If there’s one thing I’ve learned since coming on board to help Peter with CDM it’s this: The kids love a portable recorder. The Edirol R09 and Zoom H2/H4 articles are among our most popular ever.

Now Korg are supplementing their existing multi-track digital recorders with the 20GB, pocketable MR-1 (US$899 MSRP), and the 40GB, tabletop MR-1000 (US$1499 MSRP).

Shipping right now, these are “1-Bit” recorders, recording “DSDIFF, DSF, and WSD 1-bit formats, as well as multi-bit PCM format (BWF) with resolutions up to 24-bit/192 kHz”. If you’re like me and thought that more bits was better, Korg have helpfully provided a PDF entitled “Future Proof Recording Explained”, to clear up any misconceptions you had about the usefulness of more bits. The executive summary: When you’re sampling really really fast, the only numbers you need are 1 and 0, to indicate whether the sample is up from last time, or down. That’s my only insight on the science part for now, people who understand more feel free to debate the 1-bit revolution in comments.

Hardware Features

MR-1

Korg MR-1 Portable Recorder

  • 20GB internal drive (6 hours @ highest quality 1-bit, 20 hours @ CD quality, MP3 support apparently available mid-2007)
  • Ability to plug in external FAT32 USB hard drive for expanded recording space
  • USB 2 transfers
  • Rechargeable lithium polymer battery (no word on battery life) or AC power
  • Dual balanced mini phone plug inputs
  • 2 track simultaneous recording and playback
  • Includes stereo electret condenser mic
  • Backlit LCD
  • Dimensions: 64 (W) x 120 (D) x 24 (H) mm / 2.52″ (W) x 4.72″ (D) x 0.94″ (H)
  • Weight 200g (7oz) with batteries
  • More specs (including plenty with dB and Hz for the audio nerds)

MR-1000

Korg MR-1000 portable hard drive recorder

  • 40GB internal drive (6 hours @ highest quality 1-bit, 40 hours @ CD quality, MP3 support apparently available mid-2007)
  • Ability to plug in external FAT32 USB hard drive for expanded recording space
  • USB 2 transfers
  • Power by 8 AA batteries or AC power
  • Dual XLR/ 1/4″ inputs with mic pres, phantom power
  • XLR and RCA outs
  • 2 track simultaneous recording and playback
  • Backlit LCD
  • Dimensions: 192(W) x 170(D) x 56(H) mm/7.56″(W) x 6.69″(D) x 2.20″(H)
  • 1.0 kg (2.20 lbs) without batteries
  • More specs

So. Fun times ahead for the field recordists! Now someone should take all of the available options and review them somewhere exotic. Peter is in Hawaii right now. Can we get a couple of these and some decent mics shipped to him so he has something to fill those empty internetless hours before he returns?

  • cdmr

    I'll never buy another portable recorder that requires special software to convert the files to standard audio formats like aiff, wav, or mp3. That extra step doesn't seem like much when you're buying the recorder, but it becomes a real pain in the ass when you start using it.

    It's a shame because it looks like an interesting device otherwise.

  • RichardL

    As I understand it the Korg recorders will be able to record directly to WAV files in 16 and 24-bit depths and 44.1, 48, 88.2, 96, 176.4, 192 kHz.

    But the MR recorders' real unique feature is that they will create DSD recordings in the DSDIFF, DSF or WSD file formats. At this point you will need the Korg software to convert to WAV or AIFF in order to read them in almost any audio software. But that may change over time as DSD gets broader support in the prosumer, project studio market. DSD is really intended as an archival format, hence the pitch "Future-Proof Recording".

    Somewhere I also read that Korg will be putting out a firmware update later this year that will add MP3 support. Or maybe I dreamed that part.

  • e3

    anyone looking into portable recorders should also have a look at the iriver h-1×0 series. you can still get them on ebay, and they have digital io as well as microphone input. using rockbox as an operating system they can record flawless mp3 and wav. good value for money and good size.

  • David Anderson

    A friend asked me about this a while ago. The Korg white paper is, not surprisingly, a marketing spin on the data.

    The A/D converter is a Burr Brown PCM4202 (also made by TI). If you take a look a the data sheet, you'll see that the base sampling rate is 44.1khz. In DSD output mode, data is oversampled at the rate of 64x or 128x. The effect is greater bit depth but it's disingenuous to say the sampling rate is in 5.6mhz.

    My friend pointed to the charts showing the conversion of a 20khz square wave. Since all a/d circuits employ some form of low-pass filter (in addition to the Nyquist problem), any converter will have a problem with non-sine waves near the upper limit of the sample frequency.

    In short, a 20khz square wave consists of a 20khz sine wave plus a 60khz sine (1st overtone, 3rd harmonic), 100khz (2nd overtone, 5th harmonic), et cetera (ad infinitum, in theory). A 44.1 system will have a 22khz lowpass filter (approx) that will never see any of those harmonics. Only the 192khz system (with a 96khz lowpass) will see the 1st overtone. As the paper points out, it's a "notoriously difficult 'torture test'". It doesn't show the benefits of 1bit, it shows the benefits over oversampling and high frequency lp filters.

  • bliss

    Hey David, with your understanding of Korg's white paper, what are your thoughts on their new recorders? Would you consider buying one of the recorders for use on critical missions?

  • RichardL

    David, if your friend is correct then the whole thing is a bit disingenuous. I'm very curious of his source information that there is a PCM DAC inside these devices rather than something like this:
    http://www.maxim-ic.com/appnotes.cfm/appnote_numb

    That just doesn't make much sense based on the data given.

  • RichardL

    Never mind. The description of the Burr Brown ADC is in the MR-1000 description.

  • http://www.rolandreinke.com Roland

    These look good!

    How would they compare to the Maudio and Edirol products?

    If mp3 support is really in the pipeline, this should be a great product.

  • David Anderson

    All market speak is disingenuous to some degree. My friend isn't that well versed in electronics (building stuff, reading and understanding data sheets, etc) so I pointed out to him what I thought was hyperbole.

    He was particularly taken by the graphs. I don't disagree with the data presented: capturing a 20khz square wave under those sampling conditions is challenging. It cannot be escaped.

    But one is forced to wonder about the utility of that example in evaluating the device for real world use. I suspect the recorder was fed a square wave from a lab signal generator. How does that translate to the real world? How do microphones reproduce that signal? What about speakers? What about (I'll admit this is esoteric) the impact of the signal path bandwidth? Finally, what about our ability to hear results of the recording technology?

    If I were to plunk down the dough, I'd want to see some real world results. Play me a recording made on that piece of equipment and another similar one, maybe the Maudio or Edirol devices.

    Marketing is hyperbole. Take what the white paper and other marketing efforts say with a grain of salt.

  • David Anderson

    RichardL, nice link. That's a good description of sigma delta conversion and oversampling.

    Here's my major issue: they're throwing numbers around as if they are interchangeable. Making a claim that the 5.6mhz data rate is amazingly better than 192khz. The problem is that you can't make a direct comparison like that. One-bit conversion techniques, as pointed out in the Maxim page you link to, are different from multiple bit techniques. Heck, there are different techniques within the multibit converters.

    They're relying on mumbo-jumbo and smoke and mirrors to get sites like CDM to talk it up.

  • http://www.milezero.org Thomas

    Sound On Sound had a good review on one of these in the most recent issue. They said there was a significant improvement in the high end on these recordings, and that they sounded more "precise" than comparable multi-bit recorders. SOS also had some graphics on how these do exhibit different behaviors at very high frequencies. It was interesting stuff. So maybe it's not all hype.

    I understand that the Variax guitars also use this type of sampling for their instrument modeling.

  • http://sonsofjob.blogspot.com Augmentalist

    I'm to lazy to do the math, how long a recording

    in DSD format fits in 20 gigs?

    How long does a conversion from DSD to Wav take?

    i'm curious.

  • http://sonsofjob.blogspot.com Augmentalist

    duh i just read the sheet.

    sorry all.

  • http://www.jaymis.com Jaymis Loveday

    You do make a good point about conversion though Augmentalist. Hopefully as time goes past more developers will include this kind of high quality format in their workflow, but until that time people are lumped with the included software to do their conversions.

    In my experience, included software is almost always absolute crap. Who knows how horrible the workflow on Korg's software is?

  • http://www.paulsop.com Doktorfuture

    If DSD is supposed to be 'future proofed' then why do we suddenly have a 2.something and a 5.something Mhz version?

    I'm sure we'll see a 10.something Mhz version, and special multi-channel time aligned versions, yada yada.

    I mean, cummon. It's obvious.

  • AlecEifel

    OK cool, let's say that this results in better capturing and storring of audio. Can someone please tell me how this can be used with current mixer apps such as ProTools & Logic?

    …Or, this is more like for archiving hi-res "original recordings"?

    Until ProTools & Plugins understands this format, I can't justify purchasing this.

  • RichardL

    The idea is that you can archive your original recordings and your mixed down masters.

    For now you will have to convert your DSD tracks to WAV before you can mix or edit.

    It will be interesting to see if or how DSD support gets incorporated into the project workflow.

    It maybe helpful to think of DSD files as analogous to Camera RAW files from a digital SLR camera. Using Camera RAW allows the photographer more flexibility and to defer decisions until post processing.

    The history of Camera RAW workflow may provide some clues about how DSD might get added to workflows.

    The first phase, which is where we are with prosumer DSD, was when camera manufacturers supplied RAW file conversion programs with their cameras that supported RAW. These programs could manipulate RAW files and create TIFF or JPEG files that could then be further processed in other programs. These tools (eg Nikon View and Capture) usually work inefficiently and are flaky compared to mainstream tools like Photoshop.

    The next stage was when the mainstream tools like Photoshop and iPhoto added import support for Camera RAW. But that proved not particularly efficient. (Even though Photoshop added support for Camera RAW it used to take me all day to process a shoot. As a result I only used Camera RAW on projects that I could afford to invest that extra time.)

    The final phase was when sophisticated software solutions specifically designed for RAW processing workflows emerged like Adobe Bridge, Apple Aperture and Adobe Lightroom. With Adobe Lightroom I can process an entire photo shoot in a couple hours or less and it works better too. As a result I can get the advantages of Camera RAW for any shoot.

    Now the question that this analogy begs is, does using DSD for source recordings really confer advantages over using high def PCM? I know professionals who have been using DSD technology for five years and swear by it. But there are also a lot of nay-sayers. And the DVD-A vs SACD format war was particularly dirty with all of DSD's dirty laundry being laid out for all to see.

  • Stephen Burkett

    I find it a bit odd that even the 1000 model only has two inputs. Mobile recorders a quarter of its price have at least four inputs and FireWire out, which it doesn't have either. I wanna know though how a 1 bit/5.6 mHz recorder handles the distortion that skeptics of DSD point out. Would it be significantly better since it can reach a max of 110 kHz and could put it up in that range?

    And kudos to David Anderson for his insightful comments!

  • Daniel Atkinson

    Does anyone have an idea of the quality of preamps??

  • David Hamilton

    Does anyone have any insight into the differences between the DSD formats? (DSDIFF, DSF, and WSD)

  • bengo

    so whats the story on the MR1 preamps?

    i want a small pocket recorder (eg r09) and a high quality one (eg fostex fr2le)

    will this do the job of both?

    there's a real problem, the real high end pocket recorders have no reports from users, this and the nagra there is no word on the web about the quality of the preamps.

  • Mike Johnson

    Gentlemen, please let me end this debate. I own the MR-1 and I intend to purchase the MR-1000. First up. It’s the TaSCAM DV-RA1000 that is a disingenuous recording device. Forcing you to write to DVD+RW discs. I owned one of these last year. The Discwelder Bronze software was a major pain in the behind to use as well. The Tascam DV-RA1000 & MR-1 both sound better when recording DSD than they do when recording PCM at 24-bit/192kHz. The only torture test acceptable to me is this: Taking my MR-1 to my local Audio Dealer to record high end vinyl Lp’s, and then playing back on the MR-1 against the original Lp. On Mobile Fidelity’s Patricia Barber 45rpm box set, the MR-1 captured perhaps 80% of the original information to it’s internal HD. The first attempt at playback revealed a nasty harsh sound with little natural decay of the acoustic instruments. Out with the generic silver cables, and in went the Audioquest Sky w/72v dbs battery bias packs. The 2nd attempt at recording the 45rpm mo-fi discs now had a smoother balance with the natural decay the vinyl possesses. A problem with the MR-1 soon became evident. It struggles to achieve dynamic peaks and that extra punch when you record with the internal lithium-ion battery. The unit pulls 2.4 amps at 5vdc, a serious amount of current for such a small recording device. I’m going to build a custom high end regulated PSU for the MR-1, and a massive vacuum tube rectified & regulated PSU for the MR-1000. The MR-1000 PSU will be a John Broskie Aikido shunt regulated design for excellent performance. I have never been able to capture sounds like I have with the MR-1. I have finally turned my back on all CDR recorders due to the problems with almost all optical media. The only good sounding cdr’s that exist are the mobile fidelity Ultra disc & Apogee Gold CDR-74. I loved the clean sounding Taiyo Yuden CDR-74 that Microboards offered, but all of the currently available 80 minute cdr’s are unacceptable sounding! The extra 6 minutes of recording time results in massive anomalies in the lower bass & midrange compared to the best of the 74 minute cdr’s..enough said. It’s DSD fro me from now on!

  • Leigh Magee

    David Anderson wrote, "you’ll see that the base sampling rate is 44.1khz. In DSD output mode, data is oversampled at the rate of 64x or 128x. The effect is greater bit depth but it’s disingenuous to say the sampling rate is in 5.6mhz." He misunderstands the data sheet. On p. 14, in DSD128 mode, the chip samples at 5.6MHz and outputs the one-bit signal at 5.6MHz. In "Single Rate Mode", it STILL samples the analog input at 5.6MHz, but outputs a 24 bit signal at 44.1KHz. There's no spin, no disingenuous anything, just technical jargon that must be unpacked.