New Audio-gd R2R 7 Flagship Resistor Ladder DAC

Discussion in 'High-end Audio Forum' started by Currawong, Jun 30, 2017.
67 68 69 70 71 72 73 74 75 76
78 79 80 81
  1. Chopin75
    This should make sense. If your file is originally PCM it is best to play as PCM so naturally PCM should be better still. DSD conversion would be losing something, esp on the fly. It means the R2R DACs are really good in playing PCM and you can do this direct comparison on the same DAC.The next step is to get a DSD file (recorded natively in DSD, and never gets converted entirely to PCM for mixing, or recorded/mixed in analogue & transfer to DSD), and compare to its PCM version of the same resolution (via conversion on the fly and also get the actual PCM file if actually available). And see which one is better?
  2. FredA
    I will have to find a native dsd file to find out. I don't know if the dsd64 file i got this morning at the office is from a dsd master. I will try another day. For me, it is not as interesting as my whole collection is in pcm.
  3. Chopin75
    Check out native DSD- all are recorded in DSD or DXD, and it tells exactly how the recordings are made. analogue to DSD is the most analogue! I tried the Bill Evans one if you are into Jazz, pretty close to LP. It also explains exactly what native DSD means and how are they recorded.
    Channel classics and yarlung, Eudora all do not mix in PCM. If entire track is mixed in PCM the sound suffers but even then, regardless, natively recorded DSD are much better usually than PCM recording converted to DSD as most DSD are done this way and released as download or SACDs. So converting PCM to DSD on our macs for sure should not be as good! But they are better than I thought!
    And also should try natively recorded at DSD256,- should be best and rivals DXD. I would be interested in how R2R7 handle DSD256!
    Last edited: Mar 13, 2018
  4. gk2013
    Me personally never have been very impressed by DSD (complared to PCM). Currawong confirms exactly what I hear: "DSD is known to wreck the transient edges of the music, so that is what you're hearing." I cannot say DSD sounds therefore better. The nativedsd website is nice but I want to listen to music of my taste, not to music recorded in native DSD because it might sound better. I know you could upsample everything to DSD256 or 512. I tested and was not impressed.
    Give me good music in PCM on the R2R7HE and I am the luckiest man on earth :)
  5. gk2013
    Oh...and I forgot to mention the quality of the recording itself does so much more! This is key; the end product in PCM or DSD is just a fraction of it. This is in my humble view why a lot of people think DSD sounds better, the producers just pay more attention to the recording with DSD, a good recording on PCM will sound as good on my system, maybe even better. I purchased some recordings on in all different formats. PCM, DxD, DSD.. I listen to Carmen Gomes album Sings The Blues or Little Blue.
    Good music, very good sound on my R2R7HE.
    Last edited: Mar 14, 2018
  6. FredA
    Lucky you. Enjoy!
  7. Chopin75
    Hmm, my question is - is it possible that R2R7 is not optimized for DSD as it is mainly designed for PCM ? Would a signa delta chip still be better? I would not think it matters much since DSD is only 1 bit. I have no idea what you guys mean by transient edges. Maybe I need to hear the R2R DAC to know what you guys are talking about, haha. I am not sure my DS DAC can produce that sound you are talking about.
    It may explain why not many of you guys are excited using R2R7 with DSD in this thread. Perhaps one cannot really avoid the need of using 2 types of DACs:

    R2R for PCM playback - e.g. Metrum, Audio-gd, holospring. latest digital streamer& ? 1 bit Delta-signa for DSD. Interestingly Holospring is apparently superb for DSD (but uses R2R)

    Final thing, I do encourage again for you guys to try native DSD (meaning true recording done in DSD, not PCM) as I believe and have heard (though only via SACD) that native DSD sounds much better. Though Native DSD seems to mainly done in classical/jazz recordings.
    I do agree that DSD recordings are usually done much better and may account for the great sound more so than being DSD/DXD etc.
    But definitely I have noticed a DSD school vs PCM school in the audiophile worlds. I have yet to want to experience the R2R PCM playback to judge for myself :)
  8. christianvogel
    I have not tested DSD vs PCM yet, but while the DSD is one bit, the decoding often is not.

    In DAC chips, a sigma-delta modulator is used to modulate the signal from 1-bit into 4 or 5 bits. See it as a cross-over between DSD and PCM. The problem with 1-bit at DSD64 is a lot of noise (quantization errors), which get reduced by moving up to DSD128 (so some DAC chips internally upsample to this) and by moving to multi-bit. Once we are in multi-bit, we can also use the ladder of a DAC (like the Holo Spring and R2R7) to convert it to analog again. I do not know if this is how these DACs do it, because it does require that the ladder bridges are very fast.

    Some positive things I read about that is that DSD seems less susceptible to jitter effects than PCM (I would expect the opposite, since DSD needs an accurate clock). If that is true, that may also be a reason that for many people DSD sounds smoother, it simply may be more tolerant to noise in their chain. What DSD has over PCM is timing information, there simply is more of it, and it seems that the human ear / brain has crazy resolution in time (not so much in frequency, our brains compensate for frequency and dynamics very easily). In the end, this can mean that your set has an influence on which sounds best.
  9. gk2013
    I think your story contains technical assumptions this makes it difficult to understand the hard facts. This is what you see a lot on the internet. too Information is shaped to the opinion of the writer. I found this article interesting and well substantiated
  10. christianvogel
    Hi @gk2013, I have no opinion or assumptions about the sound, I have not even compared them yet.

    I had previously read the article you quoted, it is good and similar to this article:
    Also, this one for me was very informative: . It explains well how DSD is actually processed and how in practice it is often multi-bit. That was my point, that makes it also possible for a ladder DAC, if it switches fast enough. I think it was your article that shows that DSD64 has too much noise and thus it is moved up into DSD128.

    That DSD is more accurate in time, and that the human hearing has great resolution in time is a fact. That PCM is more dynamic at higher frequencies is a fact as well. How that translates to subjective experience is not of course.
  11. Chopin75
    OK, I did read some of these articles before that explain the same thing, though the bitperfect one explains more clearly. While I see how SDM (signa delta module) can convert PCM to 1 bit then to 5-6 bit (DSD-like) then back to PCM playback, if I have understand correctly, I am not sure how the DSD works. It mentions the DSD gets rid of the middle man (DSD-like) so it means it should not convert 1bit to 5 bit, even using the SDM. Actually it suggests R2R ladder doing DSD should not convert to multibit for DSD but just use the first R2R ladder for the first bit only. But then in the end it states that in reality the DSD is still converted to "DSD-like" for analogue processing due to some technical difficulties of being unstable using 1 bit? ??
    I am not sure if R2R actually converts DSD to DSD-like (with 5-6 bit) or stays as one bit. Kingwa did not explain well, nor did audio-gd, They all have limited English

    Maybe we need to do a direct comparison between these to play DSD:

    PS-audio direct stream (should be pure 1 bit processing, does not even do PCM
    R2R7 and holoaudio - R2R
    SDM - any ESS sabre or other regular signa delta chip
    TotalDAC - some sort fo separate DSD process not related to the SDM I think

    See how each sounds with DSD files ??
  12. christianvogel
    Hi @Chopin75

    The problem with 1 bit DSD conversion is that it creates a lot of quantization errors (overshoot and undershoot in the reconstruction of the original waveform). The result is a lot of noise you need to filter out. You can solve it by either higher bandwidth to move up the noise frequency (DSD128 and up) or by using a technique like SDM. There are dedicated DACs that do pure DSD though (like the PS-audio, that upsamples even DSD64 20 times). Both Kingwa and Jeff Zhu are pretty quiet about their actual implementation of this. Zhu said the Spring has a separate implementation for DSD, so they may actually both have different approaches.


    Last edited: Mar 15, 2018
  13. FredA
    About the r2r 7 and dsd, break-in could be needed as the r2r ladder has to be swiched at much higher frequency. You never know. Just guessing.
  14. seaice
    Hi all, some questions:

    - Anyone with Chord Dave vs R2R 7 experience? I am considering Dave right now.
    - Is there any mass consensus or clear individual longterm preferences regarding best NOS mode and settings of R2R 7? I have been experimenting with NOS modes in my HE-7S today and NOS modes 2 and 3 sound interesting (with pre-upsampling to 352.8 in my PC). I am not sure if R2R 7 NOS modes are identical and firmware may be different but it would be interesting to sum up all expecience.

    Last edited: Mar 15, 2018
  15. FredA
    I did try nos mode1, mode 2 and 3, with and without upsampling. I liked nos 3 better, i think, and better with max upsampling. But if you are to do software upsampling, the 6 moon configuration (jumper on att0, att1, ips0 and ips1, no jumper on s1), is even better. And to me, using the accurate fw with this combination is the ultimate.

    The Dave is supposed to be in a league of its own according to @Currrawong.
    Last edited: Mar 16, 2018
    Currawong and seaice like this.
67 68 69 70 71 72 73 74 75 76
78 79 80 81

Share This Page