|
Post by routlaw on Apr 10, 2018 11:06:01 GMT -5
Unfortunately HDMI is the worst 4-letter word in audio/video in history! What a piece of cr@p for an interface! I'm sure there is a reason for making this statement, but can you elaborate please? Why is HDMI so awful and an interface? Thanks
|
|
KeithL
Administrator
Posts: 9,929
|
Post by KeithL on Apr 10, 2018 11:09:42 GMT -5
Unfortunately, like it or hate it (as many people do), HDMI is the ONLY interface currently available for video (at least the only one at all accessible to consumers). It's also the only convenient format available for high-resolution multi-channel music. I personally prefer USB for two-channel high-resolution music.... And I tolerate HDMI for everything else.... Or, perhaps, KeithL , I should ask my question in a different way: what's going to be the "preferred" way to get Digital Data into the RMC-1. "Preferred" in the sense of offering support for the broadest array of Digital Formats and Resolutions? It's sounding like "HDMI" is the answer. to Casey Unfortunately HDMI is the worst 4-letter word in audio/video in history! What a piece of cr@p for an interface!
|
|
KeithL
Administrator
Posts: 9,929
|
Post by KeithL on Apr 10, 2018 11:57:29 GMT -5
Not at all... I agree entirely. However, part of the problem is simply that computers DO offer so many options.... The iPod, as you said "won the hearts and minds" of a bunch of people - but they weren't audiophiles. In fact, with iTunes, Apple standardized on a lossy format, based on the quite reasonable assertion that the majority of their audience was more interested in convenience than sound quality. But at least Apple computers have had built-in support for UAC2 for years, while Microsoft only finally came around to including it in the latest few versions of Windows 10 (they credited the delay to "lack of interest"). However, part of the problem is that, much like photography buffs, and those guys who love to rebuild cars all the time, a lot of audiophiles ENJOY the complexity, and the endless choices and options that go with it. The other part of the problem is that a lot of audiophiles really don't understand the technology, which tends to incline a lot of the discussions to veer sharply away from science and engineering. For example, USB, coax, and Toslink do in fact deliver EXACTLY THE SAME NUMBERS. Therefore, there is nothing "inherently better sounding" about any of them. However, it is quite possible that A GIVEN IMPLEMENTATION of one or the other may have better specs... and may even sound better. But, again to turn it around, many audiophiles have an extremely exaggerated idea about many of the factors involved. For example, high levels of jitter are in fact audible, but even a very large amount of jitter is BARELY audible.... so there are lots of other factors that make a much more significant difference. (So, if you like one DAC better than another, odds are it is NOT because one has more jitter than the other.) The general guidance that I give MY FRIENDS is this: Computer audio is great, and being able to put your entire music collection on a single USB hard drive, which fits in my pocket, is remarkably convenient. And, yes, the audio it delivers is EXACTLY the same as the bits coming off the individual CDs. And, yes, unlike vinyl, I can make a backup copy of my entire collection, or even use a simple little utility to confirm that every single bit is EXACTLY the same as it was yesterday. (So not only don't digital files wear, but you can easily make a safety copy, and you can easily confirm that they're still perfect whenever the mood strikes you.) For two channel audio, I always recommend using either a laptop or a Raspberry Pi, and USB... and making sure your DAC has an asynch USB input. The asynch USB input re-clocks the audio signal, and so removes any jitter that might be coming from your source. Asynch USB inputs do vary slightly, but the differences really are pretty small, far smaller, for example, than the differences between how various CDs are mastered. And, for home theater, we're basically stuck with HDMI for the video... And, to be totally honest, I've always found that the HDMI audio on something like the XMC-1 sounds just great. (So, HDMI audio is quite capable of sounding good enough, and the other options really aren't worth the extra aggravation... at least to me.) At the risk of sounding cynical and derogatory am I the only one who thinks all this has gotten way out of hand with complexity? I read through a couple of threads on the PS Audio forums that were very similar awhile back and likewise they were also this complex. The computer audiophile forums, same thing. Seriously when did listening to music ever get to be this convoluted, impenetrable and complex? Is it any wonder that iPods, and mobile devices with a pair of ear buds has stolen the hearts and minds of at least a couple generations of music lovers. For me it somewhat takes the fun out of it. Makes spinning vinyl and red book cd look like a walk in the park which it is by comparison. Note this is not a negative comment to anyone who has contributed to the thread, just the concept and the situation itself. Keith your knowledge and expertise is appreciated.
|
|
|
Post by Casey Leedom on Apr 10, 2018 12:42:27 GMT -5
|
|
|
Post by Casey Leedom on Apr 10, 2018 13:14:29 GMT -5
Also, the Roon Endpoint "RoPieee" for the Raspberry Pi 3 apparently only supports Native DSD (ASIO) over USB ...
Casey
|
|
|
Post by cwt on Apr 10, 2018 13:58:10 GMT -5
A lot of this subject boils down to semantics. USB is a packet-based interface. Both PCM and DSD are continuous data streams, which means that, in order to be sent over USB, both MUST be encapsulated - or "packetized". Semantics aside.... this means that the continuous data stream must be whacked up into pieces, the pieces must be stuffed into packets for shipment, and they must be extracted and reassembled at the other end. There is no way in which a continuous data stream can be sent via a bunch of packets "directly" (imagine trying to send a roll of toilet paper in a series of first-class envelopes). In principle, DoP adds an extra level of complication, because the DSD must first be stuffed into a PCM wrapper, before the PCM wrapper is packetized and packaged for transport. Because you're "packing it twice" there's more work involved... and more work to unpack it later at the other end. It would be more elegant, and take less work, if there was some way to skip that middle step; however, elegance aside, there's no guarantee that the result would be in any way be better or different. Either way, what comes out the far end SHOULD be exactly the same as what went it; and the amount of jitter on the final output should be determined by the clock used for the final reconstruction step. You may also be under a misapprehension that DSD is somehow inherently lower in jitter than PCM..... which is simply untrue. Since, in either case, the audio stream has to be split into packets, and then reassembled at the other end, the amount of jitter you end up with is going to depend entirely on the quality of the reassembly process. Appreciate the explanation Keith . When I said the lower jitter interface I was just referencing the potential of spdif ;usb and hdmi as Amir [ hddvd developer] outlines here ; www.avsforum.com/forum/173-2-channel-audio/1326576-usb-vs-hdmi-2ch-audio-receiver.html#post20263655That said I myself am happy with lpcm and was referencing those who like the idea of dsd straight to the dac . Cant apply dirac live to dsd without lpcm conversion is the way I look at it ' and its the quality of the mastering rather than the codec used
|
|
|
Post by Casey Leedom on Apr 10, 2018 14:45:48 GMT -5
And, as KeithL has mentioned multiple times, most modern good DACs don't care about jitter on the data delivery since they do Asynchronous Re-clocking. Casey
|
|
KeithL
Administrator
Posts: 9,929
|
Post by KeithL on Apr 10, 2018 14:56:30 GMT -5
Those articles you linked to all provide good information about what's involved. Both methods are simply ways of transporting a DSD audio stream through a USB cable, and delivering it at the far end, without altering it along the way. (Note that, with either method, the DSD audio itself is NOT altered along the way... so "the audio quality" of both methods is identical.) As such things go, DoP is arguably more standard so far. DSD really is a niche market, but there are more products that support DoP than that support "native DSD over USB".
The real down-side of "native DSD" is the requirement for custom ASIO drivers.
ASIO drivers have their benefits, and are often preferred in professional mixing applications for various reasons. ASIO drivers can be configured for very low latency, and often have other options that audio mixing professionals find useful, including the ability to reassign audio channels. Unfortunately, because of their extreme flexibility, ASIO drivers are also more difficult to configure correctly, and more prone to odd problems if you don't configure them properly. (Apple computers have supported standard UAC2 PCM, up to 24/192k, for years, and Microsoft has finally included support for it in Windows 10 - WITHOUT having to install proprietary drivers.). Arguably, avoiding the double-encapsulation step required by DoP is more elegant, but the requirement for a proprietary ASIO driver is far more of a nuisance. (Many people would argue that having the option of using ASIO is a benefit, but being REQUIRED to use it is a drawback.)
|
|
|
Post by Casey Leedom on Apr 10, 2018 15:03:02 GMT -5
Of course, the non-standard ASIO USB Driver issue is completely avoided with an Ethernet -> USB Bridge like the RoPieee/Raspberry Pi Roon Endpoint ... :-)
Casey
|
|
KeithL
Administrator
Posts: 9,929
|
Post by KeithL on Apr 10, 2018 15:12:17 GMT -5
I'll make it simple... just to prove that it can be.... 1) Rip all your CDs to a nice little USB hard drive - and rip them to a lossless format like WAV or FLAC. I like dBPowerAmp for ripping CDs. Also download any high-res music files you might like. 2) Get yourself a laptop computer, with the latest version of Windows 10 on it. Or get yourself an Apple computer (not my favorite choice but they do work just fine). Pick a program that can play digital audio files "bit-perfectly". For the Apple computer that would be Audirvana, or Amarra, or jRiver Media Center... For the Windows computer it would be jRiver Media Center, or Foobar2000 with the WASAPI plugin... Or get yourself a Raspberry Pi and make up an SD card with Volumio on it. 3) Get yourself any good DAC that has an Asynchronous USB Input. 4) Connect the DAC to one of the USB outputs on your computer. 5) Use any decent USB cable (Amazon Basics work great and cost under $10.) Ignore any discussion you may hear about... - jitter - high end USB cables - DSD - USB filters, tweaks, power supplies, magic rocks, etc. 6) Play your music and enjoy it already....... (Incidentally, in case you haven't looked lately, they've managed to make playing vinyl just as complicated lately ) At the risk of sounding cynical and derogatory am I the only one who thinks all this has gotten way out of hand with complexity? I read through a couple of threads on the PS Audio forums that were very similar awhile back and likewise they were also this complex. The computer audiophile forums, same thing. Seriously when did listening to music ever get to be this convoluted, impenetrable and complex? Is it any wonder that iPods, and mobile devices with a pair of ear buds has stolen the hearts and minds of at least a couple generations of music lovers. For me it somewhat takes the fun out of it. Makes spinning vinyl and red book cd look like a walk in the park which it is by comparison. Note this is not a negative comment to anyone who has contributed to the thread, just the concept and the situation itself. Keith your knowledge and expertise is appreciated.
|
|
|
Post by routlaw on Apr 10, 2018 16:03:54 GMT -5
Thanks Keith, but perhaps my initial comments painted me in a different light than was meant. Allow me to take another crack at it.
I have been implementing a computer based music server for a very long time, since the first version of Mac OS X, think turn of the century time frame. This long predated any of the aftermarket jukeboxes (J.River, Audirvana etc) and certainly DSD or any other esoteric audio files. In those days we Mac users only had iTunes for playback, it worked but had its issues as often as not. And I am no neophyte when it comes to computers, I spend most of my days in front of them working in Photoshop or Lightroom and have done this and other task on both Windows and Mac since the original version of PS 4.0 not to be mistaken for CS PS 4.0 easily adding up to some 20 years of being a computer digital imaging nerd. In other words I'm no stranger to any of this technology.
That out of the way, my comment was aimed at just how complicated this entire scenario has become and in a relatively short period of time, especially for those of us trying to obtain the best sound we can from the existing system we have. Its also worth pointing out the entire Home Theater industry is just as bad if not worse, showing no signs of abatement. I have friends who would not touch any of this (even it were for free) with a 10 foot pole, really smart people at that. Its just not worth it to them. I can't blame them.
I also have a vinyl setup too. While it can be tedious to setup a TT/Tone-arm/Cartridge system properly one doesn't have to learn an entirely new technological language to do it. I'm being a bit of a devils advocate on this. It is work to get both types (analog and digital) audio system to perform at a stellar level. But I'm not so sure vinyl is as complicated as digital, more expensive yes, and it obviously has its issues not the least of which is cleaning records just for starters.
At the end of the day, maybe I have just come full circle and find it easier, more enjoyable to spin disc (analog or digital) than to fool with a computer for musical enjoyment though I still use one, just not as much these days.
Thanks
|
|
|
Post by Casey Leedom on Apr 10, 2018 16:24:30 GMT -5
Or, we could skip all of this "USB Stuff" and just have the RMC-1 implement an Ethernet Roon Endpoint ... :-)
Casey
|
|
|
Post by Loop 7 on Apr 10, 2018 16:56:11 GMT -5
Or, we could skip all of this "USB Stuff" and just have the RMC-1 implement an Ethernet Roon Endpoint ... :-) Agree and I wish we would start to see Roon endpoint adoption at Emotiva.
|
|
|
Post by Casey Leedom on Apr 10, 2018 19:38:22 GMT -5
By the way KeithL , what's driving the USB ports on the RMC-1? Linux? And, if so, what kernel version? There's a good chance that, if you're using Linux, a good part of the work is already done for handling DSD: How Linux Audio Works vs. Windows Audio in 2017 ... Casey
|
|
|
Post by cwt on Apr 10, 2018 23:51:22 GMT -5
And, as KeithL has mentioned multiple times, most modern good DACs don't care about jitter on the data delivery since they do Asynchronous Re-clocking. Casey You mean eliminate it ? "'They employ ASRC (Asynchronous Sample Rate Conversion), which is a good method to reduce jitter. ASRC doesn't totally eliminate jitter, though." In any case I was referring to the relative advantages of different "'pipes" so to speak ;nothing more
|
|
|
Post by Casey Leedom on Apr 11, 2018 11:50:30 GMT -5
Actually, short of Digital Data simply not arriving on time, Asynchronous Re-Clocking (ASRC) can completely eliminate jitter. Look at USB as an example: the data arrives in Packets, representing multiple samples in whatever format you're sending. That Digital Data needs to get put into a Memory Buffer and then transfered to the DAC based on a completely new Clock.
Casey
|
|
|
Post by cwt on Apr 11, 2018 13:05:49 GMT -5
Actually, short of Digital Data simply not arriving on time, Asynchronous Re-Clocking (ASRC) can completely eliminate jitter. Look at USB as an example: the data arrives in Packets, representing multiple samples in whatever format you're sending. That Digital Data needs to get put into a Memory Buffer and then transfered to the DAC based on a completely new Clock. Casey Yes ; came across a 2 stage implantation using PLL and PWM reclocking to reject accumulated asynchronous jitter so its not as simple as just putting an accurate clock before the DAC apparently . www.freepatentsonline.com/8742841.htmlAlso curious as to Emo's stance on asynchronous reclocking in conjunction with asynchronous sample rate conversion for the RMC1 32bit dacs . If Keith hasn't opined on it yet ? hifiduino.blogspot.com.au/2009/06/asynchronous-re-clocker-vs-asynchronous.html
|
|
hemster
Global Moderator
Particle Manufacturer
...still listening... still watching
Posts: 51,920
|
Post by hemster on Apr 11, 2018 17:14:45 GMT -5
Unfortunately HDMI is the worst 4-letter word in audio/video in history! What a piece of cr@p for an interface! I'm sure there is a reason for making this statement, but can you elaborate please? Why is HDMI so awful and an interface? Thanks Basically it promises plug-and-play functionality but the reality is different. Many people experience handshaking issues. As Keith pointed out, HDMI is tolerated, not coveted as an interface.
|
|
|
Post by routlaw on Apr 11, 2018 19:32:58 GMT -5
@ Hemster: Until last night I don't recall ever running into any issues with HDMI, it has worked very well. But given this conversation I decided to try running DSD over PCM (which I had never done) using Audirvana as my jukebox. It worked but not well and by that I mean the signal kept dropping out, skipping and that sort of thing. I implemented all the usual suspects of rectifying the situation, rebooting, starting program over etc all to no avail. After about half an hour of that I was so agitated I almost gave up the session.
But thats not all. I could however get it to work as long as I used the highest sampling rate, in this case 176.4 kHz, but I didn't like the way it sounded, dropped down to 88.2 sounds better but skips and drops, dropped down to 44.1 and even this sounds better (I think) but again skips and drops. Fed up with this I then went back into my preferences in Audirvana and for the first time ever the DSD over PCM via the USB connection is NOT chalked out, meaning I now have the ability to try this method. And it worked quite well at all sampling rates without one hiccup.
I have no idea why the system allowed me to use DoP when this capability was never possible in the past.
|
|
hemster
Global Moderator
Particle Manufacturer
...still listening... still watching
Posts: 51,920
|
Post by hemster on Apr 11, 2018 22:11:15 GMT -5
@ Hemster: Until last night I don't recall ever running into any issues with HDMI, it has worked very well. But given this conversation I decided to try running DSD over PCM (which I had never done) using Audirvana as my jukebox. It worked but not well and by that I mean the signal kept dropping out, skipping and that sort of thing. I implemented all the usual suspects of rectifying the situation, rebooting, starting program over etc all to no avail. After about half an hour of that I was so agitated I almost gave up the session. But thats not all. I could however get it to work as long as I used the highest sampling rate, in this case 176.4 kHz, but I didn't like the way it sounded, dropped down to 88.2 sounds better but skips and drops, dropped down to 44.1 and even this sounds better (I think) but again skips and drops. Fed up with this I then went back into my preferences in Audirvana and for the first time ever the DSD over PCM via the USB connection is NOT chalked out, meaning I now have the ability to try this method. And it worked quite well at all sampling rates without one hiccup. I have no idea why the system allowed me to use DoP when this capability was never possible in the past. Stranger things have happened with software!
|
|