Freesound Forums

My first 1-bit recorder, KORG MR-1 was delivered today.
Sampling rate is the same as SACD --- 1-bit/2.8MHz.
The sound is just astounding. It has ability to describe atmosphere.

I note some of my impressions.

Good:

compact, robust, amazing sound, affordable price, nice display, easy file transfer, and more.

Bad:

hard drive system (flash memory is easier and faster to use), operation (almost the same as notorious SONY MD Walkman), built-in battery (you can't use removable AA batteries, so you need to charge it the day before,) and much more.

My evaluation rating is 7 out of 10. Not bad, but I believe other company will make better ones.

Hi,

thanks a lot for posting your impressions! Any hard-disk spinning noise?

This recorder really looks sweet, although it may take a while to get out of Japan, let alone reaching Europe! Couldn't even get an idea of its cost in Froogle
For those that didn't hear before on the gadget:
http://www.dv247.com/news/Korg%20MR-1000%20And%20MR-1%20Announced/131212

Hi, there.

Any hard-disk spinning noise?

Only when it starts spinning, yes. Once it comes to level flight, you can hardly hear any noise.
Imagine iPod.

I recorded some music instruments today, and was convinced that this is the 'next stage.'
Everybody was content to compact cassette tape twenty years ago. Now, few people want it. Similarly, I don't feel like going back to ordinary cd quality. 1-bit sound is too good. You can 'see' any detail.

Wonder if this will record bat sounds directly (up to about 80kHz). The high sampling rate sounds good, as long as there is no brickwall antialiasing filter

Wonder if this will record bat sounds directly (up to about 80kHz).

Recorder itself can handle up to 100kHz, so theoretically you can record 'bat's voice' without down-pitch-shifting. But, I don't think you can do it easily. You need very high quality microphone, very good analog devices, and very good ears that can hear 80kHz sound.javascript:emoticon('')

as long as there is no brickwall antialiasing filter

Such powerful filter makes you hear good sounds. Without it, you would be bothered with strange low frequency rumble from hell.

Anyway, 1bit/2.8MHz sound gives you inaudible but sensible beautiful harmonic overtones. Try it yourself. I think you can understand this, ermine.

Been reading a bit about 1bit/2.8MHz recording, and several worrying doubts have come to my mind...

1) you cannot directly edit the sample, you must convert it first to PCM. Fortunately the new Korgs come with converting software but...
2) DSD is a propietary format, which *is* bad IMO
3) With present-day playing gear (CD players, DVD) looks like you can't directly listen to a DSD sample. Again you have to convert it to something meaningful. That's why Heigh-hoo uploaded a PCM sample (by the way of amazing quality)

I would appreciate any comments, what do you folks think about these issues? At least in part I can understand the theorical basis than makes 1 bit the best way for archiving sounds (intuitively I see it's like an analog picture of sound), but there could be shortcomings too... (?)

Cheers and happy Christmas to all

I take it for just 'a nice recorder' for the time being. I think it will take more time for 1-bit files to be accepted as PCM or other compressed formats are now.

I have no wave-edit software to accommodate 1-bit format, so I convert them to 16bit/44.1kHz, 24bit/44.1kHz, or 24bit/96kHz. Luckily, general impression of the sound isn't spoiled so much. Down converted sound is rather beautiful than originally recorded PCM in my opinion. Just like digital photographs.

People are working on making general (non-proprietary) formats, but it's still on the way, so we have to wait.

(By the way, why does my emoticon alone come with script visible???)

1bit recording sounds excellent. Surely with time the infrastructure to support this format will be developed. Does pose an interesting question on how to display audio wavwform within an editor - wontan amplitudes etc.

Thank you for stimulating this debate heigh hoo

1bit recording sounds excellent.

Oh, you've also experienced that?
I, too, am interested in editing method of it. PCM is line, so it's easy for us to understand. DSD (1-bit), on the other hand, is "density".

I hope nice programmes will come soon.

My post was misleading: the concept sounds great, I have never heard 1 bit audio myself!

I'll be very interested to see how DAWs deal with it. Will a new visual system be devised, will it have to be converted to PCM???

The dynamic range must be obscene! Keep the user reports coming in I say

More on the subject: Acclivity pointed me to this interesting document, well worth reading: http://www.sel.sony.com/SEL/consumer/dsd/dsd.pdf

After reading it I have serious doubts that editing software will be available anytime soon... I gather that a goal of the industry in developing this technology is to make high-quality sound impossible to copy/edit/manipulate... except in very qualified and technical environments. For example, they admit that implementation of unbreakable DRM systems was a priority during development of DSD...

:enraged:

I'm not really an audiophile but I had a look at the document that dobroide and acclivity presented and became slightly annoyed when between all happy talking the authors put a little "by the way, this is another step in taking full control over customers" at the last page. I fear that dobroide is right and we can witness the introduction of DRM in recording and processing.

Let's hope customers stick to their 'old-fashioned' discs... :wink:

(Ps: No, I don't copy cd's or buy illegal ones...)

So how does 1-bit work? I thought it was better to have high bitrates. 2.8MHz sounds like an amazing sampling rate, though.

"1-bit" means "1 or 0."
If the microphone gives loud signal, A/D chip makes many 1s in the file, and many 0s for silent sound.

1011111011111111011111011100111111101101011111110011010111110101110 is rather loud, and
0000000001000000100010100000010100000000000100000000001000010000000 is very silent, I think.

1-bit recorder just records these numbers in the memory. And D/A chip changes them into natural wave for us to hear. Quite simple.

Multi-bit PCM is just like drawing a line with tiles. No matter how hard you try to simulate the real line, each tile has corners like a stairway. You shall never get rid of this fact.
By forming 16bit/44.1kHz waves, you have to abandon much fidelity. 24bit/192kHz is pretty good, but still less than simple 1bit quality. (at this rate, human ears can hardly tell the difference, though....)

1bit devices can be made at lower cost than high-quality PCM recorders, and the sound fidelity may be more than you can imagine.

"1-bit" means "1 or 0."
If the microphone gives loud signal, A/D chip makes many 1s in the file, and many 0s for silent sound.

1011111011111111011111011100111111101101011111110011010111110101110 is rather loud, and
0000000001000000100010100000010100000000000100000000001000010000000 is very silent, I think.

\quote]

Not quite as I understand it ... At each sample (at 2.8 MHz sampling rate) a decision is made "is the current amplitude greater or less than at the last sample". If greater, a 1 bit is output, if less, a zero bit. Thus, a steeply rising portion of a wave results in many 1 bits, a steeply falling part of the curve results in many zeroes. A flat portion results in alternate 1 and zero bits. So the resulting bit pattern represents what changes are taking place in the amplitude, not the actual amplitude at a point in time.

Although the sampling rate is high, only one bit is output per sample, so the total number of bits required for a recording is "only" 4 times that required for 16-bit WAV at 44.1 KHz sampling rate.

A bizarely simple concept, like many good ideas. And it seems that it doesn't matter by how much the amplitude has changed from one sample to the next ... I take it the sampling rate is so high that only a minimal change can occur ... and if it's not adequately expressed, it will be by the time a few more samples have been taken.

Mike

Cool! Thanks for the explanation.

is the current amplitude greater or less than at the last sample

Oh, is it?

Thank you for telling it.

The Wikipedia article on the subject is also interesting. People don't seem to be sure yet, whether DSD is really better.

I don't think it would be very difficult to retrieve a "PCM view" from the signal for editing purposes. In the end (i.e., in the DAC) the signal has to be converted into an amplitude over time anyway, exactly what a software editor displays.

Looks like a very good recorder, nonetheless, especially with the harddrive. With that proprietary format and also the price (€1000?) however I'll stick to my Roland R-09.

Mirko

m1rk0

I don't think it would be very difficult to retrieve a "PCM view" from the signal for editing purposes. In the end (i.e., in the DAC) the signal has to be converted into an amplitude over time anyway, exactly what a software editor displays.

Mirko

I agree, based on my understanding of the encoding. It just seemed to me there would be an arbitrary DC offset in the resulting PCM view, as one has no defined amplitude starting point.

Mike