Hi, I repost here a thread posted in digital line section, cause here seems a better place.
I have the project to go with Xover but with shareware plug in, like JRiver, Rephase... to work for my futur DIY planar (ESL or Magneplanar...hesitate)
My understanding is that there are two ways :
- One I call Asynchrone : I use a DSP which run the data in a separate box near the electronic rack : amps, sources... Acquisition measure are traken with a laptop with a micro and you load after your set up work on the DSP box whic ran the Software Xover in a plug in made for the DSP : the software and the setup are maid before on the Laptop.
One I call Synchrone : The measure acquisition, XoverSoftware, soundcard and In for datas (the stream or the music library) are in a PC, most of the time fanless and big enough to powersupply a good souncard: so PCI rather han miniPCI.
I don't believe in analog out with pci soundcard but see that some brand make good second box (sound card > numeric stream > second box with D/A and "good" analog stages for the amps before the sperakers' drivers.
I have some questions please for the ones who tried this way :
Am i right with above ?
As I have already squeezebox for digital library on a upnp server and Duett streamers with spdif output for dacs :
Is there soundcards or second/standalone box with spdif In and Sync or Async for the digital Xover and/or rooms treatments ?
I don't want to have all the time a laptop to play with my librairy but can have a little computer or a standalone professional converter near the amps. The laptop is in my mind to set up with the micro acquisition. The library has always its server and TCIP or wifi communication for the stremings flux with the Duett. So I talk only for xover, room correction.
My first hope is to have analog output to go with multiamp like hypex or else. Don't know if there are full digital crossover to go with multiple powerdac for drivers.
i understand that th volume and library management is one of the problem... but i have squeezebox Duett and want to go with Free for soft in speakers and rooms management.
DLCP is expensive for me and not very open, MiniDSP in three ways are expensive : both are async (I prefer) but not really open, MiniDsp seems better as i can understand with a light view. Any good 3 ways speakers PCI soundcard and/or second box with spdif In for my squeezebox ?
Thank you to add to my thoughs, I'm beginner here but have a good understanding with computers technology as IT... but never really mix my hobby with that...
My final understanding is asynch is better for noise and some soundcard have noise problem to be serious xover interface (better standalone box maid for noise issue (buffers, transformers, async interfaces, DSP load for FIR IIR)
Can we hope something cheaper without loose audioquality in regard to the final price ? PC is cheaper or more expensive but with more possibilities with the excellent shareware or opensource software (EasySound, rePhase, J-river)...what about hardware Interface and quality of interconnect with pc soundcards (microphonie).
Is there someting like that : DSP : AD/DA with sabre or equivalent quality for separate the chanels (Xover) and apply IIR or FIR equalization ?
sorry to be long here and thank you to read it, it's a little bit confuse in my mind. Audio quality is the first factor but TCO (Total Cost Ownership) is second (less than 800 euros/dollars). I have just a ECM 8000, no usb/phantom interface yet : i'm at the beginning.
thank you for yours answers, the thread is totally open to make the OP's original question...smarter.
Eldam
I have the project to go with Xover but with shareware plug in, like JRiver, Rephase... to work for my futur DIY planar (ESL or Magneplanar...hesitate)
My understanding is that there are two ways :
- One I call Asynchrone : I use a DSP which run the data in a separate box near the electronic rack : amps, sources... Acquisition measure are traken with a laptop with a micro and you load after your set up work on the DSP box whic ran the Software Xover in a plug in made for the DSP : the software and the setup are maid before on the Laptop.
One I call Synchrone : The measure acquisition, XoverSoftware, soundcard and In for datas (the stream or the music library) are in a PC, most of the time fanless and big enough to powersupply a good souncard: so PCI rather han miniPCI.
I don't believe in analog out with pci soundcard but see that some brand make good second box (sound card > numeric stream > second box with D/A and "good" analog stages for the amps before the sperakers' drivers.
I have some questions please for the ones who tried this way :
Am i right with above ?
As I have already squeezebox for digital library on a upnp server and Duett streamers with spdif output for dacs :
Is there soundcards or second/standalone box with spdif In and Sync or Async for the digital Xover and/or rooms treatments ?
I don't want to have all the time a laptop to play with my librairy but can have a little computer or a standalone professional converter near the amps. The laptop is in my mind to set up with the micro acquisition. The library has always its server and TCIP or wifi communication for the stremings flux with the Duett. So I talk only for xover, room correction.
My first hope is to have analog output to go with multiamp like hypex or else. Don't know if there are full digital crossover to go with multiple powerdac for drivers.
i understand that th volume and library management is one of the problem... but i have squeezebox Duett and want to go with Free for soft in speakers and rooms management.
DLCP is expensive for me and not very open, MiniDSP in three ways are expensive : both are async (I prefer) but not really open, MiniDsp seems better as i can understand with a light view. Any good 3 ways speakers PCI soundcard and/or second box with spdif In for my squeezebox ?
Thank you to add to my thoughs, I'm beginner here but have a good understanding with computers technology as IT... but never really mix my hobby with that...
My final understanding is asynch is better for noise and some soundcard have noise problem to be serious xover interface (better standalone box maid for noise issue (buffers, transformers, async interfaces, DSP load for FIR IIR)
Can we hope something cheaper without loose audioquality in regard to the final price ? PC is cheaper or more expensive but with more possibilities with the excellent shareware or opensource software (EasySound, rePhase, J-river)...what about hardware Interface and quality of interconnect with pc soundcards (microphonie).
Is there someting like that : DSP : AD/DA with sabre or equivalent quality for separate the chanels (Xover) and apply IIR or FIR equalization ?
sorry to be long here and thank you to read it, it's a little bit confuse in my mind. Audio quality is the first factor but TCO (Total Cost Ownership) is second (less than 800 euros/dollars). I have just a ECM 8000, no usb/phantom interface yet : i'm at the beginning.
thank you for yours answers, the thread is totally open to make the OP's original question...smarter.
Eldam
Hi JPlesset,
thank you for your answer you save me from the shame...my questions have not a lot of sucess !
If I have a lot of money I will go to buy Trinov which is a computer audiophile boxed with audiophile DAC outputs.
I saw Minidsp support rePhase, but don't understand if the DSP is enough powerfull. Don't understand with Minidsp if i can take measurement without their USB Micro but with a Laptop and send result targets on the DSP ? MiniDigi+MiniDSP are pleasent but two ways for xover : more is more expensive !
I never can't understand if it's possible to have digital volume management with the streamer like in my squeezeboxDuett before the D/A of a DSP/DAC unit for the Xover and room correction ? i read volume management is better in the final DACs!!!!
Last point : all these computers based solution or Pro Boxes don't have the possibility of the tunning outstages we like to aplly here...but if you put all the audiophile DAC in one box needed to manage the multiway xover before the amps and the speakers' drivers. Out connectic of PCI card seem poor in term of audiophony, noise...
good solution seem to go towards full digital with powerDAC... but best amps are not full digital like the Ncore which is analog (it's not a powerdac).
Hypex seems take care of sound with DLCP but my understanding is it has not all the possibilities of a strong PC conf (FIR+IIR...) and expensive with all the necessary refinements : preamp, remote controle : the price of a all in one with box but without the box, guaranty, powersupply...(maybe not chinese maid explain the price ?)
I believe in open world & shareware when we talk about softwares (Rephase, REW, J River, ... and want to go in a solution with free choice but with no loss of hardware quality (DAC, outputsatges....).
Sync (PC+ DAC) or Async (DSP+DAC) ?
I believe more and more that a solution like the DSP SHARC (I2S output) of MINIdsp and a I2S In a separate DAC for Xover is a solution : not the easiest in hardware management but the best for hardwarequality with DIY and cost management ? Am I right ?
So it will look like that : squeezebox Duett (library management & digital volume) with Spdif out -> SpdiF/I2S asynchrone interface -> DSP Sharc with I2S multiway outputs (3+3 for stereo 3 ways Xover) -> D/A DAC box with I2S IN : one Multichanal DAC like Sabre or multiple DAC chips and audiophile analog outputstages (powerdac are not really musicals for the moment...try one chinese one, no expensive but poor sound).
In this solution it's for me Asynchrone : measurement is taken with another software than the plug in in the DSP, for example with RePhase or Rew in a laptop with a micro : and we send the result target curve on the DSP : we can't listen result in direct like systems with firewire between PC and Xover when we seeting up the EQ...
For the last stage : DIY multiDAC or Pro Box seems better than all in one PCI soundcards when talking about analog out!
thank you for your answer you save me from the shame...my questions have not a lot of sucess !
If I have a lot of money I will go to buy Trinov which is a computer audiophile boxed with audiophile DAC outputs.
I saw Minidsp support rePhase, but don't understand if the DSP is enough powerfull. Don't understand with Minidsp if i can take measurement without their USB Micro but with a Laptop and send result targets on the DSP ? MiniDigi+MiniDSP are pleasent but two ways for xover : more is more expensive !
I never can't understand if it's possible to have digital volume management with the streamer like in my squeezeboxDuett before the D/A of a DSP/DAC unit for the Xover and room correction ? i read volume management is better in the final DACs!!!!
Last point : all these computers based solution or Pro Boxes don't have the possibility of the tunning outstages we like to aplly here...but if you put all the audiophile DAC in one box needed to manage the multiway xover before the amps and the speakers' drivers. Out connectic of PCI card seem poor in term of audiophony, noise...
good solution seem to go towards full digital with powerDAC... but best amps are not full digital like the Ncore which is analog (it's not a powerdac).
Hypex seems take care of sound with DLCP but my understanding is it has not all the possibilities of a strong PC conf (FIR+IIR...) and expensive with all the necessary refinements : preamp, remote controle : the price of a all in one with box but without the box, guaranty, powersupply...(maybe not chinese maid explain the price ?)
I believe in open world & shareware when we talk about softwares (Rephase, REW, J River, ... and want to go in a solution with free choice but with no loss of hardware quality (DAC, outputsatges....).
Sync (PC+ DAC) or Async (DSP+DAC) ?
I believe more and more that a solution like the DSP SHARC (I2S output) of MINIdsp and a I2S In a separate DAC for Xover is a solution : not the easiest in hardware management but the best for hardwarequality with DIY and cost management ? Am I right ?
So it will look like that : squeezebox Duett (library management & digital volume) with Spdif out -> SpdiF/I2S asynchrone interface -> DSP Sharc with I2S multiway outputs (3+3 for stereo 3 ways Xover) -> D/A DAC box with I2S IN : one Multichanal DAC like Sabre or multiple DAC chips and audiophile analog outputstages (powerdac are not really musicals for the moment...try one chinese one, no expensive but poor sound).
In this solution it's for me Asynchrone : measurement is taken with another software than the plug in in the DSP, for example with RePhase or Rew in a laptop with a micro : and we send the result target curve on the DSP : we can't listen result in direct like systems with firewire between PC and Xover when we seeting up the EQ...
For the last stage : DIY multiDAC or Pro Box seems better than all in one PCI soundcards when talking about analog out!
Hello, again...
You always have choice. My choices were governed by several factors, including:
1. Wife acceptance factor, including: A. No money spent, B. Minimal visual footprint, C. Minimal downtime.
2. On-hand equipment: Old PC with Turtle Beach Montego DDL sound card.
3. Unix experience.
By mixing the above, I followed the steps in this posting: Digital Crossover/EQ with Open-Source Software: HOWTO | Richard's Stuff
There are a quite broad selection of good quality sound cards (5.1, and 7.1) available that will work with the software, including external USB selections. I see quite a conversation about optimizing a Xonar card in this forum, too. That may be of interest. Since the software noted has little impact on the PC, there's really no need to build an "audiophile" computer, unless you enjoy spending extra. My system sounds better than it ever has, no audible noise in the speakers, and much more convenient to select music.
The calculations for the crossover are done with high precision, and there are a broad variety of plugin filter types available, including phase correcting types...
Good luck!
You always have choice. My choices were governed by several factors, including:
1. Wife acceptance factor, including: A. No money spent, B. Minimal visual footprint, C. Minimal downtime.
2. On-hand equipment: Old PC with Turtle Beach Montego DDL sound card.
3. Unix experience.
By mixing the above, I followed the steps in this posting: Digital Crossover/EQ with Open-Source Software: HOWTO | Richard's Stuff
There are a quite broad selection of good quality sound cards (5.1, and 7.1) available that will work with the software, including external USB selections. I see quite a conversation about optimizing a Xonar card in this forum, too. That may be of interest. Since the software noted has little impact on the PC, there's really no need to build an "audiophile" computer, unless you enjoy spending extra. My system sounds better than it ever has, no audible noise in the speakers, and much more convenient to select music.
The calculations for the crossover are done with high precision, and there are a broad variety of plugin filter types available, including phase correcting types...
Good luck!
Here is the setup I ran in my car...it might give you a few ideas...
I ran two Xonar's analog out...it was a 3-way active crossover VST (LSFilter...the best if you can find it). Since I had 4X 2-ways plus sub, I needed two soundcards. The way I had it configured, was all left channel audio was low pass, all right channel, high pass. Sub-out had its own jack. This ran into Audiomulch to do the patching, run the x-over, delay and "room convulsion" correction.
However, with J. River, you can probably do all of this without Audiomulch.
Here is the convulsion pics...
You could also create a convulsion for x-over, delay, correction in a wav file and run it in Pristine Space VST.
Sent from my iPhone...
I ran two Xonar's analog out...it was a 3-way active crossover VST (LSFilter...the best if you can find it). Since I had 4X 2-ways plus sub, I needed two soundcards. The way I had it configured, was all left channel audio was low pass, all right channel, high pass. Sub-out had its own jack. This ran into Audiomulch to do the patching, run the x-over, delay and "room convulsion" correction.
However, with J. River, you can probably do all of this without Audiomulch.
Here is the convulsion pics...
You could also create a convulsion for x-over, delay, correction in a wav file and run it in Pristine Space VST.
Sent from my iPhone...
Yes I think JRiver or rephase are great softs for that and it's a very good idea to make EQ in a car.
My feeling is the soundcards are not at the same level for D/A conversion than hifi one cause of the output stage and all the subtilities we fight against (look at digital line forum). Am I wrong or Have I to go with products like miniDSP, tweak with IIS or do the D/A with external converter (all the digital crossover and eq maid with the DSP of the soundcards) ?
I can hear with my speakers (not the futur ones) all the subtle move with electronics... That's why I think it's not only a soft choice...hardware and architecture count. maybe I'm wrong ? I don't want move for worse or just enjoyment with soft if the hardware side waste it.
that's the sense of my question but has no experience with soft sound correction. Just tweaked amp and digital sources
My feeling is the soundcards are not at the same level for D/A conversion than hifi one cause of the output stage and all the subtilities we fight against (look at digital line forum). Am I wrong or Have I to go with products like miniDSP, tweak with IIS or do the D/A with external converter (all the digital crossover and eq maid with the DSP of the soundcards) ?
I can hear with my speakers (not the futur ones) all the subtle move with electronics... That's why I think it's not only a soft choice...hardware and architecture count. maybe I'm wrong ? I don't want move for worse or just enjoyment with soft if the hardware side waste it.
that's the sense of my question but has no experience with soft sound correction. Just tweaked amp and digital sources
Hi all, Eldam,
very interested by this post,
one solution for combining cheap processing (your "Synchrone" solution) and high quality DAC would be to output from the PC in digital -> HDMI.
I found the "Richard's stuff" way is promising ( Digital Crossover/EQ with Open-Source Software: HOWTO | Richard's Stuff).
With HDMI 7.1 audio you can (or I should write "should be able to", I've not yet tried) plug into a 5.1 or 7.1 receiver as a DAC. You can choose a high end one if you wish (used ones to be found pretty cheap on internet - ebay and the such).
Another benefit of this solution is that you get volume control on all channels for free (which enables you to keep worling at line level on the DAC and therefore remove quantification noise or distorsion wich is an essential problem on the DCX2496).
Please keep us posted of your thoughts !
Merci,
Danhilu
very interested by this post,
one solution for combining cheap processing (your "Synchrone" solution) and high quality DAC would be to output from the PC in digital -> HDMI.
I found the "Richard's stuff" way is promising ( Digital Crossover/EQ with Open-Source Software: HOWTO | Richard's Stuff).
With HDMI 7.1 audio you can (or I should write "should be able to", I've not yet tried) plug into a 5.1 or 7.1 receiver as a DAC. You can choose a high end one if you wish (used ones to be found pretty cheap on internet - ebay and the such).
Another benefit of this solution is that you get volume control on all channels for free (which enables you to keep worling at line level on the DAC and therefore remove quantification noise or distorsion wich is an essential problem on the DCX2496).
Please keep us posted of your thoughts !
Merci,
Danhilu
Thank you for your boost,
The problem is I'm not sure to know what I exactly need... just learned than the Crossover + phase have to be FIR and the corrections of peaks or/and room in IIR !
I can go with "asynchrone solution" because have already data storage + volume streaming solution (with SqueezeBox Duet + smartphone + duett remote control). As say Soundcheck, I'm not against an another system i full pc: Linux or others as I'm not against a minimalist solution : a miniDigi between the streamer Duett and a multi DAC.
I read than the MiniShark works better than the MiniDiGI but don't know if it's true. And maybe not too expensive to buy one IIS DAC module ESSxxxx23 by pair of drivers for multi amp and may be one Wolfson for the bass driver. Tweaking a pc soundcard at 500 dollars... to have not a super DAC...maybe good but not super !)
HDMI... is there choice ? Oppo ? I don't know asynchrone plug like the amareno/USB with HDMI if using normal DACS (SPIDF) or HDMI to IIS ?
With IIS.... is there ussues with clocks and digital ground loop because multiple IC ?
in my mind to be serious, The dac is the masterchoice if active crossover because all the processors for Crossovers and DRC seem strong enough ?
have problems too with measurement ! .... ouf I have classics speakers for the moment to listen to music !
What do you think ?
The problem is I'm not sure to know what I exactly need... just learned than the Crossover + phase have to be FIR and the corrections of peaks or/and room in IIR !
I can go with "asynchrone solution" because have already data storage + volume streaming solution (with SqueezeBox Duet + smartphone + duett remote control). As say Soundcheck, I'm not against an another system i full pc: Linux or others as I'm not against a minimalist solution : a miniDigi between the streamer Duett and a multi DAC.
I read than the MiniShark works better than the MiniDiGI but don't know if it's true. And maybe not too expensive to buy one IIS DAC module ESSxxxx23 by pair of drivers for multi amp and may be one Wolfson for the bass driver. Tweaking a pc soundcard at 500 dollars...
to have not a super DAC...maybe good but not super !) HDMI... is there choice ? Oppo ? I don't know asynchrone plug like the amareno/USB with HDMI if using normal DACS (SPIDF) or HDMI to IIS ?
With IIS.... is there ussues with clocks and digital ground loop because multiple IC ?
in my mind to be serious, The dac is the masterchoice if active crossover because all the processors for Crossovers and DRC seem strong enough ?
have problems too with measurement ! .... ouf I have classics speakers for the moment to listen to music !
What do you think ?
Salut Eldam,
I think you have quite a lot of questions ;-)
My view is
1. I want to try active crossover - analog filter are expensive and sonicaly poor and demanding for the power amp. And a nightmare to design for a 3 way speaker.
2. "Richard's stuff" way is architecturaly elegant because you split the active crossover - on a computer - and the DAC - which you can export anywhere you want thanks to HDMI or S/PDIF. On classical active crossover solution; like the famous DCX2496, you're stuck with the DAC, with no freedom to play arround.
Plus it is cheap, because you can use any mass market PC for the digital part, and any (decent) HDMI capable Power amp for the DAC & power amp. Of course you also can decide to go with a high end DAC (with hard to prove benefit in my opinion - but this is another story).
I'm still wondering whether I want to re-install Linux on an old spare PC, so that I can try the concept in the real life withour spending a €, or I buy a new low power device (like a Celeron 1007U with max power 17W, with build in HDMI & SPDIF output).
Please keep us posted with your progress.
I think you have quite a lot of questions ;-)
My view is
1. I want to try active crossover - analog filter are expensive and sonicaly poor and demanding for the power amp. And a nightmare to design for a 3 way speaker.
2. "Richard's stuff" way is architecturaly elegant because you split the active crossover - on a computer - and the DAC - which you can export anywhere you want thanks to HDMI or S/PDIF. On classical active crossover solution; like the famous DCX2496, you're stuck with the DAC, with no freedom to play arround.
Plus it is cheap, because you can use any mass market PC for the digital part, and any (decent) HDMI capable Power amp for the DAC & power amp. Of course you also can decide to go with a high end DAC (with hard to prove benefit in my opinion - but this is another story).
I'm still wondering whether I want to re-install Linux on an old spare PC, so that I can try the concept in the real life withour spending a €, or I buy a new low power device (like a Celeron 1007U with max power 17W, with build in HDMI & SPDIF output).
Please keep us posted with your progress.
Yes I have always a lot of questions because that's my project management method...The only "filter" where W questions/answers is good ... go no-go, then buy only... because after DIY cost too much for the quality of (my) results !Thanks for the input.
1) agree, even just set up a tweeter (that I experiment for few weeks)
2) many people talk about the Richard stuff way, there are anothers too, but for me the feeling of the control remote for control library and digital volume has to be here for me and i'm happy for that with SqueezeBox. That's why my OP question... Don't see really linux plateform that manage smartphones for remote : Android or Iphone (I'm Android...)
Don't know any DAC with HDMI... I have to re read Richard stuf review..
like too the answer to me by an another french Zaireeka in another thread about Processor + tweasted pair Audio Sabre plateform... RME ADAT out -> Dantimax Optorec (a kit converting ADAT to I2S)
OptoRec/Gen
We have to avoid the factory of jitter ! another problem with I2S is the shorter wires possible and where the masterclock is...
Last edited:
danhilu, I've done the Richard's stuff route, and while it wasn't exactly simple to get working, I'm very satisfied with the results. System hasn't crashed one time, sound is excellent, and it's easy to tweak.
Yes, I used an old PC to install Linux on. Single core AMD 64-bit, with 2 gig ram. It uses at most 5% of the processor and less than one gig of ram. It's not hit swap yet.
Yes, I used an old PC to install Linux on. Single core AMD 64-bit, with 2 gig ram. It uses at most 5% of the processor and less than one gig of ram. It's not hit swap yet.
Eldam, I've read your original post again so that I understand you.
There is one point I invite you to reconsider. Doing volume control on the signal before doing the crossover leads to quantification noise. You risk to end up having a 6 bit usefull signal althoug you use expensive 24bits DACs.
Note that this is the main reason why many complain that the DCX2496 is "noisy".
Let me try to explain, and please be forgiving with my explanation, doing my best. My values are surely wrong, but you'll get the idea if you stick to it.
Let's take a track, coded on 24 bits, which has been optimised so that its peak signal is at the max value. Your peak signal has all bits sets to 1, i.e. (I think) something like 0x07FFFFF (max positive value on digned 24bit), ie decimal 8388607.
The quantification noise (means the error caused by a digital approximation of an analog signal) is (theoretically) one bit, so the noise is 1/8388607.
Which is 0,000012%. This is .. excellent. Of course not hearable, but nice.
Now let's assume you do not listen to the signal at max volume, which would blow your tweeter away. You listen at, say, 1/10th of the volume, which is a typical volume for in house. 1/10the of the volume means, on my preamp, -75dB. I never remember, dB are tricky, but I think the signal gets divided by two each time you remove 6 dB, here 12 times. 8388607 divided by 2, 12 times, You are actually removing 12 bits out of 12, is 2048 (actually instead of having 24 bits you have 12 bits). Quanrtification noise is 1/2048 is 0.05%. Which is less nice (although probably still hard to hear). And, considering the peak is never at max level, and the DAC is not perfect, you'll have even worse situation. And if now you even turn the volume down, then you're in a bad situation.
It is far better to do all computation and crossover with max "line level" signal (most of the 24 bit info), and do the DAC with line level, and do the volume control as late as possible, idealy in analog (where you do not have any quantification noise).
But then you have a problem because you want to have a volume control on 6 channels, which is not common. There are solutions (look for DCX2496 DIY volume in google) but they are expensive and DIY. My solution is to use a multi chanel amplifier (or "receiver"). And, for the cost of one fancy sound card, you get on ebay used 5+1 receiver with digital input (S/PDIF or HDMI) and volume control and power amplifier.
Note there are also digital preamps like the Sony EP9ES.
Hope my lenghty explanation is not more confusing than helping..
There is one point I invite you to reconsider. Doing volume control on the signal before doing the crossover leads to quantification noise. You risk to end up having a 6 bit usefull signal althoug you use expensive 24bits DACs.
Note that this is the main reason why many complain that the DCX2496 is "noisy".
Let me try to explain, and please be forgiving with my explanation, doing my best. My values are surely wrong, but you'll get the idea if you stick to it.
Let's take a track, coded on 24 bits, which has been optimised so that its peak signal is at the max value. Your peak signal has all bits sets to 1, i.e. (I think) something like 0x07FFFFF (max positive value on digned 24bit), ie decimal 8388607.
The quantification noise (means the error caused by a digital approximation of an analog signal) is (theoretically) one bit, so the noise is 1/8388607.
Which is 0,000012%. This is .. excellent. Of course not hearable, but nice.
Now let's assume you do not listen to the signal at max volume, which would blow your tweeter away. You listen at, say, 1/10th of the volume, which is a typical volume for in house. 1/10the of the volume means, on my preamp, -75dB. I never remember, dB are tricky, but I think the signal gets divided by two each time you remove 6 dB, here 12 times. 8388607 divided by 2, 12 times, You are actually removing 12 bits out of 12, is 2048 (actually instead of having 24 bits you have 12 bits). Quanrtification noise is 1/2048 is 0.05%. Which is less nice (although probably still hard to hear). And, considering the peak is never at max level, and the DAC is not perfect, you'll have even worse situation. And if now you even turn the volume down, then you're in a bad situation.
It is far better to do all computation and crossover with max "line level" signal (most of the 24 bit info), and do the DAC with line level, and do the volume control as late as possible, idealy in analog (where you do not have any quantification noise).
But then you have a problem because you want to have a volume control on 6 channels, which is not common. There are solutions (look for DCX2496 DIY volume in google) but they are expensive and DIY. My solution is to use a multi chanel amplifier (or "receiver"). And, for the cost of one fancy sound card, you get on ebay used 5+1 receiver with digital input (S/PDIF or HDMI) and volume control and power amplifier.
Note there are also digital preamps like the Sony EP9ES.
Hope my lenghty explanation is not more confusing than helping..
Thanks, yes it helps. No more confusing because more data helps to understand better. With no knowledge it's impossible to drive the puzzle at its end.
For the moment : the numeric volume control of the squeezebox is used in the last 10 % of maximum volume just for confort before to go to the separated DAC (= before it so wrong) and the general volume is managed after the DAC by the analog preamp ! I hope using remote control for volume between 9O and 100 (max position) don't cause to many problems.
Yes I saw six channel volume analog interface exist...some of them can be expensive but some in France make good inexpensive ones with multiple Vishay 2 channels stacked together.
I saw than Nuforce (thanks for sony ref too) make numeric preamp.
So the better if no numeric preamp is analogic volume control after the DACS. I assume too than the db level between drivers is managed in the numeric domain by the crossover interface (numeric if PC and not in analogic domain à la Linkwitz with IC and electronics parts).
Ok I understand why HDMI with soundcard if AV multiamp receiver with embeded multiDAC is used (how is managed the split of signals in it between DAC and amps and volume interface ?).
In my mind and in relation to my OP, the dacs and the amps are important to improve the classical passive solutions:
- Better if DACs can be choosen after crossover and DRC processing
- Better if amps can be choosen after the DACs (hypex or others)
What I can't understand now is that the volume of drivers for crossover is set by the soft in numeric domain. Is it equal to quantification noise ? Or do we have to manage an analogic attenuation after the DACS.
I can say better a 12 channel analogic control because i like to set up with precision left and right volume without Balance plot (i can hear it and sometimes do another set up for another disc... i am a fanatic of stereo room perception !
)
It could be simplier to go with powerDAC, i've got one from hifymeDIY with volume remote control (but think the attenuation is maid in the DAC chip...) but still thinking it's too poor solution at the moment. By the way multipowerdac amp are AV receiver for the moment !
Have we go back with opa like active crossover IC boards à la Linkwitz...himself seem pleased by processors but doesn't believe in DRC with multipoint correction in FIR domain.
Am i right ? JPlesset seems to say : don't matter, just do it, even a basic souncard with reliable set up will sound better than passive.
At the end i have to say it's very important to manage the volume and the library with remote control because if not i prefer go back to my rega planar 3 and vynils.
For the moment : the numeric volume control of the squeezebox is used in the last 10 % of maximum volume just for confort before to go to the separated DAC (= before it so wrong) and the general volume is managed after the DAC by the analog preamp ! I hope using remote control for volume between 9O and 100 (max position) don't cause to many problems.
Yes I saw six channel volume analog interface exist...some of them can be expensive but some in France make good inexpensive ones with multiple Vishay 2 channels stacked together.
I saw than Nuforce (thanks for sony ref too) make numeric preamp.
So the better if no numeric preamp is analogic volume control after the DACS. I assume too than the db level between drivers is managed in the numeric domain by the crossover interface (numeric if PC and not in analogic domain à la Linkwitz with IC and electronics parts).
Ok I understand why HDMI with soundcard if AV multiamp receiver with embeded multiDAC is used (how is managed the split of signals in it between DAC and amps and volume interface ?).
In my mind and in relation to my OP, the dacs and the amps are important to improve the classical passive solutions:
- Better if DACs can be choosen after crossover and DRC processing
- Better if amps can be choosen after the DACs (hypex or others)
What I can't understand now is that the volume of drivers for crossover is set by the soft in numeric domain. Is it equal to quantification noise ? Or do we have to manage an analogic attenuation after the DACS.
I can say better a 12 channel analogic control because i like to set up with precision left and right volume without Balance plot (i can hear it and sometimes do another set up for another disc... i am a fanatic of stereo room perception !
)It could be simplier to go with powerDAC, i've got one from hifymeDIY with volume remote control (but think the attenuation is maid in the DAC chip...) but still thinking it's too poor solution at the moment. By the way multipowerdac amp are AV receiver for the moment !
Have we go back with opa like active crossover IC boards à la Linkwitz...himself seem pleased by processors but doesn't believe in DRC with multipoint correction in FIR domain.
Am i right ? JPlesset seems to say : don't matter, just do it, even a basic souncard with reliable set up will sound better than passive.
At the end i have to say it's very important to manage the volume and the library with remote control because if not i prefer go back to my rega planar 3 and vynils
.Hi
Have please a question: if external souncard without SHARC: Where is maid Software DSP processing ? With the computer processor ? A dedicated 32 bits Sharc chip is better ?
Mini personal computer then usb to async I2S then Dacs are ok for REW or RePhase?
Hesitate between minisharc grom minidsp and computer...
Both i would use external Dac like curyman or tweastedpetaudio...or both in three ways.
Have please a question: if external souncard without SHARC: Where is maid Software DSP processing ? With the computer processor ? A dedicated 32 bits Sharc chip is better ?
Mini personal computer then usb to async I2S then Dacs are ok for REW or RePhase?
Hesitate between minisharc grom minidsp and computer...
Both i would use external Dac like curyman or tweastedpetaudio...or both in three ways.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.