Tangible or not - CD -v-MP3

[Barbirollians]

No doubt I shall be told all about FLAC and the suchlike but I had an interesting experience today on receipt of the Tower Records CD from Japan of Beethoven Piano Concerto No 3 with Annie Fischer and Ferenc Fricsay - coupled as on the LP with two Mozart Rondos .

I have over the past 18 months become very smitten with the Beethoven having downloaded the MP3 on to iTunes from Amazon there being no tangible format of the recording available in Europe . I have played it all over the place , through iPad, iPod and car radio , Sonos system and Airport Express through my living room stereo and obtained a great deal of pleasure from it .

I was , however, amazed at how much better yet it sounded on CD today - considerably more detail and airiness around the sound and the quite exquisite interplay between the orchestra and Fischer in the slow movement in particular was much enhanced .

Perhaps this explains why tangible formats are more than hanging in there in classical music .

[David-G]

Exactly! I had a similar revelation when I first heard R3 on HD Internet as compared with FM.

[richardfinegold]

Well, you already alluded to the reason. You have been listening to an MP3 download, which typically contains about 6% of the information as a CD or lp or uncompressed download. It ought to be revelatory when you get a chance to compare an MP3 with the real thing

[jayne lee wilson]

Exactly! I had a similar revelation when I first heard R3 on HD Internet as compared with FM.

But which did you feel sounded better, and why?

[Bryn]

Well, you already alluded to the reason. You have been listening to an MP3 download, which typically contains about 6% of the information as a CD or lp or uncompressed download. It ought to be revelatory when you get a chance to compare an MP3 with the real thing

Eh? Surely the degree of information would depend on the data rate of the mp3. At 320kbps it would be of the order of 50% of that from a CD (there's a lot of musically redundant data in the CD Audio spec.). Bear in mind that a lossless CODEC typically has considerably less than half the data rate of the CD Audio from which it is derived.

[rauschwerk]

I think it's pretty well established that mp3 encoding deals least well with transients. To my 70 year-old ears, string quartet sound is unaffected except at bit rates of 128k and below, whereas for harpsichord recordings I insist on FLAC or CD.

[Dave2002]

Re the first post - I have that LP - Heliodor I think. Where does one order the CD from? How much?

[Barbirollians]

I got it from a Japanese marketplace seller on amazon france Dave . It was amazingly cheap about 13 euros . Whether there are any left I do not know .

I should hope that Eloquence might release it - otherwise it lurks only in that massive DG Fricsay box .

[richardfinegold]

Eh? Surely the degree of information would depend on the data rate of the mp3. At 320kbps it would be of the order of 50% of that from a CD (there's a lot of musically redundant data in the CD Audio spec.). Bear in mind that a lossless CODEC typically has considerably less than half the data rate of the CD Audio from which it is derived.

I've read that some iTunes downloads get into the single digits, and I've heard some that were truly awful.

[mikealdren]

As other have said, I would expect the CD to sound better than MP3, especially if the latter was very compressed. However, were both recordings on the same label. Some recordings also vary dramatically between different CD remasterings.

Mike

[Ferretfancy]

Re the first post - I have that LP - Heliodor I think. Where does one order the CD from? How much?

I have it in the DG box of Ferenc Fricsay recordings Volume 1 which has many fine performances mostly in mono but the Beethoven is in good stereo.
It sounds excellent, as does the Heliodor LP, but the box will set you back about £79 for 45 CDs

[Dave2002]

YThere can be a whole bunch of reasons why the mp3 sounds worse than the CD. Originally mp3 developed out of research into low bit rate communication, and at first it was hailed as a great invention by many who used it at the end of the 1990s to send songs over the internet. Undoubtedly it was at the time, as data rates were so low as to make any higher quality reproduction prohibitively difficult over data channels at the time. The developers of various compression systems thought they had got something which worked, and at one time they used a song Tom's Diner, by Suzanne Vega as a test. She heard it,and said it was no good so they tweaked things further. Another factor was stereo. After some fiddling the developers managed to get acceptable single channel recording, but when they tried to use two channels for stereo that also didn't work well. Mp3 can now be encoded in joint stereo or as two distinct channels. Sometimes one gives better resilts that the other.

As effective data rates increased on communications channels, there was less and less need to use harsh compression, and at bit rates of 192kbps and above using mp3 many people may find the sound quality satisfactory. There are parameters which can be optimised during the encoding process, and some encoders do give results which are perceived as having better quality. Probably quite often with higher bit rate mp3 encoding, problems are simply due to poor quality control, rather than inherent failings of the encoding system. Companies may try crass things such as reencoding low bit rate mp3s or other compresed formats at high bit rates. This is hardly likely to improve things, and may sometimes make things sound a lot worse.

Some mp3 encoding might be perceived as benign, perhaps simply giving the impression of a modest high frequency cut, which may not sound unpleasant. Other encoding may give rise to perhaps a barely audible harshness, but that can become wearing over time. I know of no really serious study of the perception of compressed audio formats. Mention is often made of artefacts, but these are somewhat unspecific, and vary from one encoding method to another. One view could be to treat these as noise, and this would reduce the signal to noise ratio even if no obvious harsh distortion was apparent. This would very likely reduce the apparent volume of the sound, making it sound less vivid compared to a CD.

Turning to the CD, it is also possible that different masterings from a given original source may sound different - some signficantly better than others. A year or so back I bought some jazz CDs and I listened to some on streaming services (probably via Amazon - auto rip). When I put on the CDs I found them much more enjoyable, positively toe tapping, even though at first they sounded the same. Whether the CDs were significantly remastered compared with the original releases which were probably from 20 or more years ago I'm not sure, but they sure sounded much more enjoyable than any compressed audio versions.

The whole thing doesn't stop there though. I think equipment quality also plays a part, and one's perception may not be due to compression alone. It may be that the apparently poor quality of compressed audio, such as mp3, disappears if the encodings are good and the recordings are played through high quality dacs.

[David-G]

But which did you feel sounded better, and why?

It was just as B said: compared with FM, R3 on HD Internet showed "considerably more detail and airiness around the sound". Also, the whole stereo spatial effect was much clearer. I should add that I have a high-quality DAC.

I don't know if FM sounded perceptibly better in the old days. I have an expensive tuner, and always fancied that the FM sound was very good. But when I got the DAC, the sound of R3 on the Internet was a revelation.

[Eine Alpensinfonie]

I don't know if FM sounded perceptibly better in the old days.

It did, if my open reel tapes (recorded on a Tandberg 3341X) are to be believed.

[richardfinegold]

MP3 have their place. I am on vacation in Florida, listening to JEG in Handel, streaming MP3 via Bluetooth to a $50 BT speaker, while the wife and I sip a glass of wine before going out for dinner, and all is right with the world. Earlier I was at the beach listening to MP3 of the Krips/LSO Beethoven cycle via some Klipsch earbuds from my Android Phone, and there was no shortage of warmth, ambience, and detail. Who needs an expensive portable player that can do High Resolution Files through expensive headphones when MP3 can sound this good? If I played the same files through my regular system, however, then the deficiencies become obvious and glaring, and that is what Barbs has just discovered.
Btw, I own that Beethoven 3rd recording, having picked it up on lp in Salzburg a couple of years ago. I just sold my turntable last week and am now listing the Lps on eBay.