[CH23]

Thank you. I will check out the colours. Just need to identify the correct cable type - i think it is female abd male usb A cable. Could easily end up ordering the wrong type by mistake. Did this in the past.

Ideally Apple could atleast have a longer Camera Kit Adapter. Then I will not require a usb cable.

it's a usb 3.0 extension cable. usb A male to usb A female.

@CrocodileDundee consider yourself ignored

 

[Sebastien Chiu]

Thank you. I listened to Diablo with my new Focal Clear at a dealer. Was absolutely a match made in heaven. Incredible synergy. Perfect partner for Clear MG.

Love the pairing of my Diablo and the OG Focal Clear too!

Can someone please help me find a matching red or white USB cable to replace the blue one included in Diablo. One that is made or tested for audio.

i am in the UK.

Hope you find what you are looking for. I've been able to find some pretty clean red USB C to OTG cables.

ps.: My “why” was on a “ranting” intention, not really asking for a reason. I heard some people complaining about that cable color before, including youtubers. I expressed myself in a wrong way, sorry.

PS.: just ignore my last comment… it was supposed to be ironic.

All good, it happens!

 

[iFi audio]

Blue is the colour of USB 3.0, so that's why it's that colour

That is true, yes

Thank you. I listened to Diablo with my new Focal Clear at a dealer. Was absolutely a match made in heaven. Incredible synergy. Perfect partner for Clear MG.

Thanks! Yes, it is a nice combo and we got quite the feedback about it specifically

 

[Soundizer]

Love the pairing of my Diablo and the OG Focal Clear too!


Hope you find what you are looking for. I've been able to find some pretty clean red USB C to OTG cables.



All good, it happens!

I made a very technical observation which some might find interesting. The Diablo red colour really matches well with Apple iPhone latest RED Silicone Case.

If I can just some how colour coat the Apple Camera Kit cable aswell it would be a heavenly cosmetic synergy.

 

[iFi audio]

The Diablo red colour really matches well with Apple iPhone latest RED Silicone Case.

That is one sharp observation right there, thanks for sharing

 

[Soundizer]

That is one sharp observation right there, thanks for sharing

Please sell on Apple Store. Some people might buy it thinking it is the next gen iPhone.

 

[iFi audio]

Please sell on Apple Store. Some people might buy it thinking it is the next gen iPhone.

If only this was possible

 

[whitedragem]

G’day,.. (even if it ISN’T day time in your timezone when being here and reading this..)

its a simple salutation that states overtly I am an Aussie.

This shouldn’t matter, but us Aussies (-‘like Canadians’: if stereotypes and observations are believable) are often a patient lot/ very ‘polite’ and generally seen as nice people.
Whilst this shouldn’t matter I give this information to qualify something about myself which might give weight to my words here given...


I really enjoy talking with people who share audio as a hobby or, more so, ‘passion’..

I am here today just to share some quick FYI’s / “info updates” for ‘we’ the people.


For the community that has held me captive (enjoyable so), I have been asked questions and as I now have some perspective to tackle those questions / share an opinion, I felt compelled to be here and share my perspective or ‘give some answers’.

At the end of the day many of us are scientific in our observations (and learning methods) and more info at our disposal has proven historically to be beneficial.

I will steer clear of subjective opinion where possible (in this post) so as to keep it short/succinct and therefore potentially helpful to others..

PURPOSE OF POST
I believe a community stands tall based on the actions of its members’, so I am ‘popping by to say “hi” ‘, and sharing my recent observations with some newfound functionality I have experienced with the iFi 7.2 firmware branch.


WARNING:
updating firmwares not listed for your product is generally considered A VERY BAD IDEA, and exceptionally few manufacturers globally would dare recommend such a practice.


The firmwares I have run, to the best of my knowledge HAVE NOT BEEN DESIGNED specifically for the Diablo, and are likely to not utilise its features fully (whether one firmware might limit processing capability based on potential circuit designs or tune towards a sound in alignment with the sound profile that a SPECIFIC hardware setup might offer (ie: the ‘norm’), iFi appear to have a ‘one size fits all’ design to their firmwares, and from the perspective of using two different XMOS controller chips in the majority of their hardware catalogue, it can be seen that this approach works for them: two different firmware branches for the (two) respective controller chips.

Whether this impacts on giving features like the advertising spiel for the iFi Diablo suggests (things like processing MQA on the toslink input), the audio circuits that iFi licence and implement are NOT all the same, so like internet disclaimers given worldwide: Your Mileage May Vary!


A bad BIOS FLASH can easily brick a device and is generally NOT COVERED BY WARRANTY.
I have worked for businesses that REFUSE to perform BIOS updates for end users due to the very real risks involved. (I actually disagreed with the policy, but that doesn’t change the fact that it existed)
My first PC business started in the early 1990s’, and with that came firmware and BIOS flashing.
I have never killed a product from a bad firmware flash, but have had a couple of instances from (many) hundreds that have been ‘flakey’ and benefited from ‘another flash’.


I will leave my opinion on iFi and firmware ‘at the door’, and now rolling up my sleeves “lets begin”...


FIRMWARES TESTED

7.2 branch - B and C variants

++CH23 has been brilliant and helpful in their contributive efforts and has created tables to help other end users to understand the iFi products better. With help from CH23 (and one of the tables they created) choosing an iFi firmware that interested me was an easy process with clear details as to what each firmware contains. This post will not delve into differences between firmwares (be it 7.0 branch or 7.2 branch, nor the differences between them, other than confirming that 7.2 branch BIOSs’ I tested (B & C variants) both supported the consoles.

B is the evolved version of what we got ‘out of the box’ (is my understanding)- My Diablo was on a version 1.? firmware having never received a firmware flash previously.

C variant implements the GTO filter on the USB input when not feeding DSD or DSD via DoP..

-there might be other aspects to these firmwares but I encourage all interested parties to do their own research and confirm whether a firmware and its ‘features’ are of merit or worth to them - ie I take no responsibility for any others users actions given; even if that is because I convinced them that a $2000 USB cable would create rainbows in their living room- quite simply- I don’t know if a unicorn could even fit in your living room, so my opinion would be unresearched or not tailored to any given use case.

My USB cable makes me coffee in the morning and gets jealous when I look at other USB cables, and I am just thankful that they don’t read my words posted here.
Not all USB cables are created equally, and mine are just plain and simply better than yours’ (oops subjective opinion warning: you might not like COFFEE..)


So what comes with the 7.2 branch of firmwares that might be worth risking turning my beaut red audio box into a doorstop?

To quote the notes (as posted by the legendary community member CH23 who iFi SHOULD be remunerating/hiring); “Playstation console support”.

So, to get this out of the way: Yes; Playstation Console support was tested. As was Nintendo Switch support.

Both consoles PREVIOUSLY DID NOT WORK with the iFi Diablo, and both consoles now very much do.


Implementation is not perfect; but I am not sure how to use my time fruitfully giving feedback to iFi, so will just list some observations here:
________________________________________________________________________________
ISSUE: Diablo needs be powercycled or have USB cable unplugged and then replugged, in order to be seen by the console between console restarts/useage.

Support info: None of my other USB DACs that work with the consoles has required this (the issue is likely with iFi hardware or software (firmware)).

Consumer info: this might be a good thing and a solution might lead to change in the handshake methods between the devices and ‘get in the way’ of optimal sound.. (who knows/who cares)
________________________________________________________________________________
Regarding the above information: I am not trying to be a ‘poopster’ (no clarification with imagery regarding the word ‘poopster’ for those using translators; I don’t use the language needed, in public, to describe the word better).

Feedback (Whether positive or negative in interpretation), is generally ‘a good thing’.
Wouldn’t be doing my part in the discussion if I avoided important facts.

FIRMWARE WITH GTO or WITHOUT?

@justanut- you asked me once for my take on the GTO filter; and I couldn’t say anything to clarify my opinion on it. My bias is that it is an ace in iFis’ sleeve and a very real reason to buy an iFi part.
It is an understatement to confirm “I have been WANTING to test the Gibbs Transient Optimised filter” (for ‘quite awhile’ actually)


Setup and Testing Methodology

running initially the 7.2b firmware- everything checked out/ worked well, so I rolled another 7.2 branch firmware (thankyou to CH23 for making it possible) with the GTO filter.

As the GTO filter only works with the USB input on the Diablo component, I created a setup that allowed a feed into the USB input. To remove bottlenecks/ensure OPTIMAL delivery of the digital code, an ‘audio grade’ USB cable was used AND the iFi Purifier 3 filter technology was used outboard to the ‘noise’ filters/isolators/reducers found internally in the Diablo DAC/amp.

It must be noted that using the USB cable and iPurifier3 proved a very good transport option to the Diablo and is the first time the Diablo has given ‘exceptional’ performance via the USB input (vs high quality transportation via SPDIF toslink input).


Why care about USB quality?

It was easy to hear differences between sampling rates, and the higher than standard quality cable should (theoretically) remove potential bottlenecks that might make 44khz GTO vs 352khz GTO favour the easier data rate transmission.

I have seldom observed USB having the openess of good COAX, so I would say we were operating above excessive jitter or (fast) transient capability reduction.

Hardware:
Using an Android phone (Samsung Fold family), with USB C to USB type A (female) adapter, a Kimber USB cable (type A male to type B male), iPurifier3, then Diablo.
___________________________________________


___________________________________________
So: Phone=>USB adaptor=>USB cable=>iPurifier=>Diablo

Software:
Onkyo HF player (unlocked)


HF Player, when paid for/unlocked, can upsample files to a range of sampling rates and has four tiers of ‘on the fly’ DSD conversion available.
For DSD conversion I chose ‘high precision DSD 5.6Mhz’. This can be output as DSD or DoP.

(The software can also convert DSD to PCM and has equalisation/all sorts of DSD/PCM gain controls/ can do 3 and 6Mhz (etc)... )


Direct Transfer (not DoP) did have ‘the edge’ in sound quality when feeding into the Diablo, but it must be taken with a handful (or three) of grainy salt particulate (ideally Pink) that the subtleties between some of the sampling rates/wrapper output is of no practical concern for the vast majority even in our ‘audiofool’ community where we can often be zealots and PASSIONATE about such things...
The observation of facts and sharing them, if not given the weighting they deserve, can be misleading. To this end, sampling rate differences and wrappings was ‘hey thats different’ (threshold of observation), vs GTO and DSD and USB improvements that were appreciable in their affect


Media: many 16bit 44khz sound files in WAVE/AIFF/FLAC formats (only well engineered and mostly exceptional quality recordings have been audtioned/‘YMMV’)


So- this post is made hugely shorter by avoiding flowerly language and ‘much subjectivity’, I will list a few observations that will no doubt ‘peak’ user interest...

Initial testing was to ensure that the Diablo was decoding correctly and accepting the various output formats. The Android software ‘Onkyo HF player’ is a few years out of date (unsupported), but has been well regarded in specific circles for its ability to bypass the android system mixer and pass unadulterated ‘zeros’ and ‘ones’ (or ‘very adulterated’ 0s and 1s, as I am about to describe..)

I have had a couple of months of using Onkyo HF player to upconvert 16bit 44khz PCM to DSD (Thanks @BobParish for raising awareness that DSD avoids PCM ‘filters’ and therefore gives inherently different sound).

Previously when doing this feed, it has been through an iFi connection cable. That cable is horrible for sound duty when compared to one worth ten times the cost... Sadly for all previous testing of DSD conversion from a phone: that cable has been the only part in play. Hindsight being 20/20 I would have used a better cable and ‘no doubt’ experienced the much better analytical platform that I have been testing this last week...

Previously DSD (upscaled and converted PCM) was different to PCM (raw), and enough so that the DSD feed was consistently better. Always appeared slightly more natural /and in some ways ‘less raw’, but both options didn’t hold a candle to the TOSLINK from a Questyle QP1R..

With the USB cable replacement (and perhaps filter, apols for not testing ‘sans’ iPurifier), the difference is substantial. I would now reach for either audio transport, a dedicated android device for media playback (ie no wifi/bluetooth/cellular ‘on’) or a dedicated DAP (both being decent transports now).

The Android device using standard USB short cable was ‘average’ sound and not involving.

The test setup presently running is of a high performance nature and ‘fitting’ for the DAC being used. (the difference being connection quality /transport of the zeros and ones by changing the physical line used)

I am skating around the elephant in the room: USB transport isn’t ‘black/white. (cables and setup MIGHT ACTUALLY MATTER, but ‘not the point of this post’)


Observations:

DSD sounded great, but GTO was instantly like a great NOS DAC. The leading edge on transients was FAST and coherent and the decay was natural and lively, vs the DSD equivalent which felt like it was almost obscuring the soundstage, and definitely without the true leading edge ‘sharpness’ that great transports are capable of giving.


Remember I am trying NOT to go ‘flowery/subjective’ here. -Switching between GTO and DSD was a night and day difference. Both had been improved by the change in USB transportation method, but the noticeable sound of the ‘pure’ format (PCM passed into Diablo, and operated on with the GTO filter) was fantastic. I would say ‘objectively better’ based on comparing some of the worlds better audio transports (TEAC p700 and Denon DCD-s10) vs lowly massmarket stuff (Rotel RCD971 / Denon DCD1500(? model not in front of me)). When we go up the range in CD players a base model is like the iFi USB cable sound, the mid tier stuff (Rotel RCD971), like a DSD conversion passed over the iFi USB cable, and then any of the digital I have been experiencing via nice USB cable setup is like nice mainstream high quality stuff (CD players north of $1000 ‘back in the day’/cheap dedicated transport), with the GTO filter instantly jumping to the ‘top of the tree’ stuff (esoteric transports and top of the ladder CD players).


Is the GTO filter always better for enjoying the listen? No.

On JC Superstar (australian ed/1992?), the opening overture seemed flatter (by depth of stage) but felt further widened, and this made it more engaging. In this instance the ’air’ advantage went to the DSD process.

As a recent play with a R2R/Ladder DAC revealed, even a lossless reference file via an APT X (not Apt X HD) transmission into a Fiio BTA30 then fed using ‘the worlds cheapest fibre optic cable’ into the DAC- rendered more depth of the ‘very familiar test track’ than any previously experience.. this did make the instruments placed 70ft back in the soundstage seem ‘closer to centre stage’, but the sense of depth does that. -so my trained mind is telling me that the DSD version of the rock orchestra performing Lloyd Webber in this instance, the DSD version, whilst more exciting (under headphones), might not be the best or ‘technically better’. If I was new to the hobby I would hands down give the win to the DSD process as it was viscerally different.

The observation differences between (on the fly converted) DSD and native PCM being GTO ‘filtered’ led to the sense that DSD format was like a Chord DAC (my only reference being a Hugo (version 1)) in that it kinda homogenised the sound and created a soundfield with coherency that might not be accurate to the recording.

Swerving off the road and driving head long for the cliff of obscurity overlooking the sea of subjectivity: The Hot 8 Brass band instruments had a ‘fleshed out’ weight in a modified soundfield that could be enjoyed in ways different to the RAW soundfile. Texture was more pronounced but only cause that is what the DSD was most notably bringing..
A brightness is very much lost in the attack of the instruments/leading edge of the note.

On a poor system this ‘brightness’ would be fatiguing vs a great ‘NOS’ setup that would render these leading edges accurately and be ‘non fatiguing playback’. (Albeit a hard setup to create with the respective synergy that all parts need to bring to the table)

The difference is the GTO filter is noticably more honest to the ‘best’ playback of such sounds (I have heard), in that it clearly removes a ‘veil’ and hence feels like the sound can be more present ‘up front’. Placement for both methods was very good, and HUGELY improved by the change in USB cables (more than on the fly conversion or upsampling might bring).


Each time I listen to one format (converted to DSD / PCM GTO’d) I listen for a long while until I forget what I am listening to.
=>
EVERY TIME I CHANGED TO GTO METHOD: I instantly ‘knew’ that it was the better playback. More engaging. More musical. Cleaner and unheard air around the instruments (cause ‘all the air is there’), vs the converted to DSD variant which lost the ‘blackness’ if you will, by having ‘air pockets’/bubbles around the instruments. (Still really good and not describing these air pockets as a negative- most systems would be happy to render this quality of ‘air’)

The DSD is jump over PCM, yes, when the GTO isn’t involved.

Given all recordings have different quality levels/engineering efforts, I would certainly use these two modes to tailor my sound to make a playback session ‘easy to enjoy’.


Caveat: Didn’t use home speakers which would allow an easy soundfield placement /interaction with the perceived 3D space in a room. One method might render a better interaction with 3D (realworld) space and environments.
Under headphones (being closed back Ultrasone Edition 5, or open back AKG K612) the GTO filter was easily the preference towards a ‘higher grade’ of sound quality perception.

—That leading edge delivery I have only heard previously in ‘really great home dedicated DACs’, and for anyone who has spent a long time on this journey, it was evident that this is the Holy Grail /‘summit’ that many are seeking.

To be fair this isn’t a definitive take on DSD conversion: member @BobParish has given some great guides and recommendations as to filters to encode with. My DSD conversion here tested didn’t give filter options, just a few quality modes to choose from. Using home hardware I could highly overclock my PC (watercooling and 8 fans) to achieve a degree of (on the fly) DSD conversion that the world is ‘on the cusp’ of delivering,.. presently a ‘prerender’ to ‘large files’ and DSD 1024 /512 hi quality options are ‘no doubt’ available. The home hardware setup and the cost of that software puts it out of reach of many, and generally not practical for portable ‘head-fi’.


Further observation (relating to gaming/ movie soundfields etc): The GTO filter did not harm the soundstage perception, recede depth, or in any negative way impact on the location of sound sources in the soundstage/environment. It is a joy to utilise the GTO filter when engaging with these applications, happily netting the free benefit of upsampling the sound WELL (something the new firmware that iFi have access to can deliver.)

What became evident: GTO holds its own against portable DSD conversion.


Further anecdotal testing:
sometimes swapping between GTO and DSD offered little perceived difference, the level of subtlety ‘tiny enough’ that unless we were looking for it, probably not going to notice anything.. (again, not tested on ‘room setup with speakers’)

There was perceivable differences (subtle) between 44khz output (with Diablo doing sample rate upconversion=>GTO), and feeding out 352khz with Diablo doing simply the GTO process (which typically does upsampling as well).

From what I was hearing, it seemed that the GTO is an upscaler as well as a filter process (capable of handling ‘pre ringing issues). This is one of the best upscalers for audio I have heard implemented.

I would believe that feeding 44khz over a cheap USB cable would be easier than feeding 352khz, and so would trust in the Diablo doing the upsampling (via GTO).

352 vs DSD- I’d take either, with a preference to the GTO’d 352 depending on music file.
352 vs 44 I’d feed either into the GTO filter’d Diablo (and not care, so would probably use 44khz if ‘garden variety USB cable in chain)

Conclusion
I have NEVER heard upsampling sound sound as great as exceptionally done ‘Non Oversampling’ methods. The GTO is the first upsampling method i am truly infatuated with. (just don’t tell my USB cable or else my coffee tomorrow might be compromised!)


TLR (too long/didn’t read)

Nintendo Switch and Playstation (family) consoles now work with the Diablo (on firmware 7.2 branch, that isn’t officially out yet for the Diablo DAC/amp)

GTO filter is different and ‘awesome’ =>whatever that means to you, your own testing and observation WILL MATTER MORE and I encourage end users to consider the ‘GTO’ firmware variant of the BIOS options when making their Diablo ‘console friendly’.

-edited to fix some parenthesis and correct cable ’anatomy’ in my parts description. not proofread, apols, just a fun writeup for ‘we the people’ (audiofoolz), I prolly could remove a tonne of linebreaks etc.. but left this way to improve readability (big post)

 

[Soundizer]

G’day,.. (even if it ISN’T day time in your timezone when being here and reading this..)

its a simple salutation that states overtly I am an Aussie.

This shouldn’t matter, but us Aussies (-‘like Canadians’: if stereotypes and observations are believable) are often a patient lot/ very ‘polite’ and generally seen as nice people.
Whilst this shouldn’t matter I give this information to qualify something about myself which might give weight to my words here given...


I really enjoy talking with people who share audio as a hobby or, more so, ‘passion’..

I am here today just to share some quick FYI’s / “info updates” for ‘we’ the people.


For the community that has held me captive (enjoyable so), I have been asked questions and as I now have some perspective to tackle those questions / share an opinion, I felt compelled to be here and share my perspective or ‘give some answers’.

At the end of the day many of us are scientific in our observations (and learning methods) and more info at our disposal has proven historically to be beneficial.

I will steer clear of subjective opinion where possible (in this post) so as to keep it short/succinct and therefore potentially helpful to others..

PURPOSE OF POST
I believe a community stands tall based on the actions of its members’, so I am ‘popping by to say “hi” ‘, and sharing my recent observations with some newfound functionality I have experienced with the iFi 7.2 firmware branch.


WARNING:
updating firmwares not listed for your product is generally considered A VERY BAD IDEA, and exceptionally few manufacturers globally would dare recommend such a practice.


The firmwares I have run, to the best of my knowledge HAVE NOT BEEN DESIGNED specifically for the Diablo, and are likely to not utilise its features fully (whether one firmware might limit processing capability based on potential circuit designs or tune towards a sound in alignment with the sound profile that a SPECIFIC hardware setup might offer (ie: the ‘norm’), iFi appear to have a ‘one size fits all’ design to their firmwares, and from the perspective of using two different XMOS controller chips in the majority of their hardware catalogue, it can be seen that this approach works for them: two different firmware branches for the (two) respective controller chips.

Whether this impacts on giving features like the advertising spiel for the iFi Diablo suggests (things like processing MQA on the toslink input), the audio circuits that iFi licence and implement are NOT all the same, so like internet disclaimers given worldwide: Your Mileage May Vary!


A bad BIOS FLASH can easily brick a device and is generally NOT COVERED BY WARRANTY.
I have worked for businesses that REFUSE to perform BIOS updates for end users due to the very real risks involved. (I actually disagreed with the policy, but that doesn’t change the fact that it existed)
My first PC business started in the early 1990s’, and with that came firmware and BIOS flashing.
I have never killed a product from a bad firmware flash, but have had a couple of instances from (many) hundreds that have been ‘flakey’ and benefited from ‘another flash’.


I will leave my opinion on iFi and firmware ‘at the door’, and now rolling up my sleeves “lets begin”...


FIRMWARES TESTED

7.2 branch - B and C variants

++CH23 has been brilliant and helpful in their contributive efforts and has created tables to help other end users to understand the iFi products better. With help from CH23 (and one of the tables they created) choosing an iFi firmware that interested me was an easy process with clear details as to what each firmware contains. This post will not delve into differences between firmwares (be it 7.0 branch or 7.2 branch, nor the differences between them, other than confirming that 7.2 branch BIOSs’ I tested (B & C variants) both supported the consoles.

B is the evolved version of what we got ‘out of the box’ (is my understanding)- My Diablo was on a version 1.? firmware having never received a firmware flash previously.

C variant implements the GTO filter on the USB input when not feeding DSD or DSD via DoP..

-there might be other aspects to these firmwares but I encourage all interested parties to do their own research and confirm whether a firmware and its ‘features’ are of merit or worth to them - ie I take no responsibility for any others users actions given; even if that is because I convinced them that a $2000 USB cable would create rainbows in their living room- quite simply- I don’t know if a unicorn could even fit in your living room, so my opinion would be unresearched or not tailored to any given use case.

My USB cable makes me coffee in the morning and gets jealous when I look at other USB cables, and I am just thankful that they don’t read my words posted here.
Not all USB cables are created equally, and mine are just plain and simply better than yours’ (oops subjective opinion warning: you might not like COFFEE..)


So what comes with the 7.2 branch of firmwares that might be worth risking turning my beaut red audio box into a doorstop?

To quote the notes (as posted by the legendary community member CH23 who iFi SHOULD be remunerating/hiring); “Playstation console support”.

So, to get this out of the way: Yes; Playstation Console support was tested. As was Nintendo Switch support.

Both consoles PREVIOUSLY DID NOT WORK with the iFi Diablo, and both consoles now very much do.


Implementation is not perfect; but I am not sure how to use my time fruitfully giving feedback to iFi, so will just list some observations here:
________________________________________________________________________________
ISSUE: Diablo needs be powercycled or have USB cable unplugged and then replugged, in order to be seen by the console between console restarts/useage.

Support info: None of my other USB DACs that work with the consoles has required this (the issue is likely with iFi hardware or software (firmware)).

Consumer info: this might be a good thing and a solution might lead to change in the handshake methods between the devices and ‘get in the way’ of optimal sound.. (who knows/who cares)
________________________________________________________________________________
Regarding the above information: I am not trying to be a ‘poopster’ (no clarification with imagery regarding the word ‘poopster’ for those using translators; I don’t use the language needed, in public, to describe the word better).

Feedback (Whether positive or negative in interpretation), is generally ‘a good thing’.
Wouldn’t be doing my part in the discussion if I avoided important facts.

FIRMWARE WITH GTO or WITHOUT?

@justanut- you asked me once for my take on the GTO filter; and I couldn’t say anything to clarify my opinion on it. My bias is that it is an ace in iFis’ sleeve and a very real reason to buy an iFi part.
It is an understatement to confirm “I have been WANTING to test the Gibbs Transient Optimised filter” (for ‘quite awhile’ actually)


Setup and Testing Methodology

running initially the 7.2b firmware- everything checked out/ worked well, so I rolled another 7.2 branch firmware (thankyou to CH23 for making it possible) with the GTO filter.

As the GTO filter only works with the USB input on the Diablo component, I created a setup that allowed a feed into the USB input. To remove bottlenecks/ensure OPTIMAL delivery of the digital code, an ‘audio grade’ USB cable was used AND the iFi Purifier 3 filter technology was used outboard to the ‘noise’ filters/isolators/reducers found internally in the Diablo DAC/amp.


It must be noted that using the USB cable and iPurifier3 proved a very good transport option to the Diablo and is the first time the Diablo has given ‘exceptional’ performance via the USB input (vs high quality transportation via SPDIF toslink input).


Why care about USB quality?

It was easy to hear differences between sampling rates, and the higher than standard quality cable should (theoretically) remove potential bottlenecks that might make 44khz GTO vs 352khz GTO favour the easier data rate transmission.

I have seldom observed USB having the openess of good COAX, so I would say we were operating above excessive jitter or (fast) transient capability reduction.

Hardware:
Using an Android phone (Samsung Fold family), with USB C to USB type A (female) adapter, a Kimber USB cable (type A male to type B male), iPurifier3, then Diablo.
___________________________________________


___________________________________________
So: Phone=>USB adaptor=>USB cable=>iPurifier=>Diablo

Software:
Onkyo HF player (unlocked)


HF Player, when paid for/unlocked, can upsample files to a range of sampling rates and has four tiers of ‘on the fly’ DSD conversion available.
For DSD conversion I chose ‘high precision DSD 5.6Mhz’. This can be output as DSD or DoP.

(The software can also convert DSD to PCM and has equalisation/all sorts of DSD/PCM gain controls/ can do 3 and 6Mhz (etc)... )


Direct Transfer (not DoP) did have ‘the edge’ in sound quality when feeding into the Diablo, but it must be taken with a handful (or three) of grainy salt particulate (ideally Pink) that the subtleties between some of the sampling rates/wrapper output is of no practical concern for the vast majority even in our ‘audiofool’ community where we can often be zealots and PASSIONATE about such things...
The observation of facts and sharing them, if not given the weighting they deserve, can be misleading. To this end, sampling rate differences and wrappings was ‘hey thats different’ (threshold of observation), vs GTO and DSD and USB improvements that were appreciable in their affect


Media: many 16bit 44khz sound files in WAVE/AIFF/FLAC formats (only well engineered and mostly exceptional quality recordings have been audtioned/‘YMMV’)


So- this post is made hugely shorter by avoiding flowerly language and ‘much subjectivity’, I will list a few observations that will no doubt ‘peak’ user interest...

Initial testing was to ensure that the Diablo was decoding correctly and accepting the various output formats. The Android software ‘Onkyo HF player’ is a few years out of date (unsupported), but has been well regarded in specific circles for its ability to bypass the android system mixer and pass unadulterated ‘zeros’ and ‘ones’ (or ‘very adulterated’ 0s and 1s, as I am about to describe..)

I have had a couple of months of using Onkyo HF player to upconvert 16bit 44khz PCM to DSD (Thanks @BobParish for raising awareness that DSD avoids PCM ‘filters’ and therefore gives inherently different sound).

Previously when doing this feed, it has been through an iFi connection cable. That cable is horrible for sound duty when compared to one worth ten times the cost... Sadly for all previous testing of DSD conversion from a phone: that cable has been the only part in play. Hindsight being 20/20 I would have used a better cable and ‘no doubt’ experienced the much better analytical platform that I have been testing this last week...

Previously DSD (upscaled and converted PCM) was different to PCM (raw), and enough so that the DSD feed was consistently better. Always appeared slightly more natural /and in some ways ‘less raw’, but both options didn’t hold a candle to the TOSLINK from a Questyle QP1R..

With the USB cable replacement (and perhaps filter, apols for not testing ‘sans’ iPurifier), the difference is substantial. I would now reach for either audio transport, a dedicated android device for media playback (ie no wifi/bluetooth/cellular ‘on’) or a dedicated DAP (both being decent transports now).

The Android device using standard USB short cable was ‘average’ sound and not involving.

The test setup presently running is of a high performance nature and ‘fitting’ for the DAC being used. (the difference being connection quality /transport of the zeros and ones by changing the physical line used)



I am skating around the elephant in the room: USB transport isn’t ‘black/white. (cables and setup MIGHT ACTUALLY MATTER, but ‘not the point of this post’)


Observations:

DSD sounded great, but GTO was instantly like a great NOS DAC. The leading edge on transients was FAST and coherent and the decay was natural and lively, vs the DSD equivalent which felt like it was almost obscuring the soundstage, and definitely without the true leading edge ‘sharpness’ that great transports are capable of giving.


Remember I am trying NOT to go ‘flowery/subjective’ here. -Switching between GTO and DSD was a night and day difference. Both had been improved by the change in USB transportation method, but the noticeable sound of the ‘pure’ format (PCM passed into Diablo, and operated on with the GTO filter) was fantastic. I would say ‘objectively better’ based on comparing some of the worlds better audio transports (TEAC p700 and Denon DCD-s10) vs lowly massmarket stuff (Rotel RCD971 / Denon DCD1500(? model not in front of me)). When we go up the range in CD players a base model is like the iFi USB cable sound, the mid tier stuff (Rotel RCD971), like a DSD conversion passed over the iFi USB cable, and then any of the digital I have been experiencing via nice USB cable setup is like nice mainstream high quality stuff (CD players north of $1000 ‘back in the day’/cheap dedicated transport), with the GTO filter instantly jumping to the ‘top of the tree’ stuff (esoteric transports and top of the ladder CD players).


Is the GTO filter always better for enjoying the listen? No.

On JC Superstar (australian ed/1992?), the opening overture seemed flatter (by depth of stage) but felt further widened, and this made it more engaging. In this instance the ’air’ advantage went to the DSD process.

As a recent play with a R2R/Ladder DAC revealed, even a lossless reference file via an APT X (not Apt X HD) transmission into a Fiio BTA30 then fed using ‘the worlds cheapest fibre optic cable’ into the DAC- rendered more depth of the ‘very familiar test track’ than any previously experience.. this did make the instruments placed 70ft back in the soundstage seem ‘closer to centre stage’, but the sense of depth does that. -so my trained mind is telling me that the DSD version of the rock orchestra performing Lloyd Webber in this instance, the DSD version, whilst more exciting (under headphones), might not be the best or ‘technically better’. If I was new to the hobby I would hands down give the win to the DSD process as it was viscerally different.

The observation differences between (on the fly converted) DSD and native PCM being GTO ‘filtered’ led to the sense that DSD format was like a Chord DAC (my only reference being a Hugo (version 1)) in that it kinda homogenised the sound and created a soundfield with coherency that might not be accurate to the recording.

Swerving off the road and driving head long for the cliff of obscurity overlooking the sea of subjectivity: The Hot 8 Brass band instruments had a ‘fleshed out’ weight in a modified soundfield that could be enjoyed in ways different to the RAW soundfile. Texture was more pronounced but only cause that is what the DSD was most notably bringing..
A brightness is very much lost in the attack of the instruments/leading edge of the note.

On a poor system this ‘brightness’ would be fatiguing vs a great ‘NOS’ setup that would render these leading edges accurately and be ‘non fatiguing playback’. (Albeit a hard setup to create with the respective synergy that all parts need to bring to the table)

The difference is the GTO filter is noticably more honest to the ‘best’ playback of such sounds (I have heard), in that it clearly removes a ‘veil’ and hence feels like the sound can be more present ‘up front’. Placement for both methods was very good, and HUGELY improved by the change in USB cables (more than on the fly conversion or upsampling might bring).


Each time I listen to one format (converted to DSD / PCM GTO’d) I listen for a long while until I forget what I am listening to.
=>
EVERY TIME I CHANGED TO GTO METHOD: I instantly ‘knew’ that it was the better playback. More engaging. More musical. Cleaner and unheard air around the instruments (cause ‘all the air is there’), vs the converted to DSD variant which lost the ‘blackness’ if you will, by having ‘air pockets’/bubbles around the instruments. (Still really good and not describing these air pockets as a negative- most systems would be happy to render this quality of ‘air’)

The DSD is jump over PCM, yes, when the GTO isn’t involved.

Given all recordings have different quality levels/engineering efforts, I would certainly use these two modes to tailor my sound to make a playback session ‘easy to enjoy’.


Caveat: Didn’t use home speakers which would allow an easy soundfield placement /interaction with the perceived 3D space in a room. One method might render a better interaction with 3D (realworld) space and environments.
Under headphones (being closed back Ultrasone Edition 5, or open back AKG K612) the GTO filter was easily the preference towards a ‘higher grade’ of sound quality perception.

—That leading edge delivery I have only heard previously in ‘really great home dedicated DACs’, and for anyone who has spent a long time on this journey, it was evident that this is the Holy Grail /‘summit’ that many are seeking.

To be fair this isn’t a definitive take on DSD conversion: member @BobParish has given some great guides and recommendations as to filters to encode with. My DSD conversion here tested didn’t give filter options, just a few quality modes to choose from. Using home hardware I could highly overclock my PC (watercooling and 8 fans) to achieve a degree of (on the fly) DSD conversion that the world is ‘on the cusp’ of delivering,.. presently a ‘prerender’ to ‘large files’ and DSD 1024 /512 hi quality options are ‘no doubt’ available. The home hardware setup and the cost of that software puts it out of reach of many, and generally not practical for portable ‘head-fi’.


Further observation (relating to gaming/ movie soundfields etc): The GTO filter did not harm the soundstage perception, recede depth, or in any negative way impact on the location of sound sources in the soundstage/environment. It is a joy to utilise the GTO filter when engaging with these applications, happily netting the free benefit of upsampling the sound WELL (something the new firmware that iFi have access to can deliver.)

What became evident: GTO holds its own against portable DSD conversion.


Further anecdotal testing:
sometimes swapping between GTO and DSD offered little perceived difference, the level of subtlety ‘tiny enough’ that unless we were looking for it, probably not going to notice anything.. (again, not tested on ‘room setup with speakers’)

There was perceivable differences (subtle) between 44khz output (with Diablo doing sample rate upconversion=>GTO), and feeding out 352khz with Diablo doing simply the GTO process (which typically does upsampling as well).

From what I was hearing, it seemed that the GTO is an upscaler as well as a filter process (capable of handling ‘pre ringing issues). This is one of the best upscalers for audio I have heard implemented.

I would believe that feeding 44khz over a cheap USB cable would be easier than feeding 352khz, and so would trust in the Diablo doing the upsampling (via GTO).

352 vs DSD- I’d take either, with a preference to the GTO’d 352 depending on music file.
352 vs 44 I’d feed either into the GTO filter’d Diablo (and not care, so would probably use 44khz if ‘garden variety USB cable in chain)

Conclusion
I have NEVER heard upsampling sound sound as great as exceptionally done ‘Non Oversampling’ methods. The GTO is the first upsampling method i am truly infatuated with. (just don’t tell my USB cable or else my coffee tomorrow might be compromised!)


TLR (too long/didn’t read)

Nintendo Switch and Playstation (family) consoles now work with the Diablo (on firmware 7.2 branch, that isn’t officially out yet for the Diablo DAC/amp)

GTO filter is different and ‘awesome’ =>whatever that means to you, your own testing and observation WILL MATTER MORE and I encourage end users to consider the ‘GTO’ firmware variant of the BIOS options when making their Diablo ‘console friendly’.

-edited to fix some parenthesis and correct cable ’anatomy’ in my parts description. not proofread, apols, just a fun writeup for ‘we the people’ (audiofoolz), I prolly could remove a tonne of linebreaks etc.. but left this way to improve readability (big post)

I also love the colour. It matches colour of my Phone Case which is a red.

 

[iFi audio]

G’day,.. (even if it ISN’T day time in your timezone when being here and reading this..)

its a simple salutation that states overtly I am an Aussie.

This shouldn’t matter, but us Aussies (-‘like Canadians’: if stereotypes and observations are believable) are often a patient lot/ very ‘polite’ and generally seen as nice people.
Whilst this shouldn’t matter I give this information to qualify something about myself which might give weight to my words here given...


I really enjoy talking with people who share audio as a hobby or, more so, ‘passion’..

I am here today just to share some quick FYI’s / “info updates” for ‘we’ the people.


For the community that has held me captive (enjoyable so), I have been asked questions and as I now have some perspective to tackle those questions / share an opinion, I felt compelled to be here and share my perspective or ‘give some answers’.

At the end of the day many of us are scientific in our observations (and learning methods) and more info at our disposal has proven historically to be beneficial.

I will steer clear of subjective opinion where possible (in this post) so as to keep it short/succinct and therefore potentially helpful to others..

PURPOSE OF POST
I believe a community stands tall based on the actions of its members’, so I am ‘popping by to say “hi” ‘, and sharing my recent observations with some newfound functionality I have experienced with the iFi 7.2 firmware branch.


WARNING:
updating firmwares not listed for your product is generally considered A VERY BAD IDEA, and exceptionally few manufacturers globally would dare recommend such a practice.


The firmwares I have run, to the best of my knowledge HAVE NOT BEEN DESIGNED specifically for the Diablo, and are likely to not utilise its features fully (whether one firmware might limit processing capability based on potential circuit designs or tune towards a sound in alignment with the sound profile that a SPECIFIC hardware setup might offer (ie: the ‘norm’), iFi appear to have a ‘one size fits all’ design to their firmwares, and from the perspective of using two different XMOS controller chips in the majority of their hardware catalogue, it can be seen that this approach works for them: two different firmware branches for the (two) respective controller chips.

Whether this impacts on giving features like the advertising spiel for the iFi Diablo suggests (things like processing MQA on the toslink input), the audio circuits that iFi licence and implement are NOT all the same, so like internet disclaimers given worldwide: Your Mileage May Vary!


A bad BIOS FLASH can easily brick a device and is generally NOT COVERED BY WARRANTY.
I have worked for businesses that REFUSE to perform BIOS updates for end users due to the very real risks involved. (I actually disagreed with the policy, but that doesn’t change the fact that it existed)
My first PC business started in the early 1990s’, and with that came firmware and BIOS flashing.
I have never killed a product from a bad firmware flash, but have had a couple of instances from (many) hundreds that have been ‘flakey’ and benefited from ‘another flash’.


I will leave my opinion on iFi and firmware ‘at the door’, and now rolling up my sleeves “lets begin”...


FIRMWARES TESTED

7.2 branch - B and C variants

++CH23 has been brilliant and helpful in their contributive efforts and has created tables to help other end users to understand the iFi products better. With help from CH23 (and one of the tables they created) choosing an iFi firmware that interested me was an easy process with clear details as to what each firmware contains. This post will not delve into differences between firmwares (be it 7.0 branch or 7.2 branch, nor the differences between them, other than confirming that 7.2 branch BIOSs’ I tested (B & C variants) both supported the consoles.

B is the evolved version of what we got ‘out of the box’ (is my understanding)- My Diablo was on a version 1.? firmware having never received a firmware flash previously.

C variant implements the GTO filter on the USB input when not feeding DSD or DSD via DoP..

-there might be other aspects to these firmwares but I encourage all interested parties to do their own research and confirm whether a firmware and its ‘features’ are of merit or worth to them - ie I take no responsibility for any others users actions given; even if that is because I convinced them that a $2000 USB cable would create rainbows in their living room- quite simply- I don’t know if a unicorn could even fit in your living room, so my opinion would be unresearched or not tailored to any given use case.

My USB cable makes me coffee in the morning and gets jealous when I look at other USB cables, and I am just thankful that they don’t read my words posted here.
Not all USB cables are created equally, and mine are just plain and simply better than yours’ (oops subjective opinion warning: you might not like COFFEE..)


So what comes with the 7.2 branch of firmwares that might be worth risking turning my beaut red audio box into a doorstop?

To quote the notes (as posted by the legendary community member CH23 who iFi SHOULD be remunerating/hiring); “Playstation console support”.

So, to get this out of the way: Yes; Playstation Console support was tested. As was Nintendo Switch support.

Both consoles PREVIOUSLY DID NOT WORK with the iFi Diablo, and both consoles now very much do.


Implementation is not perfect; but I am not sure how to use my time fruitfully giving feedback to iFi, so will just list some observations here:
________________________________________________________________________________
ISSUE: Diablo needs be powercycled or have USB cable unplugged and then replugged, in order to be seen by the console between console restarts/useage.

Support info: None of my other USB DACs that work with the consoles has required this (the issue is likely with iFi hardware or software (firmware)).

Consumer info: this might be a good thing and a solution might lead to change in the handshake methods between the devices and ‘get in the way’ of optimal sound.. (who knows/who cares)
________________________________________________________________________________
Regarding the above information: I am not trying to be a ‘poopster’ (no clarification with imagery regarding the word ‘poopster’ for those using translators; I don’t use the language needed, in public, to describe the word better).

Feedback (Whether positive or negative in interpretation), is generally ‘a good thing’.
Wouldn’t be doing my part in the discussion if I avoided important facts.

FIRMWARE WITH GTO or WITHOUT?

@justanut- you asked me once for my take on the GTO filter; and I couldn’t say anything to clarify my opinion on it. My bias is that it is an ace in iFis’ sleeve and a very real reason to buy an iFi part.
It is an understatement to confirm “I have been WANTING to test the Gibbs Transient Optimised filter” (for ‘quite awhile’ actually)


Setup and Testing Methodology

running initially the 7.2b firmware- everything checked out/ worked well, so I rolled another 7.2 branch firmware (thankyou to CH23 for making it possible) with the GTO filter.

As the GTO filter only works with the USB input on the Diablo component, I created a setup that allowed a feed into the USB input. To remove bottlenecks/ensure OPTIMAL delivery of the digital code, an ‘audio grade’ USB cable was used AND the iFi Purifier 3 filter technology was used outboard to the ‘noise’ filters/isolators/reducers found internally in the Diablo DAC/amp.


It must be noted that using the USB cable and iPurifier3 proved a very good transport option to the Diablo and is the first time the Diablo has given ‘exceptional’ performance via the USB input (vs high quality transportation via SPDIF toslink input).


Why care about USB quality?

It was easy to hear differences between sampling rates, and the higher than standard quality cable should (theoretically) remove potential bottlenecks that might make 44khz GTO vs 352khz GTO favour the easier data rate transmission.

I have seldom observed USB having the openess of good COAX, so I would say we were operating above excessive jitter or (fast) transient capability reduction.

Hardware:
Using an Android phone (Samsung Fold family), with USB C to USB type A (female) adapter, a Kimber USB cable (type A male to type B male), iPurifier3, then Diablo.
___________________________________________


___________________________________________
So: Phone=>USB adaptor=>USB cable=>iPurifier=>Diablo

Software:
Onkyo HF player (unlocked)


HF Player, when paid for/unlocked, can upsample files to a range of sampling rates and has four tiers of ‘on the fly’ DSD conversion available.
For DSD conversion I chose ‘high precision DSD 5.6Mhz’. This can be output as DSD or DoP.

(The software can also convert DSD to PCM and has equalisation/all sorts of DSD/PCM gain controls/ can do 3 and 6Mhz (etc)... )


Direct Transfer (not DoP) did have ‘the edge’ in sound quality when feeding into the Diablo, but it must be taken with a handful (or three) of grainy salt particulate (ideally Pink) that the subtleties between some of the sampling rates/wrapper output is of no practical concern for the vast majority even in our ‘audiofool’ community where we can often be zealots and PASSIONATE about such things...
The observation of facts and sharing them, if not given the weighting they deserve, can be misleading. To this end, sampling rate differences and wrappings was ‘hey thats different’ (threshold of observation), vs GTO and DSD and USB improvements that were appreciable in their affect


Media: many 16bit 44khz sound files in WAVE/AIFF/FLAC formats (only well engineered and mostly exceptional quality recordings have been audtioned/‘YMMV’)


So- this post is made hugely shorter by avoiding flowerly language and ‘much subjectivity’, I will list a few observations that will no doubt ‘peak’ user interest...

Initial testing was to ensure that the Diablo was decoding correctly and accepting the various output formats. The Android software ‘Onkyo HF player’ is a few years out of date (unsupported), but has been well regarded in specific circles for its ability to bypass the android system mixer and pass unadulterated ‘zeros’ and ‘ones’ (or ‘very adulterated’ 0s and 1s, as I am about to describe..)

I have had a couple of months of using Onkyo HF player to upconvert 16bit 44khz PCM to DSD (Thanks @BobParish for raising awareness that DSD avoids PCM ‘filters’ and therefore gives inherently different sound).

Previously when doing this feed, it has been through an iFi connection cable. That cable is horrible for sound duty when compared to one worth ten times the cost... Sadly for all previous testing of DSD conversion from a phone: that cable has been the only part in play. Hindsight being 20/20 I would have used a better cable and ‘no doubt’ experienced the much better analytical platform that I have been testing this last week...

Previously DSD (upscaled and converted PCM) was different to PCM (raw), and enough so that the DSD feed was consistently better. Always appeared slightly more natural /and in some ways ‘less raw’, but both options didn’t hold a candle to the TOSLINK from a Questyle QP1R..

With the USB cable replacement (and perhaps filter, apols for not testing ‘sans’ iPurifier), the difference is substantial. I would now reach for either audio transport, a dedicated android device for media playback (ie no wifi/bluetooth/cellular ‘on’) or a dedicated DAP (both being decent transports now).

The Android device using standard USB short cable was ‘average’ sound and not involving.

The test setup presently running is of a high performance nature and ‘fitting’ for the DAC being used. (the difference being connection quality /transport of the zeros and ones by changing the physical line used)



I am skating around the elephant in the room: USB transport isn’t ‘black/white. (cables and setup MIGHT ACTUALLY MATTER, but ‘not the point of this post’)


Observations:

DSD sounded great, but GTO was instantly like a great NOS DAC. The leading edge on transients was FAST and coherent and the decay was natural and lively, vs the DSD equivalent which felt like it was almost obscuring the soundstage, and definitely without the true leading edge ‘sharpness’ that great transports are capable of giving.


Remember I am trying NOT to go ‘flowery/subjective’ here. -Switching between GTO and DSD was a night and day difference. Both had been improved by the change in USB transportation method, but the noticeable sound of the ‘pure’ format (PCM passed into Diablo, and operated on with the GTO filter) was fantastic. I would say ‘objectively better’ based on comparing some of the worlds better audio transports (TEAC p700 and Denon DCD-s10) vs lowly massmarket stuff (Rotel RCD971 / Denon DCD1500(? model not in front of me)). When we go up the range in CD players a base model is like the iFi USB cable sound, the mid tier stuff (Rotel RCD971), like a DSD conversion passed over the iFi USB cable, and then any of the digital I have been experiencing via nice USB cable setup is like nice mainstream high quality stuff (CD players north of $1000 ‘back in the day’/cheap dedicated transport), with the GTO filter instantly jumping to the ‘top of the tree’ stuff (esoteric transports and top of the ladder CD players).


Is the GTO filter always better for enjoying the listen? No.

On JC Superstar (australian ed/1992?), the opening overture seemed flatter (by depth of stage) but felt further widened, and this made it more engaging. In this instance the ’air’ advantage went to the DSD process.

As a recent play with a R2R/Ladder DAC revealed, even a lossless reference file via an APT X (not Apt X HD) transmission into a Fiio BTA30 then fed using ‘the worlds cheapest fibre optic cable’ into the DAC- rendered more depth of the ‘very familiar test track’ than any previously experience.. this did make the instruments placed 70ft back in the soundstage seem ‘closer to centre stage’, but the sense of depth does that. -so my trained mind is telling me that the DSD version of the rock orchestra performing Lloyd Webber in this instance, the DSD version, whilst more exciting (under headphones), might not be the best or ‘technically better’. If I was new to the hobby I would hands down give the win to the DSD process as it was viscerally different.

The observation differences between (on the fly converted) DSD and native PCM being GTO ‘filtered’ led to the sense that DSD format was like a Chord DAC (my only reference being a Hugo (version 1)) in that it kinda homogenised the sound and created a soundfield with coherency that might not be accurate to the recording.

Swerving off the road and driving head long for the cliff of obscurity overlooking the sea of subjectivity: The Hot 8 Brass band instruments had a ‘fleshed out’ weight in a modified soundfield that could be enjoyed in ways different to the RAW soundfile. Texture was more pronounced but only cause that is what the DSD was most notably bringing..
A brightness is very much lost in the attack of the instruments/leading edge of the note.

On a poor system this ‘brightness’ would be fatiguing vs a great ‘NOS’ setup that would render these leading edges accurately and be ‘non fatiguing playback’. (Albeit a hard setup to create with the respective synergy that all parts need to bring to the table)

The difference is the GTO filter is noticably more honest to the ‘best’ playback of such sounds (I have heard), in that it clearly removes a ‘veil’ and hence feels like the sound can be more present ‘up front’. Placement for both methods was very good, and HUGELY improved by the change in USB cables (more than on the fly conversion or upsampling might bring).


Each time I listen to one format (converted to DSD / PCM GTO’d) I listen for a long while until I forget what I am listening to.
=>
EVERY TIME I CHANGED TO GTO METHOD: I instantly ‘knew’ that it was the better playback. More engaging. More musical. Cleaner and unheard air around the instruments (cause ‘all the air is there’), vs the converted to DSD variant which lost the ‘blackness’ if you will, by having ‘air pockets’/bubbles around the instruments. (Still really good and not describing these air pockets as a negative- most systems would be happy to render this quality of ‘air’)

The DSD is jump over PCM, yes, when the GTO isn’t involved.

Given all recordings have different quality levels/engineering efforts, I would certainly use these two modes to tailor my sound to make a playback session ‘easy to enjoy’.


Caveat: Didn’t use home speakers which would allow an easy soundfield placement /interaction with the perceived 3D space in a room. One method might render a better interaction with 3D (realworld) space and environments.
Under headphones (being closed back Ultrasone Edition 5, or open back AKG K612) the GTO filter was easily the preference towards a ‘higher grade’ of sound quality perception.

—That leading edge delivery I have only heard previously in ‘really great home dedicated DACs’, and for anyone who has spent a long time on this journey, it was evident that this is the Holy Grail /‘summit’ that many are seeking.

To be fair this isn’t a definitive take on DSD conversion: member @BobParish has given some great guides and recommendations as to filters to encode with. My DSD conversion here tested didn’t give filter options, just a few quality modes to choose from. Using home hardware I could highly overclock my PC (watercooling and 8 fans) to achieve a degree of (on the fly) DSD conversion that the world is ‘on the cusp’ of delivering,.. presently a ‘prerender’ to ‘large files’ and DSD 1024 /512 hi quality options are ‘no doubt’ available. The home hardware setup and the cost of that software puts it out of reach of many, and generally not practical for portable ‘head-fi’.


Further observation (relating to gaming/ movie soundfields etc): The GTO filter did not harm the soundstage perception, recede depth, or in any negative way impact on the location of sound sources in the soundstage/environment. It is a joy to utilise the GTO filter when engaging with these applications, happily netting the free benefit of upsampling the sound WELL (something the new firmware that iFi have access to can deliver.)

What became evident: GTO holds its own against portable DSD conversion.


Further anecdotal testing:
sometimes swapping between GTO and DSD offered little perceived difference, the level of subtlety ‘tiny enough’ that unless we were looking for it, probably not going to notice anything.. (again, not tested on ‘room setup with speakers’)

There was perceivable differences (subtle) between 44khz output (with Diablo doing sample rate upconversion=>GTO), and feeding out 352khz with Diablo doing simply the GTO process (which typically does upsampling as well).

From what I was hearing, it seemed that the GTO is an upscaler as well as a filter process (capable of handling ‘pre ringing issues). This is one of the best upscalers for audio I have heard implemented.

I would believe that feeding 44khz over a cheap USB cable would be easier than feeding 352khz, and so would trust in the Diablo doing the upsampling (via GTO).

352 vs DSD- I’d take either, with a preference to the GTO’d 352 depending on music file.
352 vs 44 I’d feed either into the GTO filter’d Diablo (and not care, so would probably use 44khz if ‘garden variety USB cable in chain)

Conclusion
I have NEVER heard upsampling sound sound as great as exceptionally done ‘Non Oversampling’ methods. The GTO is the first upsampling method i am truly infatuated with. (just don’t tell my USB cable or else my coffee tomorrow might be compromised!)


TLR (too long/didn’t read)

Nintendo Switch and Playstation (family) consoles now work with the Diablo (on firmware 7.2 branch, that isn’t officially out yet for the Diablo DAC/amp)

GTO filter is different and ‘awesome’ =>whatever that means to you, your own testing and observation WILL MATTER MORE and I encourage end users to consider the ‘GTO’ firmware variant of the BIOS options when making their Diablo ‘console friendly’.

-edited to fix some parenthesis and correct cable ’anatomy’ in my parts description. not proofread, apols, just a fun writeup for ‘we the people’ (audiofoolz), I prolly could remove a tonne of linebreaks etc.. but left this way to improve readability (big post)

Lovely, thanks! Let me ask how long did it take you to write all this and conduct all your experiments?

 

[Mr BubbaHyde]

G’day,.. (even if it ISN’T day time in your timezone when being here and reading this..)

its a simple salutation that states overtly I am an Aussie.

This shouldn’t matter, but us Aussies (-‘like Canadians’: if stereotypes and observations are believable) are often a patient lot/ very ‘polite’ and generally seen as nice people.
Whilst this shouldn’t matter I give this information to qualify something about myself which might give weight to my words here given...


I really enjoy talking with people who share audio as a hobby or, more so, ‘passion’..

I am here today just to share some quick FYI’s / “info updates” for ‘we’ the people.


For the community that has held me captive (enjoyable so), I have been asked questions and as I now have some perspective to tackle those questions / share an opinion, I felt compelled to be here and share my perspective or ‘give some answers’.

At the end of the day many of us are scientific in our observations (and learning methods) and more info at our disposal has proven historically to be beneficial.

I will steer clear of subjective opinion where possible (in this post) so as to keep it short/succinct and therefore potentially helpful to others..

PURPOSE OF POST
I believe a community stands tall based on the actions of its members’, so I am ‘popping by to say “hi” ‘, and sharing my recent observations with some newfound functionality I have experienced with the iFi 7.2 firmware branch.


WARNING:
updating firmwares not listed for your product is generally considered A VERY BAD IDEA, and exceptionally few manufacturers globally would dare recommend such a practice.


The firmwares I have run, to the best of my knowledge HAVE NOT BEEN DESIGNED specifically for the Diablo, and are likely to not utilise its features fully (whether one firmware might limit processing capability based on potential circuit designs or tune towards a sound in alignment with the sound profile that a SPECIFIC hardware setup might offer (ie: the ‘norm’), iFi appear to have a ‘one size fits all’ design to their firmwares, and from the perspective of using two different XMOS controller chips in the majority of their hardware catalogue, it can be seen that this approach works for them: two different firmware branches for the (two) respective controller chips.

Whether this impacts on giving features like the advertising spiel for the iFi Diablo suggests (things like processing MQA on the toslink input), the audio circuits that iFi licence and implement are NOT all the same, so like internet disclaimers given worldwide: Your Mileage May Vary!


A bad BIOS FLASH can easily brick a device and is generally NOT COVERED BY WARRANTY.
I have worked for businesses that REFUSE to perform BIOS updates for end users due to the very real risks involved. (I actually disagreed with the policy, but that doesn’t change the fact that it existed)
My first PC business started in the early 1990s’, and with that came firmware and BIOS flashing.
I have never killed a product from a bad firmware flash, but have had a couple of instances from (many) hundreds that have been ‘flakey’ and benefited from ‘another flash’.


I will leave my opinion on iFi and firmware ‘at the door’, and now rolling up my sleeves “lets begin”...


FIRMWARES TESTED

7.2 branch - B and C variants

++CH23 has been brilliant and helpful in their contributive efforts and has created tables to help other end users to understand the iFi products better. With help from CH23 (and one of the tables they created) choosing an iFi firmware that interested me was an easy process with clear details as to what each firmware contains. This post will not delve into differences between firmwares (be it 7.0 branch or 7.2 branch, nor the differences between them, other than confirming that 7.2 branch BIOSs’ I tested (B & C variants) both supported the consoles.

B is the evolved version of what we got ‘out of the box’ (is my understanding)- My Diablo was on a version 1.? firmware having never received a firmware flash previously.

C variant implements the GTO filter on the USB input when not feeding DSD or DSD via DoP..

-there might be other aspects to these firmwares but I encourage all interested parties to do their own research and confirm whether a firmware and its ‘features’ are of merit or worth to them - ie I take no responsibility for any others users actions given; even if that is because I convinced them that a $2000 USB cable would create rainbows in their living room- quite simply- I don’t know if a unicorn could even fit in your living room, so my opinion would be unresearched or not tailored to any given use case.

My USB cable makes me coffee in the morning and gets jealous when I look at other USB cables, and I am just thankful that they don’t read my words posted here.
Not all USB cables are created equally, and mine are just plain and simply better than yours’ (oops subjective opinion warning: you might not like COFFEE..)


So what comes with the 7.2 branch of firmwares that might be worth risking turning my beaut red audio box into a doorstop?

To quote the notes (as posted by the legendary community member CH23 who iFi SHOULD be remunerating/hiring); “Playstation console support”.

So, to get this out of the way: Yes; Playstation Console support was tested. As was Nintendo Switch support.

Both consoles PREVIOUSLY DID NOT WORK with the iFi Diablo, and both consoles now very much do.


Implementation is not perfect; but I am not sure how to use my time fruitfully giving feedback to iFi, so will just list some observations here:
________________________________________________________________________________
ISSUE: Diablo needs be powercycled or have USB cable unplugged and then replugged, in order to be seen by the console between console restarts/useage.

Support info: None of my other USB DACs that work with the consoles has required this (the issue is likely with iFi hardware or software (firmware)).

Consumer info: this might be a good thing and a solution might lead to change in the handshake methods between the devices and ‘get in the way’ of optimal sound.. (who knows/who cares)
________________________________________________________________________________
Regarding the above information: I am not trying to be a ‘poopster’ (no clarification with imagery regarding the word ‘poopster’ for those using translators; I don’t use the language needed, in public, to describe the word better).

Feedback (Whether positive or negative in interpretation), is generally ‘a good thing’.
Wouldn’t be doing my part in the discussion if I avoided important facts.

FIRMWARE WITH GTO or WITHOUT?

@justanut- you asked me once for my take on the GTO filter; and I couldn’t say anything to clarify my opinion on it. My bias is that it is an ace in iFis’ sleeve and a very real reason to buy an iFi part.
It is an understatement to confirm “I have been WANTING to test the Gibbs Transient Optimised filter” (for ‘quite awhile’ actually)


Setup and Testing Methodology

running initially the 7.2b firmware- everything checked out/ worked well, so I rolled another 7.2 branch firmware (thankyou to CH23 for making it possible) with the GTO filter.

As the GTO filter only works with the USB input on the Diablo component, I created a setup that allowed a feed into the USB input. To remove bottlenecks/ensure OPTIMAL delivery of the digital code, an ‘audio grade’ USB cable was used AND the iFi Purifier 3 filter technology was used outboard to the ‘noise’ filters/isolators/reducers found internally in the Diablo DAC/amp.


It must be noted that using the USB cable and iPurifier3 proved a very good transport option to the Diablo and is the first time the Diablo has given ‘exceptional’ performance via the USB input (vs high quality transportation via SPDIF toslink input).


Why care about USB quality?

It was easy to hear differences between sampling rates, and the higher than standard quality cable should (theoretically) remove potential bottlenecks that might make 44khz GTO vs 352khz GTO favour the easier data rate transmission.

I have seldom observed USB having the openess of good COAX, so I would say we were operating above excessive jitter or (fast) transient capability reduction.

Hardware:
Using an Android phone (Samsung Fold family), with USB C to USB type A (female) adapter, a Kimber USB cable (type A male to type B male), iPurifier3, then Diablo.
___________________________________________


___________________________________________
So: Phone=>USB adaptor=>USB cable=>iPurifier=>Diablo

Software:
Onkyo HF player (unlocked)


HF Player, when paid for/unlocked, can upsample files to a range of sampling rates and has four tiers of ‘on the fly’ DSD conversion available.
For DSD conversion I chose ‘high precision DSD 5.6Mhz’. This can be output as DSD or DoP.

(The software can also convert DSD to PCM and has equalisation/all sorts of DSD/PCM gain controls/ can do 3 and 6Mhz (etc)... )


Direct Transfer (not DoP) did have ‘the edge’ in sound quality when feeding into the Diablo, but it must be taken with a handful (or three) of grainy salt particulate (ideally Pink) that the subtleties between some of the sampling rates/wrapper output is of no practical concern for the vast majority even in our ‘audiofool’ community where we can often be zealots and PASSIONATE about such things...
The observation of facts and sharing them, if not given the weighting they deserve, can be misleading. To this end, sampling rate differences and wrappings was ‘hey thats different’ (threshold of observation), vs GTO and DSD and USB improvements that were appreciable in their affect


Media: many 16bit 44khz sound files in WAVE/AIFF/FLAC formats (only well engineered and mostly exceptional quality recordings have been audtioned/‘YMMV’)


So- this post is made hugely shorter by avoiding flowerly language and ‘much subjectivity’, I will list a few observations that will no doubt ‘peak’ user interest...

Initial testing was to ensure that the Diablo was decoding correctly and accepting the various output formats. The Android software ‘Onkyo HF player’ is a few years out of date (unsupported), but has been well regarded in specific circles for its ability to bypass the android system mixer and pass unadulterated ‘zeros’ and ‘ones’ (or ‘very adulterated’ 0s and 1s, as I am about to describe..)

I have had a couple of months of using Onkyo HF player to upconvert 16bit 44khz PCM to DSD (Thanks @BobParish for raising awareness that DSD avoids PCM ‘filters’ and therefore gives inherently different sound).

Previously when doing this feed, it has been through an iFi connection cable. That cable is horrible for sound duty when compared to one worth ten times the cost... Sadly for all previous testing of DSD conversion from a phone: that cable has been the only part in play. Hindsight being 20/20 I would have used a better cable and ‘no doubt’ experienced the much better analytical platform that I have been testing this last week...

Previously DSD (upscaled and converted PCM) was different to PCM (raw), and enough so that the DSD feed was consistently better. Always appeared slightly more natural /and in some ways ‘less raw’, but both options didn’t hold a candle to the TOSLINK from a Questyle QP1R..

With the USB cable replacement (and perhaps filter, apols for not testing ‘sans’ iPurifier), the difference is substantial. I would now reach for either audio transport, a dedicated android device for media playback (ie no wifi/bluetooth/cellular ‘on’) or a dedicated DAP (both being decent transports now).

The Android device using standard USB short cable was ‘average’ sound and not involving.

The test setup presently running is of a high performance nature and ‘fitting’ for the DAC being used. (the difference being connection quality /transport of the zeros and ones by changing the physical line used)



I am skating around the elephant in the room: USB transport isn’t ‘black/white. (cables and setup MIGHT ACTUALLY MATTER, but ‘not the point of this post’)


Observations:

DSD sounded great, but GTO was instantly like a great NOS DAC. The leading edge on transients was FAST and coherent and the decay was natural and lively, vs the DSD equivalent which felt like it was almost obscuring the soundstage, and definitely without the true leading edge ‘sharpness’ that great transports are capable of giving.


Remember I am trying NOT to go ‘flowery/subjective’ here. -Switching between GTO and DSD was a night and day difference. Both had been improved by the change in USB transportation method, but the noticeable sound of the ‘pure’ format (PCM passed into Diablo, and operated on with the GTO filter) was fantastic. I would say ‘objectively better’ based on comparing some of the worlds better audio transports (TEAC p700 and Denon DCD-s10) vs lowly massmarket stuff (Rotel RCD971 / Denon DCD1500(? model not in front of me)). When we go up the range in CD players a base model is like the iFi USB cable sound, the mid tier stuff (Rotel RCD971), like a DSD conversion passed over the iFi USB cable, and then any of the digital I have been experiencing via nice USB cable setup is like nice mainstream high quality stuff (CD players north of $1000 ‘back in the day’/cheap dedicated transport), with the GTO filter instantly jumping to the ‘top of the tree’ stuff (esoteric transports and top of the ladder CD players).


Is the GTO filter always better for enjoying the listen? No.

On JC Superstar (australian ed/1992?), the opening overture seemed flatter (by depth of stage) but felt further widened, and this made it more engaging. In this instance the ’air’ advantage went to the DSD process.

As a recent play with a R2R/Ladder DAC revealed, even a lossless reference file via an APT X (not Apt X HD) transmission into a Fiio BTA30 then fed using ‘the worlds cheapest fibre optic cable’ into the DAC- rendered more depth of the ‘very familiar test track’ than any previously experience.. this did make the instruments placed 70ft back in the soundstage seem ‘closer to centre stage’, but the sense of depth does that. -so my trained mind is telling me that the DSD version of the rock orchestra performing Lloyd Webber in this instance, the DSD version, whilst more exciting (under headphones), might not be the best or ‘technically better’. If I was new to the hobby I would hands down give the win to the DSD process as it was viscerally different.

The observation differences between (on the fly converted) DSD and native PCM being GTO ‘filtered’ led to the sense that DSD format was like a Chord DAC (my only reference being a Hugo (version 1)) in that it kinda homogenised the sound and created a soundfield with coherency that might not be accurate to the recording.

Swerving off the road and driving head long for the cliff of obscurity overlooking the sea of subjectivity: The Hot 8 Brass band instruments had a ‘fleshed out’ weight in a modified soundfield that could be enjoyed in ways different to the RAW soundfile. Texture was more pronounced but only cause that is what the DSD was most notably bringing..
A brightness is very much lost in the attack of the instruments/leading edge of the note.

On a poor system this ‘brightness’ would be fatiguing vs a great ‘NOS’ setup that would render these leading edges accurately and be ‘non fatiguing playback’. (Albeit a hard setup to create with the respective synergy that all parts need to bring to the table)

The difference is the GTO filter is noticably more honest to the ‘best’ playback of such sounds (I have heard), in that it clearly removes a ‘veil’ and hence feels like the sound can be more present ‘up front’. Placement for both methods was very good, and HUGELY improved by the change in USB cables (more than on the fly conversion or upsampling might bring).


Each time I listen to one format (converted to DSD / PCM GTO’d) I listen for a long while until I forget what I am listening to.
=>
EVERY TIME I CHANGED TO GTO METHOD: I instantly ‘knew’ that it was the better playback. More engaging. More musical. Cleaner and unheard air around the instruments (cause ‘all the air is there’), vs the converted to DSD variant which lost the ‘blackness’ if you will, by having ‘air pockets’/bubbles around the instruments. (Still really good and not describing these air pockets as a negative- most systems would be happy to render this quality of ‘air’)

The DSD is jump over PCM, yes, when the GTO isn’t involved.

Given all recordings have different quality levels/engineering efforts, I would certainly use these two modes to tailor my sound to make a playback session ‘easy to enjoy’.


Caveat: Didn’t use home speakers which would allow an easy soundfield placement /interaction with the perceived 3D space in a room. One method might render a better interaction with 3D (realworld) space and environments.
Under headphones (being closed back Ultrasone Edition 5, or open back AKG K612) the GTO filter was easily the preference towards a ‘higher grade’ of sound quality perception.

—That leading edge delivery I have only heard previously in ‘really great home dedicated DACs’, and for anyone who has spent a long time on this journey, it was evident that this is the Holy Grail /‘summit’ that many are seeking.

To be fair this isn’t a definitive take on DSD conversion: member @BobParish has given some great guides and recommendations as to filters to encode with. My DSD conversion here tested didn’t give filter options, just a few quality modes to choose from. Using home hardware I could highly overclock my PC (watercooling and 8 fans) to achieve a degree of (on the fly) DSD conversion that the world is ‘on the cusp’ of delivering,.. presently a ‘prerender’ to ‘large files’ and DSD 1024 /512 hi quality options are ‘no doubt’ available. The home hardware setup and the cost of that software puts it out of reach of many, and generally not practical for portable ‘head-fi’.


Further observation (relating to gaming/ movie soundfields etc): The GTO filter did not harm the soundstage perception, recede depth, or in any negative way impact on the location of sound sources in the soundstage/environment. It is a joy to utilise the GTO filter when engaging with these applications, happily netting the free benefit of upsampling the sound WELL (something the new firmware that iFi have access to can deliver.)

What became evident: GTO holds its own against portable DSD conversion.


Further anecdotal testing:
sometimes swapping between GTO and DSD offered little perceived difference, the level of subtlety ‘tiny enough’ that unless we were looking for it, probably not going to notice anything.. (again, not tested on ‘room setup with speakers’)

There was perceivable differences (subtle) between 44khz output (with Diablo doing sample rate upconversion=>GTO), and feeding out 352khz with Diablo doing simply the GTO process (which typically does upsampling as well).

From what I was hearing, it seemed that the GTO is an upscaler as well as a filter process (capable of handling ‘pre ringing issues). This is one of the best upscalers for audio I have heard implemented.

I would believe that feeding 44khz over a cheap USB cable would be easier than feeding 352khz, and so would trust in the Diablo doing the upsampling (via GTO).

352 vs DSD- I’d take either, with a preference to the GTO’d 352 depending on music file.
352 vs 44 I’d feed either into the GTO filter’d Diablo (and not care, so would probably use 44khz if ‘garden variety USB cable in chain)

Conclusion
I have NEVER heard upsampling sound sound as great as exceptionally done ‘Non Oversampling’ methods. The GTO is the first upsampling method i am truly infatuated with. (just don’t tell my USB cable or else my coffee tomorrow might be compromised!)


TLR (too long/didn’t read)

Nintendo Switch and Playstation (family) consoles now work with the Diablo (on firmware 7.2 branch, that isn’t officially out yet for the Diablo DAC/amp)

GTO filter is different and ‘awesome’ =>whatever that means to you, your own testing and observation WILL MATTER MORE and I encourage end users to consider the ‘GTO’ firmware variant of the BIOS options when making their Diablo ‘console friendly’.

-edited to fix some parenthesis and correct cable ’anatomy’ in my parts description. not proofread, apols, just a fun writeup for ‘we the people’ (audiofoolz), I prolly could remove a tonne of linebreaks etc.. but left this way to improve readability (big post)

Wow, ok...

That was something to read right there! Thank you for putting that all together!

Cheers!

 

[mancpg]

Sorry if I've missed something but when I type in the serial number of my Diablo on the iFi website, all I get back is the 7.0 - MQA and 7.0c GTO firmwares, not any of the 7.2 firmwares. Am I doing something wrong?

 

[KaOs492]

I don't believe it's released yet

 

[mancpg]

I don't believe it's released yet

Thanks! I see the ‘not officially’ now in @whitedragem’s post now.

 

[Soundizer]

What does 7.2 add fix change?