Onkyo DP-X1 DAP

Discussion in 'IEMs and Portable Gear' started by PoochZag, Mar 18, 2016.

  1. wormcycle

    wormcycle Friend

    Pyrate
    Joined:
    Aug 13, 2016
    Likes Received:
    1,506
    Trophy Points:
    113
    Location:
    Toronto, ON, Canada
    I can only guess that with MDR7 it would be even worse for you, I was so focused on wonderful details and speed that it did not bother me.
    But today I am listening to DP-X1 with SE535. Very different experience: less treble, deeper bass, well controlled as well. Female vocals, for example Brigitte Fassbaender in Winterreise, sound just great. I would say overall more pleasant listening but I feel I am missing a bit of clarity and speed comparing to MDR7.
     
  2. wormcycle

    wormcycle Friend

    Pyrate
    Joined:
    Aug 13, 2016
    Likes Received:
    1,506
    Trophy Points:
    113
    Location:
    Toronto, ON, Canada
    Torq wrote in the his X5 III review thread:


    I was trying to check DP-X1 against those low expectations and I do not have any concerns about stability, UI, or running my preferred Android audio player USB Audio Player PRO. I am using UAPP for playing local content as well as Tidal and it works great on DP-X1.

    But the most interesting was the question of Android resampling.
    I came across the comment from one of the UAPP developers. The link to full comment is here:
    https://www.head-fi.org/f/threads/u...port-for-android.704065/page-85#post-13620045.

    In a nutshell UAPP allows you to set the Android sample rate to 41000Hz, the device native rate, and variable rate. One of the comments is that if t Android sampling is set is set to fixed, device native or 44100 ,the Android sampler is not used, instead UAPP sample rate converter is used.
    All this is very confusing to me and I am not getting answers from the UAPP dev forums, or Head-fi but it looks like some players can bypass the Android sampling.

    Another interesting comment from the UAPP dev was that on the Android systems that support variable sample rate, this is the best setting.
    What I found on DP-X1 that when I am playing the 44.1kHz, 48kHz, or 96kHz files, the UAPP shows the same sampling rate for Android and the file. For example "Android: 44100 Hz, file: 44100 Hz.", Android: 96000 Hz, file: 96000 Hz."
    So it looks like Android is at least not converting the file sampling rate.
    I hope some folks who know more about this subject, including Torq of course, can comment here.

     
  3. Torq

    Torq MOT: Headphone.com

    Pyrate
    Joined:
    Jan 15, 2016
    Likes Received:
    8,193
    Trophy Points:
    113
    Location:
    Seattle, WA
    The short version is that unless your files are at the native sample rate for your particular Android device, the vast majority of which are single-rate, then you're being forced into SRC. Most Android devices have a native sample rate of 48 KHz, which means that, one way or another, on those devices any 16/44.1 content is getting re-sampled.

    All that is changing is which implementation of SRC is being used.

    I have no doubt that using UAPP's re-sampling is better than the native SRC in Android. But now you have to use a third-party player, which some people love and some don't (for UAPP I dislike the UI, don't find it very stable, it's inconsistent in acquiring/releasing external DACs, can't handle offline-TIDAL content, usage needs to be "managed" if you're using any other player with it, and it doesn't fix the fundamental issue anyway ... just gives you a less-crappy version of it).

    In a dedicated audio player, I don't want ANY forced/default SRC ... which presents significant issues if your player is Android based. Even FiiO, who did a bunch of engineering work to avoid this, and including dual clocks and support for both 44.1 and 48 KHz timebases as "native", haven't gotten it all worked out properly yet. And of the other vaguely-current Android-based players that do get this right - they don't allow you to run external apps (so no streaming anyway).

    Way more nonsense involved than there should be for a purpose-built high-end audio player; for me at least.
     
  4. wormcycle

    wormcycle Friend

    Pyrate
    Joined:
    Aug 13, 2016
    Likes Received:
    1,506
    Trophy Points:
    113
    Location:
    Toronto, ON, Canada
    ISo as long as I stick to UAPP, which I like, that still gives me some hope :).
    95% of my music are 44100Hz rips, and Tidal, if I specify 44100, which defaults to UAPP SRC, they seem to claim they do nothing with the file that matches their fixed sampling rate.
    On Android 5.1.1 UAPP is at least stable, I agree with your other points on this app. Is there any player but Tidal that would let you use Tidal offline files?
     
  5. Torq

    Torq MOT: Headphone.com

    Pyrate
    Joined:
    Jan 15, 2016
    Likes Received:
    8,193
    Trophy Points:
    113
    Location:
    Seattle, WA
    The DP-X1 has a native sample rate of 48 KHz; if you're playing 44.1 KHz (or integer multiples thereof) content on it, it's undergoing SRC.

    The only way I know of to not have to deal with SRC on the DP-X1 with 44.1 KHz files is to use an external USB DAC (and UAPP, etc.)

    None that I'm aware of.
     
  6. wormcycle

    wormcycle Friend

    Pyrate
    Joined:
    Aug 13, 2016
    Likes Received:
    1,506
    Trophy Points:
    113
    Location:
    Toronto, ON, Canada

    I think it took me some time to come up with a really silly question but the more I read about it the less I understand. DP-X1 comes with two ESS Sabre ES9018K2M DACs capable of processing files with up to 384kHz sampling rate.
    If Android is going to convert any file into 48kHz what is the point of having any DAC on Android system capable of more than 44100Hz, or 48kHz?
    It would only make sense for the applications that can bypass the system audio and talk to DAC drectly. Apparently the native DP-X1 player does that, at least according to Onkyo support forums. Great but it does not do much for Tidal users.
     
  7. Torq

    Torq MOT: Headphone.com

    Pyrate
    Joined:
    Jan 15, 2016
    Likes Received:
    8,193
    Trophy Points:
    113
    Location:
    Seattle, WA
    That isn't what's happening.

    It has to resample any content that isn't natively 48 KHz to some multiple of 48 KHz and/or drive the sample clock at that rate. For 48, 96 and 192 and 384 KHz content, in this case, that's a simple integer multiplication. That's pretty benign, and very cheap to do, compute/power wise. But to go from 44.1 KHz, or an integer multiple there-of, to 48 KHz (or an integer multiple there-of) requires precise math and distinctly ugly, non-integer, conversion ratios.

    Mobile devices are not know for splurging their power-budget on audio-performance features - and the stock Android SRC is optimized for power utilization rather than precision.

    Except that unless there are clocks on the device that operate at the necessary, native, base frequencies, specifically 44.1 KHz or 48 KHz, then the clock, or the content, has to be scaled to take that into account. It has nothing to do with how you communicate with the DAC chip, just how you handle the clock.

    Even if it does, unless the unit has magically sprung another clock at 44.1 KHz, then it's still resampling or performing other manipulation of the native 48 KHz clock for any content using a 44.1 KHz base sample rate.
     
  8. wormcycle

    wormcycle Friend

    Pyrate
    Joined:
    Aug 13, 2016
    Likes Received:
    1,506
    Trophy Points:
    113
    Location:
    Toronto, ON, Canada
    My balanced cable for SE535 finally arrived, it is the basic kind of 6N OCC balanced cable made in China.
    For the start it seems to be better made than the SE535 stock cable. The plugs are very firm and they do not rotate like the stock cable.

    The difference using SE535 with balanced is significant, unlike what PoochZag reported for K10, and Jupiter.
    As I said, maybe the stock stereo cable I was using was not very good in the first place, but with balanced cable I heard way better imaging, the background was a touch more quiet, I hear more details or maybe I just hear them more clearly, because of the improved transient response. A balanced cable is really the must if you want to get the best out of DP-X1.

    The 2.5 TRRS jack is very good, but the the 3.5 TRS appear a bit loose with some plugs. The worst plug is ATH MSR7, the cable is junk as well and I will replace it, the stock SE535 was not much better. That is not the case with well made stereo plugs. For example the Eidolic and Switchcraft 3.5 stereo plugs lock really well.
    I am planning to make the MSR7 diy cable using Eidolic connectors, and will use the Eidolic conenctor for A&K to make a 2.5 balanced cable for T5p.
     
  9. wormcycle

    wormcycle Friend

    Pyrate
    Joined:
    Aug 13, 2016
    Likes Received:
    1,506
    Trophy Points:
    113
    Location:
    Toronto, ON, Canada
    Over the last couple of weeks I spent a lot of time with DP-X1 and three different headphones. This is the summary of it, maybe useful to someone.

    It was a long listening weekend with a disappointing start, but very satisfying at the end.The initial disappointment was caused by new pads I put on T5p.2 few days ago trying to check if the 770 velour pads would make T5p.2 sound like T1. It worked with dual mono Nuforce very well: more details, still good bass, but it was better controlled. It failed miserably with DP-X1. It sounded very bright, thin, and fatiguing.
    Thankfully. it changed when I put back my everyday T5p.2 pads, Brainwaivz hybrid HM5.

    Setup: Digital filter: SHARP, Lock range adjust: one below NARROW. Balanced setup: I think I prefer balanced to ACG but I used both.
    I also tried SHORT filter, with ACG out it makes everything sound a touch brighter.

    Headphones/IEM used:
    • ATH MSR7, SE only. I wrote about this one earlier in the same thread.
    • Shure SE535 both SE and balanced, although wrote about them earlier in this thread
    • Beyerdynamic t5p.2
    Beyerdynamic t5p.2 and comparison to two other:
    • Treble: All three headphones sound brighter with DP-X1. In case of SE535 the end result was a better sound overall. But in the case of t5p.2 with velour pads it was bad: harsh, and clearly messing up the music. It was nothing left of the London Grammar magic. With the Brainwaivz hybrids, it's very different. While still brighter than on Nuforce HA200, the treble fatigue was gone, mids and bass went went back to normal.
    • Mids: Not much to say here, nothing wrong.
    • Bass: Very good with t5p.2 and balanced out . Maybe because of a bit brighter sound overall, the is a touch less of bass, but it is very well articulated, without any bloom. Some of the Smashing Pumpkins very bassy tracks gained clarity and texture. SE536 bass was, good no matter if balanced or not.
    What is the most affected by using balanced out:
    • Imaging: there is better, not only wider sound stage as well, but the difference in imaging is more obvious. On some of my favorite string quartet tracks I could follow instruments way better using balanced out.
    • Detail: Huge change on t5p.2. I already described the improvement for SE535 but in case of t5p.2 it was a like day and night. For example Rachmanivov piano Prelude Op. 23 No. 2 starts with great intensity in 40-80Hz range. Using SE it was like some pleasant rambling. With balanced out it's very different, the notes become well articulated, I could hear different layers in the track opening, listen to the left and right hand. Some % of it maybe because the Byerdynamic stock cable is such a piece of crap.
    • Using headphones I have great problems hearing if instruments sound "real", particularly if I like the sound signature.
      But I had no doubts about piano in this case. It sounded very natural with balanced out and that applies to SE535 and t5p.2. All the range, from Andres Schiff to Bill Evans.
    Finally, I really like Onkyo DP-X1. After Torq enlightened me about the Android resampling:
    , I stopped obsessing about it and started listening:
    • ATH MSR7 are great headphones to use with DP-X1 if you are not sensitive to treble, and are focused on details. Normal gain
    • SE535 has the best synergy with DP-X1 of all three, balanced out only improves it in every aspect. Low gain is more than enough.
    • Do not bother with T5p.2 unless using balanced out, and on high gain. Then they really come to life. For critical listening to instrumental music it is clearly my choice with DP-X1.
      There is myth that t5p.2 sound great from iPhone. That's BS, they need a lot of juice to sound really good and it looks like DP-X1 SE does not have enough power to drive them properly.
     

Share This Page