mlgrado

not true. (We are speaking here of the T+A DAC, not the dac of thread record, to be clear) There are several commercial chipsets that do the exact same thing. Not necessarily with discrete components, which could very well up the quality of conversion. I exchanged some emails with that companies tech guy,... and their DSD implementation uses a delay line to feed the output switches, whose electrical parameters are the required coefficients that when summed tame the 1 bit square wave. It is if I understand correctly, a FIR digital filter implemented in analog domain. Which was originally what all the DSD buzz was about. Simple conversion with a gentle(r) filter at the end stage. In theory. He wouldn't reveal too much on the impulse response of their filter, but DID say it was shorter than the Burr Brown filter, which is a 'native' DSD implementation which you will find these days used at TEAC, iFi, Luxman and probably some other places. The thing about it, this particular conversion method works at its very best with faster DSD. The faster the bit rate, the better the impulse response. And I will leave it at that. That's the extent of my limited layperson 'knowledge'. I would think in MY mind then, the most incredible DSD playback would HAVE to be achieved by building Miska's delay line based DAC with excellent components, and use his software to convert everything to as FAST as processing can handle. I would LOVE to have his DAC and software to go even beyond DSD1024. I think in that far beyond is probably digital nirvana. OK, end of rant.

Not to keep beating a dead horse. But yes, foo_dsd_asio works with JRIVER. Its called ASIO PROXY. Works all the way to DSD512 And to answer another question you posed... it isn't restricted to any software. It can be used with any software that has ASIO compatibility. As long as the output is being routed via ASIO, it works. Which would include streaming services.

It isn't just confirmed by listening experience. It has been objectively measured as superior. ASIO PROXY has a measurably better PCM oversampling filter compared to JRIVER DSD conversion

NOT a weird choice. ASIO PROXY (foo DSD plugin) is so much more advanced than anything JRiver can do. Multiple choices of modulators, choice of DoP or Native based on sample rate. Upconversion of PCM to every DSD rate up the DSD512. ALSO upsampling or downsampling of DSD without decimation to PCM to any other DSD rate!!!!!!! And EVEN MORE IMPORTANT is the quality of the conversion. Go over to DIYHIFI.ORG which is an industry back room that several top brand, well known audio designers hang out and bitch about things. You will find a test of JRiver's DSD conversion. It sucks. Linear Phase Oversampling filter with tons of ringing. ASIO PROXY is SIGNIFICANTLY better. Oh, and speaking of Oversampling, something else you can do with ASIO PROXY is select the oversampling rate before delta sigma conversion. Or not have any at all before the modulator. SAMPLE and HOLD from NONE all the way to 8X. EXTREMELY excellent option for DSD playback via JRiver. It is the BEST option this side of HQPlayer. Now, back to what the OP is asking. JRIVER is not going to tell you what ASIO PROXY is doing. As far a JRIVER is concerned, it is outputting a bitperfect signal. The only way I know what is happening is via the iFi control panel. It shows exactly what rate is REALLY going to the iDSD DAC when using ASIO PROXY. For instance, if I am using DSD256 via DOP, it shows 705 khz as the output format. If I do that same via Native, it shows 11 megahertz. You in theory COULD do DSD512 over DoP, but the Dac would need to support 1.4MHZ PCM. But DSD512 in NATIVE ASIO works just fine. The other way to know is if your DAC has indicators. SORRY OP... I totally understand your question. I WOULD suggest this, though. If you are using the ASIO PROXY foo_DSD plugin with Jriver, and you have set the parameters, and everything is WORKING, (that is you are getting sound) I would say it is doing exactly what you are asking it to do in the settings. ANOTHER NOTE/TIP... the latest version of ASIO PROXY will not output multiples of 48khz as DSD if you use 'DSD' as the output. You MUST use DoP for those sample rates. You can only use 'DSD', which means native ASIO output, for rates that are a multiple of 44.1 khz. Not really a big deal... I just set 44.1, 88.2, 176.4, 352.8 to be converted via DoP at DSD256 rate. (Can't do DSD512 via DoP with the iFi iDSD.) Then for 48, 96, 192 and 384 I use 'DSD' or native. Not that it matters. You could choose all DoP. Probably the audio fool in me that says if I can use native streaming I must. There was A TIME when you could output ALL rates via NATIVE ASIO. But that means some non standard DSD rates... I.E. 64 x 48,000 is 3,072,000 hz. Not all DACS can handle these non standard DSD rates. One solution that ASIO PROXY will NOT use is to convert 48 khz multiples to 44.1 khz multiples THEN modulate. That is probably what JRiver does. That adds and extra step that is a potential drop in fidelity. But, if you want back the convenience factor of using Native ASIO for all PCM rates, (for instance to use DSD512 with all rates) you could turn on the JRIVER internal resampling engine and convert everything to a 44.1 khz multiple there. It would get sent out from Jriver to ASIO Proxy and now you can bypass using DoP. Myself, I would rather just use DoP and be limited to DSD256, than introduce the Jriver resampling engine into the equation. It is the poor quality of that engine that is a major reason for using ASIO PROXY in the first place, right?

I have been lurking for years now, but after reading about this DAC, and finding this thread, I really just have to say there is nothing new or innovative about this design. Sorry to be a party pooper, but, delay line with shift registers sent to weighted groups of switches is exactly what Burr Brown has been doing since the beginning with the DSD1700. So, the so called future of DAC design is , well, the past of DSD DAC design. It's the same thing Lamizator does, and is the same thing Miska does with his DAC design.

Yes, I agree. The 128 files and the extra two dollars or so, is the way to go. Any DSD, in my opinion, is best at its native recorded rate.

Okay, got it. Sorry.

You are correct. There is no conversion to 'traditional' PCM in the up or down sampling. The sample rate is always at DSD rates. Which means the filtering never approaches anything like typical decimation. You know when places like DSDFile offer both DSD64 and DSD128 versions of a file? Same deal. The DSD64 is likely sourced from the DSD128. Downsampled. In the process of filtering, yes, there is multiplication to a multi-bit form. But the sample rate never drops below 2.8mhz before being remodulated to 1 bit. This is consistent across the DSD 'universe'. It is called DSD-wide some places.

That is the first I have heard this. Pardon me for being contrary, but I believe this is somewhat incorrect. Jriver will bitstream a native DSD file as DSD, regardless of it being 64fs, 128fs, or 256fs if you use ASIO Native. I am playing back a Native DSD256 file at the moment, downloaded from NativeDSD.com. The 'bitperfect' light is on as expected. The audio path is reported as being "Direct", and "No Changes" are being made from input to output. I can do this with native ASIO streaming, or with DoP, since my Dac supports up to 768khz PCM input. Now, what I think you may be talking about, is the inability to stream via DoP at DSD256. And indeed, I cannot do this using the Jriver DoP engine. But if I use a third party engine, such as ASIO proxy, I can indeed use DSD256 via DoP. My DAC confirms this, as it indicates it is receiving a sample rate of 705khz, exactly as it should for DSD256 via DoP.

Oh, I can assure you I pay utmost attention. First of all, there is no channel imbalance at low level on my unit. The only person I have seen note it is you, actually. Second, there are virtually no clicks at all on my unit changing file formats. This is as much software dependent as anything. Results vary in different configurations. And the turn on/off pop is barely notable, here. In fact, I never would have even noticed it if some one had not mentioned it. Third, bitperfect sounds incredible, and it has nothing to do with the extra hiss. There is a slight increase in noise that is only barely, and I mean barely notable if you turn the volume pot to the max. While the signal to noise ratio may indeed be somewhat lower, it still far exceeds any standard of acceptable dynamic range. When you understand BitPerfect is a marketing term for Non-Oversampling, it is easier to understand why there may be slightly more noise. That is the nature of NOS filtering. Now, that said, some units may be experiencing a problem at the lowest sample rates in BitPerfect. (44.1 48) This is something iFi is aware of and working on. I can assure you that the 'analog' nature of the sound has nothing to do with excess noise. Fourth, it sounds like you may be experiencing an unusual amount of issues, that suggest your unit was not up to standard. I can assure you mine functions very well. Any nitpicks that may be similar to what you are describing are very, very minor. Again, either your unit wasn't performing properly, or you are overblowing things.

Wow. Enjoying DSD more than ever these days. Lots of great material to just listen to. Rather than fret over the technical details, etc. I have just been enjoying listening. Just downloaded the native DSD256 album the other day from nativeDSD. Very, very nice. That said, I can't help but still nose around the internet for information, so I wanna go back to this DXD/DSD-wide thing for a minute. I know Tom you have said the new DSD-render(or old, since it seems SACD render has been around as long as Pyramix) mode uses the same DSD-wide for edits as the Sony/Philips e-chip. I was perusing the Pyramix forums, and came across this. Graemme was asked the very question, and says it is DXD, at either 8x, 16x, or 32x depending on the DSD rate. I absolutely believe you, Tom, but it seems as always there is conflicting information out there. Thoughts? Merging Technologies • View topic - a question about DSD render mode

I posted this review on another site. Re-posted here for your convenience. BACKGROUND It was around February of 2014 when I first read of iFi while skimming the _____ forums. I was looking for a DSD capable DAC in the sub $500 range to replace the first generation Audioquest Dragonfly serving my desktop listening needs. There was only a handful of options at the time, and my interest in DSP free DSD playback further limited the choice. Actually, my choice was made for me. The iFi iDSD nano was the only DAC I could find that fulfilled this requirement in my price bracket. The only problem was I knew nothing about iFi, and I was concerned by the incredibly low price. Surely the raw sound quality would be compromised at this price point. Then again, at a mere $189 there was little risk, so with a 'What the heck?' attitude I ordered one. And wow. Wow. Wow again. I am glad I took the risk of a blind buy. In addition to being impressed with the sound quality, I became equally impressed with the crew at iFi/ AMR. Their customer interaction and support is superb. And they are truly committed to their vision of Hi-Fi, which is unapologetically different than the hi-fi norm. In the end, they are committed to providing the best sound and most useful feature set for the dollar. It should come as no surprise, then, that I followed closely the crowd design of the iDSD Micro, and am among the first 512 owners. An 'Octa-Adopter.' OCTA-WHAT??? 'Octa' as in 8x DSD, or DSD512. Yes, this DAC will playback DSD rates up to 24.6 Mhz! This is the first example of such support in a consumer level product. It also supports PCM up to 768khz. Although I know of no content currently available at these high rates, upsamling to DSD512 is possible in software, and PCM 768 allows for DSD256 playback via DoP, which means ASIO is not required for playback at that rate. Although I am not as familiar with the state of Mac computer audio, I believe this may be the first time DSD256 is available on the Mac without a need for special driver software. A DIFFERENT KIND OF HI-FI I mentioned earlier that iFi doesn't follow the hi-fi norm. What does that mean? iFi believes in minimal DSP, and believes that one should be listening to as close to the source audio as possible. DSP's such as upsampling, volume control, format conversion, etc. create unavoidable mathematical losses. The more conversions, the greater the losses. The more changes to the source signal, the more likely the changes become audible. This may especially be the case with DSD. DSP such as filtering, sample rate conversion and volume control require conversion of the 1-bit bitstream to a multibit intermediary, and remodulation back to 1 bit. As such, the iDSD Micro uses a chipset that converts DSD to analog natively with no extra digital conversion or DSP. The 1 bit DSD signal is sent to an analog FIR filter for conversion. That's it! Also, the iDSD micro has a 'BitPerfect' filter option for PCM. This eliminates the oversampling reconstruction filter used in PCM conversion. So in a DAC loaded with features, simplicity characterizes the nature of the actual audio conversion. This matches my personal audio values. THE iFi EXPERIENCE Unboxing an iFi product is a treat! Packaging is reminiscent of that other "i" company. In the box you will find a plethora (hyperbole, of course) of quality adapters and cables. Which calls attention to the unique 'OTG' USB port on the back of the Micro. It is a unique port engineered for mobile convenience. To use it with a standard desktop USB cable, an adapter is required. Two versions of the adapter are included. The adapter I chose to use is cable-less. The other adapter has a very short cable between terminations. I chose the first adapter presuming higher quality, but the cabled version may be more convenient when space behind the DAC is a concern. The 'hard' adapter combined with my iFi Gemini cable requires several inches of clearance. It is also an interesting little detail that the 'hard' adapter comes packed in an anti-static bag, like what you would expect to find enclosing delicate computer components. Also, I think it is important to add that the included USB cables are OTG cables, so if you don't already have an expensive USB cable like the Gemini, I would suggest forgetting about the adapters and going with one of the included cables. This is all I will have to say about the adapters, mobile uses, battery, etc. I will leave that to others, as I use this iDSD exclusively in a desktop environment, and cannot adequately review mobile functionality. Build quality and appearance is typical iFi. The iDSD micro is well built but take care with the switches. They feel a little fragile, and as a matter of fact, I had some trouble with a sticky switch. My over aggressive tugging, attempting to 'un-stick' it, caused the red 'Turbo' switch that controls amp output level to go flying off into the floor!! Fortunately it easily reattached, and works properly now. THE SOUND Now on to the good stuff! The sound! Crisp detailed highs, smooth upper mids, slightly warm lower mids and upper bass. Clean extension into the lows. Not too much bass; just about right. Does it deviate from neutral? That is something I am not sure I can answer. Tonal balance is the product of an entire system, and all I can tell you is how it sounds in mine, which is a custom built AMD PC running the latest Jriver Media Center software, iDSD micro, iFi iUSB power, iFi Gemini 'split' USB cable, and a modded USB cable eliminating the 5v line pre iUSB Power. The review headphones are Grado RS1i's. In comparison to the iDSD nano, the sound is the same tonally, but there is a notable increase in detail and dimensionality. On the Nano, audio images are wide, but slightly flat in comparison. The Micro has greater depth of soundstage. Never is the extra detail harsh, though. The micro is always delightfully smooth and listenable. DSD was the strong suit of the iDSD nano, and is improved in the Micro. I feel the greatest improvement, though, is with PCM material, especially using the BitPerfect filter. The promise of the Burr Brown DSD1793 segment DAC is realized more fully here. PCM sounds both silky smooth AND extremely detailed, like a hybrid of true PCM and Delta Sigma conversion, which is EXACTLY what the segment DAC is. For headphone use, which is how I exclusively use the iDSD, power is abundant and flexible. There are three settings, from Eco mode to 'Turbo' mode, which will tear paint off the walls with my Grados!!! Eco mode is already stronger than the headphone amp in the iDSD Nano, but I have settled on the middle 'Standard' mode for all my listening. The headroom it provides for the dynamic orchestral recordings that dominate my listening is welcome. This addresses the only other weakness of the iDSD Nano. The iDSD Micro has plenty of power, dynamic swing and driver control to keep up when the music gets loud and complex. I enjoy the 3D and XBass 'Analog Signal Processing'. The effect of both is subtle but notable. They never get in the way, and depending on soundtrack can really enhance the experience. For instance, the bass drum on orchestra recordings has deep authority with XBass turned on, and 3D mode really does widen the soundstage nicely, and puts the center image more 'out in front.' But I did notice that with 3D mode engaged, images on recordings I know well were placed too far to the edges for my liking, and overall imaging suffered. Instruments gain a greater sense of space, but lose their precise placement 'in space', so I do the majority of my listening with 3D mode disengaged. XBass seems ESPECIALLY useful at lower to moderate listening levels, filling in the low end nicely. At higher levels, or with music recorded with little dynamic range, the bass emphasis may be a bit much. But as most of my listening is to very dynamic music with moderate average levels, I leave XBass engaged most of the time and do enjoy the effect. Ultimately, results vary from soundtrack to soundtrack, though. IN CONCLUSION There are many more features included in this incredible product that I have not mentioned, but I believe I have covered everything that stands out to me after two days of listening. This is a special product, both in feature set and sound quality. Designed by a renowned audio engineer, with customized software and extreme functionality. Oh, and it sounds in a word, amazing. If you are looking to spend in the $500 to $1000 range, and maybe even more, you owe it to yourself to hear the iDSD micro. Highly recommended.

agreed. buyer, be ware, buyer, beware. The beauty of the ESS dac is they are comprehensive, meaning without a whole lot of work, a working model can be made. To sort of quote one of the top audio engineers in the business... if all he had to make were ESS based DAC's, he would clock in at eleven, take an hour lunch at noon, and clock out at one every day. That being said, give it a marginal power supply, high jitter TCXO or whatever the clock du jour is, along with who knows what kind of analog output stage... well, you get my drift.

Ah! See that was something I didn't know. I assumed render mode used DXD for edit points as well. So it acutally uses DSD-wide for its edits? Cool! That doesn't seem to be generally well known. I think most people assume DSD-wide is exclusive to Sonoma/Sadie... I know I did... And one last question.. mixdown. Say from 8 channels of pure DSD to 2 channel stereo. That would require the entire project to be converted to DXD/DSD-wide, correct? Or is there any native DSD mixdown? Thanks Andrew P.S. Headed over to nativedsd.com right now to take a look around. It is an incredible resource for great music. Everyone who enjoys good music recorded in super high fidelity should head there as well.

So, when you have created a DSD project, you can open it for editing in post in either DSD or DXD project mode. In DSD edit mode, you cannot monitor the edits in realtime, and only insert type edits, like crossfades, are allowable. You can basically do the same thing in DXD project mode. But a separate DXD stream is created, kind of 'in parallel', that allows you to audition edits in real time, yet, like in DSD project mode, the original DSD file is untouched except for certain types of edits that can be used in render mode,like crossfades. So, unless you need more sophisticated edits, DXD project mode and DSD project produce essentially the same result, the difference being in DXD project you can monitor what you are doing in real time.