Falcon LS3/5A crossovers Silver vs. Gold

[Rug Doc]

Tony, I’d love you to put on some electronica and report back on how they get on if you’d be so kind. I don’t listen to the typical LS3/5A staple music, but am incredibly interested to hear a pair.. soundstage and layering is hugely important to me, so I think they could do the top end beautifully, I could (would) also add a sub to add bottom end if necessary, but I’d love to hear your view on how they sound with this genre.. even well produced pop please, Daft Punk, that kind of thing.

 

[Tony L]

Like JR149s and Linn Kans LS3/5As are very good on electronica. Beautifully clean, open and clear, but they do have a very clear limit on bass excursion even though the bass they give is very tight and well controlled. Things like say Trentmøller’s Last Resort, Ulrich Schnauss, LSG etc sound absolutely superb, but you’ll likely do damage to the B110 turning it up much beyond a real nearfield or background level. The volume envelope is far less than a modern small speaker like say a Kef LS50. They definitely do it well though. Reggae too, e.g. Bob Marley’s Babylon By Bus, the first UB40 album etc sound remarkably good as the deep bass gives the little speakers some surprising weight. They can sound so much bigger than their physical size.

 

[Tony L]

To follow on to the above after playing about a bit I’d say it depends entirely on the electronica. They are superb with ambient stuff, e.g. Murcof, Boards Of Canada, the more gentle Aphex Twin, Kraftwerk etc, but they lack the headroom and kick to do full-on techno. Obviously I have an absolutely huge pair of Lockwood Tannoys downstairs, so I can bang techno 12” singles out with an almost nightclub scale and impact, so this colours my judgement, but I do think this music needs a far larger speaker than a mini-monitor. I think you’d love how Murcof, Sufjan Stevens etc sounds through the LS3/5As though.

Also worth noting I have the Leak TL12 Plus amps in, which are big punchy and spacious sounding things, but they are only 12 Watts and I may be asking a little much of them with this sort of music into such an inefficient speaker. I’ll stick the Pass Aleph 3 in at some point to see what that thinks of the crossover change, but that isn’t a lot more powerful (30 Watt single-ended class A solid state amp which will halve into 16 Ohm).

 

[lilolee]

@Tony L Nice shout on Ulrich Schnauss, who I haven't come across. Listening now to this and loving it

 

[Tony L]

Try Far Away Trains Passing By, that’s the high-watermark IMHO. The 3x vinyl is superb.

PS The one you link is really good too. Kind of Cocteau Twins-ish.

 

[Rug Doc]

Far away trains is the only Ulrich Schnauss album you need, it’s fabulous.

Thanks Tony, appreciate your thoughts. I can’t help but turn music up when I’m listening, and think the the little 5’s might struggle… when listening to aphex or Tremtemoller I really like to pressurise the room, and thinking about it, LS3/5s obviously can’t do that. What i’d love is a speaker that can do the midrange like the 3/5a but at 120db….. I tried the Impulse horns, but the midrange was too pronounced, didn’t enjoy them unless I could introduce EQ, the NS1000X I have are amazing, but I want a bit more excitement, or depth, or involvement, yes involvement I think… the Impulse were amazing at volume levels, but just lacked a bit in midrange refinement. Basically I want a BBC monitor that can do 120db….

??

 

[Tony L]

You almost certainly don’t want 120db as that is permanent hearing damage within minutes! That is close to the speaker stack at a Mogwai or Leftfield gig!

I suspect you want a good 95db with some headroom. Probably worth investigating LS5/8s. They are a fairly large active two-way with a 12” poly bass unit and powered with a BBC modified Quad 405 with a crossover circuit for each speaker. A friend has a pair and I like them a lot. I’d not swap my Tannoys for them, but they are very good indeed.

PS I stuck the Pass Aleph 3 in earlier and it suits the LS3/5As with the new crossovers better than the stock pair for sure. It sounded rather over-warm and a bit dull and boring before, now it seems to grip the bass really well and does have more headroom on tap than the little Leaks. Certainly groovier. Still not a hugely powerful amp, but an exceptionally well controlled one with a remarkably natural open and spacious midband. I still don’t think I’ve got close to hearing what this amp can do as I suspect it needs a very good Pass or equivalent active preamp upstream rather than my nearly wide-open Audio Synthesis passive (it is a low gain amp), but it was sounding very good indeed across a wide range of music. I’ll leave it in the system for a while.

 

[Rug Doc]

Thanks Tony. Will do some reading up of LS5/9’s. The current speakers have 12” drivers, so yes you might be right, but maybe the 1000x might outdo them…? I’ll do some investigation…

Nope, I like 100db plus… always have.

 

[Tony L]

If you really want gig volume and clean then you want horns. Altec VOTT, Valencias, Model 19s, Klipschorns, La Scalas, bigger blue-face JBLs, Tannoy FSMs, Dreadnoughts etc. There is no substitute. Absolutely no way to get that sort of scale and volume with small speakers, and yes I’m counting NS1000s and LS5/8s (I mistyped, not LS/9s which are much smaller) as small speakers.

 

[ToTo Man]

Nope, I like 100db plus… always have.

You, sir, must have balls of steel to push 120dB through a pair of speakers with unobtainium beryllium drivers!

Do the NS1000X have the same 90dB/1w/1m sensitivity as the NS1000M? If so then you must have an amp capable of delivering peaks of 1,024 RMS watts per channel for 120dB, which I'd be surprised if the NS1000X's bass drivers' voice coils could handle.

What amplification are you using to achieve this sort of power output??!!

 

[Rug Doc]

Okay 120db is pushing it but I use a Yamaha B2 amplifier (2 x 120w) and push both it, and the beryllium drivers reasonably hard.

I have had horns previously, 15” bass and B&C DE250’s, and would love to go there again.. that system had tangible presence!

amp - http://www.thevintageknob.org/yamaha-B-2.html

speakers -

http://www.thevintageknob.org/yamaha-NS-1000x.html

@ToTo Man - with those specs above, you’ll probably be able to tell me what output I’m achieving - I’d be very interested to know!

 

[ToTo Man]

Okay 120db is pushing it but I use a Yamaha B2 amplifier (2 x 120w) and push both it, and the beryllium drivers reasonably hard.

I have had horns previously, 15” bass and B&C DE250’s, and would love to go there again.. that system had tangible presence!

amp - http://www.thevintageknob.org/yamaha-B-2.html

speakers -

http://www.thevintageknob.org/yamaha-NS-1000x.html

@ToTo Man - with those specs above, you’ll probably be able to tell me what output I’m achieving - I’d be very interested to know!

For every 3dB increase in SPL you need twice the power, so with a speaker that's 90dB at 1w, 93dB = 2w, 96dB = 4w, 99dB = 8w, 102dB = 16w, 105dB = 32w, 108dB = 64w, 111dB = 128w. 120w would therefore be enough to give you peaks of 110dB* before clipping, which is still very loud but not as downright insane as 120dB!

* That's for 1 speaker at a distance of 1 metre in an anechoic chamber. With two speakers in a typical room you gain an extra 3dB but as you double your listening distance you lose a few dBs depending on how reflective your room is. I forget the math now but hopefully @John Phillips can bail me out!

 

[John Phillips]

For every 3dB increase in SPL you need twice the power, so with a speaker that's 90dB at 1w, 93dB = 2w, 96dB = 4w, 99dB = 8w, 102dB = 16w, 105dB = 32w, 108dB = 64w, 111dB = 128w. 120w would therefore be enough to give you peaks of 110dB* before clipping, which is still very loud but not as downright insane as 120dB!

* That's for 1 speaker at a distance of 1 metre in an anechoic chamber. With two speakers in a typical room you gain an extra 3dB but as you double your listening distance you lose a few dBs depending on how reflective your room is. I forget the math now but hopefully @John Phillips can bail me out!

OK, but ISTM you don't need bailing out.

I normally assume the combination of two loudspeakers, manufacturer's sensitivity over-estimate, room gain, and listening distance loss approximately cancel in total. So, just using the 1 metre sensitivity seems to me to be as good as you need for a broadly correct answer for mid-field listening in a typical home listening room.

So, I think that 110 dB SPL is perfectly right. It's the loudest theoretical mean SPL before the amplifier clips. The corresponding instantaneous peak level is 113 dB SPL (+3 dB). But only sine waves have a crest factor (peak/mean ratio) as low as 3 dB. It's higher for music.

If you are playing music with a (good) 15 dB crest factor, it will not get clipped on peaks up to a mean level of 98 dB SPL (i.e., 113 dB SPL - 15 dB). You will get higher clean average level for music with a lower crest factor (e.g., compressed pop); or lower clean average level for music with a higher crest factor music (e.g., classical or electronica). It can also be louder on average if you don't mind some peak clipping.

However, the above is right only if the loudspeakers themselves can actually produce 98 dB SPL continuously and 113 dB SPL on peaks without going into power compression. A good quality 12" driver should do that and more. I suspect the Yamaha drivers are likely to oblige, but it's a necessary reality check. I have listened occasionally at just a bit less than that to check that it's possible, but I go there very rarely.

 

[S-Man]

For every 3dB increase in SPL you need twice the power, so with a speaker that's 90dB at 1w, 93dB = 2w, 96dB = 4w, 99dB = 8w, 102dB = 16w, 105dB = 32w, 108dB = 64w, 111dB = 128w. 120w would therefore be enough to give you peaks of 110dB* before clipping, which is still very loud but not as downright insane as 120dB!

* That's for 1 speaker at a distance of 1 metre in an anechoic chamber. With two speakers in a typical room you gain an extra 3dB but as you double your listening distance you lose a few dBs depending on how reflective your room is. I forget the math now but hopefully @John Phillips can bail me out!

In a typical room I think you can get up to 6dB increase due to boundary effects, at least at LF, which is where the SPL limits usually are.

Also, if the signals are correlated for the 2 channels you can get an increase of 6dB (3dB for uncorrelated signals):
http://www.sengpielaudio.com/calculator-coherentsources.htm
http://www.sengpielaudio.com/calculator-leveladding.htm

 

[ToTo Man]

Concerned by @Rug Doc's listening habits I took it upon myself to blast Comfortably Numb from the 2011 'Discovery' CD remaster which has a gentle amount of compression (+3dB louder than the 1st UK Harvest CD pressing) at my loudest typical comfortable listening level and simultaneously measured the SPL at my listening seat with my digital handheld RadioShack SPL meter and UMIK-1 mic using the SPL Meter readout in RoomEQWizard.

When both meters were set to dBC weighting and Fast response, the RadioShack max was 91dBC, the REW max was 97dB, a 6dB difference! The REW meter has a response rate of 125ms so is obviously capturing a lot more of the transients. If I set the REW meter response rate to 1 second the max SPL drops to 94dB so it's still reading louder than the RadioShack meter.

If I played the original UK Harvest pressing I'd probably be listening at a max dBC value of just over 100dBC.

Even if it was suitably power rated I don't think I'd instinctually allow myself to put these sorts of SPLs through a mini monitor for mid-field listening, the thought of it just doesn't put me at ease. If a subwoofer and high-pass filter were added then I may change my mind but all else being equal I psychologically feel more comfortable knowing loud sounds are coming from a large speaker if that makes sense!

The 'LCFmax' shows the highest number recorded on the large continually fluctuating readout on the right hand side. I have no idea what the 'LZpeak' value relates to, I've read the REW help file but it doesn't give an explanation. The 'LCFmin' represents the quietest period of my room's noise floor.

 

[Tony L]

Some definitions of the terms here: www.noisemeters.co.uk

The Max is an averaged time-weighted level, the peak is the highest transient. My ‘proper’ meter has no peak reading, just a A or C Max and Min. From an audio perspective this is of a limited use, especially when it comes to trying to calculate what amp power is needed to cover those high transient peaks. My iPhone is far less accurate but does do peaks. On a good jazz recording, e.g. ECM or whatever, the transients can be 10-15 db above the max easily. One problem both with a hardware meter and a phone is you can get an entirely bogus peak reading from pressing a button either on screen or physical.

PS Try your 1st Harvest Floyd CD vs. the later remaster, it is a lot more dynamic. Not a loud cut CD, but it has a lot of dynamic range. The odd thing (if it actually is a 1st mastering) is the song start markers are in the wrong place on many tracks!

 

[Tarzan]

Tony, I’d love you to put on some electronica and report back on how they get on if you’d be so kind. I don’t listen to the typical LS3/5A staple music, but am incredibly interested to hear a pair.. soundstage and layering is hugely important to me, so I think they could do the top end beautifully, I could (would) also add a sub to add bottom end if necessary, but I’d love to hear your view on how they sound with this genre.. even well produced pop please, Daft Punk, that kind of thing.

You may like ProAc Tablette 10 for electronics etc, sorry to but in, they can go loud and have tons of musicality, at least they do in my room, system.

 

[ToTo Man]

Some definitions of the terms here: www.noisemeters.co.uk

The Max is an averaged time-weighted level, the peak is the highest transient. My ‘proper’ meter has no peak reading, just a A or C Max and Min. From an audio perspective this is of a limited use, especially when it comes to trying to calculate what amp power is needed to cover those high transient peaks. My iPhone is far less accurate but does do peaks. On a good jazz recording, e.g. ECM or whatever, the transients can be 10-15 db above the max easily. One problem both with a hardware meter and a phone is you can get an entirely bogus peak reading from pressing a button either on screen or physical.

Fascinating stuff... I've clearly been underestimating my peak listening levels! Funny thing is I don't consider the levels I listen at to be particularly loud. I always found loud noises physically painful as a child and went out of my way to avoid exposure to them wherever possible. -30dB noise reducing yellow foam earplugs were a must when attending live music events and I still need earplugs (etymotic-style) at most concerts as even just a few seconds of very loud sound is still painful. If I'm able to comfortably tolerate peaks of 108dB in my own room I dread to think how loud the peaks are at a rock gig, must easily be over 120dB?!!

A transient of 108dB on a 90dB/1w/1m speaker needs an amp capable of putting out 64 watts, but what definition of power is relevant here? RMS or IHF? These are the specs of my A-S3000, for example:

PS Try your 1st Harvest Floyd CD vs. the later remaster, it is a lot more dynamic. Not a loud cut CD, but it has a lot of dynamic range. The odd thing (if it actually is a 1st mastering) is the song start markers are in the wrong place on many tracks!

Will try later, along with a brickwalled modern pop/rock record to get an idea of the variations in Max vs Peak level.

PS - Apologies for dragging this thread way off-topic!!

 

[ToTo Man]

The run-time on (what I believe to be) my 1st UK Harvest pressing for Comfortably Numb is 6:41.60 and the peak level is -7.008 dBFS. It's my go-to version when I'm listening to the album but not when listening to compilation playlists in case I forget to turn the volume down before the next song starts and potentially blows my speakers! I keep meaning to Normalise the album in Audacity and export it as a 24-bit file but I never get around to it... I have other early CD versions of The Wall, some have the same 6:42 run time for Comfortably Numb while some are 6:24 or 6:49. The 2011 Remaster is 6:22.

The SPL Meter in REW has a 'Logger' function that plots SPL against time so I thought it would be interesting to do it for a track with good dynamic range (i.e. Comfortably Numb from the 1984 Harvest UK CD, DR=13) and a modern brickwalled track ('One Step Away' from Work Of Art's 2009 album 'In Progress', DR=6/7).

It appears I wasn't in as much of a mood for 'letting it rip' tonight as with 'Comfortably Numb' (1984 Harvest UK CD) I only obtained LCFmax = 93dB and LZpeak = 105dB. I expected it to peak in excess of the 2011 Remaster's 108dB so I can't have had the volume set loud enough, though it still sounded satisfyingly LOUD I must say!

For Work Of Art's 'One Step Away', the maximum tolerable levels were higher than I expected, LCFmax = 88dB and LZpeak = 101dB. I should've probably played this before Comfortably Numb as I have a tendency to 'creep the volume up' as the listening session progresses!

It goes without saying that the above data is very much system and room dependant. My system is relatively bass-light (only a 5dB differential between low and top end), if I were running a 10dB target slope the above readings would likely be higher.

 

[calorgas]

Thanks Tony. Will do some reading up of LS5/9’s. The current speakers have 12” drivers, so yes you might be right, but maybe the 1000x might outdo them…? I’ll do some investigation…

Nope, I like 100db plus… always have.

Keep your eyes peeled for some Rogers Studio 2 - pretty thin on the ground but would probably fulfil your brief. I had some that I got from Bob ICHM on the wam, they're like a domestic version of the LS5/8.
Here's a link to the Hennessey site - the pics are of my old pair.