'Perceive v2.0' Construction Diary

[ShinOBIWAN]

Hello all

To save re-iterating actually what's going on here take a look at this thread:

Perceive v1.0 Thread

In it you'll find details of the Perceive v1.0 such as drivers, construction, lots of questions/answers on virtually all aspects of design such as box construction, drivers etc. There is also an interesting discussion on the merits and implemention details of a full digital active XO with digital room correction (DRC) using a dedicated PC and software that provides quite startling quality.

During this time the v1.0 took some shifts and turns throughout the few months I'd spent building, testing and tweaking and compared to the original plan the following had changed:

• I'd changed the analogue active XO to full digital and saw a big improvement especially with DRC/EQ.
  
• The loading of the bass cabinet was changed from sealed to ported to help cure a very slightly muffled lower midrange/upper bass and to create a much better transition to the ATC midrange.
  
• The 6 amps that are used to tri-amp the speaker switched from Rod Elliots ESP P101 to Krell Clone and then to AKSA 55 & 100.
  
• Swapped the ATC SM75-150 for the SM75-150S 'Super' version, I got a pretty good deal here and only paid £100 for this swap
  
• And then lastly, I switched out the Seas 4 x L22's used for the lowpass to 2 x Seas Excel W22's for an incremental upgrade in sound quality.

With all these changes it simply felt like a fudge and that a complete rebuild was in order to take stock of the tweaks and modifications, especially the lowpass driver change.

As it stands now I'm using the following drivers:

• Scanspeak R2904 Ring Radiator Tweeter
• ATC SM75-150S 3" Dome Midrange
• Seas W22EX001 8" Bass

I'll post some specifics on amplification and XO details later on the thread.

Scale drawing of the v2.0:



Not really done a great deal as I've been modelling enclosures for the W22 and looking at existing designs using this driver to see what other designers thought. Got that sorted though and I started a couple of days ago on cutting.

Here's the results so far:



Lots more to do obviously but my aim is to have these finished by Christmas ready for the holidays

More to come over the next few weeks.

Cheers,
Ant

[Vil]

Shin ,

looks like another nice tread just came to life , but I am curious - you decided not go for Supravox at bass position ,what reasons ?

[jacklelad]

ShinOBIWAN,

Firstly, as a newcomer to this forum and DIY I would like to thank you for all your informative contributions to many threads I have read, especially on active crossover/PC based DRC etc.

Secondly, this looks like a great project, and I can't wait to hear about the results. I'm already thinking I will want to recreate it.

Thanks Jack

[ShinOBIWAN]

Originally posted by Vil
Shin ,

looks like another nice tread just came to life , but I am curious - you decided not go for Supravox at bass position ,what reasons ?

Hey Vil!

Your XO suggestions have been invaluable to me so its good to see you here again

Regarding the Supravox, I had something of a nightmare getting hold of two, eventually I tracked down Clofis in the Netherlands but they won't ship to the UK without CC payment. I then found out that Supravox France sells direct to the puplic but got a quite amazing quote of 650Euro's including taxes for two drivers shipped to the UK .
Can't afford that considering that I'm going to definitely need to build a sub to augment the last couple of octaves, doubley so with the supravox actually since its around -10db at 50hz. By the time I'd factored this in I was looking at another £400 for the sub driver and £250 for a decent plate amp. It was all starting to look like very diminishing returns on what was nearly double the price for the Supravox's over the W22's. I guess you've got to draw the line somewhere regardless of any design ideal's such as 'reference'.

No doubt the Supravox's are better drivers but by how much? Well I'd guess at not a lot. But if these W22's don't work out I'll rebuild the entire thing again for the 3rd time

Originally posted by jacklelad
ShinOBIWAN,

Firstly, as a newcomer to this forum and DIY I would like to thank you for all your informative contributions to many threads I have read, especially on active crossover/PC based DRC etc.

Secondly, this looks like a great project, and I can't wait to hear about the results. I'm already thinking I will want to recreate it.

Thanks Jack

Thanks for that Jack, appreciate it.

[ShinOBIWAN]

A CAD rendering of the finished speaker to give you an idea of look:

[wigginjs]

I've just got this feeling that when these things are done you could charge admission to hear them.

[RyanC]

Nice man-

One day I'll make my speaks so pretty. Right now I just got this Pro Tools HD rig that I'm learning how to use and I got my XO setup in PT with no latency and an easy switch to FIR xo's. Plus the digi 192 DA's are sounding killer compared to the motu crap I was using-

Anyway looks like they just keep gettin better! I wish I could hear them. . .

[ShinOBIWAN]

Originally posted by m0tion
I've just got this feeling that when these things are done you could charge admission to hear them.

I really hope that is true, I could then recoupe some of the money I've lost/wasted through swapping bits.

Originally posted by RyanC
Nice man-

One day I'll make my speaks so pretty. Right now I just got this Pro Tools HD rig that I'm learning how to use and I got my XO setup in PT with no latency and an easy switch to FIR xo's. Plus the digi 192 DA's are sounding killer compared to the motu crap I was using-

Anyway looks like they just keep gettin better! I wish I could hear them. . .

Hi again Ryan,

Still messing with the XO? Its addictive isn't it

I had 3 presets that all sounded great but couldn't decide which sounded best over a wide genre of music. One was great for dance, another for acoustic/classical and the last favoured bass heavy material such as home theater.
Its just a hell of a lot of fun having all that flexibility, its quite feasible to create an entire crossover within a day but I find its best to keep coming back over a period of a couple of weels to really listen and appreciate the subtle changes that can make the difference.

[RyanC]

Ahh yes-

I'm going with 2 main ones right now. One is a 96dB/oct using the UA cambridge EQ on the elliptic 6 filter. This one is not native (TDM or UAD card only) so unfoutunatly you all can't play with it cheaply. The other is the FIR filters with the waves lineq pack. I do know what you mean by different ones being better- like for Rock or Hip hop I like my big ol TC2+'s comin up to 200hz even. But for jazz it just sounds bad! Funny that you can tune it pretty flat either way and it sounds so different. Also it is amazing how diffent the lowest octive is mixed on different records.

Man Protools HD sounds good- i think it has somthing to do with the fact that all clocking is done at 256 x fs, also that the WC cable is integrated with the digital cable so the length is identical.

What is really cool is that a single fader can control levels for as many outs as you want (via grouping or multi-channel tracks) so I could easily have an all active realtime XO for a full surround system! It would probably eat up a whole HD card though.

Too bad this stuff is so spendy- I bet ProTools would make another marked improvement, and it certenly has lower latency than VST.

On an interesting note- using the FIR setup makes me not want to use IIR filters in mixing- then it occured to me- Pretty much everything you hear has gone through lots of IIR filters (even just HPF's on mic preamps). I notice on a drum set that you loose punch when you engage 2 self canceling IIR filters (eg one is +3dB at 1k one is -3dB at 1k Q is the same). Too bad mixing is not really feasable with FIR filters yet.

How often do you find yourself going damn thousands of dollars (or pounds for you) and I'm still no where near "perfect". I guess we are all going to have to settle for the pursuit of perfection. . .

RC

[Vil]

>>>Man Protools HD sounds good- i think it has somthing to do with the fact that all clocking is done at 256 x fs, also that the WC cable is integrated with the digital cable so the length is identical.


do you remember my post in #1 tread ?
low jitter clock source is at the top of most important things for DAC/ADC . 256F (super clock) means no PLL's at conversion clocking parth .

V.