Jump to content

Is Hi-SP better than SP?

Rate this topic


sub27

Recommended Posts

The original atrac was limited to 292kbps only, while the newer atrac3plus comes in different flavors like 192, 256 (Hi-SP), 352kbps.

Although both SP and Hi-SP have been designed with transparency in mind, what sounds 'best' is up to your own ears. You might find 192kbps perfectly acceptable while others hear artifacts even at 352kbps. SP can't be used on 1GB Hi-MDs and takes considerably more space (less efficient formatting of disks). Software transfer of SP will give inferior quality (fake SP - re-encoded atrac3 132kbps LP2).

Link to comment
Share on other sites

Actually classical music needs somewhat lower bitrates than is widely believed. Try to compress some classical music into ATRAC Advanced Lossless or FLAC. You'll be impressed by the size reduction in comparison to other styles, especially jazz and trance.

The cause of that is simple: harmony.

Edited by Avrin
Link to comment
Share on other sites

I acutall meant favorite tunes, especially classical stuff. But it depends entirely on the piece and the quality of the recording. Some are ok, some are simply nicer in a higher bitrate. My current fav is Violin Sonata Op. 5 No. 12 - Corelli - Jefferson in Paris soundtrack .

Edited by Sparky191
Link to comment
Share on other sites

This is a question that keeps appearing but the answer is down to individual perception. If you find a lower bitrate to be acceptable then don't be persuaded by someone else to go higher just because they think it's better! It's your ears and your enjoyment (and your disk capacity too)... :D

Link to comment
Share on other sites

The best thing to do is record two or three of your favourite tracks (that you know very well) in EVERY bitrate.

Play them all back on headphones and your hi-fi system to see how low you can go for tolerance in terms of quality. My limit is 132kbps ATRAC3, but 192kbps ATRAC3+ in terms of Hi-MD is the lowest I'd go.

One niggle I have, why have they not enabled the RH1 to record SP on 1GB discs on the unit? Would of been brill.

Link to comment
Share on other sites

The best thing to do is record two or three of your favourite tracks (that you know very well) in EVERY bitrate.

Play them all back on headphones and your hi-fi system to see how low you can go for tolerance in terms of quality. My limit is 132kbps ATRAC3, but 192kbps ATRAC3+ in terms of Hi-MD is the lowest I'd go.

One niggle I have, why have they not enabled the RH1 to record SP on 1GB discs on the unit? Would of been brill.

One shouldn't get too bogged down in the actual "Numbers" here. Over the years since SP came out advances in mathematics and physics have led to better and more efficient compression algorithms. This means you can in theory get better quality than you could have got before even though the file is smaller and more compression has been applied.

For its day SP was absolutely top dog -- Mp3 /Flac/others hadn't been even thought of then so it was either full CD WAV or ATRAC.

Most computers --if you could even get your hands on one back then were pretty terrible and not used for any sort of domestic music recording / editing / manipulation.

The best you could get was a CD with an optical out which you recorded to your MD in real time. A few combo type units (with both a CD and an MD) existed which would allow you to dub your CD ==>MD at usually 2X or rarely 4X.

Whilst there's no question whatever that SP is superior to LP2/LP4 I doubt whether it is better than HI-SP @ 256 or 352.

A decent blind test should determine this pretty quickly.

I usually save my music in ATRAC Lossless (@352) since this compresses (losslesslly) on the computer fairly efficiently --and if I want to convert to flac or whatever I can convert idividual files back to WAV (again losslessly) and then do my editing.

I only tend to store music on a computer as a backup -- I rarely use a computer for actually listening to music --got a nice deck for that purpose.

Now another possible issue that might "Obfuscate" the above conclusion is that the amplifier in the RH1 is pretty good --so recording in SP might be better than recording in SP with older equipment --however this could then equally apply to all the other modes of recording so the differential(assuming there was any) would still be maintained.

I doubt that a decent SP recording however would disappoint.

Cheers

-K

Link to comment
Share on other sites

Just try recording Miles Davis "kind of Blue" in old SP : a lot of horrible artifacts on the trumpet.

Try it with HiSP: the artifacts have totally disapear!

Hisp >> SP !

Link to comment
Share on other sites

When you convert from ATRAC Lossless to WAV, are you certain its using the lossless part? There was a bug with this in converting to other lossy ATRAC bitrates. Has that been fixed?

Hi Sparky -- I think you only even SEE the lossy Atrac files on your computer once you've transferred music to MD from SS. You can easily check this out by just transferring 1 track for which no "lossy" file exists. You'll then see the same track has both an ATRAC Lossless (the original uploaded file) and and an ATRAC 352 file which is what actually gets written to your MD.

If you transfer to WAV I don't think it will use the Lossy file --it can't in any case if the lossy file doesn't exist.

Incidently a 1GB disc holds around 5 1/2 - 6 hrs of music @ 352 compared with about 7 1/2 @ 256. I think using the highe bit rate is well worth it (especially with decent headphones). If you are just using cheap ear-buds then I don't think you'll notice any difference between 256 and 352.

Cheers

-K

Edited by 1kyle
Link to comment
Share on other sites

with the PX100 I can hear the difference between LP2 and HiSP, but not significantly between HiSP (256) and 352 and not with an VBR quality 4 Lame MP3 either (which is about half the size and which I can transfer from mac -> RH1) so guess what I'll be using?

PS: it just really sucks that MP3 can't really do gapless (unless you've got a rockboxed MP3-player and way too much time to spend on lame commandlines)

Link to comment
Share on other sites

Hi Sparky -- I think you only even SEE the lossy Atrac files on your computer once you've transferred music to MD from SS. You can easily check this out by just transferring 1 track for which no "lossy" file exists. You'll then see the same track has both an ATRAC Lossless (the original uploaded file) and and an ATRAC 352 file which is what actually gets written to your MD.

If you transfer to WAV I don't think it will use the Lossy file --it can't in any case if the lossy file doesn't exist....

On a quick test a single track

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Atrac Lossless 064k = 23,713 kb

Atrac Lossless 352k = 27,338 kb

No files in the optimized folder. Conclusion that Atrac Lossless itself contains the lossy file.

I'll generate a Wav from each and see if theres a difference.

Link to comment
Share on other sites

On a quick test a single track

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Atrac Lossless 064k = 23,713 kb

Atrac Lossless 352k = 27,338 kb

No files in the optimized folder. Conclusion that Atrac Lossless itself contains the lossy file.

I'll generate a Wav from each and see if theres a difference.

Ok WAV from CD = 35,618kb

ATRAC 064kb from CD back to WAV = 35,631kb

ATRAC 352kb from CD back to WAV = 35,631kb

WAVE from Lossless 352kb = 35,551kb

WAVE from Lossless 064kb = 35,551kb

Conclusion that the WAV from any lossless is the same. Though different from Original WAV. That said the results maybe misleading, as SS maybe (very likely) padding the file for some reason. I don't know enough about who it creates the WAV file. It would be interesting to compare the waveforms. Maybe someone could do that.

Edited by Sparky191
Link to comment
Share on other sites

Funny that somebody brought up Miles Davis; SP was terrible at encoding things like trumpets. I remember wanting to throw the disc at the wall when I heard a bunch of weird, electric-sounding artifacts on "Kind of Blue!" Hi-SP is better at this, but still not perfect - sharp clicks and attacks sound metallic and artificial to my ears. I've given up on ATRAC, personally, and only use PCM for serious listening on my discs. Once you know what to listen for, artifacts can be heard even at 352kbps with ATRAC3plus on certain signals. For portable use, though, who cares? In a noisy environment, Hi-SP is just fine, if not even overkill.

If I had an RH1, I wouldn't even bother with ATRAC3plus, and would instead use LAME -V0 or --preset insane (great sound, a universal format, and the ability to use better rippers like EAC).

Link to comment
Share on other sites

I resolved this for me personally (PC and Mac available, though mac gets most use by far and an RH1) by doing as follows:

- cds that do not require gapless to be enjoyed: Lame VBR new, quality level 4 (mostly ripped with EAC + USB-stick -> mac or directy with iTunes Lame) which I can download from mac and is about 160kbps average

- cd's that do require gapless: transfer with PC in HiSP (through SB) or 192kbps (through SS 4.0)

this saves quite some space, delivers great SQ for portable use and allows me to create discs with mac

Link to comment
Share on other sites

  • 1 month later...

Actually classical music needs somewhat lower bitrates than is widely believed. Try to compress some classical music into ATRAC Advanced Lossless or FLAC. You'll be impressed by the size reduction in comparison to other styles, especially jazz and trance.

The cause of that is simple: harmony.

Avrin, that's an intriguing claim. I don't quite buy it. Do you have some references to back this up? I'm not trying to corner you. It could be true, but as I think about it, there are other forms of music, especially the two you just quoted (jazz and trance) which have plenty of harmony. I haven't done any research on this. I know something about both computer science and music, but am not an expert in data compression algorithms. Here we're speaking of traditional non-lossy data compression optimized for music.

There are arguably more "complex" or "interesting" chords or harmonies in classical music, as a generalization, although even that claim would bring strong criticism from jazz fans.

Electronic or "trance" music, being generated by a computer, is going to be the most precisely in tune, mathematically, and have the most regular beat, as well. And the most repetition of sounds. Often no vocals. All those things lead to more compression. It also does have harmonies. They may be simple in nature, but even that would seem to argue for MORE compression, not less.

I picture an orchestra. 100 instruments playing very well in tune, but each performer contributing their own tone and very slight variation in pitch, especially on orchestral string instruments which are all fretless. The result is a powerful artistic sound appealing to the human ear, but seems like murder on a compression algorithm, to me.

There are beautiful harmonies which are not "mathematically in tune" in an absolute way, (e.g. 440 Hz) but sound fantastic, where all notes in the chord are in tune *relative to each other* like is emphasized in a barbershop quartet, but is also present in chamber music and lots of high-quality classical music. This also would seem bad for compression, to me. Plus it's present in the best jazz, as well. Non-existent in electronic music.

Here's my theory. I think what you're observing, with classical music compressing better losslessly, has far more to do with the lack of very high frequencies. The typical orchestral performance has an occasional cymbal crash, but nothing like the constant high-hat or cymbals which are pervasive in both trance and jazz. There's no drum set in an orchestra. Even the high register of a violin can't compare (in Hz) to the airy "hissy" KHz of whispy electronic sounds and percussion beating regularly throughout the music.

Mac Rhythm Guy

Link to comment
Share on other sites

Metal/Rock/Pop/Electro is also often highly compressed [dynamically, not psychoacoustically], often with a good amount of distortion, thus harder to compress further. I encode a lot [VBR mp3 most of the time] and can confirm Avrin's observations. The resulting bitrates when encoding in VBR are usually the highest with highly [dynamically] compressed metal [containing lots of high frequencies and distortion], while [usually relatively quiet and clean with occasional peaks] classical music seems the easiest to encode [resulting in relatively low bitrates while maintaining sound quality].

Microphone placement will also play a major role. A recording, which is close to mono, will be easier to encode [the benefits of joint stereo encoding].

Edited by greenmachine
Link to comment
Share on other sites

One thing that i don't see being mentioned is the source..... it only gets as good as the source, impossible for something encoded to surprass the original source. I personally like to listen to 356 or occasionally 256 when i run outta room. Imho if your gonna listen to low bit rates why not just ditch md and buy an ipod...... anyways thats for another time. I bought my RH1 for the sound quality that mds can reproduce. Also i agree with what 1kyle said. Just my .02

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...