Hugh Robjohns Top Ten Interfaces/Convertors?

[dylanhughes81]

Occasionally when reading SoS I see Hugh allude to an interface/convertor placing inside or outside his top ten as measured using Audio Precision.

Does anyone know where this list is available if at all? I did have a search around but not managed to find anything.

Thanks

[Hugh Robjohns]

I think I might have published it in the forum before, some time ago. I'll dig out the latest version and post it here sometime over the weekend for you.

[Terrible.dee]

Occasionally when reading SOS I see Hugh allude to an interface/convertor placing inside or outside his top ten as measured using Audio Precision.

Does anyone know where this list is available if at all? I did have a search around but not managed to find anything.

Thanks

Digital recording is all different degrees of crap.

There is no such thing as a "good" AD/DA converter.

[Folderol]

Occasionally when reading SOS I see Hugh allude to an interface/convertor placing inside or outside his top ten as measured using Audio Precision.

Does anyone know where this list is available if at all? I did have a search around but not managed to find anything.

Thanks

Digital recording is all different degrees of crap.

There is no such thing as a "good" AD/DA converter.

... in your humble opinion.

[Bob Bickerton]

Not so humble it would seem!

Bob

[Ariosto]

Digital recording is all different degrees of crap.

There is no such thing as a "good" AD/DA converter.

What utter rubbish!

[Folderol]

I think, what he meant so say was...

Analogue recording is all different degrees of crap.

There is no such thing as a "good" analogue recording system.

Without exception, every component that comes in contact with the signal will add noise, and all the active ones will add distortion.

As for analogue storage... just don't go there!

[Hugh Robjohns]

Nicely put.

....but really not worth the time or effort in making the case when there is no reception...

[forumuser840717]

Digital recording is all different degrees of crap.

There is no such thing as a "good" AD/DA converter.

Oh I like you. You're funny.

[BWC]

Digital recording is all different degrees of crap.

There is no such thing as a "good" AD/DA converter.

... in your humble opinion.

Not so humble it would seem!

Bob

...nor well informed. But I'm with Hugh, let's not waste our breath, as clearly this will "fall on deaf ears."

[Johnsy]

Let's not forget the poor OP though. I'd quite like to see Hugh's list too.

[Airfix]

Now we will never know.
Quantum stuff is very frustrating - the actions in response to the query have upset the quantum fields- the observation has spoiled the 'observation' -- damn

[BWC]

Let's not forget the poor OP though. I'd quite like to see Hugh's list too.

Yes, please! In fact, I'd like to see ~regular updates of it, however often it changes.

[MOF]

Drat and double drat, I was just about to wade into the debate but Hugh said there was ‘no reception’ and I realised I’d be wasting my time.

[shufflebeat]

As for analogue storage... just don't go there!

Isn't that a shed?

[ef37a]

As for analogue storage... just don't go there!

Isn't that a shed?

Ha! I recall a discussion in the hallowed pages of Wireless World about 'static digital audio storage'. No moving parts so no wow&flutter. One response (correct at the time) was "Rubbish, the memory module would have to be the size of a large car and need a small power station to power it. Never happen."

Dave.

[Folderol]

Ha!
I remember as a teenager saying that digital audio could never have enough 'detail' to compete with analogue.
We won't go into all the other things I was wrong about

[shufflebeat]

I'm still not convinced about the spinning jenny.

I realise the drip spindle has it's drawbacks but we shouldn't rush into unproven technology.

Early adopters beware.

[Hugh Robjohns]

As promised, here is my latest list of dynamic range measurements of the interfaces and converters which have passed across my test bench.

First, though, some caveats:

Point one: This list is ONLY of the products that happened to have crossed my test bench. It is categorically NOT a comprehensive list of 'The Best (or worst) Converters and Interfaces', and is not intended as a list of recommendations.

There are many entire brands and specific products that I haven't had the opportunity to measure (yet). Moreover, some of those listed here are early models (such as the UA Apollo, for example): please check the review dates as later model revisions will probably perform better than the figures below suggest.

Also, the omission of devices from this list does not imply anything about their performance in any way; I simply haven't tested them (yet).

Point two: These measurements have all been made under identical testing conditions using an Audio Precision AP x515 test set, running an AES17 Dynamic Range test with A-weighting. There are alternative ways of evaluating 'dynamic range', and there are technically 'better' test sets available than the one I'm using, although none of these tested devices' performance approach the limits of the x515's measurement capabilities.

Point three: The AES17 measurement is only one way of evaluating converter performance — there are many others measurable parameters which could be tested and might result in a different listing order.

However, the AES17 test is one which I have found to give a consistent and reliable quantitative assessment of converter performance that tallies closely with my subjective impressions. Someone else might subjectively rank products differently based on their own personal preferences and requirements.

It is worth noting, though, that to achieve a good AES17 score every aspect of the converter has to be very well-optimised: the power supplies (particularly the analogue/digital supply isolation), grounding arrangements, shielding, analogue circuitry, clocking/jitter, digital stream recovery, and much more besides. Technical or design weaknesses in any area very quickly degrade the measurement.

For me, the biggest take away is that there is a 20dB difference in performance between the worst and best models to have passed across my test bench. And although the very best performers are generally the most expensive models, and the very worst the least expensive -- much as you might expect -- a surprisingly high number of affordable devices manage to deliver extremely good figures, which is encouraging. Even the worst devices in the list are still capable of high quality results in practical home-studio circumstances.

You will notice that I have included the RND MBC Converter twice. The reason is that it achieved a higher score with its output transformer engaged than it did with it bypassed! I can't explain why that was... but it's a fact, so I logged it!

Also, please note that many products in this list are now obsolete and unavailable (I have italicised them). I retain them in the list as they illustrate the general progress in converter performance over time.

So here, in descending order, are the AES17 (A-wtd) measurements for the A-D converters/interfaces I've measured to date. The currently available 'top-ten' are highlighted in bold. I've given all the entries hot links to their corresponding SOS reviews where available.

124.0 RME ADI-2 Pro Interface (April 2017)
123.0 Lavry AD11 Converter (June 2012)
122.0 Universal Audio 2192 Master Interface (August 20120)
121.3 Lynx Hilo Interface (July 2012)
121.3 Current Focusrite ISA card Converter (~2016)
121.0 Merging HAPI (March 2015)
121.0 Antelope Eclipse 384 Interface (Feb 2014)
120.5 Grace Design M108 Mic Amp (January 2017)
120.4 RME 12Mic (March 2021)
120.4 RME AIO-Pro (July 2021)
120.0 Apogee Symphony Interface (Sept 2011)
119.5 Original Focusrite ISA multichannel A-D card (~2003)
119.4 Cranborne 500R8 500-series Rack Interface(Sept 2019)
119.3 Eventide H9000 Effects Processor(June 2019)
118.0 Focusrite Clarett+ 8Pre Interface (November 2021)
118.0 [Not tested, but original Focusrite Clarett USB believed to reside here!]
118.0 Prism Lyra 2 interface (Sept 2015)
118.0 Prism Titan Interface (Sept 2014)
118.0 Crookwood M1 (Dec 2012)
117.0 Focusrite RedNet Interface (March 2013)
117.0 UAD Apollo Interface (original model) (June 2012)
116.5 Audient ASP 880 Mic Pre (Aug 2014)
116.4 RND MBC Converter (with Xfmr) (Aug 2021)
116.0 Antelope Audio Orion Interface (June 2013)
115.0 RND RMP-D8 Mic Preamp (Sept 2020)
115.0 Burl B2 Bomber Converter (Nov 2010)
114.8 RME AIO card (original)
114.3 Tascam Series 8p Dyna Mic Preamp (Nov 2019)
113.7 Drawmer A2D2 Converter (Dec 2010)
113.0 Millennia Media AD596 500-series converter
112.0 SPL Marc One Monitor Controller (A_D_A) (Sept 2021)
112.2 Ferrofish Pulse 16DX Converter (Sept 2019)
112.0 Audient ASP800 Mic Preamp (Nov 2015)
112.0 SSL Alpha MX Converter (Feb 2013)
111.0 RND MBC (no Xfmr) (Aug 2021)
109.0 Ferrofish A16 Ultra mkii Converter (Oct 2012)
109.0 Yellotec PUC2 Interface (July 2014)
105.0 Ferrofish A32 Converter (March 2017)
103.0 Behringer ADA8200 (Sept 2013)

(Italicised devices are no longer in production)

It's interesting to note that a dynamic range figure of 124dB corresponds to 21.1 bits of useful information, while 103dB equates to 17.6 bits. Most budget interfaces and converters are achieving around 117dB these days (circa 20 bits).

Here is my list of D-A converters (top ten highlighted once again):

129.0 Apogee Symphony Interface (Sept 2011)
126.3 Merging HAPI (March 2015)
125.0 Benchmark DAC2 HGC (Oct 2013)
125.0 Universal Audio 2192 Master Interface (August 20120)
124.0 Focusrite Clarett+ 8Pre Interface (November 2021)
121.0 RME ADI-2 Pro Interface (April 2017)
121.0 Antelope Eclipse 384 Interface (Feb 2014)
120.7 Cranborne 500R8 500-series Rack Interface(Sept 2019)
120.5 Lynx Hilo Interface (July 2012)
120.1 RME ADI-2 DAC (June 2018)
119.7 Grace Design m905 (July 2014)
119.0 Focusrite RedNet Interface (March 2013)
119.0 RME AIO-Pro card (July 2021)
119.0 Crookwood M1 (Dec 2012)
118.9 Mytek Brooklyn DAC+ Converter (Dec 2018)
118.0 [Not tested, but original Focusrite Clarett USB believed to reside here!]
118.0 Focusrite Forte (Nov 2012)
118.0 UAD Apollo Interface (original model) (June 2012)
117.6 Benchmark DAC1 (July 2005)
117.5 Grace Design m900 (Feb 2021)
117.0 Drawmer HQ (Aug 2011)
116.8 RME AIO Card (original)
116.0 Eventide H9000 Effects Processor(June 2019)
116.0 Prism Lyra 2 interface (Sept 2015)
116.0 Prism Titan Interface (Sept 2014)
116.0 Antelope Audio Orion Interface (June 2013)
115.9 Cambridge Audio DAC Magic (~2010)
114.5 Burl B2 Bomber Converter (Nov 2010)
113.0 Ferrofish A16 Ultra mkii Converter (Oct 2012)
112.5 Mytek Liberty DAC Converter (Aug 2018)
112.2 Audient Nero Monitor Controller (Oct 2019)
112.0 SPL Marc One Monitor Controller (A_D_A) (Sept 2021)
112.0 Lindell DACX
111.5 Ferrofish Pulse 16DX Converter (Sept 2019)
111.2 Grace Design m903 Monitor Controller (Oct 2012)
111.0 Ferrofish A32 Converter (March 2017)
111.0 TC Electronic Clarity X Digital Monitor Controller (Dec 2015)
108.9 Grace Design M902 Monitor Controller
107.0 Yellotec PUC2 Interface (July 2014)
107.0 Dangerous Music D-Box+ (Aug 2019)
104.5 Behringer ADA8200 (Sept 2013)

Again, a dynamic range of 129dB equates to 21.9bits, while 104.5dB equates to 17.9bits. D-As almost always measure slightly better than corresponding A-Ds, not least because the hardware has a slightly easier job to do. The Apogee Symphony figure is a suspicious outlier, and I've no idea why it measured so high... I'm inclined to think the real state-of-the-art is closer to the 126dB mark.

I hope that's all of interest...

As I say, the AES17 measurement is only one aspect of a converter's or interface's performance and I would caution against using it as the sole means of making purchase decisions... but it does provide a reliable qualitative comparison between devices which might be useful.

Personally, I consider anything that scores more than about 118dB as being in the territory of top-notch mastering quality equipment. Anything scoring below 112dB is best avoided as it is decidedly disappointing by modern standards and it's not difficult to find better products at affordable prices.

[Nazard]

Many thanks, Hugh, a very interesting table to look at again.

Mine is an example where the A/D is 121.3 and the D/A 120.5, but essentially the same. The Hilo has been around for a while now, but every time the firmware is updated, there are improved features.

Funnily enough, the HiFi/Audiophile fraternity don't seem to have realised that better converters than they pay £5k or more are readily available for much less money. Or, perhaps we shouldn't want them to!