I was going to put up a post asking what 1-bit actually means, but I think I've found my explanation here.

So, now my question is, when people around here are talking about which portable recording device they're going to buy, why isn't 1-bit technology being mentioned?

If it has been, apologies, but I seem to only see recommendations for the Zoom H4. Edirol, Sony PCM etc..

--------------------
[url] http://colinjmorris.bandcamp.com/releases [/url]

Apologies... http://www.soundonsound.com/forum/dosearch.php?Cat=&fromsearch=1&Forum%5B% 5D=f11&Words=1-bit&where=bodysub&Name=&daterange=1&newerval=1&newertype=y&olderval=&oldert ype=&Limit=25&buttsubmit=Submit


But let's talk about it anyway... Has anyone bought one??

--------------------
[url] http://colinjmorris.bandcamp.com/releases [/url]

One thing I always wondered about with the 1-bit recording technology.

Given that the bitrate is effectively freeform with no upper/lower bitrate limits, where the heck is both 0dB (silence) and 0dBFS (the point of clipping)?

I'm guessing that 0dB is somehow calibrated at the factory, while 0dBFS is perhaps irrelevant as the format can supposedly deal with infinite loudness (although the microphone and recording hardware can't, of course)?

I'm also thinking that, even though the sample-rate is so high, the bitrate isn't, therefore it wouldn't take up an insanely amount of disk-space difference than the current 24-bit/96KHz? Why aren't we seeing 1-bit soundcards?

[edit]Actually, I've worked it out.

If my maths is correct:
24-bit/96KHz = 281.25 KB/s
1-bit/5.6MHz = 6835.94 KB/s = 6.68MB/s

Given that most of us use 24-bit/44.1KHz = 129.2KB/s, there's quite a difference. Not to mention the CPU strain with plugins etc.

Maybe in 10yrs time (we'll be recording and mixing in 1-bit).

[edit2] Shouldn't the sample-rate be on the horizontal axis, and bitrate on the vertical in the Korg MR-1000 article?

--------------------
http://Infekted.org ~ Access Virus news & community

Hugh Robjohns has shed light on the subject of DSD here . Very interesting reading.

Quote topten:

---

Hugh Robjohns has shed light on the subject of DSD here . Very interesting reading.

---

Yes it is, but it's probably best to re-quote it in this thread as well, for clarity.

Quote Hugh Robjohns:

---

This is something I explained in my review of the Korg MR-1000, as well as in other technical SOS articles.

That Korg article is fundamentally flawed and false in several of its arguments and claims, and who ever wrote it doesn't understand the basics of linear PCM, let alone DSD. It is clearly a marketing puff piece, and sadly contains some serious misinformation (such as in the description of oversampling), and misdirection (such as the squarewave graphs), while also spouting unproven opinions as facts! Can you tell I'm less than impressed...?

However, the basic argument is this:

You can trade off reduced word length for increasing sample rate. Philips introduced this concept back in the early 1980s with their oversampling converters (4x oversampling enabled the use of 14 bit converters to deliver full 16 bit 44.1kHz audio quality.

As the technology improved, they introduced Bitstream which used 64x oversampling and a slightly bodged 1 bit output to provide the same quality. Their competitors offered similar solutions, such as MASH.

These techniques led on to the practical development of delta-sigma conversion techniques. Essentially, the analogue signal is sampled at a very high rate (~6MHz or more) but with very low word length (typically less than five bits, sometimes only one). A digital filter (decimation filter) process is then used to translate the information from high sample rate, low word length to lower sample rate, high word length for storage or further processing.

D-A conversion works the other way, with digital filtering (oversampling) to convert the low sample rate/high wordlength signal into a very high sample rate and low wordlength signal.

The advantages are far greater linearity (way beyond that achievable with 'conventional' converter designs as used in the last century), much reduced manufacturing costs, far better stability and accuracy, and a generally better sound because of all those factors (and many others).

Everything currently produced uses delta-sigma converters, because it is not practical or cost effective to do it in any other way... but Sony had the bright idea of simplifying things further.

They wondered why bother with the decimation filter in the A-D if you're only going to reverse the process again with the oversampling filter in the D-A? Why not just leave the data in the raw high sample rate, low wordlength form and save the hassle?

And that makes sense if you are only talking about a recording medium. So that's what they do in the DSD (and the SACD) format and on the face of it, it seems a good idea...

But as with all these things there are swings and roundabouts to consider. Firstly, and probably most importantly, many argue that the current 2.8224MHz DSD sample rate isn't high enough to support true 1 bit wordlength at the intended resolution -- and that's why the bigger Korg recorder offers a second higher DSD rate. Those in the know (and I'm talking about Sony's own egg heads here) argue that even that isn't high enough to do the job properly...

Secondly, huge amounts of noise shaping is required in these formats, and that results in a lot of out-of-audible-band energy that can cause problems in some cases. It also means that the oft cited claims of huge signal-noise ratio afforded by DSD only apply at very low frequencies. The noise performance gets progressively worse with increasing frequency. Linear PCM 24/96 has a considerably better signal-noise ratio at 40kHz than DSD... should such things be of concern to you

It also makes signal processing (gain, EQ, dynamics, etc)very difficult indeed. Many DSD-based DAWs and other processing equipment actually converted back to PCM to do the work, and then back to DSD for the output -- thus completely negating all of the claimed advantages in one swoop! Later systems were developed to work more in the DSD native domain, but with less flexibility and even then the processing involved multi-bit stages -- it has too -- and even if only two or three bits further (simpler) decimationand oversampling filtering stages are still necessary.

The inimtable Dr John Watkinson wrote a very entertaining demolishing article about DSD in an edition of Resolution magazine a while back, too...

Personally, I'm quite happy sticking with 24/96 which I reckon is the best compromise in terms of audio quality, and the most practical in terms of processing, storage, and archiving.

But DSD does have its fans...

hugh

Edited to add: Don't take my word for it, read this abstract from two of digital audio's leading gurus:

Quote:

---

Stanley P. Lipshitz and John Vanderkooy
Audio Research Group, University of Waterloo
Waterloo, Ontario N2L 3G1, Canada

ABSTRACT

Single-stage, 1-bit sigma-delta converters are in principle imperfectible. We prove this fact. The reason, simply stated, is that, when properly dithered, they are in constant overload. Prevention of overload allows only partial dithering to be performed. The consequence is that distortion, limit cycles, instability, and noise modulation can never be totally avoided. We demonstrate these effects, and using coherent averaging techniques, are able to display the consequent profusion of nonlinear artefacts which are usually hidden in the noise floor. Recording, editing, storage, or conversion systems using single-stage, 1-bit sigma-delta modulators, are thus inimical to audio of the highest quality. In contrast, multi-bit sigma-delta converters, which output linear PCM code, are in principle infinitely perfectible. (Here, multi-bit refers to at least two bits in the converter.) They can be properly dithered so as to guarantee the absence of all distortion, limit cycles, and noise modulation. The audio industry is misguided if it adopts 1-bit sigma-delta conversion as the basis for any high-quality processing, archiving, or distribution format to replace multi-bit, linear PCM.

---

---

--------------------
John - Sound-Link ProAudio
President - Federation Internationale des Chasseurs de Sons

Quote Tímo:

---

Given that the bitrate is effectively freeform with no upper/lower bitrate limits,

---

In DSD, the 'bit rate' isn't freeform or limitless. It is clearly defined as 2.8224Mb/s at the standard (original rate) and 5.6448Mb/s at the (not widely adopted) elevated rate.

Quote:

---

...where the heck is both 0dB (silence) and 0dBFS (the point of clipping)?

---

Interesting (mis)use of 0dB

0dB refers to a ratio of one, or where the variable parameter equals the reference. In this case, you mean the latter but have omitted the reference making the question ambiguous at the very least!

0dBA SPL is defined as the threshold of hearing and is presumably what you mean as silence. It refers to an acoustic sound pressure level. 0dBFS is the highest sample value that can be encoded in a multi-bit PCM system. Therefore, neither can be applied to the DSD single-bit system we are discussing!

In a crude way but one that helps to understand what is going on, DSD works essentially by sending a binary one when the next sample is larger than the previous, and a binary sero when it is smaller. It is encoding the difference between adjacent samples, rather than their absolute amplitude. And it makes that decision 2.8224 million times every second, which is claimed to be sufficient to track the higest likely level of high frequency audio signals.

However, several boffins at the time DSD was launched (and even some within the Sony empire) questioned the ability of the system to function as claimed at that 2.8224MHz sample rate. And clearly, if it was as perfect as was claimed they wouldn't have felt the need to introduce the 5.6448MHz rate subsequently.... (It's still way too low according to the same boffins, but the way )

Quote:

---

If my maths is correct:
24-bit/96KHz = 281.25 KB/s
1-bit/5.6MHz = 6835.94 KB/s = 6.68MB/s

---

Not sure how you got either of those figures, but they are both wrong, I'm afraid.

The bit rate of 24/96 is 24x96000 = 2.304Mb/s per channel which is very similar to the 2.8224Mb/s rate of the original DSD format (per channel).

The bit rate of high-rate DSD is exactly what it says on the box: 1 bit times 5.6448MHz sample rate = 5.664Mb/s

Quote:

---

Maybe in 10yrs time (we'll be recording and mixing in 1-bit).

---

Unlikely. Signal processing in the DSD format is a nightmare and of the very few systems that have been developed to do it, most are not being very well supported as there is very little commercial demand.

The whole DSD concept is interesting, but the downsides are at least as significant as the upsides, and I really can't see it ever becoming mainstream as a production format let alone a consumer one.

Quote:

---

Shouldn't the sample-rate be on the horizontal axis, and bitrate on the vertical in the Korg MR-1000 article?

---

Which graph are you referring to exactly? Sample rate and bit rate are effectively the same thing -- or at least, they are directly proportional to one another.

If you mean the graph in my SOS Korg review in the DSD tint box, it shows the inherent noise level of multi-bit PCM versus DSD against frequency.

Hugh

--------------------
Technical Editor, Sound On Sound

I seem to remember a paper demonstrating that for correct operation with dither you needed at least 2 bits (sort of counter intuitive, but a lot of this stuff is), but I cannot find a reference.

I have to admit to completely not seeing the point of the whole DSD thing, except to sell me albums I already own again!

Regards, Dan.

--------------------
Audiophiles use phono leads because they are unbalanced people!

The only part of the argument that holds water is the idea of avoiding the decimation and oversampling digital filters implicit in moern multi-bit PCM formats...

...but as I said, the drawbacks of single bit working are at least as much of a problem as the advantage of circumventing the filters is an advantage.

The paper I referred to in my previous posting on this topic (the Lipshitz and Vanderkooy paper) is very interesting to read and highlights the issues involved very well.

Hugh

Hugh

--------------------
Technical Editor, Sound On Sound

Quote Hugh Robjohns:

---

Timo Quote:

---

...where the heck is both 0dB (silence) and 0dBFS (the point of clipping)?

---

Interesting (mis)use of 0dB

0dB refers to a ratio of one, or where the variable parameter equals the reference. In this case, you mean the latter but have omitted the reference making the question ambiguous at the very least!

0dBA SPL is defined as the threshold of hearing and is presumably what you mean as silence. It refers to an acoustic sound pressure level. 0dBFS is the highest sample value that can be encoded in a multi-bit PCM system. Therefore, neither can be applied to the DSD single-bit system we are discussing!

In a crude way but one that helps to understand what is going on, DSD works essentially by sending a binary one when the next sample is larger than the previous, and a binary sero when it is smaller.

---

I mean, what does it do when there is no sound present (at 0dBA SPL), or when there are no fluctuations (up or down)? I take it doesn't send any signal, or at least it tells the recorder to not expect a signal on that particular clock pulse.

Well, the way I was originally thinking about it, theoretically speaking only, was that say I wanted to record 1-bit audio into a wave editor (such as Wavelab), because there were no defined amplitudes to categorically state "right, ok, THIS is 0dBA, and everything recorded is referenced to this and limited accordingly", how it would know which level (formerly a value of bit-depth) to start recording from. That was my original query. Like it would involve some seriously massive DC-offsets, or smthg. And then, supposedly, how the speakers would interpret this in order to play it back, in addition to the fact that there was also no 0dBFS or upper/lower range to physically limit speaker cone movement if someone or something happened where the data stream just sent 1's or similar (Alert! Incoming speaker cone at warp factor 9!)

Timo Quote:

---

It is encoding the difference between adjacent samples, rather than their absolute amplitude. And it makes that decision 2.8224 million times every second, which is claimed to be sufficient to track the higest likely level of high frequency audio signals.

However, several boffins at the time DSD was launched (and even some within the Sony empire) questioned the ability of the system to function as claimed at that 2.8224MHz sample rate. And clearly, if it was as perfect as was claimed they wouldn't have felt the need to introduce the 5.6448MHz rate subsequently.... (It's still way too low according to the same boffins, but the way )

---

Yes, quite different to our current digital recording methods. Took a while to get my head around the concept.

Quote Hugh:

---

Quote Timo:

---

If my maths is correct:
24-bit/96KHz = 281.25 KB/s
1-bit/5.6MHz = 6835.94 KB/s = 6.68MB/s

---

Not sure how you got either of those figures, but they are both wrong, I'm afraid.

---

B = Byte, b = bit.
KB = KiloByte, Kb = Kilobit. You should know this!

(((24 x 96000) /8) /1024) = 281.25 KB/s.
Admittedly I was out by a factor of 10 on the DSD quote. I multiplied it by 56 million instead of 5.6 million by accident.

Therefore (((1 x 5600000) /8) /1024) = 683.59 KB/s.

Not insanely different to 24/96 after all, in line with my original thinking. I'm sure lossless data compression could be used to make the file-sizes even smaller, at least for offline archiving.

Quote Hugh:

---

Quote Timo:

---

Shouldn't the sample-rate be on the horizontal axis, and bitrate on the vertical in the Korg MR-1000 article?

---

Which graph are you referring to exactly? Sample rate and bit rate are effectively the same thing -- or at least, they are directly proportional to one another.

---

I was referring to:



From: http://www.korg.com/mr/Future_Proof_Recording_Explained.pdf

Unless I've got this completely wrapped around my head...

I mean, sample frequency rate (like 44.1KHz), over time, should be horizontally? Whereas shouldn't bit-rate, or perhaps I should have called it bit-depth (I know what I mean, even if everyone else doesn't!), be the (vertical) number of amplitude levels available (16,777,216 for 24-bit) to represent one sample? Or so I thought.

But thanks, it's generally clearer, if probably an amusing read for some, and I'll pay closer attention to correct terminology in future.

Cheers

--------------------
http://Infekted.org ~ Access Virus news & community

but if there are almost "no future" for DSD why in the world do korg release portible DSD recorders and other brands as well release new DSD recorders (Like the tascam DV- RA1000(HD))?
And I see newspapers here in Norway wright about SACD more and more. making people aware that there is a format better than original CD`s (and even better the MP3!!! ). and more and more CD and DVD players support SACD...

--------------------
GUTZ http://www.myspace.com/gutzmetal
Sublime Eyes: www.myspace.com/sublimeeyes

Quote -Nightspirit:

---

but if there are almost "no future" for DSD why in the world do korg release portible DSD recorders and other brands as well release new DSD recorders (Like the tascam DV- RA1000(HD))?

---

Maybe someone has released a chipset that does SACD which costs no more than the equivalent chipset without SACD. That's the most common reason for a sudden appearance of certain features on gear from more than one manufacturer.

I also understand that the audiophile market is much bigger in the far east so the ability to record and release recordings that haven't gone through a multibit recorder is more important there.

Cheers

James.

--------------------
JRP Music - Audio Mastering and Restoration.
http://www.jrpmusic.net

Quote Tímo:

---

I mean, what does it do when there is no sound present (at 0dBA SPL), or when there are no fluctuations (up or down)? I take it doesn't send any signal, or at least it tells the recorder to not expect a signal on that particular clock pulse.

---

No, it alternates between a one (bigger) and a zero (smaller)so that the average is no change.

Quote:

---

Well, the way I was originally thinking about it, theoretically speaking only, was that say I wanted to record 1-bit audio into a wave editor (such as Wavelab), because there were no defined amplitudes to categorically state "right, ok, THIS is 0dBA, and everything recorded is referenced to this and limited accordingly", how it would know which level (formerly a value of bit-depth) to start recording from.

---

It wouldn't. Wavelab can't accommodate DSD audio. Very few systems can. There are some versions of SADiE and Pyramix that will, and I think a version of Sonoma can, plus an Augan machine I think. Although some of those cheat and convert DSD into high rate multi-bit PCM for editing.

Quote:

---

And then, supposedly, how the speakers would interpret this in order to play it back,

---

It has to go through a D-A first, of course, but this is incredibly simple and looks a lot like Philips' old Bitstream decoders. Essentially, all that is needed is some means of averaging the datastream, so that a lot of 1s produce an increasing analogue voltage, and a lot of 0s produce a decreasing analogue voltage. D-A conversion is one of the nice aspects of DSD becaues it is so incredibly simple (which means cheap, which is why Sony wanted to promote it in the first place).

Quote:

---

...if someone or something happened where the data stream just sent 1's or similar (Alert! Incoming speaker cone at warp factor 9!)

---

There are still physical limits to the amplifier voltage rails, so it ain't infinite!

Quote:

---

B = Byte, b = bit.
KB = KiloByte, Kb = Kilobit. You should know this!

---

Sorry -- I assumed that was a typo (given the confusion over 0dB references earlier . It is very unusual to specify bitrates in terms of bytes -- they'd be called byterates then, surely?

Quote:

---

Not insanely different to 24/96 after all, in line with my original thinking. I'm sure lossless data compression could be used to make the file-sizes even smaller, at least for offline archiving.

---

As I said, the bitrate for 24/96 and DSD are very similar, but DSD does have more ultrasonic noise. Lossless data compression could be used on DSD (although I don't know of any standardised system at present), and it already is on linear PCM in the form of MLP (Meridian Lossless Packing) which is used on DVD-Audio.

Quote:

---

I mean, sample frequency rate (like 44.1KHz), over time, should be horizontally? Whereas shouldn't bit-rate, or perhaps I should have called it bit-depth (I know what I mean, even if everyone else doesn't.

---

It is a slightly confusing graph. The words are right, though. The vertical lines represent the sampling instants and the horizontal lines represent the quantising intervals. The numbers across the bottom is just a numerical way of showing the quantised value for each sample. It is confusing....

Quote:

---

I'll pay closer attention to correct terminology in future.

---

In a technically complex topic like this, incorrect terminology always causes confusion -- and we are all guilty of it at some point or other, but trying to get it right to avoid misunderstandings is a good goal to aim for

Hugh


--------------------
Technical Editor, Sound On Sound

Quote -Nightspirit:

---

but if there are almost "no future" for DSD why in the world do korg release portible DSD recorders and other brands as well release new DSD recorders (Like the tascam DV- RA1000(HD))?

---

I'm only offering my opinion. You can count the number of current DSD recorders on offer on the fingers of one hand. It would take a rugby team of hands to count up the number of linear PCM machines out there!

The major problem is that even though the average punter could record at source in DSD on the Tascam or Korg machines, they can't post-produce in that format. They'll have to convert to multibit PCM for that... in which case why not just start in that format to avoid further complications and format conversions?

Quote:

---

And I see newspapers here in Norway wright about SACD more and more. making people aware that there is a format better than original CD`s (and even better the MP3!!! ). and more and more CD and DVD players support SACD...

---

DVD-Audio provides better quality than CD too -- but really, who is interested? How many people do you know have actively sought out the ability to play SACD or DVD-A? I only know a few, and that's amongst the pro-audio world who are supposed to be interested in this stuff. And most of the hifi nuts I know are more interested in their turntables and gold-plated mains leads than they are SACD or DVD-A.

Sadly, it does look increasingly like DVD-A might die out through a lack of support, but I don't think SACD will last much beyond that either -- there just isn't the public interest.

Hugh

--------------------
Technical Editor, Sound On Sound

While DVD-A is interesting from a multichannel perspective (It supports B format), it appears that as far as word length and sample rate 16/44.1 is more then adequate as a distribution medium.

The Boston audio society demonstrated this in a paper in the AES journal before last, wherein they ABX tested SACD vs SACD via a 16 bit ADA chain.

Results at at sane listening levels were the same as random chance.

Possibly the interesting comment was that DVD-A/SACD disks tended to be recorded with more dynamic range then CDs, but that this was not due to a technical limitation but a marketing one.

Regards, Dan.

--------------------
Audiophiles use phono leads because they are unbalanced people!

Quote dmills:

---

While DVD-A is interesting from a multichannel perspective (It supports B format), it appears that as far as word length and sample rate 16/44.1 is more then adequate as a distribution medium.

---

Couldn't agree with you more. A dynamic range of 90+dB is more than adequate for all but the most fastidious of domestic listening environments, and modern delta-sigma converters are fine with 44.1kHz sources.

I'm still surprised that so few are taking up B-format technology since it is so easy to do know... but one day...

Quote:

---

Results at at sane listening levels were the same as random chance.

---

Not even slightly surprised

hugh

--------------------
Technical Editor, Sound On Sound