What are people thoughts on using bookshelves as LCRs?

[Snoochers]

I'm likely building a home theater at some point. The designs are still in the works but the front row might be 12' from the front speakers and the rear row might be 17' from the front speakers. I'm toying with the idea of having bookshelf speakers as LCRs since they're considerably cheaper. For example, the KEF R3 and HDI-1600s are much cheaper than their tower counterparts ($4200/$3750 for three vs. $8100/$7500 - CAD). The sensitivity of the R3 and HDI-1600 compared to the towers is lower by about 1 to 5 depending on the models.

I've looked at loudness using the SPL calculator at https://myhometheater.homestead.com/splcalculator.html

Using the KEF R3 as an example: Two KEF R3s from 17 feet away with 70 watts powering them can produce about 97db. This is not quite reference, but does seem quite loud for peak home theater uses. Or am I wrong?

What are people thoughts on using bookshelves as LCRs? Can they get loud enough? The bookshelves don't go as low but with subwoofers does that become irrelevant?

 

[A800]

You want to have headroom.

 

[Snoochers]

You want to have headroom.

This sounds like sense to me, but can you explain? Are you saying I want the sound to go above 97db? Is 97db not loud enough to account for headroom? Would a sensitivity 1 or 3db higher make enough difference?

 

[A800]

The problem is rated sensitivity is rarely accurate and also can't really be.
Im just saying better get a bigger amp and bigger speakers for some headroom.
Components at their limits rarely sound good.

 

[Snoochers]

The problem is rated sensitivity is rarely accurate and also can't really be.
Im just saying better get a bigger amp and bigger speakers for some headroom.
Components at their limits rarely sound good.

I suppose this is part of my confusion. The KEF R7 have only 1db more of sensitivity as rated. Will that truly make a difference? Are the R7 really not capable of being decent home theater LCRs? That seems strange to me.

 

[A800]

1db is negligible, if it's true.
According to this site https://www.hifinews.com/content/kef-r3-loudspeaker-lab-report for me the R3 is more like 85dB in sensivity (mids) although the average is around 86dB.

 

[NTK]

The R7 can take more power than the R3 too. Power is relatively cheap. I think 70 W is on the low side.

If it is for me, I'd think the R3 will be sufficient. One way to take stress off the main speakers is to cross them over to subs at a higher frequency, say to 100 Hz.

The maximum SPL, given the same speaker cone surface area and travel, goes up 12 dB per octave. (This is because the cone moves faster at higher frequencies, and thus it can put out more "work" with the same travel.) This means a loudspeaker that can produce 97 dB SPL at 80 Hz should be able to produce 101 dB SPL at 100 Hz. If you take more of the load off the low frequencies by crossing over to subs at slightly a higher frequency, it can play cleaner louder.

 

[Blumlein 88]

This can work quite well if you have subwoofer and can take the low frequencies off the LCR speakers. Depending upon the particulars it might still choke a bit on the very loudest movie scenes, but those aren't happening constantly (well not usually).

Sort of the rule of thumb is to be able to have average levels of 85 db with at least 20 db of peaks possible (105 db SPL).

 

[Snoochers]

This can work quite well if you have subwoofer and can take the low frequencies off the LCR speakers. Depending upon the particulars it might still choke a bit on the very loudest movie scenes, but those aren't happening constantly (well not usually).

Sort of the rule of thumb is to be able to have average levels of 85 db with at least 20 db of peaks possible (105 db SPL).

Do people really listen to reference level at home though? Peaks of 105db sounds pretty crazy to me.

 

[A800]

105dB and low distortion is not that much imho.

 

[MZKM]

You don’t want to calculate SPL for both. Also, the calculator uses -6dB for distance doubling whereas in actuality it’s closer to -4dB, which is a good thing.

You don’t need as good quality of speakers for home theater, you mainly want the vocal region to be good and you want it to get loud (I normally have to crank movies around +10dB compared to music). Unless it’s multi-purpose, I would go with PSA MTM-210’s and the MTM-210C.

I have 88dB speakers and at 12ft they get a bit more than loud enough, I don’t think they would cut it at 17ft.

 

[Dj7675]

I'm likely building a home theater at some point. The designs are still in the works but the front row might be 12' from the front speakers and the rear row might be 17' from the front speakers. I'm toying with the idea of having bookshelf speakers as LCRs since they're considerably cheaper. For example, the KEF R3 and HDI-1600s are much cheaper than their tower counterparts ($4200/$3750 for three vs. $8100/$7500 - CAD). The sensitivity of the R3 and HDI-1600 compared to the towers is lower by about 1 to 5 depending on the models.

I've looked at loudness using the SPL calculator at https://myhometheater.homestead.com/splcalculator.html

Using the KEF R3 as an example: Two KEF R3s from 17 feet away with 70 watts powering them can produce about 97db. This is not quite reference, but does seem quite loud for peak home theater uses. Or am I wrong?

What are people thoughts on using bookshelves as LCRs? Can they get loud enough? The bookshelves don't go as low but with subwoofers does that become irrelevant?

I have decided to try this as well. I'm going with 3 M16's for my LCR (front projector). Center will just be below the screen. First row will be @ 10-11 feet. I have a pair of M106's in our living room powered off of a Denon X3500 and volume isn't an issue here but the receiver is just powering 2 main speakers and a pair of outdoor speakers. In my theater room, the M16s will be powered by a Denon x8500 in a 7.1.6 system. I only listen at about -10 for movies, so am thinking I will probably be ok with amplification at that level. If not, I will add 3 channels of external amplification. The other speaker I was looking at was the JBL HDI-1600 as well. But the price of the M16 is just too good for its quality. If the M16 work out as an LCR I will either add M16 on the other 4 bed channels, or S16s if the M16 prove too large for mounting.

 

[RayDunzl]

Do people really listen to reference level at home though? Peaks of 105db sounds pretty crazy to me.

105dB and low distortion is not THAT much imho.

I'll agree with "low distortion" being an important quality for playback.

It's not crazy loud, though it's too loud to have a quiet conversation - had to turn it down (below).

A-weighted Leq (averaged level over the entire time) 78.5dB - that excludes a lot of low frequency power

Z - or unweighted - LZpeak - which includes all the bass, has a peak of 107.1 here, at the moment, listening to John Pattitucci jazzy stuff.

Instant SPL - Unweighted 86.22db, C-weighted 86.0db, A-Weighted

So, the little woman came out of hiding, I turned it down to this level as we admired her stimulus check that came in the mail today.

So, going by the numbers, reduced the "perceived" level by about half, or 10dB:

Still reading 97dB unweighted peaks...

 

[Blumlein 88]

In fairly large room, 11 feet from the speakers, I had 3 LSR305s across the front and a sub. It did fine, and you'd not notice much wrong. It did sound strained a bit on the largest movie peaks. I eventually replaced the L and R channels with a floorstander with 90 db sensitivity (still using sub). Those have 250 wpc amp on them which is likely plenty and then some. It does sound better in louder portions and has a more effortless ability to go loud as in you don't notice the speakers it just gets loud. I think the LSR's have a bit of compression at their loudest which was probably marginal in this situation.

So probably 3 M16's with goodly sized amps and a sub is no problem. But I'd also say you'll likely have a more effortless sound on peaks if you went with some floorstanders.

 

[Dj7675]

In fairly large room, 11 feet from the speakers, I had 3 LSR305s across the front and a sub. It did fine, and you'd not notice much wrong. It did sound strained a bit on the largest movie peaks. I eventually replaced the L and R channels with a floorstander with 90 db sensitivity (still using sub). Those have 250 wpc amp on them which is likely plenty and then some. It does sound better in louder portions and has a more effortless ability to go loud as in you don't notice the speakers it just gets loud. I think the LSR's have a bit of compression at their loudest which was probably marginal in this situation.

So probably 3 M16's with goodly sized amps and a sub is no problem. But I'd also say you'll likely have a more effortless sound on peaks if you went with some floorstanders.

When I sold my last DIY speakers I needed something so I picked up 3 308P's. In regards to passive bookshelves and amp power... am I correct that I would need to test for compression as volume increases to see if I need larger amps? How does one test for compression?

 

[NTK]

When I sold my last DIY speakers I needed something so I picked up 3 308P's. In regards to passive bookshelves and amp power... am I correct that I would need to test for compression as volume increases to see if I need larger amps? How does one test for compression?

The ANSI/CTA-2034A (the test is more than just spinorama) maximum usable SPL test test it 2 different ways.

In the maximum usable continuous output (SPLmuco) test, the loudspeaker is fed a CTA-2034A weighted pink noise and the input signal is increased slowly in steps (to ensure that the loudspeaker is close enough to thermal equilibrium). The frequency response of the loudspeaker to the pink noise is continuously measured, and when the deviation from the baseline becomes excessive (more than 3 dB), the speaker has reached its SPLmuco.

The peak maximum SPL (SPLpeak) test is done by measuring distortions using high power but short (6.5 cycles, cosine window weighted) tone bursts.

This is how CTA-2034A measures a loudspeaker's long term and short term maximum SPL capabilities.

 

[Dj7675]

The ANSI/CTA-2034A (the test is more than just spinorama) maximum usable SPL test test it 2 different ways.

In the maximum usable continuous output (SPLmuco) test, the loudspeaker is fed a CTA-2034A weighted pink noise and the input signal is increased slowly in steps (to ensure that the loudspeaker is close enough to thermal equilibrium). The frequency response of the loudspeaker to the pink noise is continuously measured, and when the deviation from the baseline becomes excessive (more than 3 dB), the speaker has reached its SPLmuco.

The peak maximum SPL (SPLpeak) test is done by measuring distortions using high power but short (6.5 cycles, cosine window weighted) tone bursts.

This is how CTA-2034A measures a loudspeaker's long term and short term maximum SPL capabilities.

Thanks for your reply. A bit over my head Could I simply set my receiver at -15db, play a test tone in REW? Then raise the volume on the receiver by 1 dB and see if there is a corresponding increase in measured spl? Keep doing that until a 1 dB increase on the receiver does not result in a 1 dB gain measured on the speakers. Would that be a measure of compression?

 

[NTK]

Thanks for your reply. A bit over my head Could I simply set my receiver at -15db, play a test tone in REW? Then raise the volume on the receiver by 1 dB and see if there is a corresponding increase in measured spl? Keep doing that until a 1 dB increase on the receiver does not result in a 1 dB gain measured on the speakers. Would that be a measure of compression?

Yes. Basically. The CTA-2034 weighted pink noise is a good signal to use. (The standard pink noise used for SPL measurements does not extend low enough in frequency and has too little low frequency contents.)

Measure the frequency response and monitor the deviations. Usually you'll see compression happens at the low frequencies first. Wear hearing protection when you test

 

[Dj7675]

Which one in REW should be used? I see pink noise and pink noise PN and some others...

 

[spacevector]

I'll agree with "low distortion" being an important quality for playback.

It's not crazy loud, though it's too loud to have a quiet conversation - had to turn it down (below).

A-weighted Leq (averaged level over the entire time) 78.5dB - that excludes a lot of low frequency power

Z - or unweighted - LZpeak - which includes all the bass, has a peak of 107.1 here, at the moment, listening to John Pattitucci jazzy stuff.

Instant SPL - Unweighted 86.22db, C-weighted 86.0db, A-Weighted


View attachment 61521


View attachment 61522

So, the little woman came out of hiding, I turned it down to this level as we admired her stimulus check that came in the mail today.

So, going by the numbers, reduced the "perceived" level by about half, or 10dB:

View attachment 61523

Still reading 97dB unweighted peaks...

What you playing?