MP3 for Archival Quality? Is MP3 a good CD backup?
MP3's will never be CD quality. The data has already been compressed....
DVD's really don't go bad, they could become scratched or whatever. If you are really that worried about it just burn back up copies. |
Thanks Deter for the reply.
Quote:
Quote:
So if I already have a good copy on second class media it won't degrade in a number of years, assuming that I have handled & stored them properly & they haven't been scratched? Whew... That is a big load off my mind... :) None-the-less, (now that I know better) I'm still going to burn all my new stuff on Verbatim media, unless of course they become downgraded in quality like TDK did from first to second class. And I will eventually burn a back-up of all my stuff, or at least the best of it, to be stored in a seperate location, just-in-case, but I won't be pressured to do so in a big huff, since I now know that my exitsing copies are going to last, even if they are RITEK & CMC MAG. Thank you for this very valuable in-put. So is the only reason why second class media is not recommended for archival purposes because of a higher burn failure rate and nothing to do with long term storage of the accurately burned media?:( |
Quote:
Compression does not automatically equate to loss. Compression is an attempt to better use available space. MP3 was developed for easy storage and transmissions over the weenie-sized hard drive and dial-up modem connections found in the early 90s*. While MPEG does involve some loss, it was always meant to be acceptable loss or transparent loss. You can see/hear acceptable loss, but you cannot really see/hear transparent loss. You achieve transparency by using higher bitrates. It's virtually impossible to tell the difference between 320kbps MP3 (using a good encoder, such as Fraunhofer's -- the developer/inventor/owner of MP3) and a 1536kbps PCM CD-Audio. Most of the "loss" is at unused frequencies stored in PCM, including those unheard to human perception. I have near-perfect pitch (former brass musician for about a decade), and have a rare form of hyperacousis. To help you understand what that means, I use this analogy: "If my ears were my eyes, I'd have 10/10 vision." I don't experience dynamic range loss as others often do, but I do still hear everything louder, including all the softer sounds and the higher pitches. I can hear compression in 128kbps MP3, sometimes even with 224-256k compression, but when you start to go into 320k-448k range, there's no loss to be heard. I can't hear any difference, aside from the usual decibel changes (up or down) between encoders or due to the decode/re-encode. Most self-proclaims "audiophiles" are talking non-sense, when it comes to digital compression. That includes
Vinyl just sounds different that CD, because of how audio mastering has changed decade to decade, over the past 100 years. CD is cooked, even when PCM. Most people just don't know how to encode MP3 very well. And that's honestly where most loss occurs. Using the wrong software, bad bitrates -- that's the real enemy to MP3 quality, not the MP3 specs. Download some quality MP3's off Amazon (99 cents each), and then compare to some illegally made torrent/download version. HUGE DIFFERENCE! I had songs from the Napster days (1999-2001), like many people, but have since trashed most of them, having bought the much higher quality studio versions off Amazon. Money well spent! * http://www.mp3licensing.com/mp3/history.html Going further back, MP3 was developed for sending music over phone lines pre-Internet. It wasn't until 90s completion of the tech that modem/Internet use became an obvious media where the format would exist. Compare the early 70s thesis to the 1992 implementation. MP3 has been a work in progress for longer than CD-Audio! Most people don't realize this! Quote:
|
Admin,
Kind of amazed that you think MP3's don't have major quality loss. Granted most people can't tell the difference. However, if you set up your system correct...Good Amp, Great Speakers, fiber to run the sound.....You close your eyes or whatever u do to listen to sound, it is a big difference. Even in the car...I am not talking about listening to stuff on your PC..... Sometimes I have to deal with only having an MP3 and not the sourse, it is harder to work with. I can always tells the difference, the top level of the sound is cut off... I have remastered maybe 150-200 CD's, sound is my thing... |
Quote:
Another issue: my experience (with iPods in particular) has taught me that MP3 players do not play back audio as well as CD players do, regardless of the technical specifications of the files being played. MP3s I've encoded at 384 kbps sound terrible on an iPod, at least as bad as 128 kbps MP3s played back on CD-R. |
I've moved this separate topic to a new thread. :)
Again, deter, it really depends on
Jargon defined: DAC or D/A = digital to analog converter ADC or A/D = analog to digital converter SNR = signal to noise ratio My car is the last place I'd listen to music to hear the full dynamic range, even with decent speakers in it. For one thing, the acoustics in a car are rubbish. The sampling, dither and aliasing are all just as important, regardless of whether it's MP3 or PCM CD-DA. Some of the hardware servers that compress MP3 at the studio level are very impressive. It's nothing any of us can afford to use, and many people often mistake their own work, using homemade software (or even open-source homemade software) as the end-all, be-all of MP3 quality. It just isn't so. Some of these studios go so far as to hire or consult with audiologists to ensure maximum quality of their music! I actually like listening to vinyl, because it has a different "temperature" to the audio. It's often a warmer sound. In addition to that, it's generally not "cooked" to death, having a wider dynamic range than the punched-up CD levels we hear so often. Technically speaking, CD-DA has a bigger dynamic range, but it's all about implementation! I'm also not one of those daft folks who insist vinyl is better than CD -- that's just silly talk. There's no easy "versus" answer between these formats, it varies from master to master. :) I can think of several MP3's that sound better than the record or CD! It's just a better mastered file. |
Yes, read this: http://www.audioholics.com/education...ion-depression
Quote:
This also fits into my similar video rant, about how people prefer crap over quality... Read http://www.audioholics.com/news/indu...or-quality-mp3 Quote:
|
I, like admin, have diagnosed super-hearing (not diagnosed as having hyperacusis as far as I know unless the audiologist just explained it to be in different terms, but I have a feeling I may have it for reasons I won't go into here, just haven't followed up on it yet), and generally don't hear a difference between CD quality and a well-encoded MP3 at a decent bitrate (I rip using Exact Audio Copy and encode w/ LAME using the V2 setting, and I've ABXed the results).
That said, MP3 is not an archival format because it's lossy. Even if you like how it sounds, it should not be used as a backup. For backups, use a lossless codec like FLAC, WavPack, Apple Lossless, or WMA Lossless. These compress only by removing redundant data, closer to ZIP and RAR than MP3, WMA, AAC, and Vorbis. I'm honestly kinda shocked I was the first one to mention lossless compression. :) As far as MP3s go, stick with a good encoder like LAME (which Amazon uses most of the time now) and use variable bitrate encoding, which uses more or less bits when needed. LAME's VBR encoding is fine-tuned using a range of pre-sets from V0 (highest) to V9 (lowest). V0-V3 are theoretically transparent most of the time, and V2 (~190 kbps in theory, but it can vary a lot in practice) is the default on a lot of programs that use LAME, like Exact Audio Copy and Foobar2000. For CD ripping in general (regardless of the output format), you should generally use Exact Audio Copy, which is the consensus best ripping program. It can use a variety of external encoders, including but not limited to LAME MP3 and FLAC. |
Quote:
As hard drive volumes continue to grow, I find it best (and sensible) to entirely skip compression and just stick with uncompressed files. I value the history and standardization WAV provides; I can't trust my files to some newcomer format that may not be here in five years. Also, I may need those audio files for a future project and want them in a format both standardized and uncompressed for maximum flexibility. |
FLAC (and other losslessly compressed files) decompress to LPCM WAVs in a few seconds, and I've never heard of anyone having a problem with the track being "wrong" in losslessly compressed form and/or when it was decompressed back to WAV. You don't lose the ability to have a WAV at all. That's kind of the point.
Sure, hard drives are bigger now, but there's really no disadvantage to using lossless compression if you want a lossless file. Why not lose at least 40% of the space? Why worry about about the format disappearing, especially when so many of of the major lossless compression formats are open source? There will always be encoders and decoders for them. That's just worrying for the sake of worrying. |
Quote:
Quote:
|
Oh, and the compressed formats have tagging. WAV and AIFF don't.
|
Quote:
|
I would remind everybody what the point of a "backup" is. A backup is simply a copy of something, in case the original is lost, damaged, or otherwise unusable or unavailable. There's no rule that a backup has to be identical to the original.
For example:
If I wanted a perfect archival-grade duplicate of a CD, I'd copy the CD 1:1 to a new CD-R, as well as rip it as an ISO to a hard drive. And given how I can buy a 2TB hard drive for about $100-150, I'm not going to suffer much when a CD image is just 700MB. If I just wanted another copy, "just in case", then MP3 is fine. FLAC is fine. Of all the lossless compression, I'd opt for FLAC. But it's still not as portable as MP3. I'd disagree about lossless encoding (FLAC et al) being future proof. I've seen so many techs come and go that I don't even remember half of them. The market is what pushes these things, and the market is still pushing MP3, albeit higher bitrates than we did 10 years ago, and from higher grade masters (not just using bad CD rehashes). I name my files intelligently, I never bother with MP3 tagging. A number of players operate from filenames anyway, or have the choice to use filenames instead of MP3 ID. Most of my MP3 files are encoded with Fraunhofer (not LAME), or outright bought from studios (Amazon, mostly). Fun debate we're having. :) |
Any reason that you choose to use Fraunhofer? It's routinely beaten by LAME in listening tests, and LAME is even what Amazon uses now. There's nothing wrong with the Fraunhofer encoder, but LAME seems like the clear winner nowadays. Well, Helix has beaten it in some recent listening tests (and it's a testament to the OSS movement that once Xing's source code was released, it was tweaked into a top lover encoder) but nobody seems to be using it.
|
Tests seem to go both ways, and the quality and age of those tests vary quite a bit. For example, this could have been an interesting test at http://listening-tests.hydrogenaudio...-1/results.htm, but the bitrates are too low, and I think the statistical analysis is flawed (simple stats, rather than mean stats). Simple stats skew results. For example, 1+50+50+50+500. The average is clearly 50, not some 3-digit number as you'd get on a basic div5 sum. A lot of these tests shoot for VBR, too, which is sort of silly, seeing how many players need CBR encodes.
I've found Fraunhofer sounds better at the higher bitrates especially, is excellent at CBR encoding, and generally has less artifacts (if any). I think LAME is optimized for file size and speed more than quality, or at least that used to be the case. In many cases, it cooks the audio. It's also spent a good bit of development on VBR encoding. However, LAME is not released as its own encoder, but rather as source, so the various implementations may be at fault for some of these things. As with anything else, LAME vs Fraunhofer really comes down to a case-by-case basis (based on content), and personal preference of the listener. My favorite heavy metal and classical music isn't going to encode the same as some rap garbage or shrill pop music. Content is too often overlooked, when it comes to bit rates and encoder types, be it video or audio. |
I hate reading article like this: http://news.cnet.com/8301-13645_3-10435038-47.html
Quotes like this annoy me: Quote:
I like some of the blog comments: Quote:
Quote:
When I work with audio, I do so in a near-soundproof room, using good equipment. But even in that environment, the best vinyl, CD or MP3 still doesn't compare to a concert hall with well-tuned string, percussion, wind and brass. Some things just can't be replicated. |
Quote:
LAME and Fraunhofer can both be found, and it can vary between ABR, CBR and VBR. It really depends on what the content is, etc. Even Apple iTunes uses certified labs for their work, Apple isn't doing the encoding to my knowledge. I have most of those submission specs, too. Most studios would love to use AAC, as there are no royalties. Every MP3 sold sends some pennies to Fraunhofer, because they own the MP3 format patents. Same for DTS and Dolby Digital AC3, royalties are paid. |
When I say uses, I mean just that their files include LAME-encoded MP3s. I'm aware that they may not be encoding them themselves. At any rate, there's not going to be a huge difference 256 kbps (which is Amazon's target bitrate, as they use that at CBR and ABR, as well as using LAME V0 for VBR) anyway.
That said, for what it's worth, Netflix has said that they do their encoding themselves from whatever the studio provides them with, so it's not always the case that the content owner is doing the encoding. AAC still has separate licensing fees comparable to MP3 even if it doesn't have the electronic music distribution royalty fee (2%) that MP3 has. I'd love more widespread AAC support (while there isn't a huge difference as bitrates get higher, it's definitely better at 128 and lower, especially with the HE-AAC profile), and frankly I'm kinda surprised that Vorbis support hasn't become a bigger deal with DRM dying out, or even before that just for so manufacturers can add more features to pimp their products. |
Yes, AAC is part of software/hardware development costs, not per-use licensing. Correct.
More on that at http://www.vialicensing.com/Licensin...AQ.cfm?faq=5#5 for anybody that's interested. Given that the Netflix blog is more than 16 months old now, and streaming is one of the few things that is "changing all the time", that may be inaccurate now. There's a lot of Netflix materials in some of the trade pubs, and I'll admit I've not read them all, although I've ripped out a number of articles for later browsing "when I have time" (which never seems to come). OGG to market products -- I an see that happening, although more mass-adopted codecs always comes with the risk of consumer backlash. That's one reason MP3 has hung on for more than a decade -- people finally understand what an MP3 is as a product (even if they have no clue about the tech). To them, any downloaded music is an "MP3", as the term is now synonymous with "downloaded music" (legal or otherwise). "OGG" would probably appear to be a Polish or Russian cuss word to most people, the tech-ignorant ninnies that make up the masses. They'd have to brand OGG as another "brand" of "downloaded music", similar to how Ford and Chevy are both cars. It's not too difference from "xeroxing" something, or "googling" something, where words are misused by consumers. It took major efforts to separate "Xerox" from "copy machine", and we're still stuck with "googling". MP3 = music, so forcing OGG (or anything else!) into the mix is an uphill battle for sure. |
Site design, images and content © 2002-2024 The Digital FAQ, www.digitalFAQ.com
Forum Software by vBulletin · Copyright © 2024 Jelsoft Enterprises Ltd.