Help Shape Fosi Audio's Upcoming K7 DAC AMP with Your User Demands!

[Fosi Audio]

Fosi Audio has always been dedicated to creating new products that cater to user demands. Currently, we are in the process of developing our next desktop dac and headphone amplifier, the K7.
We are extremely grateful for the immense support we have received from ASR members, who have provided us with valuable suggestions. These suggestions have been implemented in our recently released and upcoming products, the BT20A Pro and V3 amplifier. Now, we sincerely request your valuable suggestions again for our upcoming product, the K7 dac amp.

Functions and Features Planned to Be Implemented
1. AK4493S decoder chip;
2. USB, coaxial and optical digital decoding;
3. 3.5mm single-ended and 4.4mm balanced headphone jacks;
4. More than 1W output power;
5. DSD512, PCM 768kHz sampling rate;
6. Microphone input with adjustable gain;
7. Treble & bass tone controls;
8. Bluetooth codecs supported: LDAC, APTX adaptive.

Please share your thoughts and suggestions about this dac amp.

 

[RangerAudio]

That design is really good. I'm pretty impressed. It's a total departure from most of your designs. Nice direction, could you also do it in black and add a XLR output to the front as well? Maybe change the 4.4 mm to XLR or just add one. Very cool, I love the design so far.

 

[Fosi Audio]

That design is really good. I'm pretty impressed. It's a total departure from most of your designs. Nice direction, could you also do it in black and add a XLR output to the front as well? Maybe change the 4.4 mm to XLR or just add one. Very cool, I love the design so far.

Really appreciate your suggestion. We'll consider adding the XLR.

 

[solderdude]

Many headphones still use 6.3mm jack, perhaps add that (XLR-4 is too bulky).
Is the Pentaconn going to be balanced (differential) or just for convenience ?
Tone control or Graphic EQ, if the latter don't bother with pointless presets.
What does 'more than 1 Watt' mean ? 1.1Watt is more than 1W yet insignificant.
1W in 16ohm requires different circuitry compared to 1W in 300ohm.
Going to cater for an impedance range ?

 

[Jeromeof]

Looks very interesting, not sure you need the Bass Treble + Mic 'buttons' around the "Knobs" - I imagine you could have a cleaner / symmetric look with a similar size button to the volume button on that side and use the display to allow configuration of the bass / treble etc with a 'press' of the knob to set a value on the display etc

 

[Fosi Audio]

Many headphones still use 6.3mm jack, perhaps add that (XLR-4 is too bulky).
Is the Pentaconn going to be balanced (differential) or just for convenience ?
Tone control or Graphic EQ, if the latter don't bother with pointless presets.
What does 'more than 1 Watt' mean ? 1.1Watt is more than 1W yet insignificant.
1W in 16ohm requires different circuitry compared to 1W in 300ohm.
Going to cater for an impedance range ?

Probably cater to an impedance range of 16-300Ω. Max output of more than 1W into 16Ω.

 

[solderdude]

That will be > 50mW in 300ohm (when not current limited in 16ohm)

 

[Tom C]

Does this thread belong in Desparate Dealers forum?

 

[Fosi Audio]

Looks very interesting, not sure you need the Bass Treble + Mic 'buttons' around the "Knobs" - I imagine you could have a cleaner / symmetric look with a similar size button to the volume button on that side and use the display to allow configuration of the bass / treble etc with a 'press' of the knob to set a value on the display etc

Thanks for your suggestions.

 

[restorer-john]

Does this thread belong in Desparate Dealers forum?

Absolutely.

 

[abdo123]

Full fledged DSP with 10-20 PEQ filters or don't even bother releasing it.

 

[LearningToSmile]

Maybe not for this product if it's hard to add at this stage of development, but one feature to consider in the future that still seems to be missing outside of power amplifiers is subwoofer output.
I've checked your website and saw you already make quite a variety of power amplifiers that include subwoofer output, so I think you should have the know-how for it already.
Either a DAC, preamp, or a combo unit like this one would be good - with a defeatable high pass for mains and low pass for sub out, either set at 80Hz or adjustable in software. Not a fan of physical controls like in some of your power amplifiers, but if they were precise and well labelled that's fine too.
This is a niche that's really quite unfulfilled, as anyone wanting to integrate a sub with their powered monitors either has to rely on subwoofers with appropriate pass through, or invest in expensive multi-channel DACs.

 

[Joey]

Maybe not for this product if it's hard to add at this stage of development, but one feature to consider in the future that still seems to be missing outside of power amplifiers is subwoofer output.
I've checked your website and saw you already make quite a variety of power amplifiers that include subwoofer output, so I think you should have the know-how for it already.
Either a DAC, preamp, or a combo unit like this one would be good - with a defeatable high pass for mains and low pass for sub out, either set at 80Hz or adjustable in software. Not a fan of physical controls like in some of your power amplifiers, but if they were precise and well labelled that's fine too.
This is a niche that's really quite unfulfilled, as anyone wanting to integrate a sub with their powered monitors either has to rely on subwoofers with appropriate pass through, or invest in expensive multi-channel DACs.

active subwoofer output?

 

[charleski]

Full fledged DSP with 10-20 PEQ filters or don't even bother releasing it.

I agree that this would increase its desirability immensely. The problem, however, is the UI implementation (and I think this is why we don't see this happen). Each band of the PEQ has three variables to adjust (f,Q and gain), setting this with a simple 5-way controller is fiddly and it requires the addition of a microcontroller capable of running the UI and calculating and applying the overall filter coefficients. What would be ideal would be a system that used a PC program to run the UI and calculated impulse responses that could be uploaded to the device which would then just require a fairly simple convolver in its digital chain. This could give advanced users a great deal of DSP flexibility, but it would require a fair amount of programming expertise for an audio company.

 

[IAtaman]

I agree that this would increase its desirability immensely. The problem, however, is the UI implementation (and I think this is why we don't see this happen). Each band of the PEQ has three variables to adjust (f,Q and gain), setting this with a simple 5-way controller is fiddly and it requires the addition of a microcontroller capable of running the UI and calculating and applying the overall filter coefficients. What would be ideal would be a system that used a PC program to run the UI and calculated impulse responses that could be uploaded to the device which would then just require a fairly simple convolver in its digital chain. This could give advanced users a great deal of DSP flexibility, but it would require a fair amount of programming expertise for an audio company.

Well, not necessarily. I am an amateur and it took me less than a week to put together a simple EQ UI for CamillaDSP. I am sure professionals can do a much better job and do it much faster and you can always outsource is.

CamillaNode : A web based UI alternative to control CamillaDSP as PEQ

Hey All, TL;DR: camillaNode is an open source, free to use CamillaDSP control interface specifically built to function as a parametric EQ with AutoEQ integration, and is available on GitHub via the link here. I was looking for alternative UI to control CamillaDSP as a simple equalizer for...

www.audiosciencereview.com

 

[charleski]

Well, not necessarily. I am an amateur and it took me less than a week to put together a simple EQ UI for CamillaDSP. I am sure professionals can do a much better job and do it much faster and you can always outsource is.

CamillaNode : A web based UI alternative to control CamillaDSP as PEQ

Hey All, TL;DR: camillaNode is an open source, free to use CamillaDSP control interface specifically built to function as a parametric EQ with AutoEQ integration, and is available on GitHub via the link here. I was looking for alternative UI to control CamillaDSP as a simple equalizer for...

www.audiosciencereview.com

That requires a device that already has a full http stack implemented. They could just add a PiZero into the box, though this would still need some expertise in terms of integration.

 

[IAtaman]

That requires a device that already has a full http stack implemented. They could just add a PiZero into the box, though this would still need some expertise in terms of integration.

True. But not necessarily. Depending on the solution, all the HTTP stack can be moved to client side. I will put together a Firefox / Chrome Add-on version of CamillaNode soon that will not require Node.js. But indeed, you need a system capable of running a Linux distro. Two core Arm CPU and 256Mb RAM is more than enough for a real time DSP and a websocket server such as CamillaDSP, or even the full stack HTTP server. I am using Rock Pi S for the exact same solution, it has quad core a35 CPU, utilization is ever above 15%. It costs $27 retail, delivered from China with ethernet and other stuff. $15 as a custom solution tops. USB IN / USB Out is already there, some work might be needed for Coax and Optical but the h/w might already be available. Overall, not a gigantic investment is required to add a DSP solution to these devices in my opinion.

 

[charleski]

True. But not necessarily. Depending on the solution, all the HTTP stack can be moved to client side. I will put together a Firefox / Chrome Add-on version on CamillaNode soon that will not require Node.js. But indeed, you need a system capable of running a Linux distro. Two core Arm CPU and 256Mb RAM is more than enough for a real time DSP and a websocket server such as CamillaDSP, or even the full stack HTTP server. I am using Rock Pi S for the exact same solution, it has quad core a35 CPU, utilization is ever above 15%. It costs $27 retail, delivered from China with ethernet and other stuff. $15 as a custom solution tops. USB IN / USB Out is already there, some work might be needed for Coax and Optical but the h/w might already be available. Overall, not a gigantic investment is required to add a DSP solution to these devices in my opinion.

Yes, there are lots of microcontrollers that already have a range of decent open source software. But as I said, the problem is integration. It needs to be a turnkey zero-maintenance solution (you can't tell users to SSH in to fix things, and shipping with a buggy firmware is just going to sink the product), and that means finding and hiring someone with experience in this area. They could outsource this, but then there's a risk that after paying for a lot of design work some competitor brings out a product with strangely similar function, so it would be best to keep this in-house.

It's certainly very possible, and I'd really like to see DAC makers branch out in this direction. But there are significant hurdles that would need to be overcome to bring a viable product to market. Writing the PC-side web app needed to control the DSP is really the easiest part of the problem.

 

[kemmler3D]

It would be really great if there was user-configurable PEQ on board. I know from my time in the industry that editable PEQ on DSP chips is common, but the UI / UX is often not so good. So I guess the task would be to make a decent app/interface for updating the EQ over USB. 10+ bands would be great.

More SINAD doesn't really help anyone in today's market. But today there's often a struggle to get EQ into the signal chain without using a PC or dedicated DSP unit. So I think that would be a big deal.

What would be ideal would be a system that used a PC program to run the UI and calculated impulse responses that could be uploaded to the device which would then just require a fairly simple convolver in its digital chain. This could give advanced users a great deal of DSP flexibility, but it would require a fair amount of programming expertise for an audio company.

IME these things already exist for many DSP chips, even cheap ones that come with $50 BT headphones. The problem is that the interface of these EQ editors is very basic and designed for Chinese engineers to use and nobody else. (ask me how I know this.) So the task isn't even to create a new app for editing PEQ, but to convince the vendor of the existing app to redevelop the UI and translate it into other languages. This is also not trivial, but it's more of a business development task than software development

 

[Hatto]

Really appreciate your suggestion. We'll consider adding the XLR.

You can also consider dual TRS connectors for balanced output instead of XLR in favor of space and aesthetics.