Hi all - long time lurker, first time poster. Still relatively early on my audio journey. One of the reasons I'm here at all is James Murphy/Soulwax's Despacio sound system. I've been through most of the threads on the subject and it's clear there's not a lot of love for the design around here, but so far I haven't found any direct critiques of specific things you experts would change about the design. I've found a few things puzzling myself, but considering my limited knowledge I would love to hear what faults you all can identify in this system.
Here are a few links if you're not familiar:
DFA Monitors - Gearslutz (Scroll down for notes from the designer)
klett's Despacio Sound System project page
Inside Despacio – the world’s greatest sound system
A few oddities I noted:
1. Crossover frequencies seem odd based on component choice (hi-mid horn upper and lower XO point in particular, as well as the low mid-hi mid point in general given the low mid is using 12" drivers)
2. 12" drivers for low mid (would start to beam at about 1.3khz, no?)
3. Vented bass and low-mid cabinets (guessing that the crossover freqs are well above the inherent tuning frequency in the "minimally tuned" design; why vent at all, or why that way?)
Some critiques I've read here that I had thoughts on:
1. This system was clearly designed to be 1) old school, and 2) excessively and visibly lavish. McIntosh amps probably have little business in this sort of application, but the target audience (myself included) thinks they're pretty cool and fit with the ethos. More practical choices abound... but that wasn't the point.
2. The system was a DIY-style project from the beginning. In one of the articles above Klett talks about franken-klipsches as the genesis of this. Fair to say the were trying to avoid the latest technology. When I think of vinyl DJ's (love it or hate it) I don't think of Void or L-Acoustics.
Final thought - I've tried twice to hear this system live, and been foiled both times, the latest being the NYC show last month that was postponed for... well, you know why. Would love to hear the thoughts of anyone who has seen/heard it!
Here are a few links if you're not familiar:
DFA Monitors - Gearslutz (Scroll down for notes from the designer)
klett's Despacio Sound System project page
Inside Despacio – the world’s greatest sound system
A few oddities I noted:
1. Crossover frequencies seem odd based on component choice (hi-mid horn upper and lower XO point in particular, as well as the low mid-hi mid point in general given the low mid is using 12" drivers)
2. 12" drivers for low mid (would start to beam at about 1.3khz, no?)
3. Vented bass and low-mid cabinets (guessing that the crossover freqs are well above the inherent tuning frequency in the "minimally tuned" design; why vent at all, or why that way?)
Some critiques I've read here that I had thoughts on:
1. This system was clearly designed to be 1) old school, and 2) excessively and visibly lavish. McIntosh amps probably have little business in this sort of application, but the target audience (myself included) thinks they're pretty cool and fit with the ethos. More practical choices abound... but that wasn't the point.
2. The system was a DIY-style project from the beginning. In one of the articles above Klett talks about franken-klipsches as the genesis of this. Fair to say the were trying to avoid the latest technology. When I think of vinyl DJ's (love it or hate it) I don't think of Void or L-Acoustics.
Final thought - I've tried twice to hear this system live, and been foiled both times, the latest being the NYC show last month that was postponed for... well, you know why. Would love to hear the thoughts of anyone who has seen/heard it!
The main threads are old now.. I remember seeing threads on this system back then, and I didn't wade through all of them.. what crossover frequencies were used? I see in the later thread some 18" subs were added. I too question the use of 12" low/mids... like why bother with the massive 4" CD above them. I have my own stack with dual 15 low/mids crossing to a 2" exit B&C CD at 1khz with no issues at all, a little DSP EQ and time alignment and the results are "crisp" as my DJs like to describe the system. Subs are 18" reflex.
I also read that there is nothing in the speaker signal chain except the crossover and amps.. clearly a bit of the "EQ is BAD" audiofool thinking was involved here which clearly... I don't subscribe to because it's idiotic... but to each their own.
Some restrained use a parametric EQ can be the difference between ho-hum and WOW.. and corrective EQ in the right places can even improve phase and impulse response so ruling it out "just because" is dumb... IMO.The thing is... if you remove the amps what's left isn't much to look at. Any dance stack (which this is) based around horns looks infinitely more interesting, but personally I got no problem with that because for me it's all about the sound and I don't care what's powering a rig if it delivers the goods.
Reflex bins do sound good close up and that is how these are meant to be used.
Last edited:
The main threads are old now.. I remember seeing threads on this system back then, and I didn't wade through all of them.. what crossover frequencies were used? I see in the later thread some 18" subs were added. I too question the use of 12" low/mids... like why bother with the massive 4" CD above them. I have my own stack with dual 15 low/mids crossing to a 2" exit B&C CD at 1khz with no issues at all, a little DSP EQ and time alignment and the results are "crisp" as my DJs like to describe the system. Subs are 18" reflex.
I also read that there is nothing in the speaker signal chain except the crossover and amps.. clearly a bit of the "EQ is BAD" audiofool thinking was involved here which clearly... I don't subscribe to because it's idiotic... but to each their own.
In one of the threads he describes a 'very very general house style EQ curve', so there must be something in there. Hard to imagine that system doesn't need an EQ in a vacuum, let alone in a range of different venues and configurations. Believe that comment was specifically about compression/dynamics - in one of the interviews the Dewaele brothers talk about how dynamic variations in certain tracks were met with applause...
Frequency ranges and driver sizes:
2x21" 'off the shelf' powered subwoofers: <100hz
2x2x15" vented bass cabs: 100-300hz
2x2x12" vented low-mid cabs: 300-2k hz
1x ""1.5" throat 4" compression driver on 90° x 50° horn"": 2k-10k hz
4x bullet tweeters (looks like they're splayed out both V&H): 10k+
Where I'm confused - and maybe it's because I don't see many 1.5" throat CD's - is how the 12" low mids are the better choice than the horn at 2khz. That horn should go far lower easily, and the 12" have to run out of gas comparatively at that range.
I love the raw lacquered wood look of this. That said, I agree... without the blue lamps and the dancing VU meters, it's pretty bland and comparable to any other 'soundsystem' sound system.
I would expect it to sound slow...literally!
Agreed. For those not familiar, the whole concept is "slowly" (<120BPM, and mostly well below that) - hence the name. So the question is, what benefit of the design outweighs slowness for its application?
Agreed, I have not met a passive rig that didn't need at least some correction, and no I don't use any dynamics control on recorded inputs because they are already overly compressed in many cases, but open mics are another story.
Yep, I have some RCF 1.4" CDs here now.. they're only 2.5" voice coils but have no problem with a 1.2khz crossover. The 3" VC/2" exit B&Cs I have are crossed at 1khz again with absolutely no sign of strain.
The other crossover selections are puzzling too. The 2x15 cabs were originally the subs if I understand correctly, and that would mean those drivers are not well suited to low-mid duty so when they added the 21's(thought they were 18s) they moved the 15's to kick. OK. And I kinda get the stack of 12" mids, physically they might look more impressive to a certain clientele than a stack of 6" and dynamically those large diaphrams will impart more impact to the sound, but what's up with the 2khz crossover? You just know there is going to be a lot of beaming and comb filtering from the 12's at and above 2khz. And that massive CD is more than capable of a 800hz crossover so why aren't they using it that way? Maybe they found it had a bit too much bite crossed lower and erred on the side of smooth sound but that seems like a strange choice for a "spared no expense" kind of build where they could have any drivers they wanted.
Last edited:
I thought I knew what this was referring to but now you have confused me, what does track BPM have to do with speaker design?
The intent was/is to play bass heavy downtempo disco edits, on vinyl, really loud. Therefore the system doesn't need to go super low, and volume/sensitivity/headroom was probably more important than response & accuracy given the need to set it up at different venues.
I think I just answered my own question
I do think they've done things more for the look than the sound.
Any way up, though, I don't see the logic behind 4x 12" mids, and only 2x 21" subwoofers.
I don't really get the reason behind splitting the LF range up so much, either - a good 21" cab can do 30-80Hz just fine.
If it was mine, I'd do:
4x 21": 30-80Hz
2x 15": 80-600Hz-ish
1x Coaxial compression driver - 600Hz upwards.
FIR processing to linearise above, say, 500Hz (IIR below to minimise latency), and off we go. Nice and simple.
Chris
Any way up, though, I don't see the logic behind 4x 12" mids, and only 2x 21" subwoofers.
I don't really get the reason behind splitting the LF range up so much, either - a good 21" cab can do 30-80Hz just fine.
If it was mine, I'd do:
4x 21": 30-80Hz
2x 15": 80-600Hz-ish
1x Coaxial compression driver - 600Hz upwards.
FIR processing to linearise above, say, 500Hz (IIR below to minimise latency), and off we go. Nice and simple.
Chris
Yet somewhere in there they claim it is "the best sound system ever" so these crossover choices seem out of sorts with that goal. With all the bands they have chopped the system into, the amount of amp power behind it and the fact they typically run it at 25% the system is overbuilt and none of the drivers are being pushed hard at all, that big comp driver in particular might only see 10-20w peak.
I bet I would like the sound this system generates particularly with some nice dynamic source material.. the type of stuff I grew up with, it would be a welcome change from the overly compressed junk we have today. But I also think they missed some opportunities with the design, the system is more complex than it needs to be and the crossovers aren't quite right so there is still room for improvement.
I bet I would like the sound this system generates particularly with some nice dynamic source material.. the type of stuff I grew up with, it would be a welcome change from the overly compressed junk we have today. But I also think they missed some opportunities with the design, the system is more complex than it needs to be and the crossovers aren't quite right so there is still room for improvement.
I had not read (or I missed) that the 15's were supposed to be the subs. If that was the case, seems like they realized mid-stream that it wouldn't be enough bottom end and tacked on the 21's. There's a comment from the designer about not having all the desired drivers available for the first show and they had to make adjustments, but seven years later they've swapped out some drivers and still held on to those 21's.
What do you think they actually used as a crossover at each stack? On principle they probably avoided DSP. Only 5-way analog crossover I can find today is from Peavey (and I'm not even sure that is analog...)
@conanski, here's a taste for you (although this was mixed digitally and most of the tracks are modern edits)
What do you think they actually used as a crossover at each stack? On principle they probably avoided DSP. Only 5-way analog crossover I can find today is from Peavey (and I'm not even sure that is analog...)
@conanski, here's a taste for you (although this was mixed digitally and most of the tracks are modern edits)
IMO, the only way to get a system like this up-and-running would be with DSP. Too many bands, too much time-alignment (NB - physical alignment isn't the end of the story).
Fair enough if they've tried to do it analogue, though. They might have cascaded a couple of active crossovers to get enough bands. Might need a few parametric EQs dotted through the signal chain if any out-of-band processing is required.
All that said, I really do think a sensibly-designed 4-way system would beat this in sound quality and output.
If only I could get the same cash investment that they had...
Chris
Fair enough if they've tried to do it analogue, though. They might have cascaded a couple of active crossovers to get enough bands. Might need a few parametric EQs dotted through the signal chain if any out-of-band processing is required.
All that said, I really do think a sensibly-designed 4-way system would beat this in sound quality and output.
If only I could get the same cash investment that they had...
Chris
+1. Even a decent modern DSP processor will do no harm to the signal chain and stands to significantly improve impulse response and linearity over what they had. I know this system was designed to be all analog with vinyl pressings as the source but despite what the aficionados think vinyl is easily the least accurate and most compromised recording medium ever. I read a recent thread on another forum populated by those that have worked in the recording industry specifically in the creation of consumer vinyl, and a common theme was that it was extremely difficult to produce even a somewhat faithful copy of the master recordings on this medium, every other consumer format is better in that regard with lossless digital being the best by far.
P.S. But don't take that to mean that I don't like the sound of vinyl because I do, I still have much of the vinyl collection I started in the '70's when everything had analog masters. I have been there the whole way through the age of digital audio and I still don't think we are really utilizing it to it's potential. The kicker with this is that at least some(maybe a lot) of the music played over this system on vinyl would have been created on a digital platform to begin with so the whole "totally analog" thing is more or less a bunch of hype.
I think my DSP processed 218/215+2 stacks would give it a run and I'm sure your proposed 4-way with the coax mid/high would beat it.
Last edited:
As has been pointed out if you take away the expensive impractical amps and the marketing fluff you just have a stack of reflex speakers. Should sound decent but much less impressive compared to a fully horn loaded RLA style stack. Also they don't have any of the JBL slant plate lenses that give ultra wide dispersion which is important in these many stack systems for getting good coverage of the floor from every stack. The focus on 'audiophile' purity also seems a bit weird considering that much of the vinyl source material can't have high LF due to grove cutting limitations and sometimes also lacks HF. This is why the RLA systems had DJ controlled sub-harmonic synth and additional effect tweeters.
Paradise Garage sound system,circa late 70's. - Speakerplans.com Forums - Page 1
Paradise Garage sound system,circa late 70's. - Speakerplans.com Forums - Page 1
sorry, but you are very wrong here. Your statements can be applied to Albums with long playing times (longer than 15 minutes) at 33 rpm. And this is the reason there are 45 rpm disco singles, to have higher overall levels, and bass, and less HF distortion.
Anything released by major companies, cut in stereo after about 1965 will have been cut with variable pitch, which means that the groove spacing is dynamically adjusted according to overall level and spectral content. To enable this, the Disk transfer tape decks had a special preview head and a looped tape path before the actual playback head to provide sufficient delay for the pitch calculator. Basically an analog computer that integrated the incoming preview signal and adjusted the forward feed rate of the cutterhead to perfectly fit the grooves to each other.
On 45 rpm playback, there is less HF distortion, and also the signal gets louder because of the increased velocity.
Every time a record is cut, it is a trade-off between playing time, Bass, overall level, and distortion.
Simply because it is a broadly generalizing, incorrect statement. Severe lack of bass and poor treble would only apply to cheap collection/sampler discs such as K-tel etc, that have 25+minutes, 10 tracks per side, to maximize consumer appeal. In such a case the cutting engineer is forced to dial down overall level, and reign in the bass to fit all the stuff on one side. You don't play such a crap disc at a club. At the club you mostly play 10" or 12" singles cut at 45 rpm.
As i pointed out, it depends on the format. But even at 33rpm, there is no problem in putting 30 Hz at healthy levels on a record. The variable pitch accomodates for that. I have several such specimens. At 45 rpm the point becomes moot really. Plenty of music with strong 40 Hz, and occasionally strong notes between 30-40. Except for actual bass guitar/synth notes, bass drum hits can contain energy down to 30 Hz. Depends on the recording.
That is also sometimes the reason for album releases as double LPs. Spreading the music on two discs, you gain dynamic range, higher bass levels, and decreased HF distortion, if it allows you to avoid the inner area of the record surface (smaller radius -> more distortion)
Anyway, if you have the opportunity, take a good quality vinyl single or album, run it through an RTA while listening, and see for yourself....
This stuff is another subject I've spent more time on than is healthy....
Merits and disadvantages aside, playing vinyl was a constraint that was in place from the start. These guys have extensive record collections and plenty of young hipsters are willing to pay to see/hear them show those collections off (myself included, twice, and yet I still haven't heard them). Moreover, a lot of their records are either disco singles or their own recently cut (although digitally produced) remixes, so to Bob4's point, they're maximizing the format. But this thread isn't about whether vinyl is the right choice, or even if they did that part correctly. It's about the speakers, for the application.
Based on the responses, I have a few questions... and I admit, I've spent far too much time over the past few days (months?) dreaming of a better system than this for a very similar type of application.
1. DSP units - what's the best 'audiiophile' quality DSP unit that's also compatible with a PA/live sound environment? No personal experience but from what I've read, the typical DBX and Behringer units won't cut it from a fidelity perspective. If you just wanted to do simple EQ, XO and time alignment for 4+ bands, what would you use?
2. Further, what does DSP fix (specifically, time alignment) that running analog doesn't for this stack? From a purely geometric perspective, a bunch of 11'+ foot tall stacks surrounding a dance floor would seem to be suggest a 5'8" person in the center of the floor is perfectly positioned (assume they've baked in a slight C-shaped curve to the stacks, which seems reasonable from pictures). In a disco/dance floor enclosed and surrounded environment (not front-facing PA), what more could you ask for?
3. Seem like we all can't figure out why 5-way makes sense aside from they built 4-way and decided more low end was needed. So... if I was to go 4-way (as I've been thinking about) with, for instance, Keystone subs as the bottom, keep the 12" cabs and lower the XO to say 100-1.2khz for those 12's, and leave the horn hi mid and bullet UHF alone, with modified XO freqs. Does a tapped horn sub + reflex midbass + horn/bullet hi-mid/tweeter work, or have I gone totally nuts? Why am I nuts? I can imagine timing being different for each driver, but not so drastically out of sync that would outweigh the benefits compared to a bunch of reflex boxes... is an ~8ms spread (KS sub to bullet tweeter for instance) that critical in the intended environment - enclosed and surrounded?
4. I think we touched on why the reflex choice for midbass cabinets makes sense... but it didn't click for me. If your XO is well above the tuning frequency, what's the advantage of the reflex cabinet, loosely tuned or not? To the point about reflex sounding good close up... a) most listeners are pretty far away from the stacks, and b) same question about tuning frequency vs XO frequency, does the proximity point still hold? Why? (Maybe this is a dumb question... but if your XO is an octave above box tuning frequency, is there a noticeable difference between ported vs closed boxes? From pretty basic WinISD modeling it appears that there is a response difference, but perhaps not one that couldn't be fixed with EQ...)
An interesting observation after all of your (much appreciated!) responses: this system apparently started with discussions about the merits of Klipsch La Scala vs Cornwall, with the Cornwall bass + La Scala mid/high being favored. In a kitchen environment (quoted from designer) I'll buy it. But somewhere in translation to a disco system for 1000 dancers, those Corn-scalas went from a single reflex-loaded cabinet to 10 - TEN! - total reflex drivers. Not that they're doing it wrong (I'm one who would pay just to see these beasts anyways) but you have to wonder if aesthetics wasn't a bigger factor than they let on in the interviews...
Hope each of you are seeing light at the end of your quarantine tunnel!
Based on the responses, I have a few questions... and I admit, I've spent far too much time over the past few days (months?) dreaming of a better system than this for a very similar type of application.
1. DSP units - what's the best 'audiiophile' quality DSP unit that's also compatible with a PA/live sound environment? No personal experience but from what I've read, the typical DBX and Behringer units won't cut it from a fidelity perspective. If you just wanted to do simple EQ, XO and time alignment for 4+ bands, what would you use?
2. Further, what does DSP fix (specifically, time alignment) that running analog doesn't for this stack? From a purely geometric perspective, a bunch of 11'+ foot tall stacks surrounding a dance floor would seem to be suggest a 5'8" person in the center of the floor is perfectly positioned (assume they've baked in a slight C-shaped curve to the stacks, which seems reasonable from pictures). In a disco/dance floor enclosed and surrounded environment (not front-facing PA), what more could you ask for?
3. Seem like we all can't figure out why 5-way makes sense aside from they built 4-way and decided more low end was needed. So... if I was to go 4-way (as I've been thinking about) with, for instance, Keystone subs as the bottom, keep the 12" cabs and lower the XO to say 100-1.2khz for those 12's, and leave the horn hi mid and bullet UHF alone, with modified XO freqs. Does a tapped horn sub + reflex midbass + horn/bullet hi-mid/tweeter work, or have I gone totally nuts? Why am I nuts? I can imagine timing being different for each driver, but not so drastically out of sync that would outweigh the benefits compared to a bunch of reflex boxes... is an ~8ms spread (KS sub to bullet tweeter for instance) that critical in the intended environment - enclosed and surrounded?
4. I think we touched on why the reflex choice for midbass cabinets makes sense... but it didn't click for me. If your XO is well above the tuning frequency, what's the advantage of the reflex cabinet, loosely tuned or not? To the point about reflex sounding good close up... a) most listeners are pretty far away from the stacks, and b) same question about tuning frequency vs XO frequency, does the proximity point still hold? Why? (Maybe this is a dumb question... but if your XO is an octave above box tuning frequency, is there a noticeable difference between ported vs closed boxes? From pretty basic WinISD modeling it appears that there is a response difference, but perhaps not one that couldn't be fixed with EQ...)
An interesting observation after all of your (much appreciated!) responses: this system apparently started with discussions about the merits of Klipsch La Scala vs Cornwall, with the Cornwall bass + La Scala mid/high being favored. In a kitchen environment (quoted from designer) I'll buy it. But somewhere in translation to a disco system for 1000 dancers, those Corn-scalas went from a single reflex-loaded cabinet to 10 - TEN! - total reflex drivers. Not that they're doing it wrong (I'm one who would pay just to see these beasts anyways) but you have to wonder if aesthetics wasn't a bigger factor than they let on in the interviews...
Hope each of you are seeing light at the end of your quarantine tunnel!
To attempt to answer your questions:
1 - If quality is a concern, I'd look to the touring market. Powersoft T or X series, Linea Research, Crown ITechs, Lab Gruppen, etc etc. I'd insist on some FIR capability if I was going to make a go of this.
Any of the top-end manufacturers will have something good. I really like my Powersoft T-series amps, but others are available.
2 - See point 1.
3 - I'd go for a modern compression driver for ~800Hz-20kHz. The latest 1.4/3"s from 18Sound look excellent, plus there's the carbon diaphragm unit from Eminence, and let's not forget the coaxials which can get almost an octave lower. Keep the mid-high range point-source - they've introduced lobing and a directivity mis-match with their 2-way HF section. Sometimes those are necessary evils, but here they just aren't necessary at all.
4 - If you're operating well away from the port tuning frequency, the ported box and sealed box will be very similar in output. However, the ported box will have ventilation and the sealed box will not. Heat is a sworn enemy in the PA world, and passive air exchange is much better than nothing.
Chris
1 - If quality is a concern, I'd look to the touring market. Powersoft T or X series, Linea Research, Crown ITechs, Lab Gruppen, etc etc. I'd insist on some FIR capability if I was going to make a go of this.
Any of the top-end manufacturers will have something good. I really like my Powersoft T-series amps, but others are available.
2 - See point 1.
3 - I'd go for a modern compression driver for ~800Hz-20kHz. The latest 1.4/3"s from 18Sound look excellent, plus there's the carbon diaphragm unit from Eminence, and let's not forget the coaxials which can get almost an octave lower. Keep the mid-high range point-source - they've introduced lobing and a directivity mis-match with their 2-way HF section. Sometimes those are necessary evils, but here they just aren't necessary at all.
4 - If you're operating well away from the port tuning frequency, the ported box and sealed box will be very similar in output. However, the ported box will have ventilation and the sealed box will not. Heat is a sworn enemy in the PA world, and passive air exchange is much better than nothing.
Chris
Don't kid yourself aesthetics was a HUGH part of this design. This is a very basic system that leveraged almost none of the advancements that have been made in speaker system and driver design. Have you seen the Synergy designs by Tom Danley? You might lump line arrays in here too from a purely technical standpoint but I don't consider these to be terribly hi-fi sounding, they do lots of other things very well though.
The best PA drivers available these days are very well behaved to, and performance isn't just about frequency response. That is the thing with these stacks, they appear to have covered the whole audio spectrum very well but they didn't worry too much about off axis response which is where phase and time alignment come into play. The way they used supertweeters for example was.. well obsolete. The very best axial performance is achieved with a single point source of sound so a single HF driver that is capable of handling all the mid-highs, or a combo of drivers like a coaxial or a Synergy horn is the way to go. This type of driver arrangement produces the same spectral and impulse balance across all of it's horizontal coverage whereas this old school stack would change constantly as one walked around. If surround sound was the goal I'd bet one could produce stunning 3D effects over the whole dance floor with only 4 properly designed stacks for example. IMO there is no need for anything beyond a 4-way system here, 18's, 12's, and your choice of single or dual driver CDs for the topend.
As for processing you will want something with FIR filters although IMO that would just be the last 5% improvement or the cherry on the top. Even a lowly Behringer DCX has the processing to do what's needed here and is as transparent as an open window, where it falls short are with the analog parts included inside the box which aren't as quiet as they could be.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.