The Art of Listenng - Comparing DACs

[Bruce B]

Over the past 4-5 years I've had approximately 15 - 20 DACs that I've heard and have had time to actually listen to.
The hardest thing for a converter to reproduce are the quiet parts and high frequencies. The quiet parts are reverb tails, room ambience and note decay. Cymbals and bell tones are the other thing that is difficult for a DAC to reproduce accurately.
When you're comparing DACs over $5k, things seem to get even harder to differentiate one from the other. You need to have everything else in your chain to be at it's optimum sonic ability, which also includes the room!
The first thing you will need are several digital files, of different genres and at different sampling rates. DACs have a "native" sampling rate where they perform their best. A DAC never sounds the best at all sampling rates.
Next you will need a program like Audacity, Sound Forge or Wavelab where you can place the files on seperate tracks and zoom into the bit level to line them up perfectly for fast switching. Most of these programs also have signal generators so you can level balance the outputs to within 0.1dB A good listener can differentiate between 2 files that are as close as .2dB apart with some material.
Next you will need a a way to get at least 4 tracks or 2 stereo tracks out of your computer at the same time. The Lynx Studio AES16 can do this as well as an RME AES32. You must use the same interface to both DACs that you are comparing to rule out any variables. We use AES/EBU. You can't use USB for one and Firewire for the other.
The things to listen for are described above. Do not listen to the "music", listen for the sounds. When I was over Gary's place and we had 2 DACs going at the same time, I was telling everyone to listen to the hi-hat. Just hone in on that hi-hat beat. Don't listen to anything else. Listen for the sticks striking the hi-hat. Listen to the decay. Does the sound just die out or does it decay over a second or so.
Close your eyes and listen to the room sound. Can you "feel" how big the room is and where the boundries are? On most Classical music, the engineers will record the room for about 5-10 seconds before the first note is played. Can you hear the room boundries? Can you sense how high the ceiling is and how deep the stage is?
All of these things are tough to reproduce. When you compare converters at this level, listening to music comes last.

 

[Ron Party]

Your description mirrors my experience, Bruce. I first noticed this about 5 yearss ago when comparing RBCD on the then dCS stack with the then EMM stack. The differences were minimal; it was the decay that separated the two, with the dCS having the longer decay.

 

[RUR]

Mine, as well, Ron. When I read this:

The hardest thing for a converter to reproduce are the quiet parts and high frequencies. The quiet parts are reverb tails, room ambience and note decay. Cymbals and bell tones are the other thing that is difficult for a DAC to reproduce accurately.

I did a double-take, because I could have written those very words, based on my own experience.

Can't comment on >$5K DACs, 'cuz I haven't heard any in a comparative situation, blind or otherwise. :-(

 

[garylkoh]

Thanks, Bruce. That I agree with everything you say.

I like to use a solo acoustic guitar piece and ask myself the following questions:
- Does the decay sound natural? It tails off naturally and not in a straight line. With some DACs, when you add in an upsampler, sometimes the decay sounds longer because it does not tail off logarithmically. That sounds like a longer decay, but it sounds unnatural.
- Can you hear the body of the guitar and the slight echo between the hand and the hole. You can hear a slight "woosh" of the hand when the guitar is close miked
- Can you hear the body of the guitar against the torso of the player

With cymbals I listen to the strike of the stick on the surface - can I tell if it is hit on different parts of the cymbal. I have a couple of recordings of Charlie Watts in which I know that he uses a flat ride instead of a normal ride cymbal with a bell - you can hear the difference in the harmonics.
With a sizzle cymbal, you can almost hear each rivet in some close-miked recordings.

Of course, you can't hear all the above when you are listening to "live" music.

 

[microstrip]

The things to listen for are described above. Do not listen to the "music", listen for the sounds.
(...)
All of these things are tough to reproduce. When you compare converters at this level, listening to music comes last.

Very true. One of the best CDs I have used for this purpose is Gregorio Paniagua "La Folia" (Harmonia Mundi HMC 901050) . It is full of details, having large dynamics and astonishing decays. There is one track where you have large dynamic bells, the noise of water and you should hear a big fly. Many players can not reproduce it.

Do you know if there is any high sampling rate version of it?

 

[Bruce B]

Do you know if there is any high sampling rate version of it?

I know we transferred the HM label for HDtracks about a year ago.

 

[garylkoh]

Very true. One of the best CDs I have used for this purpose is Gregorio Paniagua "La Folia" (Harmonia Mundi HMC 901050) . It is full of details, having large dynamics and astonishing decays. There is one track where you have large dynamic bells, the noise of water and you should hear a big fly. Many players can not reproduce it.

Do you know if there is any high sampling rate version of it?

Thanks, microstrip. Sounds like something I should pick up I have Musique de la Grece Antique which is also fabulous.

 

[Bruce B]

I did a double-take, because I could have written those very words, based on my own experience.(

Unfortunately I don't get to just sit down and enjoy "music" very often. Being the engineer, I'm always picking apart the tracks and sounds. I always tell people to pick something you can hear that's repetitive like the brushes on the snare or hi-hat. Does the decay let the next hit come through? or muddy the attack/transient. As Gary said, sometimes a sizzle cymbal just sounds like digital hash and not rivets on a cymbal.

 

[microstrip]

Thanks, microstrip. Sounds like something I should pick up I have Musique de la Grece Antique which is also fabulous.

You should. Look for the fly in track 10 at 4'42 !

If you enjoy it you can also get Tarentule-Tarentelle, also a master piece with great sounding.

 

[microstrip]

I know we transferred the HM label for HDtracks about a year ago.

Yes, you transferred the wonderful "Song of Songs" by Stile Antico in Audiophile 88kHz/24bit (my first HD music download!) and some others, but as far as I know, not the La Folia.

Until them, the LP sounds better than the CD ...

 

[RBFC]

I have many of the HM discs mentioned. The XRCD of Musique de la Grece Antique is even better than the standard disc. I have to compare the SACD of La Folia with the redbook, haven't done that yet. Tarantule-Tarantelle has a bell choir torture track that is incredible when well-reproduced. All wonderful examples of recordings that exhibit natural decay in a real space.

Lee

 

[Lee]

Bruce,

Thank you for starting this forum. Do you have DACs wou would recommend in the $500, $1000, and $2000 price ranges?

In my experience, high frequencies are indeed an area of weakness of many dacs.

 

[Bruce B]

Bruce,

Thank you for starting this forum. Do you have DACs wou would recommend in the $500, $1000, and $2000 price ranges?

In my experience, high frequencies are indeed an area of weakness of many dacs.

You're welcome... I'm a teacher at heart and love to share my knowledge and my opinion, wanted or not!!

DAC's I've heard and like

$500 - Apogee Duet

$1000 - Stello DA100 Sig. or the Lavry DA11

$2000 - Lynx Aurora 8, Antelope Zodiac+ or Ayre QB-9

 

[Lee]

Thanks Bruce. Any thoughts on the Benchmark DAC1?

 

[amirm]

Fabulous intro Bruce!

 

[Bruce B]

Thanks Bruce. Any thoughts on the Benchmark DAC1?

The only Benchmark I've heard is the Pre. I was having impedance issues. I've heard good things about the HDR though. It's more of a consumer piece and trying to integrate consumer and professional pieces in the same system can be a nightmare sometime. That's why I'll never (guess I should never say never) have a DartZeel piece in here.

Fabulous intro Bruce!

Thanks Amir... I'm only as good as the people and friends around me.

 

[Bruce B]

This is the controller for a professional Class-A preamp. Has 3 digital inputs, 3 analog inputs and 3 analog outputs, where you can level match to 0.1dB Makes comparisons easy as pie.

 

[fas42]

Since we are talking cymbalic, rather than real sound , and t'is I, with the wierdo ideas about how good things can sound, I am curious what your experiences are with cymbal sound on rough and ready pop and rock recordings. Early in the piece I zoomed in likewise on "worse" and "worse" rock recordings, listening very specifically to the quality and air around the crash cymbals. The very first Rolling Stones album is a real nasty here, but a good standard, as I have mentioned elsewhere is Status Quo: essentially all systems I have heard apart from my own in good nick are fairly disasterous here. I would be interested in what thoughts you would have here ...

Frank

 

[Lee]

I have a Benchmark so it was a bit of a loaded question. I like it quite a bit. Very finely resolving, superb bass and very smooth highs. I've found it very much needs a high quality 75 ohm cable.

 

[Bruce B]

Since we are talking cymbalic, rather than real sound , and t'is I, with the wierdo ideas about how good things can sound, I am curious what your experiences are with cymbal sound on rough and ready pop and rock recordings.

I think the past 5 - 10 years has really been great for drum recording techniques. Our mix engineer went to a "boot camp" just on how to mic drums. He can spend a half a day just setting up and dialing in the mics for drums. The thing that is good for the cymbals is that now engineers are using overhead ribbon or sdc omni mics just above the drummer, like in the photo below. Our mix engineer also uses room mics to capture the room ambience as well. With those 4 mics, you get a great stereo spatial image as well as crystal clean cymbal sound away from drum noise. The photo below depicts a total of 14 microphones on this one drummer in an isolated room.