Perhaps there is a disconnect between how you are asking the question and how I am trying to answer it
Combing is the interference pattern generated by the different drivers producing the same frequency, some will be reinforced some not, so you get the peak and dip pattern. This is based on distances and not on the dispersion of the driver. The CTC distance sets the start of where the combing can occur, you seem to understand the mechanics, a line of tweeters will still comb but at at higher frequency if they are closer together. The point wesayso is making is that this also changes with distance to the array. By having the drivers small enough and listening far enough away the combing is pushed up in frequency very high almost to the point where the TC9 driver has very little natural output.
My point is that used alone the driver behaves differently, when arrayed the CTC distance and distance from the array will set the combing and the way the BMR works cannot overcome this. It is possible that the combing could be different at offaxis angles due the the beaming of the driver but I don't see a big change in that in my measurements. The BMR could be better but the TC9 is not actually bad in that regard. The off-axis is smooth and the change is quite gentle within a 30 degree range to either side. Also if the driver doesn't produce the frequency offaxis then the combing pattern won't form or be less severe. That might make the BMR worse because it will comb all the way out, without testing or comparison hard to say.
Maybe this graph of mine will demonstrate. This is 22 separate measurements across my couch 1.4m end to end and 30cm front to back. Above 2K there is not that much variation this is raw no eq or correction. In room 2.8m no room treatment of any kind. The thicker trace is the impulse average so the room contribution has been reduced and that reduces the level in line with the RT60 of the room.
A short line of 10 will also behave very differently to a full line of 36 so I don't think that will be a worthwhile test.
I don't mean to sound negative but I have run the same arguments over and over in my mind before and decided the TC9 to be the best compromise.
I hope this a better explanation of why what I said before was answering the question you asked, at least to me
Perhaps there is a disconnect between how you are asking the question and how I am trying to answer it
This is based on distances and not on the dispersion of the driver.
Hi fluid thanks for your helpful and logical answer. This is our disconnect!
I was asking if combing is affected by the dispersion of the driver. You say not.
I am not so sure!
Wesayso perfectly surmised my thoughts - with dispersion behavior at higher frequencies acting as frequency dependant shading. This is what I was clumsily trying to express.
If I had to gamble it 'could' kind of work between 0 and about 30 degree with the TC9 beyond that the output is too irregular. (Best out till 20 degree) It may still help.
It was one of the reasons for me to do the back mounting with round over as I found some proof that it showed improvement in that area. The proof consisted of measurements I found online and should have saved to disk. Sadly they are lost.
I never did separate tests on this with a single driver.
It was one of the reasons for me to do the back mounting with round over as I found some proof that it showed improvement in that area. The proof consisted of measurements I found online and should have saved to disk. Sadly they are lost.
I never did separate tests on this with a single driver.
Last edited:
It may be more straight forward to think of an extreme example fluid.
Imagine 10x 6" full range drivers which inherently beam significantly at 3000hz.
Now imagine running them from 3000hz up only.
Would they comb? Perhaps not significantly if the dispersion is so poor that their 'beams' don't overlap.
The rest of your post is absolutely common sense. Having read a lot about arrays, I too would only build a full range driver, full height, straight, non-shaded array.
It is my belief this provides the best mix of inevitable compromises.
The only question in my mind is if you are using a 2-3 inch full range driver with similar c-c distances, what is the best dispersion characteristics to have?
Imagine 10x 6" full range drivers which inherently beam significantly at 3000hz.
Now imagine running them from 3000hz up only.
Would they comb? Perhaps not significantly if the dispersion is so poor that their 'beams' don't overlap.
The rest of your post is absolutely common sense. Having read a lot about arrays, I too would only build a full range driver, full height, straight, non-shaded array.
It is my belief this provides the best mix of inevitable compromises.
The only question in my mind is if you are using a 2-3 inch full range driver with similar c-c distances, what is the best dispersion characteristics to have?
My no. 1 requirement was a clean IR, no ringing. Next on my list was a slightly rising top end with similarities in it's off axis behaviour. Meaning the focus was on a driver with fixable faults. You can't fix a ringing top end.
I would value those things in a driver. There are a lot of drivers that fall outside the above scope.
Which makes the TC9 an easy and save choice.
Looking at 20 - 0 - 20 dispersion angle was my focus as the plan from the start was to absorb the first wall reflection (where needed).
Within the array that still just about covers the top and bottom driver in an array just over 2 meters tall compared to the listening distance (shading).
I guess smaller drivers would provide less of this shading.
I looked at many drivers and if I had to start over I wouldn't pick any other.
The TG9 could be a nice choise if you can stand it's appearance. But I'm a paper guy.
I must say I was deeply impressed by the Scan Speak 10F I got from Byrtt and that isn't paper. More akin to the TG.
But I still doubt the 10F would do as well on the bottom end. It is very well behaved and a step up in almost everything else. And as a result you pay for it as it isn't cheap...
The TG has a slight advantage in the bottom end due to its TS parameters.
This isn't helping, is it?
I would value those things in a driver. There are a lot of drivers that fall outside the above scope.
Which makes the TC9 an easy and save choice.
Looking at 20 - 0 - 20 dispersion angle was my focus as the plan from the start was to absorb the first wall reflection (where needed).
Within the array that still just about covers the top and bottom driver in an array just over 2 meters tall compared to the listening distance (shading).
I guess smaller drivers would provide less of this shading
.I looked at many drivers and if I had to start over I wouldn't pick any other.
The TG9 could be a nice choise if you can stand it's appearance. But I'm a paper guy
. I must say I was deeply impressed by the Scan Speak 10F I got from Byrtt and that isn't paper. More akin to the TG.
But I still doubt the 10F would do as well on the bottom end. It is very well behaved and a step up in almost everything else. And as a result you pay for it as it isn't cheap...
The TG has a slight advantage in the bottom end due to its TS parameters.
This isn't helping, is it?
Even with 6.5" drivers I hear no combing starting from 3-4 meters from them.
Wavebourn high end Public Address line arrays (2004 year prototypes) - YouTube
Wavebourn high end Public Address line arrays (2004 year prototypes) - YouTube
A graphic representation;
from Taylor’s paper of line arrays: http://p10hifi.net/planet10/TLS/downloads/taylor-line-array.pdf
dave
A graphic representation;
from Taylor’s paper of line arrays: http://p10hifi.net/planet10/TLS/downloads/taylor-line-array.pdf
dave
Now what would it look like if the drivers were curved? I've curved 8 Pluvia 7's (per side) in my project and it got rid of the combing, as far as I could hear. I did hear the combing effect when the drivers were originally stacked straight vertically.
I was inspired by Epique CBT24K
Last edited:
Read the paper and you can construct a similar graph but with a curved origin. You will still have combing but the pattern will be different.
dave
OK now had a chance to re-read fluids very thorough explanation in more detail and I realise that you seem to be saying that at the distances we are listening to the arrays any questions about combing and dispersion really are obsolete as it will probably only make a difference if you are much closer to the array.
Sorry I am on my mobile last few days as we are away and I haven't really had a chance to properly digest or read what you were trying (very patiently!) to explain to me.
So, I think that you and wesayso have talked me around to the tc9.....I can't see a lot of other options that provide the same set of properties.
I was thinking BMR or SB65 due to closer c-c and the crazy dispersion of the BMR could possible tidy up the combing a little, and if measured at 1m up and down the array it might. But this doesn't mean it would improve things at the listening position...
Sorry I am on my mobile last few days as we are away and I haven't really had a chance to properly digest or read what you were trying (very patiently!) to explain to me.
So, I think that you and wesayso have talked me around to the tc9.....I can't see a lot of other options that provide the same set of properties.
I was thinking BMR or SB65 due to closer c-c and the crazy dispersion of the BMR could possible tidy up the combing a little, and if measured at 1m up and down the array it might. But this doesn't mean it would improve things at the listening position...
To put it simply, I fell in love the the Pluvia 7 sound in a BLH kit (originally meant for a Fostek126En) and realized that 1 driver could not handle the SPL I wanted to hear. I replaced the 126En's with a Peerless NE123W and the sound was much impoved with no 'beaming'. But after I tried the Pluvia's, it revealed that paper cone sound the NE123W's had. Paper cones break up in the upper mids IMO...
EDIT: Woops, I see I posted too late.
I think we are all saying the same thing in different ways, on axis the distances dominate, off axis if the response is the same the distances will dominate, if the offaxis rolls off there won't be as much output to cause the interference pattern to generate. So you could call that frequency dependent axis dependent shadingHi fluid thanks for your helpful and logical answer. This is our disconnect!
I was asking if combing is affected by the dispersion of the driver. You say not.
I am not so sure!
Wesayso perfectly surmised my thoughts - with dispersion behavior at higher frequencies acting as frequency dependant shading. This is what I was clumsily trying to express.
Seems we agree, if the driver doesn't produce the frequency it can't comb.
Me too, but don't mention that to the CBT fanatics
Smooth and consistent, that was the part I wasn't so keen on in the particular BMR you mentioned. The dispersion was very wide but a bit uneven.
No problem, I am glad that my explanations are making some sense to you, it is often hard to explain exactly what you mean on a forum. You can see the combing in the very high frequencies in the measurements but you can't hear it. Smaller CTC or longer listening distance would push it higher for prettier looking graphs
From an objective measurement stance seeing another driver used would have been interesting. The TC9 really is hard to beat when you examine all of the specs and take cost into consideration.
Sacrificing performance elsewhere to fix a problem that isn't really a problem doesn't make much sense to me.
Here's a question probably for fluid and wesayso as you guys have done tons of research and have real world experience. I understand the tg9 may have slightly better upper end than the tc9, but there are two versions of the tg9 the 4 and 8 ohms models. Both with quite different graphs. Are you able to give pros/cons on these closely related products?
Extension-wise the TG9 and TC9 are similar, but the fibreglass cone sounds smoother to me, similar sound signature nonetheless. Both work fairly well in a ~2 litre sealed cabinet. Not much difference between the 4 and 8 ohm apart from the impedance, just my own opinion though.
Can't really answer that for Fluid & Wesayso though...
Can't really answer that for Fluid & Wesayso though...
From seeing measurements and reading a lot of listener reviews I "believe" the 8 ohm version to perform a little better. No hands on experience though.
There's a thread on the TC9, that actually includes a lot of history on the whole family (If you can get trough it). It includes the predecessor of the TG9 and their relation to Scan Speak's 10F. It even includes the Kevlar coned sibling that's largely under the radar. Some interesting links with measurements on almost all of them, though the 4 ohm version is mostly left out of that discussion.
The 4 ohm was popular in Car Audio but seemed to always loose out to the 8 ohm in direct comparisons I have read. It has been a while since I've looked hard enough but I do faintly recall the TG9 FD10-08 having the best measurements.
Back in 2011 I had a pretty hard time deciding between the two (TC9 - TG9).
Having had an Aramid coned mid bass in my car and comparing that to the paper coned that replaced it made me choose the paper coned driver. A very intuitive decision at the time, not based on any facts.
Even Lynn Olsen's write up helped me sway towards the (possibly more polite) paper cones. Can't say I regret my decision, though there have been times where I thought: "what if"...
The money saved wasn't the primary reason for me. At that time I had enough disposable income to go either way. But I thought about clinical sound compared to more laid back and natural sound and together with prior experiences this led me to the paper coned TC9.
Looking at measurements alone the Scan Speak 10F is the winner above 200 Hz in this entire family. The TG9 rates second and the TC9 third in my humble opinion. I don't think you can go wrong with any within that family.
So for the 4 ohm version all I can say about it is based on other people's opinions. Buy one of each and do a comparison?
Edit: 50 eyeballs staring at me also didn't attract me all that much... You however are used to huge eyeballs with your latest Synergy build
There's a thread on the TC9, that actually includes a lot of history on the whole family (If you can get trough it
). It includes the predecessor of the TG9 and their relation to Scan Speak's 10F. It even includes the Kevlar coned sibling that's largely under the radar. Some interesting links with measurements on almost all of them, though the 4 ohm version is mostly left out of that discussion.The 4 ohm was popular in Car Audio but seemed to always loose out to the 8 ohm in direct comparisons I have read. It has been a while since I've looked hard enough but I do faintly recall the TG9 FD10-08 having the best measurements.
Back in 2011 I had a pretty hard time deciding between the two (TC9 - TG9).
Having had an Aramid coned mid bass in my car and comparing that to the paper coned that replaced it made me choose the paper coned driver. A very intuitive decision at the time, not based on any facts
.Even Lynn Olsen's write up helped me sway towards the (possibly more polite) paper cones. Can't say I regret my decision, though there have been times where I thought: "what if"...
The money saved wasn't the primary reason for me. At that time I had enough disposable income to go either way. But I thought about clinical sound compared to more laid back and natural sound and together with prior experiences this led me to the paper coned TC9.
Looking at measurements alone the Scan Speak 10F is the winner above 200 Hz in this entire family. The TG9 rates second and the TC9 third in my humble opinion. I don't think you can go wrong with any within that family.
So for the 4 ohm version all I can say about it is based on other people's opinions. Buy one of each and do a comparison?
Edit: 50 eyeballs staring at me also didn't attract me all that much... You however are used to huge eyeballs with your latest Synergy build
Attachments
Last edited:
That made me laugh wesayso!
Thank you again for your insight and links. I will have a good read of that thread.
Bizarrely to me the 4ohm version has a slightly better response graph and I thought is the model used in the ids25 but I could be wrong.
I have a feeling I will be heading back to the tc9 though.
If only the sb65 was a little cheaper, I would do 30 a side of those...
I am leaning towards an oval cross section enclosure with a chamber for each driver and bending ply to skin the outside at present.
I have looked at schedule 40 pipe and aluminium box section, but both don't quite work with the tc9 frame.
Line array thoughts
Hi Bushmeister,
Here are a few thoughts which I hope will help.
(1) Go with a full floor to ceiling line array, or at least 80% of the floor to ceiling height if you want all the benefits.
(2) There is no substitute for Sd. In my experience there is a clear minimum Sd required to generate life like dynamics. This of course varies with room size / acoustics, required SPL and music types.
The attached is 16 drivers per side, each driver has an Sd of 70 cm square, total of 1,120 cm square per side.... Thats just over two 12 inch drivers per side. Even with the boundary reinforcement benefits from on wall location this was not enough in a medium sized lounge.... I had to add in a pair of 15 inch subs crossed over at 80Hz to achieve life like dynamics using simple " girl & piano" type music.... It still fell way short of classical music or big band dynamics.
(3) Forget about comb filtering issues... Its a technical / measurement issue which is not audible in real world listening rooms at listening distances greater than 1.5 times the line array height (assuming point 1 above is implemented).
(4) Re driver selection.... Many factors come into play but if you really boil everything down I believe there are two vital keys to subjective sound quality: Time domain accuracy (70%), accurate CSD plots reveal all you need to know here. Even on & off axis response all the way out to 75 degrees off axis (look for this on the polar response graphs) is the remaining 30%.
(5) No driver can be Eq'd accurately unless it has a very good polar response... Within plus minus 5dB all the way out to 75 degrees of axis. If you Eq any driver using its on axis measurement you will be automatically be making the same changes at all off axis angles.... This spells trouble in most cases!
Hope this helps and all the very best
Derek.
CIA – New SM 15 monitors | Custom Install Audio
Hi Bushmeister,
Here are a few thoughts which I hope will help.
(1) Go with a full floor to ceiling line array, or at least 80% of the floor to ceiling height if you want all the benefits.
(2) There is no substitute for Sd. In my experience there is a clear minimum Sd required to generate life like dynamics. This of course varies with room size / acoustics, required SPL and music types.
The attached is 16 drivers per side, each driver has an Sd of 70 cm square, total of 1,120 cm square per side.... Thats just over two 12 inch drivers per side. Even with the boundary reinforcement benefits from on wall location this was not enough in a medium sized lounge.... I had to add in a pair of 15 inch subs crossed over at 80Hz to achieve life like dynamics using simple " girl & piano" type music.... It still fell way short of classical music or big band dynamics.
(3) Forget about comb filtering issues... Its a technical / measurement issue which is not audible in real world listening rooms at listening distances greater than 1.5 times the line array height (assuming point 1 above is implemented).
(4) Re driver selection.... Many factors come into play but if you really boil everything down I believe there are two vital keys to subjective sound quality: Time domain accuracy (70%), accurate CSD plots reveal all you need to know here. Even on & off axis response all the way out to 75 degrees off axis (look for this on the polar response graphs) is the remaining 30%.
(5) No driver can be Eq'd accurately unless it has a very good polar response... Within plus minus 5dB all the way out to 75 degrees of axis. If you Eq any driver using its on axis measurement you will be automatically be making the same changes at all off axis angles.... This spells trouble in most cases!
Hope this helps and all the very best
Derek.
CIA – New SM 15 monitors | Custom Install Audio
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.