ASUS Essence or ESI Julia Sound card for PC

[Hugh Robjohns]

Every month for years I have devoured any SoS reviews of gear with converters in them and cannot honestly see any incremental improvement in sound quality mentioned?

I don't review many interfaces, but of those (and the converters) I do review I have seen a small and slow but steady rise in technical performance. A decade or so ago it was normal to see AES17 dynamic range figures for typical budget equipment around the 105 to 112dB mark, while high end stuff was usually 116 to 118dB.

Today budget gear often betters 117dB and high end stuff is pushing 125dB or more.

So the incremental improvement is there... The problem for an end user is that converter performance is so far above that of the typical home studio room acoustic noise floor (or that of any analogue sources) that it's virtually impossible to actually hear the benefit in most cases on typical monitoring systems!

You might then logically ask why I bother to measure the AES17 figures? Good question... I do it because the AES17 performance is quite a good indicator of the overall technical standard of the design. The performance of a great converter chip can be degraded significantly if the circuit board layouts, power supply, and clock designs aren't absolutely top notch, or the build has had corners cut to save money. Comparing what the chip maker claims for dynamic range with what the end product achieves is educational and informative.

A good interface will only be a good interface if careful attention has been paid to all areas, and if the manufacturer can achieve good AES17 figures they will have had to have got everything else right too! And while you might not hear or care about system noise floors, you will care about aliasing and jitter artefacts which can subtly degrade the musical quality in far more insidious ways, but which are managed through good design!

H

[ef37a]

I was not aware of that Zoom AI Pete, have I missed an SoS review? It is actually a fair bit more expensive than the KA6 as it lacks a good deal of the NI interface's connectivity (does at least have MIDI!)

I shall revise my suggestion to noobs for a mid price AI and include the Zoom but in many cases the extra line ins and outs plus its driver/latency performance still puts the KA6 on its own AFAIK for the money?

Hugh, I bristled when I first read your comment on dynamic range. "Who tf cares or can HEAR a better than -100dB noise floor. Then I read further and calmed down!

You are right of course, getting the best DR possible does show a commitment to generally good overall engineering.

Any chance of a KA6, Zoom, di-da.... AI shootout (using those tasty, weird 1,5kW monitors)?

Dave.

[Pete Kaine]

So the incremental improvement is there... The problem for an end user is that converter performance is so far above that of the typical home studio room acoustic noise floor (or that of any analogue sources) that it's virtually impossible to actually hear the benefit in most cases on typical monitoring systems!

Absolutely this as Hugh says and it's why I'm still more than happy to recommend a KA6 to anyone working in a lower end setup. End of the day in the audio market not only are returns diminishing past a certain point but also the weakest point in the chain can also mask everything else.

If your in a position to drop £1000 on an interface and haven't done your speakers and acoustic treatment first, then the moneys wasted regardless, as you wouldn't notice any of the improvement sat the higher price point interface.

I was not aware of that Zoom AI Pete, have I missed an SoS review?

The was discussion of one going to to occur a while back but I'm unsure if it ever appeared. One for someone else to answer I think.

It is actually a fair bit more expensive than the KA6 as it lacks a good deal of the NI interface's connectivity (does at least have MIDI!)

Only because the KA6 is on sale right now, when prices are both at full price then the is a couple of quid between them.

That said the I/O flexability on the KA6 is still better when compared to the base UAC2 as you say and that volume dial will always remain rather handy. The UAC 8 is great too, but well into another price bracket.

I shall revise my suggestion to noobs for a mid price AI and include the Zoom but in many cases the extra line ins and outs plus its driver/latency performance still puts the KA6 on its own AFAIK for the money?

Yeah, as you say the I/O flexability on the KA6 is still better and that volume dial will always remain rather handy. I'd never complain about using either of them.

I'll note that I can hear the difference in testing quite easily on my setup, but I'll drive home the point that I'm in a fully treated room, 2K+ of speakers and 1K worth of cans... so not exactly what you'd call a low end setup (well I'm sure some people here can/will, but hey, it's not what the average user of the KA6 is going to call a low end studio :p ) and I'll be the first to admit that if I took the padding out of the room the interface choice would proceed to make zero difference to the sound quality in there.

[ef37a]

Thanks Pete, would really like to see a review (and the KA!) .

I SHOULD have said that the KA is better value for money, even at full Wonga IF you want the extra holes!

Dave.

[ef37a]

Pete,
I am confused and a bit annoyed! I have just read about the Zoom UAC-2 on the Scan site and fixed upon the claim that BECAUSE USB 3,.0 is SO much faster than USB 2.0 that gives the UAC-2 its very low latency. Annoyed because only about a year ago I was asking why there were no USB 3.0 AIs about and that surely the speed advantage would show in just this way?
I was shot down several times by people claiming that "due to the nature of the USB protocol, USB 3.0 will not make any difference to latency" They went on to add that "Companies like RME have shown that USB 2.0 is all we need."

Is someone telling me porkies Pete?

I have also downloaded the manual for the Zoom and cannot for the life of me see a reference to a balanced line input? Does the AI have such? If not the price gets even less attractive. I also think an ADAT input port at least was a missed trick.

Dave.

[Pete Kaine]

Pete,
I am confused and a bit annoyed! I have just read about the Zoom UAC-2 on the Scan site and fixed upon the claim that BECAUSE USB 3,.0 is SO much faster than USB 2.0 that gives the UAC-2 its very low latency. Annoyed because only about a year ago I was asking why there were no USB 3.0 AIs about and that surely the speed advantage would show in just this way?
I was shot down several times by people claiming that "due to the nature of the USB protocol, USB 3.0 will not make any difference to latency" They went on to add that "Companies like RME have shown that USB 2.0 is all we need."

Is someone telling me porkies Pete?

Yes, it would appear to be our site in this instance. Half of the info looks to have been culled from previous internal testing and everything below feature details looks to have been lifted from the technical mail out from Zoom itself. I'll give my content guy a quizzing when he's back in office as it's all a bit jumbled having just read through it.

As a note, the Zoom hits the same RTL results on both USB 2 and USB 3.
6ms is pretty much what I'd expect from the rather more common Xmos controller that man + dogs use. Obviously the RME & MOTU kit lands south of that, so I need to get that corrected.

I have also downloaded the manual for the Zoom and cannot for the life of me see a reference to a balanced line input? Does the AI have such? If not the price gets even less attractive. I also think an ADAT input port at least was a missed trick.

The front two jacks are XLR/TRS.

Checking the "features at a glance" section.

"Two combo balanced XLR/TRS input connectors accept both mic- and line-level signal"

[The Korff]

The whole "it's USB 3 so it's got lower latency!" spiel has crossed our desks here before too, and from a number of different manufacturers. As far as I can remember, it's proven to be nonsense every time (and we've pointed as much out in the mag on a few occasions!).

[ef37a]

The whole "it's USB 3 so it's got lower latency!" spiel has crossed our desks here before too, and from a number of different manufacturers. As far as I can remember, it's proven to be nonsense every time (and we've pointed as much out in the mag on a few occasions!).

Thank you. If AI companies want to point up the benefits of USB 3.0 why don't they mention the extra 2watts of bus power? This makes for a much better specified bus powered device allowing, depending upon design decisions, more connectivity, a second and more potent headphone output or, because they need more juice, lower noise pre amps.

Pete, props for honesty! However, the spec' for inputs 1/2 is still very vague to me as far as the LINE input function? Is it balanced? A TRS connector does not make it so of course. Then a keyboard is shown connected by TRS. Very few have balanced outs. Is the input Z only 5K? And surely the line input does not have 60dB of gain?

Please don't think I have a "down" on the Zoom! I shall recommend it wherever people want a very high specc' converter and low latency at a decent price but I hate incomplete and confusing specifications...Especially on good stuff!

Dave.

[Sam Inglis]

The whole "it's USB 3 so it's got lower latency!" spiel has crossed our desks here before too, and from a number of different manufacturers. As far as I can remember, it's proven to be nonsense every time (and we've pointed as much out in the mag on a few occasions!).

There is one exception, predictably. RME have demonstrated that if you develop your own USB controller and write your own drivers, you can get excellent performance out of USB3. I'm testing a Fireface UFX+ at the moment, and its low-latency performance over USB3 is identical to performance over Thunderbolt (on Mac OS anyway, I don't have a Windows machine with Thunderbolt). But as far as I can tell, more or less everyone else is using off-the-shelf controllers and driver code, which as yet doesn't seem to offer any improvement over USB2.

[Pete Kaine]

Pete, props for honesty! However, the spec' for inputs 1/2 is still very vague to me as far as the LINE input function? Is it balanced? A TRS connector does not make it so of course. Then a keyboard is shown connected by TRS. Very few have balanced outs. Is the input Z only 5K? And surely the line input does not have 60dB of gain?

A TS connctor is unbalanced, and in general a TRS connector is balanced and it's also combo XLR (should also be balanced) port, so I'd take home from that would be that it's balanced, but yes it could be clearer I surpose.

Also, I'll refer to the previous quote again : "Two combo balanced XLR/TRS input connectors accept both mic- and line-level signal"

as it does state that it's both line level capable as well as balanced. Admittedly the quote is from the Zoom spec page and something I need them to bring over for our own listing.

[Pete Kaine]

The whole "it's USB 3 so it's got lower latency!" spiel has crossed our desks here before too, and from a number of different manufacturers. As far as I can remember, it's proven to be nonsense every time (and we've pointed as much out in the mag on a few occasions!).

There is one exception, predictably. RME have demonstrated that if you develop your own USB controller and write your own drivers, you can get excellent performance out of USB3. I'm testing a Fireface UFX+ at the moment, and its low-latency performance over USB3 is identical to performance over Thunderbolt (on Mac OS anyway, I don't have a Windows machine with Thunderbolt). But as far as I can tell, more or less everyone else is using off-the-shelf controllers and driver code, which as yet doesn't seem to offer any improvement over USB2.

Aye, but the point everyone's getting at is that whilst your perfectly correct with your findings as they are, if you now go and hook up that same interface to a USB2 port it's going to perform pretty much the same as if you'd used the USB 3 port as far as latency is concerned.

If you hook up a load of digital inputs then maybe once you pass 60 odd channels of high quality audio you might need more bandwidth as found on the USB3 buss, but the the perfomance is under discussion here, rather than bandwidth capacity.

[Sam Inglis]

You may well be right, Pete -- and I'll test it with a USB2 cable too -- but it's still mighty impressive.