inside/compared BSS - 9088 / DBX - PA / & DCX2496 |
Post Reply | Page 123> |
Author | |
nineleaves
Young Croc Joined: 31 August 2007 Status: Offline Points: 1327 |
Post Options
Thanks(0)
Posted: 02 April 2009 at 4:23pm |
justr remembered about my intention to post this up a while ago,
heres a comparison between the electronics in the bss soundweb 9088, the DBX - driverack pa & behringers dcx2496. so long as the internal dsp math is good (not difficult !), and not comprimised to run on crippled/overburdened dsps ..then the final word in the quality of the results will be the converters and opamps. here they are compared, make your own mind up ;) in summary: Behringer - DCX2496 - (96k/24bit) £200 //ADC (AKM) - dr. 117db, thd/n. -105db // DAC (AKM) - dr. 120db, thd. -100db // OPAMP - (4580) noise .8 nv, thd/n - 0.0005% (-106db) // DSP - 24bit sharc, 198 mflops, 66mhz, 80 bit capable// DBX - Driverack PA - (48k/24bit) £400 // ADC (AKM) - dr. 110db, thd/n. -103db // DAC (AKM) - dr. 112, thd/n. -94db // OPAMPS (4580/5532) - noise 0.8-5nv, thd - 0.0005% (-106db) // DSP - freescale 24bit, 150mips, 150mhz, 56 bit capable barrel shifter// BSS - 9088 - (48k/24bit) £2,250 // ADC (AKM) - 48khz, dr. 116db, thd/n. -105db // DAC (CRYSTAL) - 48 khz, dr. 106db, thd. 0.001% (-100db) // OPAMPS (OPA275) - noise 6nv, thd. 0.0006% (-104.5db) // 2x DSP - freescale 24bit, 80mips, 80mhz, 56 bit capable barrel shifter// (key to abbreviations: dr. = dynamic range - effective useable range of loudness from hiss floor to max output / thd/n. = total harmonic distortion + noise - how far below max output the magnitude of distortion elements and noise sits, and effectively masks/colours the sound. / noise, nv - level of voltage hiss produced by opamps ) (all phone piccies this time, sorry ;) ===== Behringer - DCX2496 ===== Chips are: * AKM - AK5393VS http://www.asahi-kasei.co.jp/akm/en/product/ak5393/ak5393_f04e.pdf Enhanced Dual Bit 24-Bit ADC # Sampling Rate: 1kHz to 108kHz # Full Differential Inputs # Dynamic range: 117dB # Thd/N: -105dB # Noise: -117dB (for mic input C ?) * AKM - AK5383VS http://www.asahi-kasei.co.jp/akm/en/product/ak5383/ak5383_f03e.pdf Enhanced Dual Bit 24-Bit ADC # Sampling Rate: 1kHz to 108kHz # Full Differential Inputs # Dynamic range: 110dB # Thd/N: -103dB # Noise: -110dB * 3X AKM - AK4393VF http://www.asahi-kasei.co.jp/akm/en/product/ak4393/ak4393_f02e.pdf Delta sigma stereo 24-Bit DAC # Sampling Rate: upto 108 khz # Full Differential outputs # Dynamic range: 120dB # Thd/N: -100dB # Noise: -120dB * ANALOG DEVICES - SHARC ADSP21065L http://www.analog.com/static/imported-files/data_sheets/ADSP-21065L.pdf 24 BIT DSP # 198 mflops # 66 mhz # 24bit with dual 80bit capable accumilators. * njr - jrc4580 http://semicon.njr.co.jp/njr/hp/fileDownloadMedia.do?_mediaId=152 # Dual Bipolar, Operational Amplifier # Low Noise: 0.8 nV/Hz # Low Distortion: 0.0005% # High Slew Rate: 5 V/us # Wide Bandwidth: 15 MHz newer versions also include a noise shield on the transformer, presumably to further improve matters: ===== BSS - DRIVERACK PA ===== Chips are: * AKM - AK5383VS http://www.asahi-kasei.co.jp/akm/en/product/ak5383/ak5383_f03e.pdf Enhanced Dual Bit 24-Bit ADC # Sampling Rate: 1kHz to 108kHz # Full Differential Inputs # Dynamic range: 110dB # Thd/N: -103dB # Noise: -110dB * AKM - AK4356VQ - (Discontinued ?) http://pdf1.alldatasheet.com/datasheet-pdf/view/54937/AKM/AK4356.html Delta sigma 2/6 channel 24-Bit DAC # Sampling Rate: 96 (6ch mode) / 192kHz (2ch mode) # Full Differential outputs # Dynamic range: 112dB # Thd/N: -94dB * MOTOROLA - DSPB56367PV150 - (discontinued) http://www.freescale.com/files/dsp/doc/data_sheet/DSP56367.pdf?pspll=1 24 BIT DSP, NOW DISTRIBUTED BY FREESCALE; PART DISCONTINUED BUT RELATED CHIPS HERE: # 150 mips # 150 mhz # 24bit with 56bit capable barrel shifter. * WINBOND - W78L32P-24 - (discontinued) 8 BIT MICROCONTROLLER * njr - jrc4580 http://semicon.njr.co.jp/njr/hp/fileDownloadMedia.do?_mediaId=152 # Dual Bipolar, Operational Amplifier # Low Noise: 0.8 nV/Hz # Low Distortion: 0.0005% # High Slew Rate: 5 V/s # Wide Bandwidth: 15 MHz * ti - ne5532a http://focus.ti.com/lit/ds/symlink/ne5532.pdf # Dual Bipolar, Operational Amplifier # Low Noise: 5 nV/Hz # Low Distortion # High Slew Rate: 9 V/s # Wide Bandwidth: 10 MHz ===== BSS - SOUNDWEB 9088 ===== chips on the conversion side are pretty nice tbh :) then again id be pretty pissed if they cut corners at that price; mind i dont use these personally. chips are: * AKM - AK5392 http://www.asahi-kasei.co.jp/akm/en/product/ak5392/ak5392_f01e.pdf Enhanced Dual Bit 24-Bit ADC # Sampling Rate: 1kHz to 54kHz # Full Differential Inputs # Dynamic range: 116dB # Thd/N: -105dB # Noise: -116dB * CRYSTAL - CS4390 http://www.cirrus.com/en/products/pro/detail/P32.html 24-Bit, 48 kHz Stereo D/A Converter # Sampling Rate: 1 KHz to 50 kHz # Dynamic range: 106dB # Noise: -115dB * CRYSTAL - CS3310 http://www.cirrus.com/en/pubs/proDatasheet/CS3310_F1.pdf Stereo Digital (analog domain) Volume Control # Dynamic range: 116dB # thd/n: 0.001% (-100db) # 0.5db step size # 127db total trim range * ANALOG DEVICES - OPA275 http://www.analog.com/static/imported-files/data_sheets/OP275.pdf # Dual Bipolar/JFET, Audio Operational Amplifier # Low Noise: 6 nV/Hz # Low Distortion: 0.0006% # High Slew Rate: 22 V/s # Wide Bandwidth: 9 MHz * ANALOG DEVICES - ADG411 http://www.analog.com/static/imported-files/data_sheets/ADG411_412_413.pdf Precision Quad analog SPST Switches * 2X MOTOROLA - XC56303PV80 http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=DSP56303&fsrch=1 24 BIT DSP, NOW DISTRIBUTED BY FREESCALE; PART DISCONTINUED BUT RELATED CHIPS HERE: assumed to be: # 80 mips # 80 mhz # 24bit with 56bit capable barrel shifter. * CUSTOM ASIC - SOUNDWEB * and other support chips i couldnt be asked to document. Edited by nineleaves - 18 July 2009 at 5:19pm |
|
nineleaves
Young Croc Joined: 31 August 2007 Status: Offline Points: 1327 |
Post Options
Thanks(0)
|
updated this post, comparing the 3 crossovers on hardware spec.. have a look and ponder whats going on here ;)
(((pssst ! ..the cheapest one is throwing up the better overall numbers ;) ))) Edited by nineleaves - 17 July 2009 at 8:15pm |
|
ceharden
The 10,000 Points Club Joined: 05 June 2005 Location: Southampton Status: Offline Points: 11776 |
Post Options
Thanks(0)
|
Well, just to add some real-world experience, about a year ago I 'upgraded' the system controller in a local venue from a Soundweb 9088 (Mk1 not iis) to a DCX2496 and not looked back. The system really benefited from the increase in sample rate (and any other advantages it has).
|
|
tommysb
Registered User Joined: 24 April 2006 Location: Finland Status: Offline Points: 1036 |
Post Options
Thanks(0)
|
Very interesting. Makes using a Motorola based chipset for DSP processors seem like not such a bad idea, and potentially could get v good performance with good D/A and A/D
|
|
nineleaves
Young Croc Joined: 31 August 2007 Status: Offline Points: 1327 |
Post Options
Thanks(0)
|
yup, tho using motorola 56000's are dsp chips by default & their deployment in these applications isnt a new thing by any measure; motorola ran into problems mind & failed with the launch of the sucessor 96000 series prior to becoming freescale semiconductors.
the analog devices Sharc processors are also veterans & the chips are right workhorses for dsp. its a long an contentious argument about who has the 'best' dsps on the market. but it dont count for much if the process of converting those analog waveforms to numbers and back again is sub-optimal :) |
|
Freddy Hopper
Registered User Joined: 31 July 2007 Status: Offline Points: 430 |
Post Options
Thanks(0)
|
in layman s terms.. who is the best and how much better |
|
TAIL LIFT
|
|
MarjanM
Old Croc Joined: 10 February 2005 Location: Macedonia Status: Offline Points: 7816 |
Post Options
Thanks(0)
|
And why no DCX pics?
|
|
Marjan Milosevic
MM-Acoustics www.mm-acoustics.com https://www.facebook.com/pages/MM-Acoustics/608901282527713 |
|
nineleaves
Young Croc Joined: 31 August 2007 Status: Offline Points: 1327 |
Post Options
Thanks(0)
|
dcx pics now uploaded, not all mine.. as i cant find the ones i took a while back.
'better' is an issue of complex technical debate & a matter of biased personal oppinions. but in general, most such boxes perform in similar shades of 'good'. just going on brand, people may prefer to put on display a bss or a dbx, however the little berry box can be clearly seen to not be using cheaper components, dispite its price point. it even shares a number of them with the other boxes too. what this roughly shows, is you certainly arent getting less for your money from them. in short, im a shameless fan of the dcx, you get alot more for your money even in hardware, than you do from most other manifacturers. but also added to this, the dcx2496 has the higher 96khz sampling rate, better all round combination of analog components; which offer overall less distortion & a lower noise floor. rather concerningly, the 2 grand soundweb 9088 uses components which will cause at least 10db more distortion, and raise the noise floor by at least 16db. and although it trims the volume of channels in the analog domain, the choice of trimmers introduces not insignificant extra distortion. i do appreaciate that this could make soundweb owners a little uncomfortable, though.. In tests using a dcx2496 & soundweb side by side with identical settings (using crossovers & parametric eq to correct a system), i also found the dcx2496 to sound alot clearer & could see no good reason to spend £2,000+ on such a box. |
|
nineleaves
Young Croc Joined: 31 August 2007 Status: Offline Points: 1327 |
Post Options
Thanks(0)
|
you may also see on the web that theres afew dubious types selling mod kits to 'improve' the dcx.. whilst some of these mods will help to a miniscule extent, its mostly about suckering impressionable people. so wouldnt pay too much attention to that as an indicator of 'out of the box' performance.. its just people shilling for a buck.
also to improve the clarity of sound & better utilise the range of the limiters, etc - the dcx2496 (and others) tends to drive its output signals a little 'hot', resulting in less bit depth being used at volume than is available. a crewd, but effective mod is to modify the leads going from the crossover to your amps like so: open up one end of the cable & add a 18k resistor to each of the connections, for both the - & the + (white & red). this will attenuate the output so the crossover can be driven harder, with its max output being closer to that of the amps. ..try smaller resistor values if max xover output is reached before max amp output - leave a little headroom on the crossover, so that it doesnt clip before the amp does ! ..as digital clipping sounds bad.. and will do nasty things to drivers ! Edited by nineleaves - 18 July 2009 at 6:21pm |
|
jonminns
Old Croc Joined: 13 April 2008 Location: United Kingdom Status: Offline Points: 1877 |
Post Options
Thanks(0)
|
just to add to this
i bought a 9088 a few weeks ago and had it out for the first time friday night. saturday night i did an a/b with the dcx. there really wasn't any noticeable difference i thought there was slightly more noise from the soundweb however the quality of the limiters overshadows this. they are excellent and make the dcx's sound awful! i still think the dcx is one of the best dsps that you can buy for a sensible price, but 8x8 for around £250 is a win for the web jon |
|
nineleaves
Young Croc Joined: 31 August 2007 Status: Offline Points: 1327 |
Post Options
Thanks(0)
|
fair play,
i have however noticed that alot of folks tend to totally misuse limiters & use them as a 'substitute' for correct gain structure. typically folks fail to set the senstivity of the xover to even vaguely match max output of mixer... and then run them so far into the red, that youd think they where power on led's. the soundwebs ones where certainly far more forgiving of this sort of abuse :) im the sort of chap who just sets up limiters to lightly catch the last few gremlins that get through, under those circumstances ive not noted any concerning quirks. but i guess its horses for courses :) |
|
jonminns
Old Croc Joined: 13 April 2008 Location: United Kingdom Status: Offline Points: 1877 |
Post Options
Thanks(0)
|
i set up limiters to catch clipping, or maybe hold the signal at -3dB on an amplifier too powerful etc. the gain structure will be set to be flat out at 0dB, with full limiting in effect at +3dB
i don't misuse them, but djs do, and the web limiters didn't really sound like they were working hard. the behringer ones are still good, much better than the drpa or bbe units interested in using the leveller feature of the web to cut reject anything above 0dB at the input for djs |
|
Post Reply | Page 123> |
Tweet |
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |