How To Choose The RIGHT Speakers For Your Home Theater

If you'd like to book a private acoustic consultation with Matt or interested in purchasing the best Home Theater equipment please send your request through our website:
www.poesacoustics.com
We proudly carry Perlisten, KEF, Q Acoustics, JVC and Sony projectors, Artnovion acoustic treatments, Trinnov, Denon, Marantz, Samsung and Sony TV's, Sonos.
follow Matt on Facebook / poesacoustics
In this video I discuss what I think are the most important attributes for choosing speakers. I take a performance based approach to designing home theaters. As such, I choose speakers that have the technical capabilities necessary to achieve those performance metrics in my design. Using this same approach, you too can choose the best speakers for your needs.

[@SwirlingDragonMist]

I remember you touching on this concept in a video with Gene, and It stuck with me as being a very good insight I hadn’t ever heard anyone else touch on. I very much align with the quality over quantity philosophy, yet have drifted into the seduction of high channel count, so it really helped me to reconsider the tradeoffs that potentially lay ahead for me in that pursuit.

[@kirkcunningham6146]

I went with projected sound, currently have the JBL Studio 5 series with the 590's as my mains, the 570 for the center and 530's for surrounds. Awesome speakers. Made very well and very good quality. Love the sound for two channel and Home Theater. Dedicated Yamaha Pro-Audio AB amplifiers.

[@decaf77]

I’ve been following your content with Audioholics and now on your own channel. I really appreciate that you’re sharing your knowledge, I’m learning so much, new sub!

[@PoesAcoustics]

Awesome! Thank you!

[@Nightjar726]

It would be cool if you guys did get your hands on some Dutch&Dutch 8c or Kii3 speakers to measure the subjective differences of wide and narrow dispersion since these have options in their DSP to do this. They also have an option to have the speakers be perfectly time aligned and not. Would be great to hear your or Audioholics impressions on those. Great vid man

[@Tulipo08]

Where were you 2 years ago when we were building our theater! Great advice and resource of knowledge. Subscribed!

[@PoesAcoustics]

Awesome! Thank you!

[@DTA_Crypto]

Great video thats where im at now saving for my end game I KNOW I WANT PERLISTEN For my speakers but I keep going back and forth between S series bookshelves vs R series R5t towers? I have a open living room concept (3000 cubic ft) now I'm thing like you suggest series for the front r series for the surrounds and atmos?

[@miscreant1739]

@PoesAcoustics could you please offer your thoughts on open baffle speakers? I see some audiophile types hyping them up but I'm pretty skeptical and I can't find good info from sources I trust.

[@nicolasben-fredj8162]

What are your thoughts on front wides are they worth factoring in, both with and without regard for budget?

[@RickRalstonAsumendi]

Before I get to my point of contention, I want to say that I do agree with most of what you said in this video. Your advice not to spread the budget thin across too many speakers rings very true for me. I try to turn people away from garbage speakers all the time because they want "Atmos under $1000". At the end of the video, you claimed that Arendal and the Perlisten R series are not even in the same ballpark. I want to challenge you to quantify your ballpark and explain why you chose those numbers. Here is my thinking for why they ARE in the same ballpark. We don't have the same quality of data for all models, so we'll have to do some interpolation. Let's give the R4b the benefit of the doubt and say it measures exactly as well as the S4b in terms of frequency response linearity and directivity. If we compare to the Arendal 1723 Bookshelf S THX, the preference scores with sub (assuming both are bass managed) are 0.5 point apart (8.1 vs 7.6). I consider this the same ballpark based on a comment that Dr. Sean Olive made in the linked thread: "The statistical confidence of the predictions does not allow comparisons beyond 0.5- 1 point so it is moot to get down to 1 decimal place." audiosciencereview.com/forum/index.php?threads/i-cannot-trust-the-harman-speaker-preference-score.31454/post-1118260 Now let's consider output. THX rates both of these bookshelf speakers as "THX Certified Ultra", so clearly they are putting them in the same ballpark. Perlisten claims a maximum output of 107.6 dB @1m (100-20kHz). I don't see where Arendal claims a peak output, but the RMS power handling is in-line with the R4b and it did well in Erin's compression testing up to the 102dB test limit. Can you help me understand what puts the Perlisten R series in a separate ballpark than the Arendal 1723 THX series?

[@PoesAcoustics]

You claim a lack of comparable data but that is because you don’t trust the source of other data. It’s completely fine. And in fact, Erin’s data isn’t totally consistent with other measurments using equally valid methods, including Klippel, but better mics. The Klippel NFS needs to find the acoustic center and doesn’t always get it right. It appears that the data Erin obtained may have actually been below or above the true acoustic center as he shows some on-axis stuff that is actually above and below the central vertical window. I don’t really know why or what went wrong, but I have other sources of data that I trust in that range. Where did you get a preference score for the speaker. Erin doesn’t calculate it and doesn’t provide the raw data. Did someone trace it out? Sean and I have talked about this before. So here is the thing. The preference score took basically the biggest factors and assembled them in a way that allowed a fairly decent prediction of preference. You previously mentioned some great examples of where the scores fall apart. There are speakers where the output is actually so limited that even in a small to medium room they wouldn’t have sufficient output for home theater use, but otherwise measure very well. Two of the speakers you listed had excessive distortion by 200-300hz at under 100dB of output at 1 meter. As I mentioned in the video. I think quantifying is very hard. Something that reflects a numerical small change can lead to a subjective big difference for some but not all listeners. Some are more sensitive to certain things. For whatever reason the differences in smoothness of the listening window of speakers leads to small relative changes in the preference score. That actually makes some sense but we also know that some folks have shown themselves to be more sensitive to these resonances than others. Amir is a good example. The Perlisten response is smoother when looking at the listening window, early reflection, and power response. You are putting a ton of faith in this model. Sean and I also talked at one point about various findings related to similar scoring speakers sounding quite different. He made a comment that the model doesn’t account for distortion and once speakers achieve a certain degree of excellence the distortion likely does start to matter. Well comparing these two, the distortion of the Perlisten is better. It’s lower at both tested levels and doesn’t rise as much. It’s output compression behavior is clearly better too. It’s true that the Arendal did very well, but isn’t it pretty obvious the Perlisten did better? One problem I often have with these videos is that I have access to measurement data I am not allowed to share. I can make general comments like I do, but I can’t say, hey I know this because here is the actual data. Until someone like James, Amir, or Erin out their own version out there and it’s accurate I have no good way to back it up. At the same time, if I ever divulged that data I would lose the trust of these manufacturers and access. To quantify in the ballpark seems moot. We can’t do that. It is my opinion that they aren’t in the ballpark based on my preferences and experience with the product. You disagree based on a preference score that I don’t believe is adequate to characterize better beyond a certain point. Short of Sean putting both speakers into his listening test and seeing which one emerged victorious, we can’t really know if my or your assertions are most fair. Keep in mind I never said that spending more gets you substantially more. My point was simply that within the confines of well engineered products, spending more can still eek out more performance. The last thing I would say is that when I made that comment I wasn’t thinking about the S4B or R4B. I was thinking about the towers. I happen to think that the bookshelf models are the least good and interacting of the speakers.

[@RickRalstonAsumendi]

@@PoesAcoustics I am going to address each point, but I want to keep the discussion in context. My gripe is with the hyperbolic claim that these two speakers are "not in the same ballpark". To be clear, I am not trying to claim the Arendal 1723 Bookshelf S THX is the exact equal to the Perlisten R4b. The only Spin data I have seen on the R4b is significantly smoothed compared to the Klippel NFS data from Erin. It would not be fair to directly compare scores computed with the smoothed Perlisten data to those computed with Klippel NFS Arendal data. I thought it would be more than fair to assume the same level of Klippel performance between the R4b & S4b, despite the R4b not being as good per your own admission. The vertical contour for Erin's S4b data looks to be aligned with the acoustic center in the tweeter range, so I'm not sure I see the evidence of the inconsistency you mentioned. There is an ASR forum member that compiles Spin data from a number of sources. I think Erin might actually send the raw data directly to him, but I am not positive on that point. Erin has actually started to compute the scores on more recent reviews in the "On-Axis Response Linearity" plots. pierreaubert.github.io/spinorama/ I take a different approach to the score than you do. I disagree that they "fall apart" for the speakers I mentioned under the other video. I think of it as a tonality score and decouple it from the factors that it does not evaluate like output and distortion. You picked out the two smallest and most limited speakers from my list, but did not address the larger speakers like the Kef R3 or Ascend CMT-340 SE2 that can easily reach 10dB below reference in a medium room. Small changes in listening window response leading to large perceptual differences sounds like conjecture that would need to be validated with more research. I could agree the S4b is slightly smoother. The frequency deviation (300Hz-10kHz) is 2.0dB compared to 2.6dB on the Arendal. Some of that difference is from a bit more upward tilt on the Arendal tweeter. Again, are we really talking about a "different ballpark"? Where are you putting the fences? For sure, the S4b does better than the 1723 Bookshelf in distortion and compression. It better as a THX Dominus rated speaker! Without the data to compare myself, I am going to lean on THX here to say the R4b and 1723 Bookshelf are in the same "Ultra" class for output capability within distortion and compression limits. If we can't quantify the ballpark, then the comparison is subjective and no more reliable than any sighted listening test. You seem to be contradicting yourself by saying that we are now eeking out the last bit of performance with the Perlisten R series rather than achieving substantial gains to land ourselves in a different ballpark. I can agree with this characterization. I only chose the bookshelf models for comparison because the S4b is the only Perlisten speaker with publicly available Klippel NFS data. Arendal also has standmount monitors and towers with higher output capabilities.

[@PoesAcoustics]

@@RickRalstonAsumendi the spin data is not substantially smoothed. That isn’t true. The data you mention is the factory data I assume. It is true anechoic data with no smoothing. I never claimed my comment was an objective fact. It was meant as my opinion. Proof for both variance in detecting resonances and the variation can be found in the studies that looked at this topic. I don’t have them right at my finger tips but I certainly recall both variance in what was reliably possible and that the small differences were detectable by some which were quite fine. In any case. This seems like a pointless discussion. The video was trying to make a point. I think the point is valid. You seem to want to pick it apart. The Ascendo has a terrible distortion resonance in the tweeter, unless we are looking at different products. Amir noted he felt it was audible. It was their at pretty low levels. Where do you see evidence it will do -10dB of reference. The r3 can sure. But why are we setting the bar so low? -10dB isn’t a standard. It’s your own moved goal post.

[@RickRalstonAsumendi]

@@PoesAcoustics I tried to make it clear from the start that I largely agree with the video and just had a bone to pick with the hyperbole of the Arendal comparison. I have come to respect and value your opinion because you generally do your best to support it with objective facts. The point of the discussion for me is that exclusive comments to the effect of "your experience is not in the same ballpark if you can't spend $X" should be backed up by more than opinion. Maybe smoothed is the wrong term, but they are not the same resolution as evidenced by comparing the Klippel NFS data on the S4b to Perlisten's data. Regardless of which one you trust as "right", it is more fair to compare scores calculated from the same measurement source. I don't think Amir has reviewed the SE2 version of the CMT-340. Ascend bought a Klippel NFS and has been making big improvements to their lineup. Anyway, the topic of target volume is a totally separate disagreement we have on another video. The context of the list I provided was your challenge to name speakers that had comparable tonality to the S4b that were just limited in output. In my experience, -10dB is a common target among home theater enthusiasts and represents a much more attainable cost for quality sound.

[@Nightjar726]

Hang on man. And I say that with all due respect. But as far as wider vs narrow dispersion and the sweet spot. How is that true with bass freq ? Since bass freq are super wide and cover 360deg. How then does an upright bass sound smack dab in the middle ( if it was recorded that way) when bass is let’s say the widest dispersion? Legit question. I can see what you are saying with higher freq. but how does imaging work with bass? Is it the leading higher freq that make up an uprights bass reproduction which give us the spatial cues ?

[@theHyghwayman]

😲 svs not the best speaker All joking aside, thank you brother for sharing your knowledge as I've always wondered what sensitivity I should look for. Happy New Year

[@PoesAcoustics]

I don’t happen to think so either. I just needed an example of a speaker that I knew would not work great in a larger theater. What may surprise many is that a lot of speakers wouldn’t work in a larger theater space. Quite a few otherwise good speakers wouldn’t have the output needed for a 3500 cubic foot theater or larger. Something with a 12’ or greater listening distance.

[@powerguymark]

@@PoesAcoustics this video just came up in my KZbin feed. I've enjoyed your commentary on audioholics for quite some time so I subscribed to your channel. I've never heard you say anything that just didn't seem to ring true, although I'm sure there are many people that take you out of context at times. Based on your comments I believe my first attempt at a dedicated home theater has a lot of happy accidents. It is a 40 by 30 x 7 space, but I'm only using about 2100 cubic feet 20x15x7. Since I really didn't have room for an overhead I'm using an 84in TV. I mounted a 12-foot blackout curtain behind the TV and angled 16' blackout curtains outward in a V shape with four seats approximately 11' from the screen. I bought 3 pairs of vintage Polk audio LSI9's with a LsiC 704 center channel in my 9.1.4 base layer. I'm driving my LCR with an Emotiva BasX A3, my L&R surround, surround wide and rear are also each being powered with a Basx A3. That's providing 200wpc into my 4 ohm LSI9's freeing up my AVR to drive 3 pairs of SVS prime satellite Atmos speakers which are flush-mounted a aimed towards my primary listing position. I'm using two small SVS PB 1000 pro subwoofers connected to a dedicated 20 amp circuit for future upgrade when I can afford it. I have all the base level tweeters mounted nearly equidistant at ear level. So far I am very pleased. But I would like your thoughts. I am a complete novice when it comes to understanding room treatments (reflection absorption diffraction) the existing ceiling tiles that I did not take down I simply covered up with 1" black foam tiles just to make the room darker. TYIA

[@chandan6119]

how can one identify that a particular speaker can do reference level (at least clean 95-100db peaks)in a 3300ft treated room ? i know very high sensitivity, 92db plus speakers can achieve that easily, but they tend to be horn speakers or very costly speakers, which may sound harsh compared to dome tweeters. but there are speakers like monitor audio bronze 100 bookshelf's which are rated at 87db sensitivity and they are advertised as 110db max spl for pair (mostly measured at 1 meter),does this speaker able to achieve that clean peaks with sufficient power ? some manufacturers doesn't even advertise peak spl levels. in majority of reviews they only show frequency response and distortion at lower volume levels. only Erin shows compression and distortion levels at 102db in his reviews. is there any other way to identify clean max spl of a particular speakers using spec or reviews? I am looking for high output speakers for my lcr below 1000$ per pair with soft dome tweeters, due to harshness of horn tweeters. I couldn't able to get the max spl of many well reviewed speakers. also i am thinking of going for bookshelves like monitor audio bronze 100 for front lcr rather than going for towers to reduce cost and i have 2 18inch gaint subwoofers, is that good idea? can a bookshelf produce good 100db peaks in midrange and highs?

[@PoesAcoustics]

I’ll do a future video on how to calculate the maximum in room SPL capability of a speaker. Truth be told you can only know if you measure. However there are ways to use the sensitivity, maximum power handling, and a room gain fudge factor to estimate it. I’ll actually be developing a protocol for this that we will use in our Audioholics reviews and doing some testing on a number of speakers to confirm the results. The simplest solution is to choose a THX rated speaker. That has gone through rigorous testing for certification that ensures a speaker can achieve a certain SPL with minimal compression and distortion, has good frequency uniformity and good directivity. Keep in mind that reference levels refers to maximum SPL of 105dB at the listening position from each speaker. The lower the RT60 the less room gain. Larger the room and father back the listening position, the louder it must be. While reference levels are very loud, they are the standard at which movies are mixed. The big problem with turning the volume down is that voices and quiet effects get buried in the noise floor. If you turn it down you really need to add compression.

[@chandan6119]

@@PoesAcousticspresently, i use online spl calculator or home theatre gurus method to calculate the max spl at 12feet distance. It would be helpful if you can make a video regarding this. i can use the sensitivity and max rms wattage specs of speaker for calculation,but is that accurate?what if the distortion is very high or it may have compression issues. I was wondering why my speakers are so quiet in my newly treated theater room(rt-0.3ms),it felt very loud in my living room before. As you said,i want quiet scenes and effects to stand out with peaks with realistic volume. i was wondering does a 87db bookshelf monitor audio bronze 100 can achieve atleast 95db peaks at 12feet distance. Theoretically it can do it with enough power,even specs say so.but ,im not so sure . Also please do a video regarding bookshelf vs towers for dedicated theaters using big subwoofers. I have been following your videos from audioholics,now your own channel also has good videos and are really informative.