Atrac3 LP2(132) vs Atrac3Plus HiLP(64)

[Avrin]

Well, LP2 (ATRAC3 132kbps) really rocks when recorded to a Hi-MD. At least its trackmark shift error is almost insignificant (as compared to any ATRAC3plus bitrate). And it is not padded on a Hi-MD.

And almost everyone is going to survive its sincere 17.5 kHz cutoff, which allows it to encode the lower (hearable) frequencies with really good quality.

Edited November 10, 2006 by Avrin

[Justin42]

I see people saying this more often, but I disagree. If you're transcoding a 128kbps mp3 to a low bitrate Atrac file, you're just making it worse. You're right in saying you can't improve a 'crappy' original, but you can most certainly make it even worse.

It's actually more like taking a picture of a picture. If the original picture is grainy/low quality, and you photograph it with a poor camera, the resulting picture will be even grainier/less quality. However, if you photograph it with a high-end camera, the resulting picture won't be better than the original, but at least it won't be much worse than it either.

I just wanted to second this; taking a 128kbps MP3 to LP2 is definitely going to make a crappy sounding MD. You want to encode as high as possible to minimize the loss of quality.

Try it for yourself-- if you can't hear the difference, that's one thing (everyone's hearing is different), but there is definitely a dropoff between MP3->HiSP and MP3->LP2.

[1kyle]

I see people saying this more often, but I disagree. If you're transcoding a 128kbps mp3 to a low bitrate Atrac file, you're just making it worse. You're right in saying you can't improve a 'crappy' original, but you can most certainly make it even worse.

It's actually more like taking a picture of a picture. If the original picture is grainy/low quality, and you photograph it with a poor camera, the resulting picture will be even grainier/less quality. However, if you photograph it with a high-end camera, the resulting picture won't be better than the original, but at least it won't be much worse than it either.

Making the file Bigger won't improve it. Even a WAV file made from your 128 kbs mp3 file won't sound better than the original - you will still lose a bit.

Whether converting the 128 kbs mp3 file to a similar bit (132 ATRAC LP2) file is worth while would depend on the accuracy of the conversion algorithm and the quality of the MP3 file. For Lowish quality sound it might work acceptably. I would in general avoid transcoding a compressed (lossy) file to another compressed (lossy) file in any case.

If you have the original in say FLAC, WAV or Compressed Lossless ATRAC then converting to MP3 would be far better. --Again the quality of the original source might render this as overkill.

Your ears should be the only real judge. --We can theorize about the maths and physics but it's what your own ears hear is how you'll probably proceed in any case.

Cheers

-K

[Sony_Fan]

ATRAC3plus @ 128 is available in SonicStage but it can not be transferred to Hi-MD disc. LP2 sounds a lot better than MP3 @ 128, but @192 it becomes almost identical if not a little better than LP2.

[carlesraf]

for me HI-Lp sounds betthet than Lp2.....and atrac3plus at 48kb sounds quitte equal to LP4

[jpteasy]

I've been following this discussion with interest. Since the hiss of cassette days I've always wanted my music to be crystal

clear. When MD came on the scene in the UK I was so happy.

Obviously over time Sony has modified and improved the MD experience and I think it has reached a point now where

MD/HiMD covers all audio tastes. Ranging from PCM to 48 Atrac3plus.

Unless you are an audiophile I think that PCM is overkill unless you are lucky enough to have one of the Onkyo HiMD

separates or use HiMD through a high-end stereo system.

I did a test using an Oasis track recorded digitally from a Hitachi portable DVD player. The track was Mucky Fingers and

involves alot of percussion coupled with Noel Gallagher's high pitched vocals.

The formats I used were; PCM, 352kps Atrac3plus, Standard SP, HiSP, HiLP and 48kps Atrac3plus. (I'm thinking of trying

another test with the old MDLP formats to see the difference).

PCM and 352kps were indistinguishable to my hearing (I'm 27). You could hear clearly the lisp when singing words with an 's'

and his breathing when pausing etc. plus all the background information.

The difference I heard between PCM/352 and Standard SP/HiSP was just a slight loss of some of the background information,

but it was still there.

Between HiSP and HiLP there is a noticebale difference however, it had merely dulled the 's' sounds and the pauses. The

cymbal sounds had changed slightly, but nothing major. The majority of the background information disappears, but you

quickly forget that this is missing and unless you listened to an uncompressed version previously would probably not notice.

Between HiLP and 48kps you immediately begin to hear the artefacts appearing in the high pitched sounds and the distortion

of the 's' sounds. There is a lack of background information which limits, for me, the enjoyment I can get from the music.

In conclusion, in a previous post I mentioned my project to place onto MD all my favourite music in a variety of formats.

Currently MDLP isn't used, but after reading this discussion I will be doing tests to see if it is worthwhile using it.

For my recording I use; Atrac Advanced Lossless (on the PC @ 256 and 64) Standard SP, HiSP and HiLP. I rarely use my

NetMD, preferring to use my MZ-R35 or MZ-RH710 so these 3 formats suit me perfectly.

Just to clarify though for some who have mentioned other bitrates, when transferring from Sonicstage 4.3 you have a larger

choice of bitrates you can send to your HiMD. You just can't use those bitrates when recording using the unit.

PS Don't forget the beauty of MD/HiMD is that you don't have to compromise on quality or convenience like you do with an

MP3 player. Just take an extra disc or two in your preferred format.