• The Vendor's Bazaar forum is for commercial offers and transactions. Only unmoderated members can post here.

    diyAudio provides this forum for the convenience of our members, but makes no warranty nor assumes any responsibility. We do not vet any members. Use of this facility is at your own risk. Customers can post any issues in those threads as long as it is done in a civil manner. All diyAudio rules about conduct apply and will be enforced.

Reference DAC Module - Discrete R-2R Sign Magnitude 24 bit 384 KHz

Jump to Latest
#4,821
pos said:
Nice, I am honored! 😀
Did you try the recently introduced Albrecht cosine windows? (rephase 1.0.0 from last June)
Click to expand...

No, most of my filters were done before that.... In short, what's the advantage of the Albrecht Cosine windows ? (I don't really know much about digital filters....)

Any idea which windows type gives the best sound ? (I know best is hard, more like most neutral and musical sound....)
 
#4,824
soekris said:
No, most of my filters were done before that.... In short, what's the advantage of the Albrecht Cosine windows ? (I don't really know much about digital filters....)

Any idea which windows type gives the best sound ? (I know best is hard, more like most neutral and musical sound....)
Click to expand...

Selecting the window algorithm let you choose between shape/slope accuracy and noise/ripples floor.
Typically for a given number of taps a rectangular or very "sharp" window will give a resulting shape that more closely track the target one, but it will reach the noise/ripple floor quicker.

The advantage of the Albrecht windowing algorithm is that it let you accurately choose that noise/ripple floor (calculated for a brickwall filter), and possibly down to very, very low values.
 
Last edited:
#4,828
I don't think Albrecht Cosine windows are very much used for FIR windowing, but I found the paper and tried them and... they work like a charm 🙂
But of course other windowing algorithms can also be used.
The main advantage of these windows is the flexibility, and the possibility to reach very low ripples floor when enough taps are provided (the price being a shallower shape/slope when the number of taps is not large enough, there is no free lunch).

TNT, I don't think there is anything "magic" in all these antialiasing filters, and all the aspects can be checked and judged pragmatically.
 
#4,829
TNT said:
Yes - but not up to standard as e.g. Paul so You have not been able to exploit the full capacity of your exquisite HW. There is some evidence for this as per the uptake and use of other filters then the supplied ones.

//
Click to expand...

I would disagree on that.... The current standard filter set have different requirements, they have to be neutral and meet certain specs on t.ex. anti aliasing and hf noise.

Other people can do filters based on personal preference and taste, that's why I did make it possible....
 
#4,831
Søren,

Thanks so much for your continued hard work and an excellent DAC product range.

Would you perhaps consider making
* suitable cases/enclosures available?
* a easy way to duplicate/split the I2S signals when using 2 boards in balanced mode?
* A XMOS based USB-2-I2S board + SE/balanced mode (see you are using XMOS in dac1101)?

Can two i2S boards be used to drive each channel (balanced) by duplicating the USB D-,D+ signal? Downside?
 
#4,832
I can make an effort yes... but I call them as I see them.

But I'm puzzled of the marketing strategy. He started by saying that he will produce the best sounding DAC in the world. He never came back on telling us if *he* found that he had indeed done that - or any other evidence. Its such a bold statement that I cant just be ignored - but maybe you SSassen eat that kind of sh*t without being bothered.

Then all of sudden he declares that he cant master one very important aspect of the design... again for you this might not be problem because you might be gullible and not so careful what you buy - but thats not me. So you can call A-Ho as much as you like but I don't care as I have listened to the differences and know what it does. But the designer hasn't even listened to the "competing" filters, not even the versions which indeed fulfills hes sound design criteria.

You sir, seem to indeed accept things without examination...


//
 
#4,833
TNT said:
I can make an effort yes... but I call them as I see them.

But I'm puzzled of the marketing strategy. He started by saying that he will produce the best sounding DAC in the world. He never came back on telling us if *he* found that he had indeed done that - or any other evidence. Its such a bold statement that I cant just be ignored - but maybe you SSassen eat that kind of sh*t without being bothered.

Then all of sudden he declares that he cant master one very important aspect of the design... again for you this might not be problem because you might be gullible and not so careful what you buy - but thats not me. So you can call A-Ho as much as you like but I don't care as I have listened to the differences and know what it does. But the designer hasn't even listened to the "competing" filters, not even the versions which indeed fulfills hes sound design criteria.

You sir, seem to indeed accept things without examination...
//
Click to expand...

That's funny, I actually bought two Soekris DACs after doing my due diligence on the examination part.

One thing I find very comforting to know is that (unlike you it seems) Søren is motivated by the technical challenges and approaches this from an engineering point of view where raw numbers, simulation, measurements etc. determine performance and the direction in which development needs to go. For filters that simply means that passband ripple, phase-lead/lag or constant phase are determining factors, not how it subjectively sounds.

Being an EE myself I can only second that, nothing is gained by taking the MyFi approach of non-ABX listening tests, opamp rolling, cap swapping and more of that audio-fool nonsense.

As for Søren's statements, in my opinion he can post reminders about the uniqueness of this DAC in this topic every day in red, 64pt bold fonttype of his choosing, as he's the only one offering such a excellent performing and well rounded R2R design to the DIY market at this point.

Measurements that he has posted previously show that this DAC indeed can hold its own against the best sigma-delta has to offer, but without the nastiness that comes with the DS principle, not a small feat to accomplish I can tell you.

But judging from your comments you have no engineering background, hence the above will simply be lost on you, you do not fully realize what amount of work is involved here, or the technical challenges Søren has very gracefully overcome, as you lack the hands-on experience with designing something like this.

With that fact being established it also disqualifies you from making any sort of comments about Søren's design choices or claims he's making.
 
#4,834
SSassen said:
For filters that simply means that passband ripple, phase-lead/lag or constant phase are determining factors, not how it subjectively sounds.
Click to expand...
:yes:
And that is the way 99% of the DACs out there are made, for good reasons.
For example minimum-phase antialiasing filters are not there to sound different, but to lower latency.

That is the way "stock" filters should be implemented. And then Søren also let the user enter his own filters (contrary to 99.9% to DACs out there), so the subjective part can be dealt with there, where it belongs.
 
#4,835
Hi SSassen,

I am not a technical man. Buy I totally agree with you.
Soren already produced a very good cost/performance R2R DAC for all DIYer.

Soren discuss technical issue in the forum and try to improve the board perform better. At this cost, we cannot find such good performance R2R DAC.
Although I need mod my ver 1 board to improve performance.

I enjoy it because I am a DIYer.

SSassen said:
That's funny, I actually bought two Soekris DACs after doing my due diligence on the examination part.

One thing I find very comforting to know is that (unlike you it seems) Søren is motivated by the technical challenges and approaches this from an engineering point of view mwhere raw numbers, simulation, measurements etc. determine performance and the direction in which development needs to go. For filters that simply means that passband ripple, phase-lead/lag or constant phase are determining factors, not how it subjectively sounds.

Being an EE myself I can only second that, nothing is gained by taking the MyFi approach of non-ABX listening tests, opamp rolling, cap swapping and more of that audio-fool nonsense.

As for Søren's statements, in my opinion he can post reminders about the uniqueness of this DAC in this topic every day in red, 64pt bold fonttype of his choosing, as he's the only one offering such a excellent performing and well rounded R2R design to the DIY market at this point.

Measurements that he has posted previously show that this DAC indeed can hold its own against the best sigma-delta has to offer, but without the nastiness that comes with the DS principle, not a small feat to accomplish I can tell you.

But judging from your comments you have no engineering background, hence the above will simply be lost on you, you do not fully realize what amount of work is involved here, or the technical challenges Søren has very gracefully overcome, as you lack the hands-on experience with designing something like this.

With that fact being established it also disqualifies you from making any sort of comments about Søren's design choices or claims he's making.
Click to expand...
 
#4,840
Søren:
From v.99 firmware it is quoted:
* FIR1 filters doubled in size, so max 2032 coeff @ 44K/48K, 1016 @ 88/96K, 508 @ 176/192K, 127 @ 352/384K
* You can now select four sets of filters, named Linear, Mixed, Minimum and Soft.
* FIR1 coefficients are now 1.31 format, filters should be good down to around -150 db[/QUOTE]

Are these the same for V1.05 firmware.
BTW: I notice you have a v1.05e uc (1021_uc_105e.skr) firmware also. What is the difference between these 1.05 and 1.05e and 1.06. ?

Thanks in advance.
 
You must log in or register to reply here.
Top Bottom