|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
92.24.4.177
Relative newcomer to PC audio here but reading around, I believe that MS Media Player is low-quality for those that purse "bit perfect" quality ideally - such as FLAC etc
Does anyone have any tips to best configure Media Player to optimise it for the best sound quality playback? or just give it a miss and download a better quality player?
thanks
Follow Ups:
If you take these steps, there's no reason why Windows Media Player can't sound decent, on any hardware (obviously, the better the audio device, the better the sound):
1. Install Vista. [reason: XP has a component called kmixer.dll that can degrade sound quality]
2. Configure Vista so that the default sampling rate is 44.1kHz and 24 bits. Leave all audio enhancements disabled.
3. In WMP, configure your audio settings to enable 24-bit playback.
4. Install WMP Tag Support Extender (http://wmptagext.sourceforge.net/)
5. Install either the Illiminable Ogg Directshow Filter (http://www.illiminable.com/ogg/) or the Core Flac Directshow Filter (http://coreflac.corecodec.org/index.html). Don't install both.
6. If you rip CDs, try to do it with another program to ensure you get perfect rips (I recommend dbPowerAmp http://www.dbpoweramp.com)
WMP also has the advantage of being one of the very few players that can decode HDCD (Microsoft owns HDCD, that's why).
I prefer to use WMP on all my systems, apart from one (my dedicated music player). It has the advantage of being pre-installed with Windows, a good set of features, and sounds quite decent if you follow the above steps.
The ASIO plugin mentioned by another poster is worth trying if you have an audio device with an ASIO driver. But since you are a beginner, leave it alone for the time being. There is a potential that sound will improve, but that's very dependent on your specific audio device.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
I just tried it again, on a few albums with joined tracks, and the transitions were completely seamless.
One of the advantages of WMA is that it was designed to be gapless from the start, as opposed to MP3 which is not gapless by default.
The WMA lossless CODEC plays gapless with foobar2000 and if the files are converted back to WAV they join seamlessly. The problem comes with the WM Player (version 11). Perhaps system related?
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Edits: 08/18/09
*** The problem comes with the WM Player (version 11). Perhaps system related? ***
They play just fine on WMP 11 on every system I have tried so far - on both my laptops and 3 desktops, on both XP and Vista. WMP is my standard audio player, so I would definitely notice it if it wasn't working.
As far as I know, WMP has supported gapless since version 9. Prior to that I can't remember. It's even gapless on high resolution files (up to 192kHz).
I will try again and see if the problem still exists. Perhaps the problem was specific to my configuration or perhaps I was using an old version of WMP the last time I tested.
I don't cut software much slack. If a serious problem persists for several releases, I just ditch the software as soon as I can find a suitable replacement. (Those who have money invested in M$ stock had better hope that Windows 7 isn't a dog, otherwise M$ will be history, just like GM.)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
M$ history????
Come on, they'll be fine after...
Money for Monopolies.
Shillings for spyware.
Bucks for bloatware.
Taxes for technology.
Bribes for billionaires.
Treasure for tax-evaders.
And best of all:
Alms for audiophiles.
Rick
Why should anyone:
1. install Vista (becoming phased out and rather bloated)?.
2. look for decent sound and not better sound for more money and time and effort?
The WMP ASIO and Upsampling Plugins sound 2nd rate at best in my experience.
Foobar and cPlay are free. Others are cheaper than Vista.
Yes, but, it only decodes HDCD from CD ROM, NOT hard drive.
That's okay, dbpoweramp will decode HDCD as well.
A few years ago, I posted a way in which you can trick WMP into decoding HDCD onto a file. I also posted some technical information on HDCD.
Based on that, some guy reverse engineered the HDCD decoder. That was pretty cool!
Ah ha. Well, you beat me to it. I was just getting ready to edit my post:
Edit: 'Unless you knew a way around this faux pas'. :)
The Microsoft HDCD decoder, as well as the reverse-engineered one, only implement 2/3 aspects of HDCD, namely
1. proprietary dithering and noise shaping (this doesn't require decoding, so even non-HDCD players benefit from this)
2. peak expansion (this is the *only* thing that the Microsoft HDCD decoding does, by the way, that's why it's so easy to reverse-engineer)
So far, no software HDCD decoder implement dynamic filter switching. The author of the reverse engineered HDCD decoder wisely left the HDCD frames intact so that some future enterprising soul can decode them, but unfortunately the filter parameters are not published and not easy to reverse-engineer.
Arguably, of course, dynamically switching playback filters isn't a very good idea and of questionable benefit. Also, I doubt any recent HDCDs were recorded using the proprietary HDCD ADC which uses dynamic filters.
Wouldn't dynamic filtering have to be done at the DAC hardware level? Otherwise in spite of whatever filter used by the software, a DAC will add it's own filter!
The AV receiver contains a DSP, and HDCD decoding is implemented as a library routine on the DSP.
Typically, the decoder implements dynamic filtering by upsampling HDCD to 88.2 (and sometimes 176.4).
The main problem is that there is no software HDCD decoders available for the PC that implements dynamic filtering.
I would send you the e-mail I received from the ex-principal of Pacific Microsonics, but that would be violating the confidentiality of his communication.
Are you calling me a liar, now?
Because, I was told, in no uncertain terms, by one of the inventors of HDCD that there never was any implementation of an HDCD decoder, whether in hardware or software that had more than one playback filter.
That is as plain as I can make it. Any questions?
There is little benefit from swapping receive filters, since any desired encode/decode transfer function can be achieved by appropriate choice of encoding filter(s) with a fixed decode filter. There are two key requirements: the two filters need to be designed as a system, i.e. the decode filter must be standardized (something SP didn't do) and the channel between them needs to be linearized. (Something that HDCD addresses using other mechanisms.) So it's no wonder that multiple decode filters were dropped out.I must confess I still don't understand why HDCD used a complicated and non-linear companding scheme to achieve greater effective bit depth when they could have gotten as good (and probably better) results by going to subtractive dither, with the stolen low order bits being used for the necessary synchronization function.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Edits: 08/19/09
It's just an old fashioned way of manipulating sound.
Analog compression and dynamic range extension used to be 'fad' processess. So is HDCD.
People have heaped praise on the PM100 and 200. The ones I have heard are very average on CD material. In fact, I pulled the PM100 from my Assemblage DAC3 and replaced it with the DF1704. I have swapped several times and each time I did the same.
Anyone want to buy the 'superlative' PM100 in MINT condition? People have paid over the odds for this.
HDCD supposedly uses subtractive dither:
http://web.archive.org/web/20030602211206/www.hdcd.com/partners/proaudio/index.html
http://web.archive.org/web/20030605091427/www.hdcd.com/partners/proaudio/overview.html
Thanks. I had seen these pages previously. Also there is some discussion of these issues in the patents.
There is little question that noise is added in prior to the initial analog to digital conversion and subtracted out digitally prior to encoding but while the signal is still in 20 bit form. This is a known technique of improving low level converter linearity, i.e. building a better sounding 20 bit converter. The same results could be obtained by other converter implementations, and has little to do with "subtractive dither" unless the extra signal is subtracted out in the decode process, i.e. after the original 20 bit signal has been converted to 16 bit, stored, played back, and then reconverted to 20 bits. And of course this doesn't have anything to do with noise shaping techniques used to gain psycho-acoustic benefits by spreading the 16 bit noise around in a less obvious form. Perhaps there is an interaction between the low level encoding and decoding processes that are described in the patent and the dither noise, but it wasn't immediately apparent from two cursory readings of the patents.
I happen to be a fan of subtractive dither. It is subject to complete mathematical analysis that does not depend on any psycho-acoustic assumptions. If implemented properly (for example in a digital encoder and decoder that converts between 24 bits and 16 bits and back) it produces results that are indistinguishable from a 24 bit channel with 16 bit's worth of random noise added in, even if one gets to examine each bit in and out of the two systems in as much detail as desired. There is no 1st order interaction (distortion), no 2nd order interaction (noise modulation) or any higher order interaction between the errors of the channel and the input signal. TPDF dither is predicated on the psycho-acoustic result that 3rd order and up correlations are inaudible, and is therefore subject to all the vagaries, inaccuracies, and controversies associated with models of human hearing. In addition it has about a 7 dB penalty over subtractive dither, i.e. slightly more than one bit. If one doesn't mind the hiss, listening to 96 kHz PCM subtractive dithered to 8 bits is quite pleasant, with absolutely no artifacts from the reduced bit depth. (Not that you would want to go down this far, except for test purposes.)
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
If subtractive dither is so good, why is it not commonly used?
Does it need hardware support or can it be done in sw (like a foobar plugin)?
Subtractive dither requires that the dither signal be created at the time of encoding and be made available at the time of decoding. It has to be a random signal (or a pseudo-random signal). If a random signal is used then it has to be included on the medium (or elsewhere) and a separate signal is needed for each stereo channel. Storing the required dither to compress a 24 bit signal down to 16 bits would add an extra 8 bits, for no net gain. So a pseudo-random signal is needed. These can be of arbitrary complexity, as much as a cryptographic algorithm that couldn't be broken by espionage agencies, or something much simpler. However, this algorithm must be agreed up, i.e. standardized, otherwise subtractive dither won't work. Given that SP controlled the standards for RBCD and didn't do this then we don't have it. In addition, there is the necessity to synchronize the dither on encoding and decoding. These all involve details of encoding, but this can be done, indeed the low level bits stolen by HDCD can be used to accomplish this.
Subtractive dither was first disclosed in the early 1970's by Larry Roberts, in the video context as part of an MIT Master's thesis. It was the original dither. What got implemented by SP the better part of a decade later was a bastardized form.
In addition to the requirement of standardization there may be patents involved.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
There is NO DYNAMIC FILTERING on playback of HDCD. See the posts immediately below.
Audio at a greater resolution than CD quality can also be HDCD encoded.
http://web.archive.org/web/20030814031310/www.hdcd.com/partners/proaudio/hdcdmix.html
At least one DVD-A (Pet Sounds) was supposedly mastered with the HDCD flag!
http://www.audioasylum.com/cgi/t.mpl?f=dvda&m=6822
Higher resolution (24 bit) audio does not require peak extend or Low Level Range Extend. HDCD dithering and noise shaping is done at the encoding side. So if dynamic filtering was never implemented, was HDCD for hirez audio/DVD-A just BS?
> > At least one DVD-A (Pet Sounds) was supposedly mastered with the HDCD flag! < <
That's because ANY signal coming out of the HDCD encoder box had the flag encoded. It was just a way of advertising their process.
> > So if dynamic filtering was never implemented, was HDCD for hirez audio/DVD-A just BS? < <
The biggest problem was that there were two distinct parts to Pacific Microsonics -- a crack team of brilliant engineers, and then a sleazy marketing team full of scammers. So you had some great products with a lot of misinformation and marketing bull spread around.
For example, they had an HDCD sampler disc that was give away to thousands of people at one CES. There were some "comparison" tracks that were supposed to let you hear how good HDCD was. But it wasn't a fair comparison at all. They could have done that by using their A/D converter and then recording two otherwise identical tracks, one with the HDCD features turned on and the other with the HDCD features turned off.
But instead, they cheated.
The "non-HDCD" tracks were made with a Sony 1630 A/D converter (I think that was the model number). For many years it was the only converter available to studios. It sounded like garbage and was one of the reasons that CD had such a bad sonic reputation in the early years. Then they compared that to the "HDCD" tracks that were made with a Keith Johnson-designed converter with ultra-low jitter clocks, all discrete analog circuitry, super high quality components, et cetera, et cetera.
Well of course the "HDCD" tracks sounded miles better. That's like saying that an Audio Rsearch Ref 3 preamp and Ref 110 power amp sounds better than a Sony receiver! But that had nothing to do with the actual HDCD scheme of compansion, which is mostly what HDCD was.
Then there was the whole "gain scaling" scam whereby an HDCD disc was played 6 dB louder than a normal disc. Everyone caught on to that one quickly and most of the manufacturers include easy ways for the end-user to defeat the "gain scaling" feature that Pacific Microsonics mandated.
Back to your question. Is there any reason to perform compansion when you have 24 bits? I don't think so!
The first thing to remember is that ANY signal sent through their machine will have the LSB's altered so that the HDCD light comes on. Basically this is just an advertisement and not an indication that anything needs to be decoded. In fact in the Operator's Manual for the Model Two encoder (the last one ever made), it specifically says that low-level extension and peak extend ONLY apply to 16-bit signals. So that doesn't leave much for the machine to do on a high-res release...
The A/D converter was better than anything else available at the time. It took nearly ten years for there to be a lot of other A/D converters that sounded really good (Apogee, Lavry, et cetera). And they *did* have the two dynamic filters during record, at least at 44.1 kHz. But there was no reason to use two filters for high sample rates.
They also applied proper dithering to their signals. Again, this is important for 16-bit data, but pretty meaningless for 24-bit data, as the noise from the analog circuits are far greater than any dither would be.
The bottom line is that a high-res "HDCD" recording would sound good (and turn on the HDCD light!) simply because the encoder was a good one, but not because of any specific HDCD features.
Thanks for enlightening me about the full truth of HDCD. So the main deal about HDCD (for CD) seems to be companding like dbx or Dolby.
(This was already posted in another forum years ago: http://forums.slimdevices.com/showpost.php?p=182413&postcount=13 )
16 bit is supposed to be enough for most music and so I don't know why this companding was ever needed especially when they forced non-HDCD CDs to have a 6db level cut. Better ADC of course is helpful.
I was always tempted to buy this Pet Sounds DVD-A just for the HDCD encoding although I don't care too much for their music. Now I know it's probably not worth the money & the space it will take in my shelf.
Probably the single best explanation of HDCD I have seen.
The only thing that he overlooks is that for at least five years the HDCD A/D converter was by far the best sounding one available. So for a relatively long time the best sounding CD's *were* HDCD's, but not because of the compansion process. Instead they were good sounding because of Keith Johnson's outstanding design work on the basic A/D converter.
> > Also, I doubt any recent HDCDs were recorded using the proprietary HDCD ADC which uses dynamic filters. < <
There is no other machine that will record HDCD besides the two models that Pacific Microsonics made. These were discontinued many years ago, but they are still very good sounding converters.
And they are the only HDCD encoders on the planet. So if a recording was made with HDCD, it was made with the dynamic filters. This is the one HDCD feature that *cannot* be turned off.
But it is also the one HDCD feature that cannot (and never has been) be "decoded".
> > So far, no software HDCD decoder implement dynamic filter switching. < <
Neither has any hardware decoder.
> > The author of the reverse engineered HDCD decoder wisely left the HDCD frames intact so that some future enterprising soul can decode them, but unfortunately the filter parameters are not published and not easy to reverse-engineer. < <
The parameters would actually be trivial to reverse engineer if they existed. But they don't. See the link below for more information.
Charles, I am a fan of your work and own one of your CD players. However, I have noticed lately that the tone of some of your posts is...well...annoying. "Wrong!" or "Also wrong" is not a good way to do a subject line. It has a tone I do not like. Interestingly, the actual message text is often not bad in tone, but you've already torqued me (and probably others) before I get there! :) I respectfully suggest you tine it down a bit, if not because it's the right thing to do, but because I don't want to see the sales of your great equipment to suffer. I would have sent this in a private email, but your aa acount doesn't accept them. Thanks and best wishes.
...is that Christine and I went around and around and around and around on this topic a few years back. (I'm sure that you can find it with the "Search" function.)
Despite all the (admittedly circumstantial, yet overwhelming) evidence to the contrary, she clung to the belief that there were two playback filters for HDCD.
I decided that the only way to settle the matter was to go straight to the horse's mouth. I know someone that knows one of the ex-principals of Pacific Microsonics, so I got in touch and asked him directly.
I have posted the facts about this on several forums since then, including the Audio Asylum, but Christine seems to have conveniently ignored them, or overlooked them, I have no idea which. But I am tired of her righteous attitude.
Sorry if a private argument (that took place in a public forum) offended you.
For me, HDCD is now irrelevant. It had its days when CD was bad and there was no hirez. It does sound much better than CD on some material, but why bother now with what is essentially sound manipulation on a disc?
With the advent of DVD-Video and 96/24 audio (let alone DVD-Audio and SACD), there wasn't really any need HDCD.
Ironic, because farther down the road with the advent of MP3's & over compressed CD's, there doesn't seem to be a need for any better sounding format(s).
G
So they just buy the cheapest crap they can find.
Remember when you couldn't buy a telephone because the phone company owned them and you just paid a rental fee per month? Those were damned near indestructible because the phone company knew that was the best way to save money in the long run.
Try and buy a decent phone today. They are all made in China. One drop and they either die or crack. Even without any abuse they will die after a short time anyway. If you ever opened one up you would laugh at the construction compared to the insides of the old Bell telephones.
Same with audio. Most people are happy with cheap crap. Crappy equipment, crappy downloads, doesn't matter as long as it's cheap.
The people who visit these forums are an aberration. There are 300 million people in this country, and about 300 thousand audiophiles. That means that only one person in a thousand will pay extra money to get better sound.
So it's a small market. It's enough to support small companies, but while a small company can build great audio equipment, a small company cannot market music on a national level. Only big companies can do that, and to be big they must sell to the general public. Who only cares about price and that's why we have to listen to cheap crap for our music sources.
"while a small company can build great audio equipment, a small company cannot market music on a national level."
It seems to me that you may have it backward. As a boutique hardware manufacturer you of course are stuck with poor economics of scale. As you well know the amount of product marketing, research, engineering, manufacturing engineering and tooling is about the same if you are building a hundred or a hundred thousand and the parts cost is higher.
However, most of us don't buy new gear all that often once we have a satisfying system and duplicating devices is of limited interest. Few have more than two or three systems.
Music however is no longer dominated by the cost of pressing and distribution. The internet has removed most of those barriers and the remaining remaining speed barriers are falling rapidly. I don't see much of a barrier to small businesses going into the music business.
Capturing and distributing public domain music in decent resolution, ie > 50KS/s at a low cost/copy could be great for everyone including the performers. The key is quality/cost and that should be very deliverable. For one thing at the right cost I wouldn't mind having perhaps a dozen performances of most pieces. I already often have several to choose from and like it. Sometimes it's even fun to listen to them sequentially and experiences the differences.
Distributing high-quality, high-value music would be a perfect matching to go along with high-performance gear because the more really good recordings a person has access to, the more they will want something extra special to play them on.
Rick
The problem is that the small number of customers for low volume music includes the same low proportion of audiophiles. So there may not be more than one or two customers total if the volumes are really low. The effort to add an album to a web site isn't recovered at these volumes.The key is for the musicians to recognize and want the better sound. Then there will be the motivation to make it happen. Unfortunately, many musicians don't really care about sound, given that they are listening to musical ideas and not sound waves. (Or they have gone deaf prematurely.) Hence the challenge.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Edits: 08/21/09
I kind of like direct, no-nonsense style.
But, I guess it's possible that someone else may prefer his opponent's inconspicious subject lines, with conspicious appeals to one's own authority and healthy dose of bad blood inside the post.
Unfortunately, any mention of HDCD and Charles immediately goes frothy.
There was a prolonged exchange between Charles and myself on this subject several years ago that was rather painful. You should see some of the private emails I received from various industry insiders saying how shocked they were by the way Charles was treating me.
Well, fortunately, something good came out of it, because the technical detail I provided enabled someone to reverse engineer the Microsoft HDCD decoder, and as a result the whole audio community now benefits from a freely available HDCD decoder.
For that, I would like to thank Charles for the part he played in enabling this to happen.
> > You should see some of the private emails I received from various industry insiders saying how shocked they were by the way Charles was treating me. < <
Why not let people decide for themselves how I was treating you? Below is a link to the original thread where you pretended to know something that you didn't.
Not that it should surprise anybody.
The more illuminating post was the one I linked above explaining that there is only one playback filter for HDCD. That was entire bone of contention in that old, silly, lengthy thread. And now it is settled, definitively.
I have privately communicated with at least 4 different people who have reverse engineered HDCD, all of the independently. Not one of them has mentioned you once.
Certainly it is possible that you had something to do with *one* of them, I have no way to know, but as I said not one of them mentioned you or any other outside influence. But since there are three others that without question had nothing to do with you, I'm not sure what you are so puffed up about.
.
Is there a description of how the reverse engineered decoder works on the web?
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
If you PM me, I can try and retrieve the email exchange between myself and the author.
From memory however, the decoder works by scanning the bitstream looking for HDCD packets, then decodes the peak expansion flag. If peak expansion is turned on, then all samples above a certain value get expanded via a lookup table. By feeding the Microsoft HDCD decoder with a doctored set of samples, the lookup table can be easily reconstructed. Further tests were then done to compare between the two decoders to ensure they were bit identical with respect to each other.
Link below.
Remembering that the Asio plugin only works for Audio files.
It will not work for DVD MPEG etc.
Under control pannel, there is "add/remove programs".
Select MS media player and press "remove".
:-)
As others noted, there are many other much better players for free.
> Under control pannel, there is "add/remove programs".
> Select MS media player and press "remove".
Very funny.
Or the OP can just ignore it and install some other player. Lots of music players to try in the Windows environment.
I use J. River Media Center 12. Foobar2000 has an enthusiastic following.
On this forum, Cplay/cmp is very popular. However, it might be something to try after the OP gets up to speed with the basics of ripping, tagging and playback.
Bill
Skip WMP.
Many better sounding free players like cPlay out there to try, and IME there is usually a trade off with pretty pictures and sound quality...
I wouldn't use Windows Media Player. There may be a way to extricate its player functions from the poor quality (and somewhat arcane) operating system mixer and sample rate converters, but why bother. The player itself is useless for one reason alone: it does not play "gapless". This means any album that has music that plays continuously across tracks (e.g. many live recordings) will have glitches at the start of each track. That this has remained the case for multiple releases of the product is proof enough to me that the player is junk and the people maintaining it are not doing a competent job. (These remarks based on WMP version 11 running on WXP SP3.)
Another thing to watch out for is that the Burn function that is built in is also incompetent. It includes the option (which you should not check) to do volume leveling (normalization) on a track by track basis. This degrades the sound of any 16 bit material unacceptably and destroys the volume balance made by the mastering engineer that makes for a smooth listening experience when an album is played straight through. Worse, it inserts two seconds of silence between each track, so that the CD that is burned is not the same as the original CD that was ripped.
If you must run WMP for some stupid reason, then it is possible to make it play FLAC files by downloading and installing the appropriate directshow filter.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Stop all unnessary features.
I suggest you compare it with others like Foobar
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: