iFi Audio Pro iDSD discussion thread
Feb 14, 2015 at 12:03 AM Post #391 of 2,788

Franatic

100+ Head-Fier
Joined
Jun 25, 2014
Posts
322
Likes
79
   
Wow!  Thank you. This is exactly the nerdy stuff that most of us like.  Keep 'em coming...and good luck at the Bristol show!
atsmile.gif


+ 1 Great job prodding ifi into giving us info, EVOLVIST!
beerchug.gif

 
I love my idsd micro. My only issue is that ifi doesn't make an idsd without the headphone amp. Many of us are using it in main systems direct. I would love to see an idsd pro version without the headphone amp. That could either save us money or get us additional dac features.
 
Keep up the great work ifi....loving my idsd micro....as a dac only. My Woo WA2 is my headphone amp and preamp.
 
Feb 14, 2015 at 7:02 AM Post #392 of 2,788

BillsonChang007

Headphoneus Supremus
Joined
Jul 21, 2012
Posts
5,956
Likes
434
Location
Johor, Malaysia
   
Hi,
 
As promised....please bear in mind that we can only discuss so much as we have to walk a fine line between informing our customers and not giving away our secret recipes.
wink_face.gif

 
Kind regards
 
iFi Skunkworks.
 
 
 
 
To filter actively or to filter passively, this is the question (part 1)

Some interesting questions have been raised around the subject of the filter circuitry for the iDSD Pro and the amplification stages. It is a complex subject.
 
So please bear with us as we do our best to inform you on the course we have taken and the reasons behind.
 
There is a whole lotta tech speak to crunch through - and it may seem heavy going. So if you suffer from insomnia, you may wish to bookmark this.
 
Without further ado…
 


Why filter the signal at all?

First, we must understand that a DAC-Chip will produce substantial supra-sonic output (which we justly may call “digital distortion”) in addition to the Audio signal we want. Just how this looks like depends on many factors. Rather than looking all options, let's focus on the BB DSD chip we use in iFi products.

It has a core that runs normally at a speed of around 11.3MHz to 12.3MHz. This is a very high speed. This is the speed at which the elements in the core switch. As the process is switching, it creates higher frequency components reaching much higher than main switching frequency.

This switching is one of the processes that produces this noise outside the audio band, the others are related to sample theory and produces what is often called “images”. Simply said, the actual audio signal is “mirrored” around the sample rate in a frequency plot, hence the name mirror images.
 

 
Source: http://defenseelectronicsmag.com/site-files/defenseelectronicsmag.com/files/archive/rfdesign.com/images/digitalpll-Figure03.jpg

Most audio circuitry cannot handle such supra-sonic signals, they are too fast and cause distortion to our audio signal. So what we require with any DAC is a filter that so to speak “strains out” the unwanted noise and leaves the audio signal. And ideally it gets all the noise and leaves the audio totally untouched.

In the real world ideal filters do not exist. A filter that removes the supra-sonic noise will have impact on the audio range and will introduce either phase-response variation, transient-response variation or both.
 
 

 
Source: http://www.electronics-tutorials.ws/filter/fil82.gif

We have to sail between the Scylla of insufficient filter selectivity (too much noise gets through) and the Charybdis of excessive filter impact on the audio (we filter out most noise and pad a lot of distortion the music). But filter to some degree we must, so we must chart a course that will hopefully take through instead being wrecked on rocks of Charybdis or sucked into the chaotic whirlpool of the Scylla.
 

 
Next time: Part 2: How to filter Digital Audio

and then after reading this...
I have a sudden urge of working in iFi's technical team after going through all the neccesary studies and degrees and stuff xD
 
Feb 16, 2015 at 10:21 PM Post #394 of 2,788

ClieOS

IEM Reviewer Extraordinaire
Joined
May 11, 2004
Posts
20,063
Likes
9,490
Location
Mid Johor, Malaysia
If this is "mini" version of iDSD, what are we getting in the end???

iDSD Regular would be some kind of big stereo system then?

 
It is called the 'Pro' series now. No more 'mini' and certainly no 'regular'. If they start making anything bigger and more expensive, it might as well be part of AMR's line-up instead.
 
Feb 17, 2015 at 6:34 AM Post #397 of 2,788
Joined
Mar 20, 2013
Posts
9,350
Likes
5,450
Location
Location: HQ in the UK, but spread far and wide.
To filter actively or to filter passively, this is the question (part 2)
 
 
How to filter digital audio
If we exclude digital filters (which are a whole other can of worms), to create filters, we require electronic parts that change with frequency in predictable ways.
 
 
1. Simple Filter - goes so far
 
The simplest filters combine just a single resistor and either an inductor or a capacitor. The problem is that such simple filters are not very selective. They impact the signal little, but also do not do much for the noise.
 

Source: http://upload.wikimedia.org/wikipedia/commons/thumb/e/e0/1st_Order_Lowpass_Filter_RC.svg/2000px-1st_Order_Lowpass_Filter_RC.svg.png

(see also: http://en.wikipedia.org/wiki/Electronic_filter#Passive_filters)

We normally need better selectivity than we can get that way.
 
 
2. Passive Filter - good in theory, not easy in practice
 
One way is to use combinations of multiple resistors with inductors and capacitors. This is called a passive filter.
 
With real capacitors and inductors such a filter can remain effective to as high as 100MHz. So it can very effectively filter not just the images but also the switching noise.
 

Source: http://upload.wikimedia.org/wikipedia/commons/thumb/0/0e/Lumped_elements_ladder_filter_order_4.svg/450px-Lumped_elements_ladder_filter_order_4.svg.png

Quality inductors take space and cost money and it is often difficult to find inductors that do not cause substantial distortion themselves. This is why this method is rarely seen today, even though it approaches an ideal filter quite closely and of course was used in early generation digital equipment.
 
 
3. Active Filter - another way to skin a cat
 
The other way is to use an active filter. Here we only use resistors and capacitors and “simulate” inductors using amplification. This means we rely on the amplification function of the active element to shape the filtering.
 
- one key advantage other than getting rid of inductors is that active filters allow the designer much, much freedom in trade-offs between selectivity and filter impact on the signal.
- another advantage is that these filters are widely documented and are easily calculated and modelled.

Source: http://www.sussex.ac.uk/Users/pjly20/ras100_clip_image002_0000.jpg

(see also: http://en.wikipedia.org/wiki/Active_filter )

There is a downside. Most of the common amplifier elements loose amplification as frequencies rise. Many audio Op-Amps will have no remaining gain at frequencies above a few MHz. So at very high frequencies the filter no longer filters as it should.

So such active filters are easier and cheaper to implement than adequate passive filters and can filter the so-called images well. But they struggle to filter the switching noise from the DAC as well, or indeed often at all.
 
 
Decisions decisions, what is one to do?
 
In the end it is the choice of the designer how to make sure all the different requirements are met and how trade-offs between mutually exclusive requirements are arranged.

In the iDSD nano we rely only on passive filtering, but the filter selectivity is not as extreme as with other iFi DACs.
 
In the iDSD micro, iDAC2 micro and the DAC in the Retro Stereo 50 we use a combination of passive and active filtering, using very wide bandwidth Op-Amps.
 
In the iDSD Pro, budget/size constraints are less of an issue and thus we can implement a fully-passive filter.
 
Think of the nano iDSD as BMW M1, micro iDSD/iDAC2 as M3 and Pro iDSD as the M5.
 
Each has its own price and engine/performance characteristics to suit different users.
 
(As a small side note, as we develop our code in-house, we "remap" the onboard XMOS (among other things) to push them beyond the theoretical read: datasheet specs. As is evident from the  Quad-DSD256 and Octa-DSD512 on the iFi platform. So our M-line is beyond the factory version.)
 
Before you ask, for those car buffs among you, the BMW i8 is the reserve of AMR.
 
iFi audio Stay updated on iFi audio at their sponsor profile on Head-Fi.
 
https://www.facebook.com/iFiAudio/ https://twitter.com/ifiaudio https://www.instagram.com/ifiaudio/ https://ifi-audio.com/ info@ifi-audio.com
Feb 17, 2015 at 9:15 AM Post #398 of 2,788

EVOLVIST

500+ Head-Fier
Joined
Apr 1, 2014
Posts
968
Likes
180
 
3. Active Filter - another way to skin a cat

There is a downside. Most of the common amplifier elements loose amplification as frequencies rise. Many audio Op-Amps will have no remaining gain at frequencies above a few MHz. So at very high frequencies the filter no longer filters as it should.

So such active filters are easier and cheaper to implement than adequate passive filters and can filter the so-called images well. But they struggle to filter the switching noise from the DAC as well, or indeed often at all.
 
 
In the iDSD Pro, budget/size constraints are less of an issue and thus we can implement a fully-passive filter.
 
 
 
(As a small side note, as we develop our code in-house, we "remap" the onboard XMOS (among other things) to push them beyond the theoretical read: datasheet specs. As is evident from the  Quad-DSD256 and Octa-DSD512 on the iFi platform. So our M-line is beyond the factory version.)
 
 

 
The "secret sauce" is all over the place in here, especially where I've bolded and/or underlined. How to implement a fully-passive filter...well, you've said it's up to the engineers, which leads us to your "As a small side note," quote, which really isn't a small side note at all, reading between the lines. Indeed, your closing paragraph, in parenthesis, no less, isn't really parenthetical, either, as your "small side note" would lead us to believe! Haha! 
 
I don't think I'm reading too much into it, this time. You're equating the leap in implementing Quad and Octa-DSD, the iFi way, with new technological leaps* in how you're now implementing fully-passive filtering, with no OPA, etc. etc... I mean, this is how a layman reads it.
 
Keep 'em coming! Got any Bristol photos?
 
*Or, rather, that iFi is wont to borrow from past technologies which worked very well, why reinvent the wheel? Instead, borrow from the old masters, but implement in such a way that technologies - old and new - dovetail in such a way that something wholly unique is created.
 
Feb 17, 2015 at 9:24 AM Post #399 of 2,788
Joined
Mar 20, 2013
Posts
9,350
Likes
5,450
Location
Location: HQ in the UK, but spread far and wide.
   
The "secret sauce" is all over the place in here, especially where I've bolded and/or underlined. How to implement a fully-passive filter...well, you've said it's up to the engineers, which leads us to your "As a small side note," quote, which really isn't a small side note at all, reading between the lines. Indeed, your closing paragraph, in parenthesis, no less, isn't really parenthetical, either, as your "small side note" would lead us to believe! Haha! 
 
I don't think I'm reading too much into it, this time. You're equating the leap in implementing Quad and Octa-DSD, the iFi way, with new technological leaps in how you're now implementing fully-passive filtering, with no OPA, etc. etc... I mean, this is how a layman reads it.
 
Keep 'em coming! Got any Bristol photos?

 
Hi,
 
The above is not the secret sauce....we tell you how we do it but if someone really wants to do implement this, it isnt that easy.
 
The source links are to wiki et al!
 
If you read the micro iDSD Crowd-Design thread index, you get the gist of how much we divulge...
popcorn.gif

 
iFi audio Stay updated on iFi audio at their sponsor profile on Head-Fi.
 
https://www.facebook.com/iFiAudio/ https://twitter.com/ifiaudio https://www.instagram.com/ifiaudio/ https://ifi-audio.com/ info@ifi-audio.com
Feb 17, 2015 at 9:35 AM Post #400 of 2,788

EVOLVIST

500+ Head-Fier
Joined
Apr 1, 2014
Posts
968
Likes
180
Hi,

The above is not the secret sauce....we tell you how we do it but if someone really wants to do implement this, it isnt that easy.

The source links are to wiki et al!

If you read the micro iDSD Crowd-Design thread index, you get the gist of how much we divulge...:popcorn:


No, no...I didn't mean to imply that you've spilt any secret sauce. The sauce is behind the scenes. This is just the whiff of the sauce in the air, which lets us know it's there. :wink:

Yeah, believe me: I've read the crowd design articles many times. I love the insight. But even if you were to tell me exactly how it's all done, I wouldn't want to hear it. One still marvels at the magician, even though we know there's scientific law behind it all.
 
Feb 17, 2015 at 6:55 PM Post #401 of 2,788

Rich Brkich

New Head-Fier
Joined
Mar 24, 2007
Posts
3
Likes
1
So, with the iDSD Pro/Mini I see mention of balanced analog outputs thanks to 4 DAC (dual differential) topology which will be used. For those folks who are going to want to hook this up to a nice single ended (i.e RCA) preamps/amps gear based HiFi rig (or headphone amp), will there also be unbalanced analog outputs???? How will they be generated so that they get to take advantage of the balanced DAC/analog stage topology???
 
P.S. Sorry if this has been discussed before - I looked through most of this  27 page thread and did not see anything about this, but could have missed it.
 
Feb 18, 2015 at 12:08 AM Post #402 of 2,788

sdolezalek

New Head-Fier
Joined
Jan 9, 2015
Posts
40
Likes
16
I'm not sure I quite follow the BMW analogy.  The BMW M1 is a classic that now typically sells for half a million dollars or so (which is at the very other end of the spectrum from how I think of the Nano).  I get the M3 and M5  comparison, but I was really thinking that the whole iDSD series was like the I1, I3 and I8 (the new experimental edgy products) and that the AMR versions were really the M3, M5 and M7? versions -- souped up but thoroughly tested flagship products...
 
In fact, I figured my I-8 and my iDSD Pro might arrive on my doorstep at about the same time late this spring.  Either way, the Pro still sounds like very much a product worth waiting for.  
 
Feb 18, 2015 at 2:06 AM Post #403 of 2,788

JuleZ3C

Head-Fier
Joined
Jun 20, 2014
Posts
58
Likes
29

Users who are viewing this thread

Top