 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
99.16.141.181
In Reply to: RE: Internet tuners in general as a digital source; Magnum Dynalab MD801 in particular posted by Langdell on August 25, 2012 at 16:44:01
...then a Logitech Squeezebox Touch is a very good choice. They can be had for $240 on Amazon. Internet radio is just streaming MP3 (or sometimes AAC+), so it's lossy compression, thus no need for a device with state of the art sound quality to handle it.
Follow Ups:
Thanks for that helpful input. On sound quality, one professional review of the MD 801 I've found basically says it makes Internet radio sound significantly better than most other Internet tuners. And I've found one or two posts on other forums suggesting the same thing. So that sounds like a pro in favor of the MD. Also, while Internet radio was the main feature I'm thinking of, after reading of the MD's other capabilities I'm thinking those are features I would like to have too. Plus its FM tuner could probably replace and better the one in my system, although there's not much FM in my area worth listening to anyway, so that's just a minor point.
"Record Are Your Best Entertainment Value!"
Edits: 08/25/12
...once a music file is compressed in a lossy way, such as with MP3 or similar, there is no way to get back to the original data of the file. Some components on the market might have circuits or software that the manufacturer claims "fixes up" such files, but in reality, that is not possible in the general case. The results might be subjectively pleasing on some files, but there is no technical justification for such an approach in a general sense.That said, if you need high quality FM, then Magnum Dynalab has a very good reputation. But for sheer number of stations, FM has an impossible task trying to compete with internet radio. Last I checked shoutcast.com for example, there were something like 50,000 stations. That's mind boggling.
Edits: 08/25/12
Thanks again! Although mostly an analogue guy, I am familiar generally with digital audio and the principle that a lossy format is never going to sound as good as a higher rez format. Heck, that's one reasone I stick with vinyl :-). But I have also heard, definitively to my ears, different components make a difference in sound with the same format that, in theory, maybe should not exist. E.g., I've had XM radio for several years both in car and at home; in my main system I have the Polk XM tuner with Burr-Brown DAC, which does sound significantly better than lower quality XM tuners I've tried through my system (not saying it makes XM sound awesome, but better than with lesser tuners).I do sometimes play Internet radio on my system using iPhone or iPad as a source. So one comparison I would be making with any Internet tuner I got is to see if it sounds significantly better than that. If not, I'd probably want to return it. (And, yeah, the FM function of the MD is a relatively minor point. In my area there's not much worth listening to on FM, unless you like commercials. But I still like having an FM tuner in my system, and the MD's probably better than the one I have, and it would be one less box and preamp input.)
"Record Are Your Best Entertainment Value!"
Edits: 08/26/12
You got an iPhone and IPad? With a $10 app, those turn into really cool remotes for Squeezebox devices. I actually got my Squeezebox 2 out of the closet and started using it again because of the iPhone/iPad remote apps. Before, searching was like texting with an old phone - you know - cycling through the number keypad to get to the letters. Screw that.
Like another poster said, most IR is MP3 quality so the ole "polishing a turd" saying comes to mind. If a component serves them up better sounding than another, then be happy with that. I don't play IR for it's quality, but for it's content. I do have 90% of my CD's ripped as FLAC files that the SB2 can serve to a DAC (that I don't have one anymore since going to a single box player), but I still preferred to pop in the actual CD when I had the SB2/DAC setup. As I literally cannot store all of my CD's in my living room, I am tempted to give the ripped CD's another try.
The Squeezebox Touch, based on what I've read, should get your foot well in the door with served digital music, without bringing the computer in the room with you. Lastly, your FM stations can be heard through the Squeezebox - though I can't comment on the quality.
marc g. - audiophile by day, music lover by night
While the degradation by a lossy CODEC is irreversible, a good streaming player can prevent further degradation of sound quality.
The streaming processing off the Internet can have jitter problems, since the player starts with a stream that has jitter measured in seconds that must be reduced to sub nanosecond levels. (If not jitter problems, then glitches due to buffer recenterings, which appears to happen with Foobar2000. The number of recenterings will depend on clock accuracies, amount of delay, etc. The 48000/16 FLAC stream for Czech Radio has good sound, apart from these occasional glitches and the occasional disconnect, usually during the announcements between music.)
Another problem with Internet streaming is clipping distortion introduced by the filter banks inside the CODEC. This will happen with hot program material (loudness wars). The reconstructed waveform can have peaks that are louder than the original. A proper CODEC will encode these peaks correctly, but the decoder may get this wrong. Typically, if floating point arithmetic is used the problem will show up when converting to fixed point to go to the DAC. A digital gain reduction of 2 dB will avoid these problems, but if done across all formats would degrade measured DAC specifications, so tin-eared objectivists wouldn't go for this. This is related to the problem of intersample peaks causing clipping during anti-imaging filtering done by an oversampling DAC, and again the cure is lower digital gain, which comes at the cost of measured performance. (My Mytek Stereo192-DSD DAC has a number of optional settings to deal with these "sounds good" vs. "measures good" issues. If one is hoping for the best possible sound out of one's system one has to deal with these issues of system interaction.)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
"Another problem with Internet streaming is clipping distortion introduced by the filter banks inside the CODEC. This will happen with hot program material (loudness wars). The reconstructed waveform can have peaks that are louder than the original. A proper CODEC will encode these peaks correctly, but the decoder may get this wrong. Typically, if floating point arithmetic is used the problem will show up when converting to fixed point to go to the DAC. A digital gain reduction of 2 dB will avoid these problems, but if done across all formats would degrade measured DAC specifications, so tin-eared objectivists wouldn't go for this."
I gather that optimum decoding of lossy formats is not as simple as it seems. A while back, Stereophile did some measurements of MP3-encoded sine waves and found distortion in the decoded output. According to Chris Helmrich of Fraunhofer, the distortion is caused by the conversion of floating-point results to 16-bit integer by simple rounding, and if dither is used, the distortion goes away. I wonder how many implementations get that right?
With a decent software and a decent DAC the conversion can be 24 bits, which will put the distortion problems far away. However, as you pointed out, there was (and presumably still is) software out there that is not decent.
I checked the conversion of a 44/32(int) 1 kHz sine wave to MP3 using LAME (as supplied with dBpoweramp). Converting back, the default was to output 16 bits and it was properly dithered. Most of the MP3 artifacts were below the dither noise at most frequencies. Any harmonics were below the -150 dB limit of my FFT. If I set the default to output 24 bits or 32 bits, there were still MP3 artifacts and still no harmonics, but the dither noise was gone.
It is still an open issue whether dithering on conversion from 32 bit floating point to 24 bit fix point is advantageous. There is no obvious theoretical answer to this question and any audible differences are difficult to hear. (One notorious and outrageously outspoken audiophile referred to the effect of dithering 32fp to 24 bit as "completely trashing the sound" compared to the source 24 bit file.)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Interesting result there.
Speaking of the buffering issue, I did a couple of experiments with the Touch. Playing internet radio, I unplugged it and it continued to play for 10 seconds. Also, I was listening to the 320k station "New England Jazz Express" using SqueezePlay, the SB Touch software emulator, on the computer that runs Logitech Media Server. The sound was dropping out like crazy, playing silence about half the time. I turned on the Touch in the next room and it hung in there - not perfect, but only very occasional and brief glitches.
It could be different buffer size, buffer strategy or different clock speed. Also, given there is feedback control, the problem with the computer could be higher interrupt latency which could upset the feedback.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
nt
"Record Are Your Best Entertainment Value!"
Post a Followup:
| FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
||||||||||||||
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2013, Audio Asylum, all rights reserved ]
[ Take5 ]
