I have just updated the Malmsjö Acoustic Grand website with two demos in new genres, one jazz, and one impressionist piece. I have included full-length versions as well as excerpts in 250kBit/s, 128kBit/s and 96kBit/s.
All \"java-buttons\" are now removed because they did not function on all computers.
After reading your suggestions I bought the Mp3 Producer Professional which has Fraunhofer encoder. After comparing the output at 256, 128 and 96, I decided to use the Fraunhofer on the 256kbit/s sample but stay with the Xing on the two lower rates. I read up a little bit on these encoders, and found that where Fraunhofer tries to reproduce the frequences all the way up to 20K, the Xing and the LAME for instant cuts at 16k. The Fraunhofer seemed to distort HF more at the lower bitrates when applied to these samples.
Below is a website with a very thorough test of different encoders. It does suggest that the Franhofer would be the best encoder at 128kBit/s. My own listening tests did not confirm that on these recordings. The Fraunhofer HF sounded distorted at 128kBit/s but the Xing mostly ignore everything above 16k.
Hans, I visited the site you mentioned where mp3 encoders are rated. I can\'t believe the rating of Xing at 128! The test results, if they are accurate, suggest the Xing encoder is equivalent to or better than Lame, especially above 16 k. My experience completely contradicts this result. I moved to LAME from Xing (at 128) because with Xing many of my files were unlistenable.
Anyhow, it\'s a shame that most mp3 sites still use 128 as the standard, since it is demonstrably inferior to current cd standards.
I listened to the new high bit-rate classical offering. It sounds every bit as good as I thought it would. And like Sam, I\'ll be ordering your sample whenever it\'s ready!