Hi Bjorn
Thanks for that - that at least pretty much confirms what everybody here has been "suggesting". I don't like to make definative accusations without evidence - however it does appear to me now probable that the figures displayed in the AudioXpress article are not fully representative of a standard anechoic response. We have heard it from the horses mouth as it were.
Catapult - The fact that the port is a rear firing port places it further away from the microphone, impacting it's effect on the response doesn't it? Anyhow thanks for being insulting.
Planet - "Give catapult the benefit of the doubt". Sorry but I don't like to operate on doubt. I like to operate on facts, which I think we now have unless Joe himself would like to come on here and explain the reason for any difference (Prehaps - Bjorn would like to invite him
). The port output is low and as you have said will likely not add musch to this response. I also find it interesting that a sealed enclosure of 12 litres (See Seas standard woofer specs) would have done just as well as the Nearly 70 Litre Thor in terms of low frequency performance - so no need to waste all that space
I would also like to suggest that Seas add an addendum to their description of the Thor on their site explaining these disparities between the graphs. Possibly even add the graph that you have displayed here Bjorn?
Thanks for that - that at least pretty much confirms what everybody here has been "suggesting". I don't like to make definative accusations without evidence - however it does appear to me now probable that the figures displayed in the AudioXpress article are not fully representative of a standard anechoic response. We have heard it from the horses mouth as it were.
Catapult - The fact that the port is a rear firing port places it further away from the microphone, impacting it's effect on the response doesn't it? Anyhow thanks for being insulting.
Planet - "Give catapult the benefit of the doubt". Sorry but I don't like to operate on doubt. I like to operate on facts, which I think we now have unless Joe himself would like to come on here and explain the reason for any difference (Prehaps - Bjorn would like to invite him
). The port output is low and as you have said will likely not add musch to this response. I also find it interesting that a sealed enclosure of 12 litres (See Seas standard woofer specs) would have done just as well as the Nearly 70 Litre Thor in terms of low frequency performance - so no need to waste all that space I would also like to suggest that Seas add an addendum to their description of the Thor on their site explaining these disparities between the graphs. Possibly even add the graph that you have displayed here Bjorn?
This is a very small effect. Remember that the sound coming from the open end of the TL originates at the back of the driver and has already traveled the full length of the TL. You can think of the open end as a second source that is already far behind the driver. The relatively small depth of the cabinet will add some to this path length but it will be a small percentage.
The bass frequencies coming from the open end will travel in all directions, the response should be almost the same as a front firing open end. I think the placement of rear mounted ports, or open ends of TL's, is done primarily to help suppress midrange leakage and maybe to try and get more room coupling of bass frequencies with a rear wall.
Byrd said:Catapult - The fact that the port is a rear firing port places it further away from the microphone, impacting it's effect on the response doesn't it? Anyhow thanks for being insulting.
Sorry for butting in, but you are being unjust to Catapult. You are using the term in this last post of yours "standard anechoic response", and yet you were suggesting that the port response should be "added" to the real anechoic response measurements. You see the contradiction there? In the real anechoic room, the port output will travel around the speaker and reach the mic along with the direct wave from the driver(s). The same thing is true for dipole measurements. The distance from port to the mic will be taken into account automatically, no need to worry about it. If you are after "standard anechoic response" measurement, you only need one measurement done at an anechoic room. If you have an anechoic room, there is no need to separately measure the port output and do anything with it, which has been already said, if what you are after is anechoic response. Port location and distance from (assumed) mic position, can only be a consideration to be added to the measurements when you are doing nearfield measurements and trying to calculate from those the "expected" anechoic response.
Fez - I suppose it can be seen from both points of view. My point is that a rear mounted port will show a different response plot to forward mounted. with the same speaker even though both of the driver and port are radiating the same signal in both instances.
The question is how does this relate to room response. A rear mounted port would generally provide a higher level of bass and to give the viewer of the graph that information, the port can be adjusted to relfect that.
The question is how does this relate to room response. A rear mounted port would generally provide a higher level of bass and to give the viewer of the graph that information, the port can be adjusted to relfect that.
Depends on the room, depends on the placement. I don't think that a single conclusion about bass is possible. The only plot that can be applied universally, and then attempted to be adjusted for a specific room, is the anechoic response.
For a TL and a BR, when the open end or port is producing a significant amount of bass output the driver's output typically is significantly attenuated.
In my experience, testing a speaker and getting accurate and usable response plots is much more difficult then designing and building the speaker. The only test data that I pay attention to is data accompanied by a detailed description of the test set-up, the raw data, and then the steps used to post process the data, including all settings applied in the measurement software, to produce the final plot for presentation. A single plot without this back-up information can be very misleading. Testing is tricky and requires great skill and understanding.
A question to TL guys
This is how Joe D describes he did nearfield driver measurements:
"In practice, the near-field response of each woofer is measured
separately with the other woofer shorted, and then both responses are added to get the total low-frequency response
of the woofer pair."
My understanding is he measures the nearfield of each woofer driver separetely while blocking the other woofer by short circuiting its terminals, then adds the two together. I don't know much about TL's, but wouldn't this cause error? By not driving the other driver, the effective volume behind the driver being driven will be doubled from what it is in real usage, wouldn't it?
This is how Joe D describes he did nearfield driver measurements:
"In practice, the near-field response of each woofer is measured
separately with the other woofer shorted, and then both responses are added to get the total low-frequency response
of the woofer pair."
My understanding is he measures the nearfield of each woofer driver separetely while blocking the other woofer by short circuiting its terminals, then adds the two together. I don't know much about TL's, but wouldn't this cause error? By not driving the other driver, the effective volume behind the driver being driven will be doubled from what it is in real usage, wouldn't it?
Yes! Excellent point.
It would be like putting two drivers in a closed box and measuring the response of each separately. Each driver would see double the volume it would physically see if both were operating together. For a TL, the bass response would be much better then reality if it were measured this way.
It would be like putting two drivers in a closed box and measuring the response of each separately. Each driver would see double the volume it would physically see if both were operating together. For a TL, the bass response would be much better then reality if it were measured this way.
Compare the calculatued response plots in posts 13 and 57.
Post 13 - two drivers running together in the Thor enclosure. Rolled off bass as has been reported by a few people who have constructed the project. Also, closer to the SEAS measured response.
Post 57 - one driver running in the Thor enclosure. Nice deep flatter bass, closer to the measured plots presented in audioXpress.
Maybe both inputs are correct, calculations and measurements, for the conditions in which they were obtained. Unfortunately if this trail of logic is correct, the actual speaker response is closer to post 13 which is not that great for a TL enclosure.
Post 13 - two drivers running together in the Thor enclosure. Rolled off bass as has been reported by a few people who have constructed the project. Also, closer to the SEAS measured response.
Post 57 - one driver running in the Thor enclosure. Nice deep flatter bass, closer to the measured plots presented in audioXpress.
Maybe both inputs are correct, calculations and measurements, for the conditions in which they were obtained. Unfortunately if this trail of logic is correct, the actual speaker response is closer to post 13 which is not that great for a TL enclosure.
An odd error considering. But then again, giving him the benefit of the doubt, he didn't / doesn't appear to have a huge amount of experience with TL behaviour, which is rather different to most other speakers.
A FatThor? Ouch. Ever get the feeling that would be one BIG cabinet?
Here's a thought: I roughly re-calculated the Ariel MK4 cabinet for the Seas drivers Thor uses. Same line-length and folding geometry as Ariel, cross-section scaled up for the larger drivers, 1.5"x2" (WxD) port at the base for each line. Not the best TL response available. But, if you want to D'Appolito these drivers, this is probably one of the better ways to do it without getting a cabinet of epic proportions and a swift divorce. Add 6db to overall response for the paralleled drivers, as usual, and remember to factor baffle-step loss into things as well.
A FatThor? Ouch. Ever get the feeling that would be one BIG cabinet?
Here's a thought: I roughly re-calculated the Ariel MK4 cabinet for the Seas drivers Thor uses. Same line-length and folding geometry as Ariel, cross-section scaled up for the larger drivers, 1.5"x2" (WxD) port at the base for each line. Not the best TL response available. But, if you want to D'Appolito these drivers, this is probably one of the better ways to do it without getting a cabinet of epic proportions and a swift divorce. Add 6db to overall response for the paralleled drivers, as usual, and remember to factor baffle-step loss into things as well.
Attachments
I am sure that one of you enterprising gentlemen (or perhaps 2 or more in collaboration) could write an article for AX or Elektor with an improved design -- I am plowing through Mr. King's mathematics this morning. Here's the link if not elsewhere provided: http://www.quarter-wave.com/
MJK said:
Well, I already went to the dump, got the Volvo washed and filled, went to Shop-Rite, picked up Barrons and the Weekend FT, harvested the green tomato crop and am working on getting the basil and sage in before the first frost -- it helps when the little lady get's you out of bed at 6:30 a.m. on Saturday.