[Joe Bloggs]

Hi all,

I'm here to re-introduce the HiBy FC6 to the head-fi community.
https://store.hiby.com/products/hiby-fc6

First introduced early this year, it immediately found favour from AndyEF of DongleMadness, making its way to #1 of his top dongles list over hundreds of contenders in one swell swoop. It was a decision based pretty much on sonics alone, as the price of the unit is unremarkable, and features such as high drive power or balanced output were lacking. It was later also found somewhat lacking by another reviewer in terms of e.g. frequency response flatness measurements,

but I have isolated the reason why the FC6 was so preferred by AndyEF among others subjectively, and it has something to do with a measurement result that was generally objected to. These factors have been distilled and all instilled into the new Darwin Ultra filter introduced at the turn of this week. It contains all the factors that make a filter (or, indeed, an NOS system) subjectively preferable (for better or for worse in terms of measurements), while functioning fully as an oversampling AA filter, complete with all the advantages that brings to fidelity. In short, it's going to be a filter that sounds as good as can be for analog-lovers while measuring as well as that permits.

But before that, we ran into other practical problems with the FC6, in the form of USB receiver incompatibility with some smartphones (where it was determined that the smartphones in question didn't actually adhere to the USB audio standard fully, but guess what, if you point fingers at Qualcomm, you'd get them turned back to point at yourself after every bone on your fingers have been broken, right?). Also some problems with the HDR module for OS mode on a certain member's earphones. (HDR improves dynamic range / accuracy at low signal levels at audible frequencies, but sometimes at the cost of extra ultrasonics, which may become audible through IMD depending on the earphones).

Herewith I would like to introduce the release of the first firmware update aimed at addressing all of the technical issues given.

https://www.dropbox.com/scl/fi/rpvs...date.zip?rlkey=qicpr2z2jiiszb6l63wgpqgyw&dl=0
(update: I've been told that an older version of the DFU installer actually works better. I have repackaged the firmware with that version as an option below. The firmware itself, is of course the same.)
https://www.dropbox.com/scl/fi/btbs...ller.zip?rlkey=lmm6wpzoowhbhomtfpne4ff5q&dl=0

Changelog:
0. Updated installer settings for DFU and flash tools. DFU now defaults to English, flash tool now opens with all files ready to go, no need to dig in folders.
1. Fixed issue where click-sounds occurred at intervals when using FC6 in non-exclusive USB mode on Snapdragon 8 smartphones.
2. Providing on/off switch for HDR (which was fixed to on for OS in previous firmware), which may be used in both NOS and OS modes for maximum flexibility. Default setting (which are given every time the OS / NOS is toggled) is off for NOS, on for OS. (this should fix @povidlo 's problems)
3. OS filters have been updated to v1.6beta (so not quite v2 yet, Darwin Ultra hasn't made it onto the list yet. But that's ok, because Darwin Ultra was modelled on several aspects of how the FC6 sounded by itself anyway Makes for a bit of a practical problem in terms of fully labelling a next version FC6 firmware fully v2 certified actually)

Let me know how it goes. I will be closely monitoring this thread as well as pm's for any issues both with the installer (which I have optimized) and with the firmware itself.

With any luck, the FC6 with this and the next firmware (which should come soon; the full v2 filters have already been adapted for the FC6, just waiting for some new programming which should be quite simple) will fully shake off its teething bugs regarding usability and simply sound ******** awesome (if measuring with some quirks which should optionally be eliminated in the next release), period.

 

[XERO1]

 

[Joe Bloggs]

Unfortunately, Hiby still decides to keep the big treble roll-off for whatever reason.

Huh… that roll off looks like a poorly implemented filter (you typically see this roll off but above 20k)…

Yes, well, it's the analog filter, not the Darwin v2 (or V-anything) responsible for this. But these measurements (and the associated listening impressions published) actually formed the basis of the published Darwin Ultra filter in Darwin v2, which is being very well received. Note that frequency response that isn't flat is not in and of itself distortion. In any event, this rolloff will optionally be defeated in the next firmware. (and I'll bet dollars to donuts it will sound worse to professed chasers of analog / R2R sound that way)

 

[povidlo]

@Joe Bloggs new firmware works well so far, thanks.

I like it. Seems like the sound improved slightly, especially in terms of imaging and clarity. Still euphonic and analogue yet dynamic and engaging.

I just had a few follow up questions.
-During MQA playback, HDR icon is not displayed. Is this a display bug or is HDR not supported with MQA? If so, any other limitations during MQA playback?

-Since you mentioned Darwin Ultra is somewhat similar to how FC6 sounds already, which of the current version 1.6 filters would be closest to Ultra?

-Does NOS support filters, or are they only in play when OS is selected?

 

[Joe Bloggs]

1. HDR is not supported / needed for MQA. Indeed, strictly speaking, the icon should be off anytime material is being played back at 7xx kHz.
2. It's hard for me to say. They all share the basic characteristics that I chose as part of Darwin Ultra. A further update is coming up which should incorporate an actual Darwin Ultra preset for the FC6 as well.
3. The filters, are AA filters for use with oversampling, to filter out the ultrasonics generated by the conversion process. They run in the digital domain, in the oversampled intermediate format. Since there is not oversampling in NOS mode, there is no oversampled intermediate format for digital filters to run on. That's not to say that the issues presented by the conversion process do not exist in NOS, however. (they exist for anyone and everyone, every model / brand)

 

[povidlo]

1. HDR is not supported / needed for MQA. Indeed, strictly speaking, the icon should be off anytime material is being played back at 7xx kHz.
2. It's hard for me to say. They all share the basic characteristics that I chose as part of Darwin Ultra. A further update is coming up which should incorporate an actual Darwin Ultra preset for the FC6 as well.
3. The filters, are AA filters for use with oversampling, to filter out the ultrasonics generated by the conversion process. They run in the digital domain, in the oversampled intermediate format. Since there is not oversampling in NOS mode, there is no oversampled intermediate format for digital filters to run on. That's not to say that the issues presented by the conversion process do not exist in NOS, however. (they exist for anyone and everyone, every model / brand)

Thanks for reply.

For 3., I know that filters are usually applicable to OS only. However, it was mentioned in your general post regarding new filters that NOS will now support three different filters as well somehow? I guess that's coming with v2 firmware or am I missing something?

Any ETA for v2 roll out for fc6?

 

[Joe Bloggs]

Thanks for reply.

For 3., I know that filters are usually applicable to OS only. However, it was mentioned in your general post regarding new filters that NOS will now support three different filters as well somehow? I guess that's coming with v2 firmware or am I missing something?

Any ETA for v2 roll out for fc6?

It was mentioned in the general posting, that in addition to the actual NOS filter in (RS6 / RS8 / RS2 / FC6), the RS6 / 8 would have filters in its filterset that implement other NOS implementations. This is done by programming the oversampling engine in a novel manner such that it is functionally equivalent to NOS. Simply put, OS filters "look" at a grid of several hundred samples around the current sample to determine the sound of the current sample (for purposes of the antialiasing filtering, not really to change the sound audibly, but it's hard to explain), whereas NOS implementations only "look" at the previous sample (one implementation), sometimes also the next sample (another implementation), but sometimes not even looking at any sample other than the current sample (3rd implementation) in determining the final sound. You can do this in hardware (which the NOS switch does), but the OS engine can also be made to act just like these NOS implementations, by simply telling them to narrow down their "view" of samples to the 1, 2 or 3 samples right now.

Technical "difficulty", is that the FC6 has only so much room in its UI for filters, otherwise it gets cumbersome to navigate. And it already has the NOS filter in hardware, as well as hardware characteristics that make the other filters more "NOS-like" without me even trying with the actual programming. Our current plan, is again to aim for more differentiation between the filters, in that the next version of fast/slow normal/late rolloff filters will pursue a more conventional "standard" sound, leaving NOS to be NOS, and Darwin Ultra will be doing its own thing. There would remain 5 filters in the next version FC6 filter set, and none of them will be the NOS-OS filters, as I fear that including 3 NOS options for filters when you turn on the OS switch, would simply be too confusing for most people.

 

[onoff]

How it looks v2 roll out for fc6?

 

[Joe Bloggs]

How it looks v2 roll out for fc6?

The filters have been finalized, however some reprogramming of the device is needed around them. "Soon".

 

[sergibarroso]

To me, the worst drawback of the FC6 still how sensible it is to cellular interference. This, to me, is a deal breaker because, most of the time, I use it with my phone. And I guess this is a hardware design issue

 

[Joe Bloggs]

To me, the worst drawback of the FC6 still how sensible it is to cellular interference. This, to me, is a deal breaker because, most of the time, I use it with my phone. And I guess this is a hardware design issue

From the models of earphone etc. given in pm, this is not a normal issue, you're welcome to discuss with your seller (which could be us, if you bought on store.hiby.com) about sending this in for repair / replacement.

 

[onoff]

Tomorrow is a month since your last post, new blood to FC6 would be nice as a new fw, can't wait...thanks for your efforts

 

[Nivalof]

It was mentioned in the general posting, that in addition to the actual NOS filter in (RS6 / RS8 / RS2 / FC6), the RS6 / 8 would have filters in its filterset that implement other NOS implementations. This is done by programming the oversampling engine in a novel manner such that it is functionally equivalent to NOS. Simply put, OS filters "look" at a grid of several hundred samples around the current sample to determine the sound of the current sample (for purposes of the antialiasing filtering, not really to change the sound audibly, but it's hard to explain), whereas NOS implementations only "look" at the previous sample (one implementation), sometimes also the next sample (another implementation), but sometimes not even looking at any sample other than the current sample (3rd implementation) in determining the final sound. You can do this in hardware (which the NOS switch does), but the OS engine can also be made to act just like these NOS implementations, by simply telling them to narrow down their "view" of samples to the 1, 2 or 3 samples right now.

Technical "difficulty", is that the FC6 has only so much room in its UI for filters, otherwise it gets cumbersome to navigate. And it already has the NOS filter in hardware, as well as hardware characteristics that make the other filters more "NOS-like" without me even trying with the actual programming. Our current plan, is again to aim for more differentiation between the filters, in that the next version of fast/slow normal/late rolloff filters will pursue a more conventional "standard" sound, leaving NOS to be NOS, and Darwin Ultra will be doing its own thing. There would remain 5 filters in the next version FC6 filter set, and none of them will be the NOS-OS filters, as I fear that including 3 NOS options for filters when you turn on the OS switch, would simply be too confusing for most people.

Hello Joe.
What exactly is the difficulty or roadblock in the case of adding additional NOS "filters"?
We have two modes OS and NOS and there are several sub-modes/"filters" for each. I honestly don't see what the user could be confused with here.
It's just that unless new NOS "filters" are added, it seems like a conscious downgrade to the rest of the r2r lineup. I'd like to get full functionality for the asking price, which is one of the highest for dongles.

And how are things going with the firmware update?

p.s.
English is not my first language. I use a translator. So please do not read my message with a negative tone. I do not have any claims to you and I am not attacking you. Just showing interest.

 

[Joe Bloggs]

Hello Joe.
What exactly is the difficulty or roadblock in the case of adding additional NOS "filters"?
We have two modes OS and NOS and there are several sub-modes/"filters" for each. I honestly don't see what the user could be confused with here.
It's just that unless new NOS "filters" are added, it seems like a conscious downgrade to the rest of the r2r lineup. I'd like to get full functionality for the asking price, which is one of the highest for dongles.

And how are things going with the firmware update?

p.s.
English is not my first language. I use a translator. So please do not read my message with a negative tone. I do not have any claims to you and I am not attacking you. Just showing interest.

Hi Nivalof,

Thanks for your interest.

Firstly, please understand that there are no "NOS filters" here, you can select between oversampling filters, and you can also turn on or off NOS, which bypasses any filter (for all the good and bad that implies).

The idea was to provide new filters that neutralize the rolloff of the FC6 towards 22kHz, as well as a new Darwin Ultra filter that provides the desired characteristics of NOS (insofar as they are desired) without the undesirable characteristic of not having a filter.

The technical holdup is that these filters would measure badly in a swept sine test of frequency response and distortion, at or near full signal swing.

Whether this would have any impact on actual music listening, is not yet clear to us. It shouldn't, as the region where the swept sine test of these filters would perform badly, is at over 10kHz, where (you'd hope! ) you never encounter music signals with these frequencies at anywhere near full swing.

I will be happy to get the team to upload test firmware of these new filters here, in the next few days, to gather feedback and then see about putting them up officially.

 

[Nivalof]

Hi Nivalof,

Thanks for your interest.

Firstly, please understand that there are no "NOS filters" here, you can select between oversampling filters, and you can also turn on or off NOS, which bypasses any filter (for all the good and bad that implies).

The idea was to provide new filters that neutralize the rolloff of the FC6 towards 22kHz, as well as a new Darwin Ultra filter that provides the desired characteristics of NOS (insofar as they are desired) without the undesirable characteristic of not having a filter.

The technical holdup is that these filters would measure badly in a swept sine test of frequency response and distortion, at or near full signal swing.

Whether this would have any impact on actual music listening, is not yet clear to us. It shouldn't, as the region where the swept sine test of these filters would perform badly, is at over 10kHz, where (you'd hope! ) you never encounter music signals with these frequencies at anywhere near full swing.

I will be happy to get the team to upload test firmware of these new filters here, in the next few days, to gather feedback and then see about putting them up officially.

Thanks for the detailed answer!
So it turns out that the NOS will remain as it is, and the Darwin Ultra filter is designed for OS and has the advantages of the NOS mode, while not having the disadvantages of this mode, am I correct?
And as you wrote above, again, if I understand correctly, there is no point in including Darwin Ultra in the FC, as it is already based on the current FC position, am I correct?