Signal Level in miniDSP x-over - Page 4 - diyAudio
Go Back   Home > Forums > Commercial Sector > Manufacturers > miniDSP

miniDSP Low cost, modular Digital Signal Processor (DSP) kits for the DIYer from miniDSP.

Reply
 
Thread Tools Search this Thread
Old 17th March 2013, 06:46 PM   #31
boris81 is offline boris81  United States
diyAudio Member
 
Join Date: Oct 2009
Location: white plains, ny
I see now, I had the wrong idea. Thank you, George for the detailed explanation!
I only have basic understanding of DSP, I'm sorry if I confuse anyone by posting my wrong assumptions.

I understand that attenuating the input level will decrease the resolution of the system since signal will initially be converted using less bits. At an input level of 0 the conversion will use 24bits, at -6dB 23bits, at -12dB 22bits and so on.
I understand now that keeping the input at 0 and lowering the output level instead can keep the digital resolution of the system at 24bits.

Last edited by boris81; 17th March 2013 at 06:49 PM.
  Reply With Quote
Old 17th March 2013, 08:59 PM   #32
dmills is offline dmills  United Kingdom
diyAudio Member
 
Join Date: Aug 2008
Location: High Wycombe
Ahh, kind of.....

The word 'resolution' is kind of meaningless in a correctly dithered quantizer, as it has a noise floor, but is linear and does not abruptly go silent at 1LSB (Thanks to the dither). Any quantizer that is not correctly dithered is just plain broken.

As you turn the input down, the output gets closer to the noise floor, just the same as happens in an analogue system, the word length sets the dynamic range, but a narrow tone can be heard down well into the 20-20K noise floor if the conversion is done right.

Dither (Like the necessity of reconstruction filtering) is generally poorly understood among audio types and silly diagrams of stairstep waveforms give rise to many myths.

Here is an undithered piano at 3 bits, sounds like **** does it not?
http://media.soundonsound.com/sos/fe...io/piano_3.mp3
Note that the thing is silent when there is no signal, but sounds NASTY and that the decay on the notes mutes abruptly as the signal drops below the smallest quantisation threshold. This is the only one to which a notion of 'resolution' can be applied, because it is funamentally broken (No dither).

Here is the same thing with TPD dither applied before quantisation (still at 3 bits).
http://media.soundonsound.com/sos/fe...ed_piano_3.mp3
It hisses all the time at roughly 1 LSB, but now the notes decay smoothly down into the noise floor.

Finally, here is the same thing but with some fairly crude noise shaping applied to the dither noise, http://media.soundonsound.com/sos/fe...ed_piano_3.mp3, still hissy, but far better.

Regards, Dan.
  Reply With Quote
Old 17th March 2013, 10:23 PM   #33
gpapag is offline gpapag  Greece
diyAudio Member
 
gpapag's Avatar
 
Join Date: Nov 2002
Location: Athens-Greece
Quote:
Originally Posted by boris81 View Post
I see now, I
Boris, no need to worry. We are all on the same boat. If we knew a lot on DSP, we would build and program our own creations .

Both input and output attenuations work in the digital domain.
Attenuating at the input will lower the resolution on the signal going to all the outputs.
Attenuating at the output will lower the resolution on the signal going to this output only.
But as Dan explained, it is not a very big issue, at least compared to distortion from digital overdriving (clipping or skewing distortion) which we try to avoid by attenuating the signal.

George
__________________
["Second Law is a bitch." - SY] ["The Road To Heaven:Specify the performance & accept the design. The Road To Hell:Specify the design & accept the performance"-Bruno Putzeys]
  Reply With Quote
Old 17th March 2013, 11:43 PM   #34
dmills is offline dmills  United Kingdom
diyAudio Member
 
Join Date: Aug 2008
Location: High Wycombe
Ahh, no it will NOT lower the resolution, at least not if it is done correctly, what it will do is push the signal closer to the noise floor in exactly the same way as attenuating a analogue signal will.

Digital audio is somewhat couterintuitive that way, and notions like resolution borrowed from massively subsampled imaging systems do not apply in any meainingful way.

Digital clipping is of course wince inducingly nasty (Mostly because of aliasing), but with reasonable design and sane headroom margins should never be an issue.

Regards, Dan.
  Reply With Quote
Old 18th March 2013, 02:48 AM   #35
diyAudio Member
 
Join Date: Apr 2005
Location: Melbourne
Quote:
Originally Posted by gpapag View Post
It is still in my User downloads section.

George
Edit: And then he sees the PM.

Last edited by KenTripp; 18th March 2013 at 02:51 AM.
  Reply With Quote
Old 18th March 2013, 10:21 AM   #36
gpapag is offline gpapag  Greece
diyAudio Member
 
gpapag's Avatar
 
Join Date: Nov 2002
Location: Athens-Greece
Quote:
Originally Posted by dmills View Post
Ahh, no
Regards, Dan.
I agree Dan. There is a lot of posting regarding attenuation in analogue vs digital domain in this site. No need to do it again in this thread.

Michael has done a good job on gain structure here:
What is Gain Structure?

George
__________________
["Second Law is a bitch." - SY] ["The Road To Heaven:Specify the performance & accept the design. The Road To Hell:Specify the design & accept the performance"-Bruno Putzeys]
  Reply With Quote
Old 18th March 2013, 08:46 PM   #37
boris81 is offline boris81  United States
diyAudio Member
 
Join Date: Oct 2009
Location: white plains, ny
Thank you Dan and George!
I'm sorry again for posting my wrong assumptions about the system resolution and taking this off-topic. Your explanations are helping me get a much better understanding and hopefully will others reading this as well.
  Reply With Quote
Old 18th March 2013, 09:06 PM   #38
diyAudio Member
 
Join Date: Sep 2009
Location: Australia
Quote:
Originally Posted by minidsp View Post
Hoping this info helps
Thanks Devteam and George!!
  Reply With Quote
Old 2nd April 2013, 11:47 AM   #39
gpapag is offline gpapag  Greece
diyAudio Member
 
gpapag's Avatar
 
Join Date: Nov 2002
Location: Athens-Greece
After dealing with the gain through miniDSP x-over board with the beta version of the 2x and 4x plug-in, I had to look at the gain of the miniAmp
miniAMP | miniDSP

This board is very handy during the experimentation phase.

As with the x-overs I had to check the gain of the miniAmp for to see the signal level up to overload.
Jumpers allow for –3db, +3db, +9db and +12db gain setting. They refer to I2S input signal level, so I had to find out what these numbers mean.

Feeding the input of the stacked miniDSP x-over with the 1281Hz sinusoid as before and having the settings flat, I was monitoring the analog output (10kOhm load) and the speaker output of the miniAmp (8Ohm load).

Below is the max levels as shown by the output RMS meters of the miniDSP x-over for max undistorted output of the miniAmp.

George
Attached Images
File Type: jpg miniAmp overload signal level.JPG (71.0 KB, 165 views)
__________________
["Second Law is a bitch." - SY] ["The Road To Heaven:Specify the performance & accept the design. The Road To Hell:Specify the design & accept the performance"-Bruno Putzeys]
  Reply With Quote
Old 2nd April 2013, 01:54 PM   #40
diyAudio Member
 
vacuphile's Avatar
 
Join Date: Apr 2011
Location: Seaside
That is spot on
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
signal level sakshi Analog Line Level 2 31st December 2012 11:40 AM
Signal levels shown in MiniDSP software look too low - is it working? DrNick miniDSP 2 10th December 2012 08:12 AM
Signal level 39tomcat Analog Line Level 2 23rd March 2012 03:43 PM
How do I turn an high current signal into a line level signal?? sardonx Tubes / Valves 8 27th August 2005 11:16 AM
how to bring down line level signal to preamplifier level deji Solid State 15 15th April 2004 04:13 PM


New To Site? Need Help?

All times are GMT. The time now is 07:04 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2