Okto dac8 stereo DAC Review

[thats_it]

Weird that Okto maunf is so secretive on internals (DAC chip, op-amps, architecture). Bring up the Amazon or Ali page on just about any Chinese D/A (or similar product) manuf -- Fiio, Topping, etc -- and the "flyer" given you the whole spiel, colorful pictures and graphs to boot ... more than you ever wanted to know.

I've read (head-fi.org) that this Okto uses the ESS ES9028Pro chip. A bit dated now. The 9038Pro or 9038Q2M are newer.

 

[Veri]

I've read (head-fi.org) that this Okto uses the ESS ES9028Pro chip. A bit dated now. The 9038Pro or 9038Q2M are newer.

The 9038Q2M is leagues worse than the 9028Pro desktop chip, so no that would not do. Also this is the absolute best specced/measured device on this whole site including lots of 9038Pro chip-based devices. Newer does not mean better. The Okto Stereo is just about on the theoretical maximum performance levels a DAC can go.

 

[hampestampe]

Weird that Okto maunf is so secretive on internals (DAC chip, op-amps, architecture). Bring up the Amazon or Ali page on just about any Chinese D/A (or similar product) manuf -- Fiio, Topping, etc -- and the "flyer" given you the whole spiel, colorful pictures and graphs to boot ... more than you ever wanted to know.

I've read (head-fi.org) that this Okto uses the ESS ES9028Pro chip. A bit dated now. The 9038Pro or 9038Q2M are newer.

But i does not make sense to develop a new product every time a new dac chip is released, especially not for a small company. The latest version is not necessarily the smartest choice especially if you already have got a great class-leading product. The risk of implementing a new chip is greater than the benefits, I think. Is there any numbers in the data-sheet, you think would make the product better, with the 9038 pro chip?

 

[John1959]

But i does not make sense to develop a new product every time a new dac chip is released, especially not for a small company. The latest version is not necessarily the smartest choice especially if you already have got a great class-leading product. The risk of implementing a new chip is greater than the benefits, I think. Is there any numbers in the data-sheet, you think would make the product better, with the 9038 pro chip?

I read somewhere (don't remember exactly where) that Pavel himself mentioned the 9038Pro theoretically could have some slight benefits at some point and he would look into it. But he also mentioned some drawbacks. Had something to do with more current and noise I believe. I am sorry that I can't more precise. Anyway, it seems doubtful to expect a significant improvement in this particular application by using the more recent 9038Pro instead of the already excellent 9028Pro

 

[hampestampe]

I read somewhere (don't remember exactly where) that Pavel himself mentioned the 9038Pro theoretically could have some slight benefits at some point and he would look into it. But he also mentioned some drawbacks. Had something to do with more current and noise I believe. I am sorry that I can't more precise. Anyway, it seems doubtful to expect a significant improvement in this particular application by using the more recent 9038Pro instead of the already excellent 9028Pro

Okto Research /Pavel gave the aswer in post #313 in this thread, he wrote like this
Quoute:
"ES9038PRO is the same chip with different output scaling. Simply put: much more output current. That allows for a higher dynamic range, but it can also induce added distortion. One of the reasons is that the physical dimensions of the circuit package are the same for both, including thickness of the leads and especially bonding wires between the leads and the chip itself. More current on a given resistance equals more voltage drop, hence a more significant source of various forms of distortion. Same applies to a PCB. There might be some benefit that we are going to explore. But as mentioned in my previous post #225, even a better real-world result would not lead to a better measured value using the APx555, no matter whether the DAC has an actual SINAD of 126dB or 130dB. Indirect and complex methods would need to be used as mentioned by @JohnYang1997"

 

[thats_it]

Per the RMAF 2011 talk by Martin Malinson, ESS does audiophile-ear test their products (as Mallinson notes in the video, the ESS ceo of that time could discern SUBJECTIVE differences that were NOT measurable ... but that, as he claims, took the ceo training to learn. Mallnson himself notes that he CANNOT audiophile-ear test the dacs he designs.)

RMAF 2011 Noise Shaping Sigma-Delta Based Dacs, Martin Mallison, CTO, ESS Technology

 

[thats_it]

The 9038Q2M is leagues worse than the 9028Pro...

Not sure about that ... I've seen the 9038Q2M in > $1,000 DACs. I hardly ever see the 9028PRO being used ... maybe 'cause that thing is -- as I noted earlier -- a now-OLD device.

 

[Veri]

Not sure about that ... I've seen the 9038Q2M in > $1,000 DACs. I hardly ever see the 9028PRO being used ... maybe 'cause that thing is -- as I noted earlier -- a now-OLD device.

The Q2M series are mobile class chips with limited computing power compared to the Pro desktop-class ones. At least in the oversampling filter you can see that all 9038Q2M implementations have worse THD+N vs frequency ratio performance, because they can't do any better there. The 9028Pro vs 9038Pro is another discussion but I believe they both have their pros and cons..

 

[HiFidFan]

Not sure about that ... I've seen the 9038Q2M in > $1,000 DACs. I hardly ever see the 9028PRO being used ... maybe 'cause that thing is -- as I noted earlier -- a now-OLD device.

As if price is an indicator of performance

As others have mentioned, the 9038Q2M has some trade-offs. It is not the better/perfect chip.

 

[mp006ltk]

And it still says, “Our current lead time is 6-8 weeks” on their website. Very misleading to a new customer potentially unaware of the real time scales...

If you don't want to wait I have one for sale on USAM.

 

[mp006ltk]

The Q2M series are mobile class chips with limited computing power compared to the Pro desktop-class ones. At least in the oversampling filter you can see that all 9038Q2M implementations have worse THD+N vs frequency ratio performance, because they can't do any better there. The 9028Pro vs 9038Pro is another discussion but I believe they both have their pros and cons..

In all my years in this hobby, it would seem that the analog side of a DAC makes a much bigger difference in sound then the DAC chip itself. The chip in my current DAC (Meridian Ultra) isn't even new. It's a 2 channel 24/192 AD chip per side with a fancy mux chip to leverage both channels driven by some serious DSP processors. And all I can say is that it works. The sound is eargasmic.

 

[thats_it]

In all my years in this hobby, it would seem that the analog side of a DAC makes a much bigger difference in sound then the DAC chip itself. The chip in my current DAC (Meridian Ultra) isn't even new. It's a 2 channel 24/192 AD chip per side with a fancy mux chip to leverage both channels driven by some serious DSP processors. And all I can say is that it works. The sound is eargasmic.

My personal reference digital-playback device is a 1989 Philips CD player-- much diy modded -- with a TDA1541A Double Crown dac and the SAA7220B (that's "B" version) digital filter ... many, many mods including a self-designed, discrete cascoded I/V output stage. It doesn't measure terribly well (tho' Naim did push the measurements aspect of the classic Philips chipset to quite a respectable degree in the late 90s) ... but it sure sounds great.

 

[thats_it]

As others have mentioned, the 9038Q2M has some trade-offs. It is not the better/perfect chip.

Actually, toots, you just PROVED that it is the "better/perfect" chip because of the "trade-offs". It's very small and energy efficient (great for battery/portables) ... it is relatively cheap ... it incorporates ALL of the latest ESS Sabre topology (including internal programing) ... and (lucky for ESS), the AKM competition is significantly curtailed 'cause someone probably wasn't careful with their Marlboro.

 

[HiFidFan]

Actually, toots, you just PROVED that it is the "better/perfect" chip because of the "trade-offs". It's very small and energy efficient (great for battery/portables) ... it is relatively cheap ... it incorporates ALL of the latest ESS Sabre topology (including internal programing) ... and (lucky for ESS), the AKM competition is significantly curtailed 'cause someone probably wasn't careful with their Marlboro.

I did? Glad I could help. Do I win something?

 

[Veri]

 

[phoenixdogfan]

So I'm now at the nine week anniversary of my order placement. My number is 291. Anyone have any idea of how much longer a wait?

 

[DWPress]

So I'm now at the nine week anniversary of my order placement. My number is 291. Anyone have any idea of how much longer a wait?

Two weeks.

Seriously, mine was 222 and I ordered it Oct 2nd. Just cleared US customs yesterday. In between when I ordered and now they've moved factory locations, Covid, Covid, and the last batch of faceplates had bad anodizing which they refused. Did I mention Covid? Hopefully you will find it worth the wait like I have.

 

[HiFidFan]

So I'm now at the nine week anniversary of my order placement. My number is 291. Anyone have any idea of how much longer a wait?

Have you reached out to Okto?

 

[ColonelJoy]

Well, I guess I'm extra happy I went for the used Okto that popped up, as it arrived today, so I can't speak to expectations on ordering a new one.

Previously I streamed Qobuz from a node 2i into an excellent Hegel H590 Dac which sends nice power through Silversmith Fidelium speaker cable to Magnepan 1.7i...I was not anticipating this level of refinement utilizing the Okto via Volumio controlling the Oktos rasberry pi streamer---refinement concerning detail and tone...Kind of Blue sounds like a new recording all together...what a hifi product.

 

[phoenixdogfan]

Two weeks.

Seriously, mine was 222 and I ordered it Oct 2nd. Just cleared US customs yesterday. In between when I ordered and now they've moved factory locations, Covid, Covid, and the last batch of faceplates had bad anodizing which they refused. Did I mention Covid? Hopefully you will find it worth the wait like I have.

So you're telling me 3-4 more months? Whatever happened to 19 weeks?