially to you Paul for sharing your first hand experiences with this sort of speaker. 

[Danny Richie]

Paul,

Regardless of how many you put in a line, you will not increase the output levels in the top octave. You will still have the same output as a single unit. So if one unit will output 80db with a 1 watt input (as seen in the measured responses) and you are getting 106db in the lower ranges then you already have a variance of 36db.

And whether you wire them all in parallel or series/parallel it will not change the comb filtering behavior.

And Paul, do you have any measurements of this speaker of yours?

And there is no way that you can't get peaks and dips in the response, especially in the top octaves. It is the simple physics of it.

And the reason I suggested the rear firing tweeter is that the output level in the top end goes away. From 7kHz to 20kHz it looses 20db. That is going away.

To me, and this is just me, it is too choppy to listen to a single unit much less what the response will look like with multiples. I am used to +/-1 to 2db over the whole range.

[Browntrout]

We all have different approaches to hifi and perhaps speakers more than any other section show this due to their very characterful nature.
 One of the most important parts of discussing anything is to be able to understand each others approach and accept that what one person likes very much another might not like or think is going to work.
 I personaly will take experience over simulated ideal world calculations and this is why I asked the original question in the first place. My limited experience of speakers is one of a very small but focused window into an event and I look forward very much to a more encompassing experience where I am in the music rather than looking into it, if it is possible. 

[Rudolf]

Just to illustrate Dennis' point:



This is a simulation of an Visaton FRS8 line array, showing the vertical response at 0, 30 and 60 degree. First diagram is for two drivers spaced 20 cm with ear height just between them. For the second diagram (to the right) I added one driver above and one below (each at 20 cm distance too). Third diagram is for 6 drivers and the last (bottom right) for 8.

IMHO it shows how adding drivers will smooth the response at lower frequencies somehow, but the peaks at (in this case) 3.5 kHz and multiples of it stand out all the time.

Listening distance is assumed at infinity.

[opnly bafld]

Paul,

Tried 4 drivers in parallel, bottom driver at the floor and top driver at about ear level, the lower 2 drivers are facing backward (this probably helped since there would be fewer mids/highs), no filters, baffle tilted back a few degrees. Works surprisingly well, soundstage was good if a little lower than normal (this could likely be improved with some tweaking). Sounds much better than it has a right to.

Also tried basically same set up, but moved up one driver since there are 5 on the baffle. Top 3 drivers facing forward with bottom driver facing back, top driver @ one foot above ear level and bottom driver one foot off the floor. Soundstage was good, a little higher, but not quite as focused, also bass was thin because of loss of floor gain.

Went for a triple play and used the top 2 drivers with the bottom 2, so not really a line array because as mentioned the bottom drivers face backwards and there is a foot gap between the 2 pairs of drivers. This arrangement sounded the best to my ears with a good soundstage and good bass.

All of these configurations sounded much better than I expected them to since there were no filters (other than mechanical- provided by the driver magnets).

I used 3 drivers for quite a while with the top full range driver near ear level and the bottom 2 bass drivers as close to the floor as possible, with a single amp all drivers in parallel, biamped with the bass drivers in parallel, and inductors inline with the bass drivers in both cases.

I have benefited from the experiments, from now on I will always wire the bass only drivers in parallel (will probably triamp for a 6-3-3 ohm load), but I still plan on using only one driver full range. 

Lin