In my downtime I decided to do a in-depth comparison between the A/D converters I have at hand. With the amount of “Night & Day” reviews you read about it’s easy to question if converters are drastically different from box to box, It got me questioning at-least. The technical converter specs over the last 10years or more out perform most sound sources, processing, & environments. I have done my share surgically analyzing the THD+N, frequency response, Jitter, & so on, But that never answered the question if one can audibly hear the difference? Time for a real-world test I say.
I 1st decided on a few methods/guidelines to rule out any false impressions.
- Analog generation source.
- Same signal to all converters.
For the sources I recorded some late night noodling of Analog synth/drum, Rhodes, & Udu drum. I also took a random cross selection from my Vinyl collection (via a kab 1200/apt holman for the curious audio geeks). This gave me a wide range of content to listen/judge.
I split & recorded the signal into the analog line inputs of a Lynx, API, TC, & a MBox. All levels are roughly matched to within .1dB.
The files below are the 24bit/48khz recording from the four converters. Download the files, can you hear a Night/Day difference? Can you tell what is what?
Loading ...A.wav
FYI – When the votes die down I’ll reveal what file is from what converter, I’m curious to see what is most preferred.
Just flicking through them on Soundcloud I really can’t tell any real difference…it’s all just daytime to me!
Don’t hear much of a difference, but after a few listens, D’s my fave, with B a close second. A was ok, didn’t like C. Did I pass?
Well as of today before this reveal.
12 voted for A (API).
14 voted for B (TC).
5 voted for C (Lynx).
3 voted for D (MBox).
I’m not to surprised with these results based on my more analytical D-Scope measurements beforehand. I don’t expect a different outcome, but I’ll do a multi-track test when time permits.
thanks for the efford, this was revealing. i voted for a, but that might just psychological since it was the “first” example. in reality i think i didnt hear the difference.
would you say the yamaha 02r preamps and converters (16 bit model) play in the quality range of, say, the m-box?
I would guess the analog components of the 02r are of better quality. Being a 16bit device optimum levels while tracking is important. 16bit audio gives you 96dB of dynamic range, still greater then most sources you would be recording. With that in mind as long as you have your gain correct & record nice hot/full signals you should be able to get better results from the 02r.
What were your D-Scope measurements for each interface?
I didn’t commit the exact values to memory. All do have a flat 20/20khz, but as far as THD+N/IMD Distortion @ +4dBu goes I’ll list them in order of cleanest.
1. Lynx
2. API
3. TC
4. MBox
I don’t know how others judge, but I find for blind testing like this I must rely on quick first impressions for form an opinion, then I look for details to either confirm or denounce that impression. In listening to no more than 15 seconds each, I found I most preferred A.
Trying to more deeply examine the differences tends to make them seem less different to me.
I listened and compared using my monitors (KRK VXT8) and headphones (AKG k240), and could not hear much of a difference.
I’m constantly confused by the converter shootouts. At first, I listened to several of these types of shootouts and got the same results; very little difference. Any actual difference could not be legitimately separated from ear/brain bias.
On my own I recently went from using a Focusrite Sapphire to an Apogee Symphony I/O, and the difference was absolute mind-blowing! Immediately noticeable! But, I tested with an acoustic guitar recording, two separate takes where I played the same riff, same mic location etc… Just swapped USB/FW cables. The difference in the takes and distance from the mic was WELL within the constraints of the quality in sound I heard.
Same thing seems to happen with mic preamps. You listen to these shootouts and one test shows a 1073 to sound exactly like a VLZ3. But then you buy one and do it yourself and it’s a whole different story.
Anyway… I’m left confused.
Proper gain matching with identical signals/takes is crucial.
one has to say that those were soundcloud? samples and i dunno what cables were being used what sequencer you recorded it, what src was being used, what codec mutilated those files into streams.
to my ears the lynx sounded best. thats good cause i got a pair of 16vts here wired with vovox purum into a rme aes32. i recently did a shootout with prism, dad, meitner and lynx. barefoot mm27 in an acoustically treated room. the dad -> meitner -> prism -> lynx. the lynx are still the best bang for the buck regarding you get 32 channels for a price you wouldnt get with the a-league of addas.. nice test
C surely does best in high frequency , but sounds kind of flat ,must be due to capacitors with really small farad ,whereas all the other ones use bigger capacitors which is bad on frequency response due to loading time of capacitor .
a capacitor is a filter because of its working principle .
so the flat one is true ,but maybe other sound more musically ,but what for you got an equalizer.
i use a akg k701
Front or Back TRS inputs? API a2d
Back. These bypass the 312 & directly feed the ADC.
liked D the most. although the glitch in beginning 😀 there are obvious differences but I think they can be considered as small. D has the most pressure and 3d sound imo. cool test!
Where’s the reveal?
Just above. http://9n2.ac0.myftpupload.com/1313/?p=3702#comment-1615