Recording Format

[md-max]

Since I got my Hi-MD unit a month or so back, I've been using Hi-SP - the quality sounds perfectly OK to my ears and just over 2 hours of music (using old 80 min discs) somehow 'feels' about right.

At the risk of opening a can of worms - why bother with PCM, which would greatly reduce the amount of music? And isn't Hi-LP just terrible quality?

My experience from years of all sorts of recording, initially on cassette (remember those?) and then years of (now sadly dead) Sharp MD, is that as long as the quality isn't absolutely terrible then your ears very quickly get used to whatever you're listing to and it (deep breath!) doesn't really matter that it's not the maximum possible quality.

Come on, disagree with me!

[dex Otaku]

For important things - PCM

For most things - HiSP is sufficient

I never record anything even remotely important [including speech] with less than HiSP.

The only thing I've used HiLP for is testing purposes.

Were ATRAC SP usable in HiMD mode, I would probably choose it over HiSP, simply because of SP's maturity.

And, were I ever to record live castanets, I'd stick with PCM.

[KJ_Palmer]

I use them all, really, on my various equipment (I have old and new MD), and depending on what sort of compilation I'm making. I've even used 48kbps occasionally for filling in a few gaps at the end of compilation - of course it's bad quality, but still listenable on a train...

It's mostly Hi-SP though, as you say 2 hours on a standard blank is just about right for a day's listening, or 8 hours on a 1GB disk for travelling or longer mixes. I've played with PCM, but the quality increase over Hi-SP is barely if at all perceptible (to me), and as I have the original CD's there's no need to store them on another media...

[dex Otaku]

Ah. See, the question was about recording formats. I assumed that meant for recordings, not for downloads, since they are not the same thing.

[Michael1980]

Well, PCM is for those who record and then pass it back to the PC to encode in whatever format they prefer. When I had my 600 that is all I used.

[ROMBUSTERS]

if im recording line in then i'll only use PCM or HiSP (probably more HiSP than PCM just for space)

For playback I'd either use LP2 most often, some HiSP but not much seeing as you loose a huge amount of space moving up to it from LP2, and even sometimes HiLP just because depending on where I am and what I'm doing (i.e. on the bus) I cannot tell the difference

[8track]

Any of the new or old LP formats sounds just awful. LP4 is downright painful.Yes even netMD.IMHO as they say.

I realy like the sound of an analog recording dubed to ARTAC (SP) or the new Hi-SP.To me they sound about the same . After using PCM for a a bit I'm right back to SP or Hi-SP.PCM is great for dubing quality (don't get me started again) digital recordings but for some reason it dosn't seem to work as well with non digital sources to my ears. Why? I don't know. For some reason it (PCM) just dosn't seem to capture the warmth of the original recording as well. You can talk bitrates all you'd like but to me ATRAC is the best sound digital has had to offer so far for live or dubbing....

.....But then again sound is just a matter of taste. It's in the ears of the listener.

Edited March 8, 2005 by 8track

[leenuss]

For some reason it (PCM) just dosn't seem to capture the warmth of the original recording as well.

Are you saying lossy formats like Atrac3 sounds better than non-lossy formats like PCM? Haha, that's just stupid!

Hi-SP Atrac3+ sounds really good, but I would not record in it as long as there is true PCM. I record in PCM, transfer the track to my computer and do editings (cutting and maybe adding a limiter). Then I'll archive my .wav-recordings on DVDRs and put a lossy copy back on MD.

[ROMBUSTERS]

Are you saying lossy formats like Atrac3 sounds better than non-lossy formats like PCM?      Haha, that's just stupid!

Hi-SP Atrac3+ sounds really good, but I would not record in it as long as there is true PCM. I record in PCM, transfer the track to my computer and do editings (cutting and maybe adding a limiter).  Then I'll archive my .wav-recordings on DVDRs and put a lossy copy back on MD.

←

actually Atrac is a 24bit audio system whereas PCM is a 16bit. Although Atrac is lossy it can actually (at times) sound better when recording from a live source. Although this is always personal opinion

[dex Otaku]

actually Atrac is a 24bit audio system whereas PCM is a 16bit. Although Atrac is lossy it can actually (at times) sound better when recording from a live source. Although this is always personal opinion

←

Actually, ATRAC is a discrete-transform encoding system, and has no bit depth in the sense that PCM does.

ATRAC is capable of encoding 20+bit PCM and of decoding to 20+bit PCM. As are basically any lossy formats whose codecs have been made to natively convert >16bit PCM, and including MP3, OGG, AC3, DTS pro [a hybrid codec], and many if not most other recent codecs.

Funny point, though:

PCM on HiMD is 16-bit, 44.1kHz stereo.

If the portables contain a 1-bit ADC [as most equipment do] then they should be capable of converting to >16-bit PCM directly, and encoding to ATRAC/3/plus formats from >16-bit PCM - meaning, in theory, that the lossy formats, albeit with compression artifacting, might actually be capable of recording with a higher dynamic range than PCM mode can.

[8track]

Are you saying lossy formats like Atrac3 sounds better than non-lossy formats like PCM?      Haha, that's just stupid!

Hi-SP Atrac3+ sounds really good, but I would not record in it as long as there is true PCM. I record in PCM, transfer the track to my computer and do editings (cutting and maybe adding a limiter).  Then I'll archive my .wav-recordings on DVDRs and put a lossy copy back on MD.

←

For recording fom an analog source... yes that's exactly what I'm saying. WAV, PCM,ect are not lossless. They do not sound like the the original recording when dubed form a good analog source. It sill sounds like a CD. ATRAC has a masking quality to it that I prefer to what I find to be a rather cold and dead sounding CD,PCM,WAV,ect. CD quality is not the end all best sound there is. To quite a few of us it's rather dissapionting. Becouse of this I still by music on LPs unless it is something that was mastered digitaly. To me ATRAC has more of an analog sound becouse of the masking that takes place. Don't let the numbers fool 'ya. It's the formatting/prosessing that matters.

Not to get in a analog vs digital debate. But there is a reason most music producers still master thier recordings on tape instead of digital. Sound quality. You can use your computer for an audio device if you'd like to. You can also make frozen waffles in a toster, but that dosn't make it a waffle iron. It's not bad just not the real deal.

Edited March 8, 2005 by 8track

[Asia]

PCM all the way!

[dex Otaku]

For recording fom an analog source... yes that's exactly what I'm saying. WAV, PCM,ect are not lossless. They do not sound like the the original recording when dubed form a good analog source. It sill sounds like a CD.

This depends on what you mean by "lossless."

If, by lossless, you mean "the data that came in is exactly the data that comes out" then PCM is completely lossless.

If you mean, however, that there is a loss incurred elsewhere in the recording chain, such as colouring imposed by cheap, low-power analogue to digital conversion, then you're dead-on correct.

On the other hand, keep in mind that even if PCM mode is limited to 16-bit, and the ATRAC/3/plus modes encode >16-bit, the first stage before data reduction is still converting that analogue signal to PCM. i.e. if you're recording from analogue sources, you're depending 100% on the accuracy of the unit's ADCs; the compressed signal comes from the same source as the PCM signal.

On one hand, this could indicate that your preference might be based on liking how ATRAC colours the sound with its artifacts, assuming that the encoder is fed the 16-bit stream.

On the other hand, if the encoder is fed a >16-bit stream, that might explain your preference in terms of, despite ATRAC's being lossy, its actually being capable of recording, in some sense, a higher resolution signal than PCM mode is.

Note that these are merely theories of mine. I have no way of testing them, since I lack real measurement equipment &c. It would be nice if Sony's engineering dept. would answer questions like this.

There is a reason most music producers still master thier recordings on tape instead of digital. Sound quality.

I know of only one studio in this province that still uses analogue multitrack tape. Most have moved to hard-disc based systems or have been using ADAT [and Tascam DA-X88 systems, which are what CBC use in their studios as well as in their $500k 56-track mobile van] and the like for at least the past 7-10 years.

By the same token, many of the digital studios I've seen keep equipment like 1/2" open-reel 6-track Dolby SR analogue recorders on hand to run the final master through an analogue phase just to get the "phat" sound of slight tape saturation in their recordings.

Analogue equipment [thinking such as Otari 2" 24-track machines] are extremely expensive and finicky to maintain, and most of the people I've met who own multitrack units have gone digital to avoid the maintenance issues alone.

[ROMBUSTERS]

heres an odd thing to throw into the mix: don't all Atrac encoders work by first converting the audio stream into PCM and then working from there?

[dex Otaku]

heres an odd thing to throw into the mix: don't all Atrac encoders work by first converting the audio stream into PCM and then working from there?

←

Yes. Read above.

[ROMBUSTERS]

Yes.  Read above.

←

and this is also done via any line-in?

whats the point of Atrac being able to go higher than PCM's abilities then? Does this only apply to premastered Atrac MDs?

[dex Otaku]

and this is also done via any line-in?

whats the point of Atrac being able to go higher than PCM's abilities then? Does this only apply to premastered Atrac MDs?

←

Keep in mind that I'm dealing with theory, though I'd say fairly plausible theory. Go ahead, A440, and blast me for dealing in theory again.

ATRAC SP since v.3 on decks has supported direct SP encoding with at least 20-bit PCM audio. To my knowledge, ATRAC has gone on to support direct 24-bit conversion in later incarnations. I am making the assumption that in more recent years, the same has been extended to all ATRaC/3/plus modes, or at the very least, HiSP on HiMD, since the higher-resolution input stream wouldn't really benefit lower-bitrate modes].

If I'm wrong on this, I'd appreciate someone pointing me to an info source that can correct me on it.

The 16-bit 44.1kHz limitation on PCM mode is likely there for several reasons:

1. Because it's a well-established standard that meets basic recording requirements

2. to limit the quality of the recordings to be less than that of professional equipment [if you think this is specious, do some reading about DAT and why there are 2 different 32kHz modes in addition to 44.1kHz, which was locked out on most consumer equipment, and 48kHz, which was considered "analogue" mode.]

3. to limit the quality of illegally copied, copyrighted recordings by more or less incurring a generation loss [assuming sources of >16-bit material are available]

4. to keep the recording length of the media reasonable

5. to lower power requirements [a lower bitrate means slower buffering, lower disc rotation speeds, lower write speeds, &c. all of which affect power consumption]

With consumer equipment, ADC conversion is generally done with 1-bit ADCs simply because they are cheaper. They are also scalable in the sense that you can convert the 1-bit stream to whatever sampling rate and bit depth you like, as long as your "chipset" has the power to do the math. [This is remarkably similar to how DSD works, incidentally.]

It's a pretty safe assumption that our portables all use 1-bit ADCs.

The bit-depth of the PCM output is pretty much arbitrary, to my understanding. If your encoder can deal with >16-bit streams, why not use them?

The limitations imposed on PCM recording are mostly practical. Using a higher resolution source stream for lossy modes doesn't use more space, so it's not impractical as long as the initial conversion to a set PCM bit-depth is capable of dealing with >16-bit data.

My apologies for being so far off-topic.

[8track]

"might be based on liking how ATRAC colours the sound with its artifacts"

Bingo! It dose'nt have that "way too clean" sound to it. It has a little bit of an edge. You just have to have a little distortion in there somewhere or it simply dosn't sound right to me. PCM is just not as "full" sounding to me. It's just too cold and clinical. There is something missing. This may be hard to understand for folks who don't remember the pre-CD days. I wish I could explain myself better, but I think you get the idea....yes I like tube gear too....8)

[ROMBUSTERS]

ok now that the limitations of PCM are pointed out, back to Atrac

You implied direct Atrac encoding however according to the Atrac Developers Interview:

"— Is the technology behind these codecs very different?

Suzuki: The basic technology is much the same. All codecs take as their input a PCM sound source. They perform signal processing on a fixed number of samples which are converted into frequency data and then compressed (encoded). At the other end, these steps are performed in reverse, so the data is decoded. After restoring the frequency data, signal processing reproduces the PCM sound source. With each codec, there's a difference in the number of samples used for frequency conversion: the higher the compression ratio, the larger the number of samples."

All audio is taken in as PCM and then using a fixed number of frames (probably dependant on the PCM values) it encodes it.

Any clearity on this one?

[dex Otaku]

"might be based on liking how ATRAC colours the sound with its artifacts"

Bingo! It dose'nt have that "way too clean" sound to it. It has a little bit of an edge. You just have to have a little distortion in there somewhere or it simply dosn't sound right to me. PCM is just not as "full" sounding to me. It's just too cold and clinical. There is something missing. This may be hard to understand for folks who don't remember the pre-CD days. I wish I could explain myself better, but I think you get the idea....yes I like tube gear too....8)

←

To the surprise of many, I agree with you on this completely.

My own priorities are such that .. well, in all honesty, to achieve the same quality as PCM on HiMD with analogue equipment, you'd be spending a few thousand dollars on equipment rather than a few hundred.

[Fast Eddie]

For music on the move I often use LP3 105 but this wasn't included in the POLL.

[ROMBUSTERS]

For music on the move I often use LP3 105 but this wasn't included in the POLL.

←

i dont mind the sound quality of LP3 either (although it certainly varies depending on the song)

[betamaxDATminidisc]

I prefer format PCM recording that he's order:

1 PCM

2 Hi-SP

3 SP

[garyc]

SP all the way

[ROMBUSTERS]

SP all the way 

←

heh

they really need to make SP avialable for computer transfers. I know its never been true SP but does anyone know what exactly the limiting factor here is? Why havnt they added it?

[bug80]

heh

they really need to make SP avialable for computer transfers. I know its never been true SP but does anyone know what exactly the limiting factor here is? Why havnt they added it?

←

The last thing I've heard is, that SP doesn't support DRM, and that's the reason why.

But, true SP... how great would that be? Sony....????

[imkidd57]

Definitely SP for any decent stuff in domesticity; LP for anything in travelocity.

[dex Otaku]

The last thing I've heard is, that SP doesn't support DRM, and that's the reason why.

But, true SP... how great would that be? Sony....????

←

SP is an encoding format; OpenMG is a container format with DRM. Put SP into an OpenMG container, and you get SP with DRM. Just like how they can take an MP3 and wrap it in DRM with encryption, and allow Sony players [that support MP3 decoding] to play it.

It's really not that hard to figure out.

Whatever Sony's issue with SP is, it's not DRM.

[ROMBUSTERS]

SP is an encoding format; OpenMG is a container format with DRM.  Put SP into an OpenMG container, and you get SP with DRM.  Just like how they can take an MP3 and wrap it in DRM with encryption, and allow Sony players [that support MP3 decoding] to play it.

It's really not that hard to figure out. 

Whatever Sony's issue with SP is, it's not DRM.

←

well maybe in the case of the original MD it was, everything in MD land is done in equal sectors. Each sector in a standard MD is large enough for a portion of SP audio. maybe they couldnt get the player to read an entire portion and then sacrifice 2 other sectors (on either side of it) for DRM. It would lower the disc capacity by 2/3

[sector][sector][sector]

[DRM][sPAudio][DRM]

Edited March 11, 2005 by ROMBUSTERS

[dex Otaku]

well maybe in the case of the original MD it was, everything in MD land is done in equal sectors. Each sector in a standard MD is large enough for a portion of SP audio. maybe they couldnt get the player to read an entire portion and then sacrifice 2 other sectors (on either side of it) for DRM. It would lower the disc capacity by 2/3

[sector][sector][sector]

[DRM][sPAudio][DRM]

←

I've been trying to find a straightforward answer to this, and can't seem to.

My real questions are as to how DRM works on netMD, which, for backward-compatibility reasons, can't alter the stored stream significantly.

According to the IEEE paper on minidisc.org:

<blockquote>"Variable size of the write data causes some difficulty in managing address control, therefore, we introduced the cluster unit, which constitutes the minimum unit of write or over-write of data to the disc. One cluster consists of 36 sectors. 32 sectors of one cluster are used for main data recording and 4 sectors are used for data linking or additional sub data.</blockquote>

My previous thoughts on how netMD's protection works were that it simply used the subcode to mark the stream as an SCMS 'copy' and whatever needs to be changed for "TRPROT", which could be in both the written audio/subcode clusters and the UTOC. I can't be sure of this, of course, as I haven't found any material on exactly how netMD's DRM works.

This way, no change to the actual format of the audio stream would be needed, and backward-compatibility is ensured while maintaining the weak version of DRM that is basically still just SCMS.

Since Sony elected to not allow uploading/ripping, this is really the only form of copy protection they need with netMD, as digital copying can only be done using SP/DIF [which, as we all know, can be stripped of copy protection quite easily, but hey, it works to keep the majority of people from copying things until they come up with a completely new system - HiMD].

In any case, I don't think your method above is what's used, as the capacity hit would be taken regardless of what bitrate the audio is encoded in.

In the end, to me it still appears that SP writing should be possible with both MD and HiMD.

Also, with HiMD the conformance of blocks/cluster is no longer really necessary as HiMD uses a true filesystem, enabling true random access that is not limited by or dependent upon the physical layout of the disc as with MD; rather it's limited by HiMD's own indexing, done in the DRM table, track info table, and the audio "virtual partition" themselves, on top of a FAT filesystem. Editing to single-frame accuracy should be possible, though electing to use blocks conformant to the 32kB cluster size of the FAT partition format used by 1GB discs is also possible - though that would be a pretty coarse resolution, even with high bitrates.

It seems to me that HiMD should be capable of writing any bitrate of any format as long as the hardware is aware, i.e. able to recognise, said format. The bitrate of the audio should be completely independent of the medium, unlike with MD.

This is evidenced by HiMD's 2nd-gen being capable of playing both CBR and VBR MP3s.

The more I think about this, the less I'm making sense to myself, but in the end this is what my thoughts boil down to:

Writing SP from a computer to netMD should be possible, because DRM was not part of the system from the ground-up, and altering the bitstream significantly to take it into account would break compatibility with the discs on all non-netMD equipment.

Writing SP on HiMD should be possible, because the use of separate DRM, track info, and audio portions on an actual filesystem [allowing true random access] means that the medium need not be physically tied to the bitrate of the audio on it [as with MD]. Basically, format and bitrate are arbitrary and not dependent on the physical disc structure any more; as long as the hardware's decoder can recognise whatever you throw at it, it should be playable at the very least, and given the proper processing and seeking logic, editable as well [potentially even with VBR streams].

Finally - my opinion of why Sony don't allow SP writing from computers:

The quality was too high to risk using with a weak DRM system [sCMS at the least].

The same does not apply to HiMD with HiSP or PCM because HiMD is designed with stronger DRM from the ground up. The omission of SP mode from HiMD is probably just from a desire to move to the new atrac3plus format, which, with a higher-resolution transform, at least has the potential to be better than SP.

Mind you, HiMD decks will likely have full optical I/O, defeating this once again.

Gah. Theory as usual.

If anyone can point me to more reading material on this, I'll be happy to prove myself wrong.

[bug80]

I still don't get why Sony doesn't offer SP downloads while most units have an optical in. The result is basically the same, but the method only takes more time.

[dex Otaku]

I still don't get why Sony doesn't offer SP downloads while most units have an optical in. The result is basically the same, but the method only takes more time.

←

Optical in = realtime copying; no one in their right mind is going to spend thousands of hours copying things by SP/DIF in order to pirate them, whereas they would spend the much shorter time required to rip+distribute with a computer.

Basically, it doesn't prevent anything - it just makes the option far less attractive.

[Labman]

If the portables contain a 1-bit ADC [as most equipment do] then they should be capable of converting to >16-bit PCM directly, and encoding to ATRAC/3/plus formats from >16-bit PCM - meaning, in theory, that the lossy formats, albeit with compression artifacting, might actually be capable of recording with a higher dynamic range than PCM mode can.

The dynamic range of 16bit 44k recordings is regarded as 96 dB, i.e. the highest recordable signal level will be 96 dB above the lowest, zero record level. I've seen the origianl minidisc format being quoted with a dynamic range up to 120 dB.

Whether you believe this or not is another argumant, but its all pretty irrelavent because everyone forgets about the noise floor of the equipment, be it miniidsc or not.

If your quoted dynamic range is 96 dB re zero recording level of 0 dB and the noise floor of the equipment is say 20 dB (this would be the hiss you hear at high playback level) then your dynamic range becomes 76 dB, regardless.

[MZ-NH1]

I use Hi-SP, only for the reson that if I am going to be recording a long conert, I want the whole thing to fit onto a 1GB disc. If there were 5GB media about, then I would quickly make the move to PCM

[bug80]

Optical in = realtime copying; no one in their right mind is going to spend thousands of hours copying things by SP/DIF in order to pirate them, whereas they would spend the much shorter time required to rip+distribute with a computer.

Basically, it doesn't prevent anything - it just makes the option far less attractive.

←

Most probably this is the main reason for Sony for not supporting SP downloads. To me, it's just like selling a car without the highest gear, because people might drive too fast with it. By now, it should be clear to Sony that MD isn't the first choice for Britney Spears downloading kids who just want a playback device anayway.

[twine]

I've got a question.

Does recording from a analog source such as radio plays on audiotape also sound better with HI-SP or SP than PCM?

I would like to save my radio plays (the three investigators) on minidisc.

[Michael1980]

I don't know about SP.

I mean, why did they introduce Atrac3Plus 256 instead of Atrac 292?

[Michael1980]

I've got a question.

Does recording from a analog source such as radio plays on audiotape also sound better with HI-SP or SP than PCM?

I would like to save my radio plays (the three investigators) on minidisc.

←

Not unless you like the sound characteristics of Sony codecs (and some do).

No.

Storing radio programs in PCM is a waste of space, record in PCM only if you are thinking of converting that PCM to a lossy format of your choice (Mp3, AAC etc).

[yamagatacamille]

Team SP!! because my two E75s said so!

[Michael1980]

Funny how 10 year old tech (Type-R) is still the pride of md lol.

Just shows how good is was then.