[skunark]

I was under the impression that Firewire had some realtime latency spec in it (perhaps for video). Wrong?

It's still considered acceptable to drop a frame in FW, so that's still not audiophile quality.

[JohnR]

It's still considered acceptable to drop a frame in FW, so that's still not audiophile quality.

Ah. Didn't know that, thank you.

[firedog]

https://sites.google.com/site/hifacemods/home/hiface-dac

The JKDAC has the following characteristics:

It's a modified I2S Hiface capable of 24/192
integrated with a 24/192 Sabre ES9022 DAC
operates exclusively from 2 internal LiFePo4 (Lithium Nano Phosphate) batteries
completely portable - will run for >8 hours from fully charged batteries
Only USB input - standard USB type B input socket
Only RCA analogue outs
Outputs at 2V line level for driving preamplifier or amplifier
RCA outputs will happily also drive most headphones
external battery charger plugs into box
aluminium enclosure (Size in mm L:120 W:78 H:43)
Price €500

A steal if it is any good

[jamesg11]

Yeh, thinking about getting the non-dac mk3 - small thing, but note that hiface won't allow integer mode, if using Audirvana.

[Noseyears]

Guys, you forgot to mention the upcoming Burson Dac...i think it will be priced for less than 1k. (not sure)

I only found a spanish link though.
http://www.hifilive.es/2011/06/10/burson-audio-da-160/

(google translate helps  )

[ted_b]

Guys, you forgot to mention the upcoming Burson Dac...i think it will be priced for less than 1k. (not sure)

I only found a spanish link though.
http://www.hifilive.es/2011/06/10/burson-audio-da-160/

(google translate helps  )

Srajan has reviewed its DAC/preamp brother (Same DAC, etc) and given it a Blue Moon award in the past 18 months.  And he has already previewed and explained the differences here. 
http://www.6moons.com/audioreviews/burson6/1.html
It just didn;t seem "new news" enough for me, sorry.  Great value DAC.

[wilsynet]

It's still considered acceptable to drop a frame in FW, so that's still not audiophile quality.

I'm not aware of any comparable real time streaming protocols that can fully compensate for a lost block or frame.

Using SPDIF, a DAC on losing a block does not ask the CD transport to re-transmit.  If a frame is lost in HDMI, your television does not ask the cable box to send again.

I'm wondering which protocols and transport mechanisms meet audiophile standards.

[skunark]

I'm not aware of any comparable real time streaming protocols that can fully compensate for a lost block or frame.

Using SPDIF, a DAC on losing a block does not ask the CD transport to re-transmit.  If a frame is lost in HDMI, your television does not ask the cable box to send again.

I'm wondering which protocols and transport mechanisms meet audiophile standards.

I2S would be ideal.   SPDIF packet size is 48 bytes, but any clock locking issues you would lose a bit since it is serial, the debate would be is losing a bit more noticeable than jitter...   USB audio, which has a packet size of 192 bytes (1ms worth of music) is a sizable amount if lost. USB DACs still need to convert the USB packet to SPDIF/I2S packets.   HDMI carries up to 28 bytes per channel (56 bytes per 2 channel source), there's encryption and byte-encodings on this data with a small amount of error correction, so if you did lose a frame/packet, etc, 56 bytes would be lost.  Video might actually recover since the frames are repeated for a 24p source (5 times for 120hz TVs).  HDMI will still have to recover the audio clock and ship the data either I2S or SPDIF to the dac.

In the end, if you can take a known good file that contains the lossless pcm data, feed that to a soundcard that has a built in dac, and you just sidestepped a lot of unnecessary protocols to playback audio.   Even if the soundcard feeds it to another chip via I2S or SPDIF, it's still a minimalistic solution.   

[srb]

You can overanalyze the technical aspects of transport and DAC connection protocols to death.  The fact is that I have heard great sounding audio from a variety of digital playback systems with transport and DAC connected via S/PDIF, USB and FireWire.  They all met my audiophile standards.  In most cases, the big difference was the DAC itself.
 
Steve

[OzarkTom]

The final price on the Burson DA-160 will be $850, is scheduled to be out this month.

[Srajan Ebaen]

Unless John Delmo made a mistake communicating with me, it'll actually be $950 and not available before the end of this months at the soonest. They're revisiting something in the circuit (presumably the USB interface to perhaps become async and 192 - but that's just a guess) and release will be delayed until this whatever has been added/changed/improved to their satisfaction. 

[wilsynet]

I2S would be ideal.   SPDIF packet size is 48 bytes, but any clock locking issues you would lose a bit since it is serial, the debate would be is losing a bit more noticeable than jitter...   USB audio, which has a packet size of 192 bytes (1ms worth of music) is a sizable amount if lost. USB DACs still need to convert the USB packet to SPDIF/I2S packets.

Wouldn't losing 1ms of data result in a noticeable drop out?  I don't notice any noticeable drop outs with my USB DAC, perhaps the position that Firewire (and I assume USB) are not audiophile grade is based on a theoretical finding rather than a real world one.  Or are frames really dropping all the time and we're just not noticing?

There are a handful of DACs which support external I2S, and then usually the source is some other vendor proprietary device.  I2S would be ideal, but the ecosystem does not seem to exist.  Given the lack of ecosystem, I'm wondering if there is some technical reason that few have ventured down this path.

In the end, if you can take a known good file that contains the lossless pcm data, feed that to a soundcard that has a built in dac, and you just sidestepped a lot of unnecessary protocols to playback audio.   Even if the soundcard feeds it to another chip via I2S or SPDIF, it's still a minimalistic solution.

The minimalist solution seems to be sound card to analog output to your preamp.  Is there a sound card that you would use that has an audiophile grade analog output stage?  If the suggestion is sound card to SPDIF, is async USB and firewire (firewire is already async) to the DAC, which then converts to I2S is a better solution than the implications of clock recovery that's required for SPDIF?

Anyway, I'm sort of with SRB.  Given the lack of perfection in the broadly adopted protocols, perhaps the whole necessity is the mother of invention thing has yielded some pretty good (albeit imperfect) inventions.

I'm sure the IBM Token Ring guys thought Metcalfe's Ethernet was pretty absurd with its complex collision resolution scheme.  But they found a way to make Ethernet perform and it has stood the test of time.

[Noseyears]

Unless John Delmo made a mistake communicating with me, it'll actually be $950 and not available before the end of this months at the soonest. They're revisiting something in the circuit (presumably the USB interface to perhaps become async and 192 - but that's just a guess)

I will wait for the reviews and see how everything goes. Maybe you can make a comparison between Dacs on the same price range on the market? it would be interesting. 
BTw, thanks for the link Ted. 

[skunark]

Wouldn't losing 1ms of data result in a noticeable drop out?  I don't notice any noticeable drop outs with my USB DAC, perhaps the position that Firewire (and I assume USB) are not audiophile grade is based on a theoretical finding rather than a real world one.  Or are frames really dropping all the time and we're just not noticing?

There are a handful of DACs which support external I2S, and then usually the source is some other vendor proprietary device.  I2S would be ideal, but the ecosystem does not seem to exist.  Given the lack of ecosystem, I'm wondering if there is some technical reason that few have ventured down this path.

The minimalist solution seems to be sound card to analog output to your preamp.  Is there a sound card that you would use that has an audiophile grade analog output stage?  If the suggestion is sound card to SPDIF, is async USB and firewire (firewire is already async) to the DAC, which then converts to I2S is a better solution than the implications of clock recovery that's required for SPDIF?

Anyway, I'm sort of with SRB.  Given the lack of perfection in the broadly adopted protocols, perhaps the whole necessity is the mother of invention thing has yielded some pretty good (albeit imperfect) inventions.

I'm sure the IBM Token Ring guys thought Metcalfe's Ethernet was pretty absurd with its complex collision resolution scheme.  But they found a way to make Ethernet perform and it has stood the test of time.

FW and USB audio was never designed for audiophile, it's amazing to see the "real world" success with it.   Definitely use your ears, if you feel it sounds great that's all what matters. 

It's expected that Video will lose a frame from time to time, the source and the sink clocks are never locked, but as stated most video formats repeat the frame 2,3 or 5 times, some will do it even 10 times, so it's less of an issue on newer TVs, but it's probably hardly noticeable on older TVs.  A 3:2 pull-down/up is a great example of a missing frame and you can see tearing on a landscape pan and most folks don't even notice that.   

[wilsynet]

If SPDIF was designed for audiophiles, wouldn't they have attempted to eliminate jitter?  By contrast, both asynchronous USB audio and FireWire were designed explicitly to eliminate jitter as much as possible.  So let me say then that they were indeed designed for audiophiles.

The only argument I've heard is that USB and FireWire streaming audio frames are too large and a lost frame is a problem -- except there does not appear to be much evidence that this happens in practice.  Perhaps I'm all wet here, would love to understand the dropped frames problem with audio a bit more carefully.  Totally understand that HDMI video drops frames (I've seen it) but I haven't personally had audio drop out from USB from a computer that was running well.

[Napalm]

In the end, if you can take a known good file that contains the lossless pcm data, feed that to a soundcard that has a built in dac, and you just sidestepped a lot of unnecessary protocols to playback audio.

And make sure you're also using this:

http://en.wikipedia.org/wiki/Real-time_operating_system

otherwise it's an exercise in futility if the OS is not guaranteed to always feed the soundcard when required.

Nap.

[efhjr]

I'm very interested in the TEAC UD-H01 USB DAC. Accepts up to 192kHz, headphone output, priced around $500US.

It's been tough to find much information about this unit outside of "Hey, check out this new DAC from TEAC" announcements. It seems it will be put on sale this month, but I don't know if it'll make it to the US. I'm ready to make the leap to 192kHz, and the Shiit Bifrost and TEAC are high on my list.

Full info here: http://www.digitalaudioblog.com/2011/06/teac-ud-h01-usb-dac.html

[Napalm]

Full info here: http://www.digitalaudioblog.com/2011/06/teac-ud-h01-usb-dac.html

Funny. I clicked on the link without noticing that you mentioned the price already, so here I go seeing the first part of the article, noticing a mention of ESOTERIC in all caps, then I see the pic with the nice brushed aluminum and my first thought is "sh**, this must be yet another $15000 scam" then I looked for the price.....

Nap.

[OzarkTom]

EAR-Yoshino DAC by Tim de Paravicini
 (est. £4000),