[xLoud]

I am going for Silver version, not the Black Label Edition. I am not a regular traveller, I travel may be once in a month and a backpack always accompany me. So, portability is never going to be much of an issue at all.

 

[iFi audio]

 
Anyone knows why the Standard filter isn't recommended for listening in the manual, but for measurement instead?

 
Sound quality is subjective. And on top of that, perception of reproduced sound quality is also a learned response. So one may initially prefer a result that matches the learned expectation of "good sound" as opposed to the more real or natural sound.
 
We do not recommend the standard filter, because in our experience it sound less natural, less like real music and more artificial. But we also recognise not everyone agrees. Hence the inclusion of the filter switch, to give the choice to an owner, rather than to force our preferred choice.
 
To discuss the reasoning behind this, any filter is a trade-off between impulse response and frequency response. Specifically for CD and DSD64 sample rate... the problems are severe. CD should use at least an 11th order low pass filter at 20kHz with 0dB attenuation at 20kHz. And SACD/DSD theoretically needs a 7th order filter at 20kHz. Those are very severe slope filters right at the upper edge of the audible range. Neither filter is an easy proposition as analogue filter, so digital (FIR) filters are often employed.
 
Further filters come in multiple "flavours". Namely we have infinite impulse response - IIR (ONLY analogue filters offer true IIR) and finite impulse response - FIR (only digital filters have this). Further, digital FIR filters may be designed with either symmetrical impulse response (the filter has equal pre- and post-ringing when presented with an impulse) or asymmetrical impulse response (the filter has more or less pre-ringing that post-ringing).
 

 
This ringing "must" be added if such steep slope filters are employed. Even analogue filters only will show ringing. Now let us be 100% crystal clear. Adding ringing to an impulse is adding distortion. And a LOT of distortion. Most traditional digital filters add >10% distortion for a significant time after an impulse event.
 
At AMR we produced a short paper focusing on filter options in the DP-777 DAC, however a read of this may also be beneficial here:
 

1. http://www.amr-audio.co.uk/large_image/Tech%20Paper%203%20-%20Ringing.pdf

 
Different manufacturers have different views and promote different solutions. Again, we have no desire to comment on specifics. Our preferred view however is that less added distortion is better in all domains, amplitude, frequency and time domain, if we desire a realistic, natural sound.
 
Previously Head-Fi'er Earphonia measured the square wave response of the iDSD micro and posted it here. Comparing the response of the different filters is likely instructive:
 

1. http://www.head-fi.org/t/728236/ifi-idsd-micro-dsd512-pcm768-dac-and-headphone-amp-impressions-reviews-and-comments/3045#post_11387805

 

 
 
Stereophile USA often posts Impulse responses of DAC's. As example we shall refer to the AMR DP-777, as it offers many different filters. It showcases most possible filter responses:
 

1. http://www.stereophile.com/content/abbingdon-music-research-dp-777-da-processor-measurements

 
Many other DAC's can be found and compared.
 
The audibility of these filter differences is a fairly hotly debated topic, something that often starts flame wars - which we have no desire to get involved with. But we hope that the above helps to understand a little better why there are choices and what they are.
 
In the end the proof is in the listening and one should always use the filter setting one finds most enjoyable. If this is the "Standard / Measure" filter, well all the more power to you.

 

[hemipowered007]

Sound quality is subjective. And on top of that, perception of reproduced sound quality is also a learned response. So one may initially prefer a result that matches the learned expectation of "good sound" as opposed to the more real or natural sound.

We do not recommend the standard filter, because in our experience it sound less natural, less like real music and more artificial. But we also recognise not everyone agrees. Hence the inclusion of the filter switch, to give the choice to an owner, rather than to force our preferred choice.

To discuss the reasoning behind this, any filter is a trade-off between impulse response and frequency response. Specifically for CD and DSD64 sample rate... the problems are severe. CD should use at least an 11th order low pass filter at 20kHz with 0dB attenuation at 20kHz. And SACD/DSD theoretically needs a 7th order filter at 20kHz. Those are very severe slope filters right at the upper edge of the audible range. Neither filter is an easy proposition as analogue filter, so digital (FIR) filters are often employed.

Further filters come in multiple "flavours". Namely we have infinite impulse response - IIR (ONLY analogue filters offer true IIR) and finite impulse response - FIR (only digital filters have this). Further, digital FIR filters may be designed with either symmetrical impulse response (the filter has equal pre- and post-ringing when presented with an impulse) or asymmetrical impulse response (the filter has more or less pre-ringing that post-ringing).




This ringing "must" be added if such steep slope filters are employed. Even analogue filters only will show ringing. Now let us be 100% crystal clear. Adding ringing to an impulse is adding distortion. And a LOT of distortion. Most traditional digital filters add >10% distortion for a significant time after an impulse event.

At AMR we produced a short paper focusing on filter options in the DP-777 DAC, however a read of this may also be beneficial here:

• http://www.amr-audio.co.uk/large_image/Tech%20Paper%203%20-%20Ringing.pdf

Different manufacturers have different views and promote different solutions. Again, we have no desire to comment on specifics. Our preferred view however is that less added distortion is better in all domains, amplitude, frequency and time domain, if we desire a realistic, natural sound.

Previously Head-Fi'er Earphonia measured the square wave response of the iDSD micro and posted it here. Comparing the response of the different filters is likely instructive:

• http://www.head-fi.org/t/728236/ifi-idsd-micro-dsd512-pcm768-dac-and-headphone-amp-impressions-reviews-and-comments/3045#post_11387805

Stereophile USA often posts Impulse responses of DAC's. As example we shall refer to the AMR DP-777, as it offers many different filters. It showcases most possible filter responses:

• http://www.stereophile.com/content/abbingdon-music-research-dp-777-da-processor-measurements

Many other DAC's can be found and compared.

The audibility of these filter differences is a fairly hotly debated topic, something that often starts flame wars - which we have no desire to get involved with. But we hope that the above helps to understand a little better why there are choices and what they are.

In the end the proof is in the listening and one should always use the filter setting one finds most enjoyable. If this is the "Standard / Measure" filter, well all the more power to you.

This is why I am now and will be an ifi fanboy as long as you're making products. Thank you so much for always responding to questions here, it's great to hear from the creators of how and why certain functions were implemented. Also, thank you very very much for creating the black label, obviously you have a lot of major fans of it. Great job ifi!

 

[dobigstuff]

 
Thanks for sharing your experience. Indeed, things do sound more accurate in Bit-perfect mode. But oddly, my headphones had the reverse effect with yours, with Standard sounding bigger, bolder and a bit more up-close than BP. I wonder if it all comes down to pairing synergy.  

Topspin70:
Just my input:  I had been using the iFi iDSD Black as a Dac.  Back on 3/3/17 for the first time I hooked it to my computer (nothing special HP Envy).  I downloaded the Driver and connected the iDSD Black.
I use jRiver.  The Blue indicator input Light showing on the iDSD Black confirmed DSD 256 11.3 MHz.   
 
With the DSD setup, I could actually hear the differences in the Digital Filters.  (When I use Digital coax out from my Marantz NA 7004 into the iDSD Black there is no significant audible difference in the filters). But listening to DSD here is what I perceive:  Bit Perfect seems warmest nothing is too harsh or bright (Can listen all day).  Minimum Phase adds punch, smooth Mids and Highs and a bit wider soundstage.  Standard adds a little bit more punch, smooth Mids as well Tight Bass clean highs.
(Headphones:  HD 700 w/Black Dragon v2 Cable)

 

[Dobrescu George]

  Right now I have Fiio X3 2nd Gen(For outdoor use), iBasso D14(Used as a DAC with PC at home) paired with Schiit Magni 2(AMP). I was looking to upgrade the DAC, then decided to replace all of them iFi iDSD Micro. iDSD is powerful enough to use it with my HifiMan HE-400i, and I can use it as a portable dac/amp with my mobile. 
Am I wrong to replace all of them with iFi iDSD?

 
Well, I'm pretty sure that iDSD is a great device! You can still use X3-2 as your transport so you can get best of both worlds! 

 
Quote:

  I have the HE-400I  and Black label can drive it with ease on Eco or Normal power setting.  IMO a little big for portable rig unless you use in a back pack .I don't have that big of pants pockets. Great device that I use on my computer at home and can drive anything you throw at it with ease and built like a tank. I was lucky enough to find on sale new for $399 which made even a better deal. Can't go wrong as it is the Swiss army knife of DAC/Amp combo that happens to be transportable. I love the bass boost switch and I am not a bass head.  
 

 
I don't think anybody in this worlds has pants large enough to be able to use iFi iDSD BL in their pockets. A side bag will do, even a small one that's meant for accessories or a camera bag should do well 

 

[Dobrescu George]

  Today I played around with the filter settings of iDSD micro, monitored by simple USB oscilloscope (Velleman PCSU200). So line output of the micro iDSD goes direct to USB oscilloscope input (1 Mohm input impedance).
 
Line Output setting: Direct
Power Mode: Normal
Polarity: +
RCA Cable: Wireworld Luna 7 (50 cm)
Adapter: Monster RCA to BNC
 
 
Here are the screenshots if you guys interested, showing only the left channel:
 
 
12 kHz Sine wave - PCM bit depth and sampling rate: 24 bit - 48 kHz
 
Bit-Perfect

 
Minimum Phase

 
Standard

 
 
 
12 kHz Sine wave - PCM bit depth and sampling rate: 24 bit - 192 kHz
 
Bit-Perfect

 
Minimum Phase

 
Standard

 
 
 
12 kHz Square wave - PCM bit depth and sampling rate: 24 bit - 192 kHz
 
Bit-Perfect

 
Minimum Phase

 
Standard

 
 
 
12 kHz Sawtooth wave - PCM bit depth and sampling rate: 24 bit - 192 kHz
 
Bit-Perfect

 
Minimum Phase

 
Standard

 
Thank you very much for providing those measurements! 

The DAC Filters now make a lot of sense 

 

[Dobrescu George]

   
Sound quality is subjective. And on top of that, perception of reproduced sound quality is also a learned response. So one may initially prefer a result that matches the learned expectation of "good sound" as opposed to the more real or natural sound.
 
We do not recommend the standard filter, because in our experience it sound less natural, less like real music and more artificial. But we also recognise not everyone agrees. Hence the inclusion of the filter switch, to give the choice to an owner, rather than to force our preferred choice.
 
To discuss the reasoning behind this, any filter is a trade-off between impulse response and frequency response. Specifically for CD and DSD64 sample rate... the problems are severe. CD should use at least an 11th order low pass filter at 20kHz with 0dB attenuation at 20kHz. And SACD/DSD theoretically needs a 7th order filter at 20kHz. Those are very severe slope filters right at the upper edge of the audible range. Neither filter is an easy proposition as analogue filter, so digital (FIR) filters are often employed.
 
Further filters come in multiple "flavours". Namely we have infinite impulse response - IIR (ONLY analogue filters offer true IIR) and finite impulse response - FIR (only digital filters have this). Further, digital FIR filters may be designed with either symmetrical impulse response (the filter has equal pre- and post-ringing when presented with an impulse) or asymmetrical impulse response (the filter has more or less pre-ringing that post-ringing).
 

 
This ringing "must" be added if such steep slope filters are employed. Even analogue filters only will show ringing. Now let us be 100% crystal clear. Adding ringing to an impulse is adding distortion. And a LOT of distortion. Most traditional digital filters add >10% distortion for a significant time after an impulse event.
 
At AMR we produced a short paper focusing on filter options in the DP-777 DAC, however a read of this may also be beneficial here:
 

1. http://www.amr-audio.co.uk/large_image/Tech%20Paper%203%20-%20Ringing.pdf

 
Different manufacturers have different views and promote different solutions. Again, we have no desire to comment on specifics. Our preferred view however is that less added distortion is better in all domains, amplitude, frequency and time domain, if we desire a realistic, natural sound.
 
Previously Head-Fi'er Earphonia measured the square wave response of the iDSD micro and posted it here. Comparing the response of the different filters is likely instructive:
 

1. http://www.head-fi.org/t/728236/ifi-idsd-micro-dsd512-pcm768-dac-and-headphone-amp-impressions-reviews-and-comments/3045#post_11387805

 

 
 
Stereophile USA often posts Impulse responses of DAC's. As example we shall refer to the AMR DP-777, as it offers many different filters. It showcases most possible filter responses:
 

1. http://www.stereophile.com/content/abbingdon-music-research-dp-777-da-processor-measurements

 
Many other DAC's can be found and compared.
 
The audibility of these filter differences is a fairly hotly debated topic, something that often starts flame wars - which we have no desire to get involved with. But we hope that the above helps to understand a little better why there are choices and what they are.
 
In the end the proof is in the listening and one should always use the filter setting one finds most enjoyable. If this is the "Standard / Measure" filter, well all the more power to you.

 
Thank you very much! 

This is a very well-though answer to a question that's really been on my mind for a lot of time! 
 
Seems that you guys thought things through and I really appreciate this! It is nice to have the option and I feel that giving the option to the user will always be a blessing! 

 

[Topspin70]

   
Sound quality is subjective. And on top of that, perception of reproduced sound quality is also a learned response. So one may initially prefer a result that matches the learned expectation of "good sound" as opposed to the more real or natural sound.
 
We do not recommend the standard filter, because in our experience it sound less natural, less like real music and more artificial. But we also recognise not everyone agrees. Hence the inclusion of the filter switch, to give the choice to an owner, rather than to force our preferred choice.
 
To discuss the reasoning behind this, any filter is a trade-off between impulse response and frequency response. Specifically for CD and DSD64 sample rate... the problems are severe. CD should use at least an 11th order low pass filter at 20kHz with 0dB attenuation at 20kHz. And SACD/DSD theoretically needs a 7th order filter at 20kHz. Those are very severe slope filters right at the upper edge of the audible range. Neither filter is an easy proposition as analogue filter, so digital (FIR) filters are often employed.
 
Further filters come in multiple "flavours". Namely we have infinite impulse response - IIR (ONLY analogue filters offer true IIR) and finite impulse response - FIR (only digital filters have this). Further, digital FIR filters may be designed with either symmetrical impulse response (the filter has equal pre- and post-ringing when presented with an impulse) or asymmetrical impulse response (the filter has more or less pre-ringing that post-ringing).
 

 
This ringing "must" be added if such steep slope filters are employed. Even analogue filters only will show ringing. Now let us be 100% crystal clear. Adding ringing to an impulse is adding distortion. And a LOT of distortion. Most traditional digital filters add >10% distortion for a significant time after an impulse event.
 
At AMR we produced a short paper focusing on filter options in the DP-777 DAC, however a read of this may also be beneficial here:
 

1. http://www.amr-audio.co.uk/large_image/Tech%20Paper%203%20-%20Ringing.pdf

 
Different manufacturers have different views and promote different solutions. Again, we have no desire to comment on specifics. Our preferred view however is that less added distortion is better in all domains, amplitude, frequency and time domain, if we desire a realistic, natural sound.
 
Previously Head-Fi'er Earphonia measured the square wave response of the iDSD micro and posted it here. Comparing the response of the different filters is likely instructive:
 

1. http://www.head-fi.org/t/728236/ifi-idsd-micro-dsd512-pcm768-dac-and-headphone-amp-impressions-reviews-and-comments/3045#post_11387805

 

 
 
Stereophile USA often posts Impulse responses of DAC's. As example we shall refer to the AMR DP-777, as it offers many different filters. It showcases most possible filter responses:
 

1. http://www.stereophile.com/content/abbingdon-music-research-dp-777-da-processor-measurements

 
Many other DAC's can be found and compared.
 
The audibility of these filter differences is a fairly hotly debated topic, something that often starts flame wars - which we have no desire to get involved with. But we hope that the above helps to understand a little better why there are choices and what they are.
 
In the end the proof is in the listening and one should always use the filter setting one finds most enjoyable. If this is the "Standard / Measure" filter, well all the more power to you.

Double thumbs up to you guys as always. Such a detailed explanation, and done with great respect to your customers, acknowledging the diverse tastes and opinions (scarier than taste IMO) that exist out there! For me it's very educational. I can't ask for more. 
 
As for the filters, it appears my learned expectation of good sound is wide enveloping soundstage and tight detailed yet smooth sound, which is what I hear from the Standard filter. So thanks for including that option in the BL. 

 

[Topspin70]

  Topspin70:
Just my input:  I had been using the iFi iDSD Black as a Dac.  Back on 3/3/17 for the first time I hooked it to my computer (nothing special HP Envy).  I downloaded the Driver and connected the iDSD Black.
I use jRiver.  The Blue indicator input Light showing on the iDSD Black confirmed DSD 256 11.3 MHz.   
 
With the DSD setup, I could actually hear the differences in the Digital Filters.  (When I use Digital coax out from my Marantz NA 7004 into the iDSD Black there is no significant audible difference in the filters). But listening to DSD here is what I perceive:  Bit Perfect seems warmest nothing is too harsh or bright (Can listen all day).  Minimum Phase adds punch, smooth Mids and Highs and a bit wider soundstage.  Standard adds a little bit more punch, smooth Mids as well Tight Bass clean highs.
(Headphones:  HD 700 w/Black Dragon v2 Cable)

Oh I see. So our set up and usage are quite different. I'm using the BL as a one-box solution, and playing mainly red book files and tidal (presumably PCM). But your description of the filters during DSD playback is consistent with what I hear from my music, especially "Standard adds a little bit more punch, smooth Mids as well Tight Bass clean highs", which is what I found appealing. @iFi audio detailed post shed more light into the differences. Truly educational. 

 

[earfonia]

   
Thank you very much for providing those measurements! 

The DAC Filters now make a lot of sense 

 
I did more test on the filters here if you're interested:
http://www.head-fi.org/t/833933/ifi-micro-idsd-black-label-in-depth-review#post_13221621

 

[Dobrescu George]

   
I did more test on the filters here if you're interested:
http://www.head-fi.org/t/833933/ifi-micro-idsd-black-label-in-depth-review#post_13221621

 
Woah 

Thank you very much for pointing to this post! 

It would be good to make an entire thread for iDSD BL with those measurements since I was looking for this information! Awesome findings and tests you did there! Now a lot of things make more sense! Also, the test seem to make iDSD BL a great device! 

 

[Topspin70]

While on the subject of the BL functions, I like to share my experience regarding the power mode options.
 
Many reviews and comments talk about their varying ability to drive phones to certain loudness, but not many talk about their effects on quality (unless I missed all those that did). Through my 300ohm HD800S, Eco is more than enough to drive it to enjoyable levels with volume at 1 o'clock. But with Normal, with volume at 11 o'clock to achieve similar loudness, the sound is much more dynamic, notes more defined and textured, bass more pronounced and musical nuances more evident.
 
Overall feeling is that SQ is more detailed and realistic using Normal as compared to Eco, even if it means using less of the volume pot range, which I believe @iFi audio usually encourages owners to use more of it cuz it is designed to perform best that way. My conclusion is consistent with that of my iCAN SE which I set at mid gain. Naturally, how far we go on the volume knob will change with different headphones, but the experience I'm sharing is that quality matters as much as loudness when choosing a power mode to drive your phones. 
 
The album is used for A/B-ing is Laura Nyro Live at Loom's Desire. Try it and just listen to those piano notes, the nuances of her voice as well as the quality of the back up singers' voices. Goosebumps! Have a go and enjoy!

 

[Dobrescu George]

  While on the subject of the BL functions, I like to share my experience regarding the power mode options.
 
Many reviews and comments talk about their varying ability to drive phones to certain loudness, but not many talk about their effects on quality (unless I missed all those that did). Through my 300ohm HD800S, Eco is more than enough to drive it to enjoyable levels with volume at 1 o'clock. But with Normal, with volume at 11 o'clock to achieve similar loudness, the sound is much more dynamic, notes more defined and textured, bass more pronounced and musical nuances more evident.
 
Overall feeling is that SQ is more detailed and realistic using Normal as compared to Eco, even if it means using less of the volume pot range, which I believe @iFi audio usually encourages owners to use more of it cuz it is designed to perform best that way. My conclusion is consistent with that of my iCAN SE which I set at mid gain. Naturally, how far we go on the volume knob will change with different headphones, but the experience I'm sharing is that quality matters as much as loudness when choosing a power mode to drive your phones. 
 
The album is used for A/B-ing is Laura Nyro Live at Loom's Desire. Try it and just listen to those piano notes, the nuances of her voice as well as the quality of the back up singers' voices. Goosebumps! Have a go and enjoy!

 
I noticed this as well actually. The texture and clarity and dynamism / realism is better on normal for me as well 

 
The music I used was a bit more energetic, guitar textures are felt really fast with certain types of music and I gotta say, I favor normal and no iEMatch engaged for best results - the sound is sweet this way 

 
I love iDSD BL + Meze 99 Classics with the 3D setting engaged and even with X-Bass setting engaged for certain songs. Gives the listener a true rush this thing! 

Also asked my brother who is more of a casual music listener to give iDSD a try with his music and his IEMs. He fell in love with BL after one minute and told me that this is one of the most incredible DAC/AMPs he heard to date as well 

 Slowly converting him to the enthusiastic music listener part of things. Any minute now 

 

[Topspin70]

It's hard not to love the BL so I don't think you need to try too hard  @Dobrescu George
 
BTW Anyone here leave the BL on all the time, while connected to the powered USB? I understand from an old ifi post that the unit goes to sleep after 15 mins when there's no signal. Can I presume there's no harm leaving it on all day then?

 

[earfonia]

   
Woah 

Thank you very much for pointing to this post! 

It would be good to make an entire thread for iDSD BL with those measurements since I was looking for this information! Awesome findings and tests you did there! Now a lot of things make more sense! Also, the test seem to make iDSD BL a great device! 

 
You're most welcome! BL is a great DAC! I really like it.
 
 

  While on the subject of the BL functions, I like to share my experience regarding the power mode options.
 
Many reviews and comments talk about their varying ability to drive phones to certain loudness, but not many talk about their effects on quality (unless I missed all those that did). Through my 300ohm HD800S, Eco is more than enough to drive it to enjoyable levels with volume at 1 o'clock. But with Normal, with volume at 11 o'clock to achieve similar loudness, the sound is much more dynamic, notes more defined and textured, bass more pronounced and musical nuances more evident.
 

 
Agree on this. I suspect that at Normal mode, micro iDSD uses higher voltage power supply for the amp, therefore it improves the slew rate of the amp. Just my guess.