KeithL
Administrator 
Posts: 10,273
|
Post by KeithL on Jul 5, 2019 10:37:32 GMT -5
Hopefully not to confuse the issue further..... but there are a lot of considerations beyond "not running out of CPU power". In general, most players do not convert the entire file before playing it, but rather read and convert it piecemeal as it is played. (This is a good argument in favor of players that do actually read the entire file into memory first... or those with the option to do so.)
Also, in general, there are a lot of different processes going on inside a computer, and some can wait, while others cannot. For example, failing to have data available at the exact instant a USB packet must be sent can result in a dropout.
(And, even if you have a huge buffer, this can still happen, if the buffer itself is unavailable for a split second when you call for it.)
This is where all those settings for "buffer size", and "interrupt priority", come into play. (Sometimes a smaller buffer works better because, even though it can supply less data, the time when it is unavailable because it's getting re-filled is also shorter.)
Modern audio technologies, like asynchronous USB connections, are designed to minimize the impact of this sort of thing, but don't always entirely eliminate it.
However, on the other side of the coin, the processors in many dedicated music player devices these days are actually not very powerful at all. And devices like DLNA servers, which may be run on reasonably powerful computers, may be asked to perform several complex tasks at the same time. (Like read a file, decode it, re-encode it in a different format, and stream it over a network.)
Here's an interesting little free utility that monitors "DPCs"... (In this context, Delayed Procedure Calls relate to USB-related situations when USB audio data may be delayed "because the processor is busy doing something else".)
Thesycon, who provided it, writes USB audio drivers - among other things.
I wouldn't take what it shows too much to heart... because many modern DACs should be largely immune to the problems it shows. However, it's interesting to see how many occasional "problems" show up - even on processors that would seem not to be working very hard at all.
(And, if you see a huge number of red lines... it might be worth trying to reduce them a bit.)
The only theory that fits that situation is that the server's CPU is spending time uncompressing and buffering the data when it should be just shipping bits. I gave that a passing thought, and what I realized is that the time it takes to un-compress a file and convert it to another format that's no compressed is far less than the time it takes to play the file. I can un-compress and convert an entire ripped CD faster than it takes for 1 song to play. And, while that's happening - my computer's CPU is still doing a lot of other things unrelated to audio. When I check CPU usage during those times, my cheap little PC has plenty of capacity left. So, it seems to me the rate limiting step is playback - not un-compressing, and today's CPU's have plenty of extra capacity. Maybe if CPU's were as slow as when I started into computing in the late 70's and RAM were effectively nothing, I could buy the argument. But, today - even low end CPU's are very fast and RAM is huge. That said, the guys from Mythbusters did manage to add luster to fecal matter in 1 episode. Mark |
|
KeithL
Administrator
Posts: 10,273
|
Post by KeithL on Jul 5, 2019 10:49:37 GMT -5
I listened to MQA files for several days (first unfold but not second).
I would have to say that my experience with them was similar to my experience with other sorts of "remasters"... Some sounded better, some sounded worse, and some sounded about the same.
The catch is that, by definition, ALL MQA files have been "remastered". All have been processed, and deliberately altered, with the intent of making them sound better (which is my definition of remastering).
There is no opportunity to tell how well MQA works in the context of "accurately reproducing the original"
We have no opportunity to hear the actual original performance.
All we have are an "original CD" - which we can safely assume already isn't perfect...
And an MQA processed remaster - which we know has been altered yet again - and which all we can say about is "whether we think it sounds better or not".
And, since most popular albums have been reissued many times, we don't even specifically have an opportunity to make a direct comparison between "the MQA copy and the non-MQA copy of the exact same CD".
(We have not been provided with an MQA encoder we may use to process our own CDs - and compare the before and after versions.)
We have the exact same issue with "high-definition files". Some 24/192k versions of albums absolutely sound better than "the CD version"...
However, is it because they're 24/192k, or because they've been remastered, or both? Your guess is as good as mine.
Hi Mark - What you say is absolutely logical. So I'm left with no good explanation. However... Your experience with the Micro-Rendu streamer was true despite the fact that we never reached consensus on any logical explanation. When I tell you that the .WAV files (created from Apple lossless encoding files) sounded better than the source ones, I'm very confident that my hearing did not deceive me, either. Does the FLAC codec also reduce sonic quality? I don't know - I haven't done back-to-back comparisons. Many who have done such comparisons claim no difference. But others disagree. The only claim that I've NEVER heard is that FLAC sounds better than WAV. That absence alone may be significant? Boom I believe you when you say you heard it. Each individual's perception is their reality. The other thing I was thinking about it that MQA is compressed and there are 2 unfolds (de-compressions). So far, I've not heard a single MQA with both unfolds but I have heard the 1st. I've so far not heard a single piece of music that I thought sounded worse. I have heard some say they think MQA is worse. Maybe some are. Maybe the versions they heard had different mastering from the original master tapes and the mastering was poor. Who knows. Mark |
|
|
|
Post by drtrey3 on Jul 5, 2019 11:19:12 GMT -5
I read those articles. I could not hear the difference between flac and wave files on my setup in blind or sighted comparisons. So I use flac.
Trey
|
|
|
|
Post by Boomzilla on Jul 5, 2019 12:15:48 GMT -5
Boom, just box up all your CDs and a couple 4TB SSDs and send them to me. I'll rip them for you and make sure all the tags are right. I only charge $150 an hour. Sounds like the consultant's motto: "For a small, phenomenal fee, anything is possible." |
|
klinemj
Emo VIPs 
Official Emofest Scribe
Posts: 15,093
|
Post by klinemj on Jul 5, 2019 12:42:58 GMT -5
Boom, just box up all your CDs and a couple 4TB SSDs and send them to me. I'll rip them for you and make sure all the tags are right. I only charge $150 an hour. Sounds like the consultant's motto: "For a small, phenomenal fee, anything is possible." Send half to DYohn and half to me, and we'll get them done a lot faster for you. Same low rate, just faster results. What could be better than that? Mark |
|
|
|
Post by Boomzilla on Jul 5, 2019 14:03:38 GMT -5
Send half to DYohn and half to me, and we'll get them done a lot faster for you. Same low rate, just faster results. What could be better than that? The supermodel of my choice could show me how to do it while sitting on my lap!  |
|
klinemj
Emo VIPs
Official Emofest Scribe
Posts: 15,093
|
Post by klinemj on Jul 5, 2019 14:18:29 GMT -5
Send half to DYohn and half to me, and we'll get them done a lot faster for you. Same low rate, just faster results. What could be better than that? The supermodel of my choice could show me how to do it while sitting on my lap! The price might be a bit higher...but...likely worth it! Mark |
|
DYohn
Emo VIPs
Posts: 18,491
Member is Online
|
Post by DYohn on Jul 5, 2019 14:27:37 GMT -5
The supermodel of my choice could show me how to do it while sitting on my lap! The price might be a bit higher...but...likely worth it! Mark Yes, I was about to say, not for $150 an hour she won't.  |
|
|
|
Post by Boomzilla on Jul 5, 2019 14:30:39 GMT -5
As one of my amigos likes to say "Smooth as a supermodel's bottom - and it'd be worth the slap to find out!"
|
|
|
|
Post by Priapulus on Jul 5, 2019 15:45:23 GMT -5
I use Tag&Rename <https://www.softpointer.com/> to edit the tags on all my music files, so all the meta-data (including folder.jpg) is permanently part of the music file. It works with almost all music formats. Then it is always there no-matter what music player you use. Google images is the best place to find folder.jpgs. It takes a while to edit, but you only have to do it once to get the meta-data the way you like it. I find most music meta-data is incomplete, inconsistent and error filled. So I feel editing is necessary. Sincerely /b I rip using dBPowerAmp; then cleanup the tags with Tag&Rename. Sincerely, /b |
|
