Dx200 and DSD512, How, Why ?

Discussion in 'Portable Source Gear' started by Whitigir, Sep 11, 2018.
Page 1 of 6
2 3 4 5 6
Next
 
Last
  1. Whitigir
    Why ?


    Basically sigma-delta DAC will apply oversampling process at 2x or 4x or 8x. Theoretically, 32 bits into 1 bit at 2x sampling is 64. This is DSD64.


    Majority of the newest desktop class DAC will do 8X or 256 oversampling such as disclosed on the Datasheet of AKM4497EQ. All of your PCM under this process will become DSD256 by the DAC modulator and filters (modern DAC allows optional filters), then it comes through Low-pass filters and become analog signals.

    79AEAC7B-24F4-4C5B-AFA0-6668C4720FD4.jpeg
    Therefore, anything with desktop DAC already is processing at such rate, and if you convert your PCM into anything like DSD64 / 128/ 256, it will not show much of an improvements if you can observe at all. However, DSD256 will show improvements only if you are using offline conversion.


    What is DSD ?


    Simply put, DSD is a processed PCM signals, and all it need doing is to pass through Low-pass filters to become analog signals. The different types of DSD is due to the original bits and the processing of oversampling. A DSD record being done by a Studio and ship to you under SACD format is an offline conversion of the original PCM. Why ? Because PCM allows better editing for the track. Then this DSD files are already processed with sigma-delta modulations from the studio, and what they deemed best, and your equipments job is to simply pass it onto the last chain of it process (low pass filter)

    There are Native DSD recording but very very little, and some studios are doing analog tape recording and then convert to DSD later. However, the majority of digital recording shall fall into the offline conversions (I could be wrong as this is not a business I know too much about).


    Why off line vs live conversion ? Live conversion takes up a lot of resource, and everything counts. Hence the implementations of a DAC, different design and different parts being used on different manufacturers will sound different eventhough if they are using the same kind of DAC. Due to this limitations, the live conversion has too many errors, like timing errors, phase , noises errors....etc...


    You can’t add informations ?


    That is correct for existed files! But you can EQ when playing back PCM can you ? It can be altered (this is why different softwares will affect your outcome. Different modulators, different algorithms). Yes, you can look at it as a different tuning, but remember, the latest Delta-Sigma can only do 256X over Sampling and not 512 or 1024!


    However, it is not about adding but it is about how much information your system has lost during the process. This is the reason why we all crave for better power supply, better clocks.....and even better software. By trying very hard, we minimize the lost of informations during this processing but at cost.




    Then, let’s say everything else is optimized, what is the best to do ?


    In the event where you have a superbly built computer, live conversion will save storage.


    IMO, Offline conversion of DSD is better, because the computer can take so long to process with the algorithm, it negates the Jitters, the errors that could happen to a live conversions. So basically put, to get a perfectly done Process like an offline conversion of DSD256, your system need to be very precise and expensive. Where as offline conversion with a great program that has great algorithm can sound easily at it best (notice, bad program and algorithm can also bring bad results). The only thing you are losing is storage! That is correct, a DSD512 track can take up 1.7Gb or more.


    What program out there is for Offline DSD512 conversions ? There are 2 that I know, 1 is affordable and another one is much more expensive: Xivero, and AUIconverter.


    https://samplerateconverter.com

    https://www.xivero.com/xisrc/


    I can recommend Xivero as a much more affordable and observable improvement in performances



    Equipments: Just look for anything that is able to play DSD512 Native ! Portability wise, the only 2 I know are Ibasso DX200 and The upcoming Lotoo Paw Gold Touch ! And more to come in the future ? Hopefully


    ***the advantages of DSD512 Offline conversions*** are more pronounced with portable equipments due to it limitations of It Form/functions

    Recommend to use only quality CD FLACS or better!!

    I personally love this DSD512 on the dx200 and so many others, so it was better to have another thread for it own rather than burying it deep inside dx200 discussion thread. Over here, we get to discuss about the how, the why, and the perceived impressions of Dx200!
     
    Last edited: Sep 11, 2018
  2. Whitigir
    How to use J-River and play your PCM into DSD512 toward Dx200 ? Windows10

    1/ install dx200Ti driver for windows
    http://www.ibasso.com/down.php
    9DA178AF-F169-481C-BE56-F8586D481A61.jpeg

    2/ get your dx200Ti into Mango mode by holding power button for 5 seconds, switch to Mango

    3/ swipe right and pick usb setting >>> usb-DAC

    73E5D86F-B016-477B-A82D-EE8AFC818A73.jpeg F0221DC2-7641-495D-B70B-CCFFC3EB3656.jpeg C9FEF860-CDDD-4A13-BADC-AAFFBEAE9F25.jpeg

    4/ open up Jriver >>>tool>>>options >>>> Audio Devices >>>Ibasso as shown
    679700BF-5DBB-4E7B-999C-738712DC3442.jpeg
    5/ go to DSP and output and pick options as shown. Then check your Asio4All Ibasso driver to see if it is running the correct frequency as show

    No need to check large buffer option, your choices

    4863866F-A0DA-47F9-91BE-51EE35713F44.jpeg
    225B8B53-3CD3-4CBF-9FBB-04F31A819202.jpeg
    D12D7A29-7E6D-4B94-8559-4EEF4A27FE88.jpeg

    I used Dx200 as external DAC/Amp into my Stax system and ofcourse it sounds wonderful
     
    Last edited: Sep 11, 2018
    teknorob23 likes this.
  3. Whitigir
    Reserved 2
     
  4. natemact
    Though only truths were spoken by the OP, this is HF after all and it's expected polite questions about the benefits of upsampling to DSD512 will be asked. Considering it should be universally agreed upon that upsampling can not possibly add information to a file, I believe this is where most people will get hung up with what's being proposed. While this is true, upsampling can not add info, I will only reinforce the OP's mission statement by seconding that Delta Sigma dacs process PCM and DSD files differently. It's this rule that allows for such claims that upsampling any file up to DSD512 will improve it's sound quality when played back on the DX200.

    I'd go into describing the benefits but why bother when you can easily download a trial copy and hear for yourself?

    BTW I was biased to not believing I would hear a benefit prior to hearing a benefit. Please try for yourself and sorry about the file sizes : )

    DX200 gets very warm/hot when playing DSD512 so de-case and do not charge while playing. As per @Paul - iBasso best practice is to play from a plugged in, fully charged DX200. This will save battery life as the playback of DSD512 files will decrease battery life, both short, and long term.
     
    teknorob23 and Whitigir like this.
  5. newtophones07
    Subscribed and thanks for explaining/typing all of this
     
  6. Lurker0
    This statement by itself is useless, because most of people don't understand how digital to analog conversions works. I'll try to explain the difference in another language.

    Filter

    Many people are confused by what does "filter" means, when we speak about DACs. There are many different steps in D/A process that are called filter.

    Consider the signal, incoming to a filter, as a sum of a valuable signal, and a noise, or garbage. The filter is supposed to remove garbage and to pass through the valuable signal. An ideal filter removes 100% of garbage and passes through 100% of valuable signal, but it does not exist. A real filter passes through a part of garbage, and removes some of the valuable signal.

    From this point of view, any DAC at whole is a filter, where the valuable signal is the original analog signal, and the garbage is that what makes it digital. But different stages in D/A conversion are filters as well.

    PCM vs DSD512

    Conversion from PCM to DSD is a filter as well. The higher output bitrate, the better the filter.

    Oversampling of PCM in a delta-sigma DAC is a filter which produces DSD-like output.

    To make analog signal from DSD, one more filter is required, but much simpler than the filter required to convert directly from PCM. In a DAC, this final filter is one and the same for both DSD and oversampled PCM.

    As @Whitigir correctly stated in the first post, most of the DACs make DSD256-like bitstream from any incoming PCM at the first stage. So, now it should be obvious, that making DSD512 may sound better, because it has passed a better filter before going into the DAC.

    The conclusion: PCM, being converted to DSD512, contains less garbage and more from the original analog sound, than the same PCM, oversampled in a DAC.
     
  7. Mario Samek
    Last edited: Sep 12, 2018
  8. Whitigir
    That is correct, yes! Try it out and enjoy. It has 1 minute free trials, and to purchase it is only $20 or so

    Thank you @Lurker0 for the valuable information
     
    Last edited: Sep 12, 2018
    Mario Samek likes this.
  9. iridium7777
    how does that work? where does the garbage from the original flac/cd file go? more importantly -- what is this garbage? how does a piece of SW determine what's garbage on a CD vs. what's the original analog sound and only extract that? don't you think taking out the "garbage", as you call it, would have been more efficiently removed during the actual mastering of analog sound down to the 16/44 file?


    if you can show any data analysis of what a sound wave looks like from a 16/44 flac into a DSD512 it would be very useful.
     
  10. Whitigir
    If you are asking those basic questions and are into measurements, it is a better place to post them up in sound science sub-forum

    Everything affect the sound quality, but playing PCM, your Sigma Delta is doing a live conversions of it own, and it is dependent onto a plethora of everything’s that build up around it (which becomes your own DAC), where as offline conversion of DSD512 is processed at it best. Even live conversions, if you have a well built PC with good software, it would be a much more improvements than your own PCM into the Delta Sigma itself.

    If, and only if, your DAC is very high quality with excellent clock and parts that revolving around it, and then your PC as well. This is the only time doing Pure PCM and let your DAC does it job would make more senses...heck, even then, when you upsample and apply different filters into the PCM, it still would be an improvement. The best software I am observing right now is the HQPlayer, and the man done such a wonderful job
     
  11. Mario Samek
    Hello,@Whitigir please reply me, correct best setting for iBasso DX200 in to Xivero XiSRC is (output file format) DFF or DSF-DSD512? Thank you.
     
  12. Whitigir
    DSF will allow for easy tagging and title changing and so on for future categorizing, where as DFF is like Wav without this good option
     
    Mario Samek likes this.
  13. iridium7777
    i would have thought that someone could at least point me to a reference that shows an analysis between a 16/44 file and that same file upsampled to DSD512, without me going and asking in a different forum. but fine.

    most importantly, you didn't answer my question at all: i asked him what this garbage was and where does it go? my question, as far as i'm aware, didn't bother with the detail of whether this garbage is removed live or offline, unless you're trying to tell me that only offline conversion removes the "garbage" but not the live one?

    a 16/44 file has some finite representation of an analog wave, or as Lurker puts it analog sound wave + some garbage. supposedly the conversion methods that you reference:
    • take that 16/44 file, and bit-by-bit discern the garbage from the analog sound wave
    • strip out the 'garbage'
    • up sample the remaining pure sound wave bits and presumably correctly interpolate between this finite data set to fill it to a file size that's 16x larger (well, not really 16x because presumably the 'garbage' took some space and you've only upsampled the real/intended sound wave, but you know what i mean here).


     
  14. Whitigir
    I did told you that, but I will say it with more details

    1/ noises from voltage rails, regulated, components, are all electrical parameter errors.....garbages!

    2/ phase noises, Jitters from clocks....etc...etc...from mechanical vibrations...Jitters from Digital cables....to whatever else....is garbages!

    3/ Bad algorithms, on the fly processing, so buffering speed, high latency....etc.... garbages !

    Unless you have excellent built source, and excellent built DAC. You will encounter so many garbages in your digital sound processing. @Lurker0 did tell you that garbages are attached to your informations, and it will be filtered away.

    On a portable device you can skip out the Interconnect and impedance characteristics of digital cables as it is on the PcB from one part to another which meets or exceed engineering parameters for parasitic losses and errors. The only thing left you need to worry about is your power supply, noises from it rails, parasitic loss within itself, tolerances, temperature-co-efficiency, clocking system....,ultimately, the larger the clock, the better it perform, and there are absolutely 0 OCXO to fit on any portable player devices like a DX200 for example. Therefore, especially if a portable player is capable of playing back DSD512, it will especially be the best of what that player can do, and not internally processing
     
    Last edited: Sep 12, 2018
    natemact likes this.
  15. Mario Samek
    Hello, look the same track without DSD (FLAC 24/96) and with DSD512 :)

    [​IMG] [​IMG]
     
    Last edited: Sep 12, 2018
    Whitigir likes this.
Page 1 of 6
2 3 4 5 6
Next
 
Last

Share This Page