Korg MR-1000 / 1bit-5.8MHz DSD portable recorder

[MaxEars]

Thanks very much David

You're right ... I've been wanting to try DSD sound for these recordings ... I needed to get some feedback on my thought processes... very helpful all...

Fundamentally they (MRs) would not be the right tools for the job .... at this time there is no 4 track ... so get real ... and forget it... for now! ( : As you say any unit I buy now could be sold later .... but i usually don't sell gear ( :

I do love SACD and DSD and wold like to support these formats ... hopefully the MRs are a great success and we can look forward to multi-channel portable units some time in the future

I'll do as you suggest and move my further enquiries to a new thread ... you have given me much to consider .... with the 788T a fifth mike could be used ...

Best regards and thanks for your thoughtful replies

Max

[Shadow_7]

Originally Posted by MaxEars

The MR2000 is not an option as I will need DC power out there in the middle of wherever

There are a number of power options if you need power in the middle of wherever. Larger batteries that plug directly into the devices DC port. And battery plus inverter options to supply A/C power. Or even a distant generator. They can be fairly small and somewhat quiet these days.

While there may not be much drift between two units, it adds a step in post. Which is one more opportunity to get it wrong. And any drift may cause phase issues and/or intonation issues that may not have existed in the source. Plus any sort of clap for alignment has a negative affect if recording wildlife. The animals run away, the birds stop chirping, and the crickets take a pause.

[bobkeiser@ieee.o]

I think you should get a pair (of MR-1000s) and try it! Don't let the naysayers hold you back.

[Plush]

You may indeed have some fun and some success with using two 1000 machines. However, your success will not come from keeping them in sync or from any added "quality" from dsd recording. They will not stay in synch.

The way one keeps 2 stereo machines in synch is with SMPTE time code and these machines do not have time code. As for the dsd, by the time you convert the Korg dsd recording to pcm with the included Audiogate software, there will not be much difference between the dsd and the pcm.

If you require an excellent machine for multitrack, I am suggesting you buy a Nagra 6 (6 channel) machine made in Switzerland or other machine like the Sound Devices 4 channel made in Wisconsin.

[ghellquist]

Originally Posted by Plush

The way one keeps 2 stereo machines in synch is with SMPTE time code and these machines do not have time code.

Or even better, word-clock. SMPTE does not work on small digital recorders.

And I agree on all the recommendations. Another suggestion might be two Sound Devices 702. These are proved rugged, with good preamps and can easily be synched. You might want to buy one and rent the second.

I have tested my 722 against a MR1000, and my ears cannot believe this vogue about DSD recordings. Clearly the MR1000 has better sounding headphone amps, easily heard even without recording, but once a at home they sound very similar. A good mark for the lower price MR I believe, but nothing to be very excited about.

As for jungle recordings, well, the traditional advice surely would be Sennheiser MKH mics. Supposedly better suited to humid environments than most other condensors.

As for underwater recordings (also asked for), my guess is that this is something completely different.


GUnnar

[matyas]

Originally Posted by ghellquist

Or even better, word-clock. SMPTE does not work on small digital recorders.

GUnnar

Wordclock has nothing to do with positional sync, which is what you'd need in this context. Wordclock is for sample sync, an altogether different beast. When synching multiple digital machines (like, say, the Sony 3348 that I am occasionally called on to use), you may need to use wordclock to deal with jitter, but that won't keep transports locked. You do need SMPTE for that. I don't own a Korg unit (someday, I may), but I'm pretty sure that it has neither word nor SMPTE.

[ghellquist]

Originally Posted by matyas

Wordclock has nothing to do with positional sync, which is what you'd need in this context. Wordclock is for sample sync, an altogether different beast.

In digital recorders it is the same thing really. When sampling is done at exactly the same speed in both machines they will stay in relative positional sync forever once recording is started. This can only be achieved reliably with "sample clocking", usually done through a word clock connection. The only thing left to cater for is exact starting time of the recording, easily compensated for by sliding samples in post production. A clap to synch to is an easy answer.

Word clock allows samples from several units to stay in synch down to a very low jitter (less than sample time). This will allow for a stable stereo picture even if you record the left and right channel on two different machines. This of course goes for surround stability as well.

SMPTE is generally not supported by todays crop of boxes, so we can quickly rule that one out. It is more or less a thing of the past with ADAT-s or (shudder) reel-to-reel.

Several of the boxes today support Time Code, used mostly on film recordings. The implementation of time code is done by putting the start time of the recording into a special field in the .wav file (the format used is called bwav). Time code here does not compensate for diffent sample speeds so units will slowly drift out of synch.

Slow drift between the clock in two machines can to a certain degreen be compensated for in software by resampling. The function could be called time stretch or something like that.

The 7xx boxes allows you to connect one cable between two boxes which transfers both word clock and start/stop commands between them. It is not perfect but it brings home the bacon. In contrast the MR1000, to my knowledge, has no way of synching two machines.
Gunnar

[Ruphus]

Originally Posted by Plush

As for the dsd, by the time you convert the Korg dsd recording to pcm with the included Audiogate software, there will not be much difference between the dsd and the pcm.

While I suppose the rest of the post to make good sense, I dare to object to the above quote.

To my ears DSD recordings made with the Korg stay different from PCM recordings yet after conversion.
Especially HF contents seem to remain much more smooth and pleasent.
No sizzle with DSD / PCM converts, as far as I could tell.

Ruphus

[matyas]

While you go on to explain what you mean a little bit at the end, I do take issue with the statement that wordcolck and SMPTE are the same thing on digital machines. If certain machines support position markers and transport commands over wordclock, then that's a non-standard use of the technology and it's not something I've ever seen. I've never seen a system that sends transport or position control over word. Conversely, SMPTE is far to slow to sync digital clocks. As far as it being a thing of the past, anyone who works with video or with synching analog gear to digital will tell you otherwise. I'd much rather have proper frame synch than try to line things up in post - I've done that and I don't like it.
As far as newer portable machines not supporting SMPTE, it doesn't really surprise me, but I just don't want anyone coming away with the impression that wordclock and SMPTE do the same thing.

Originally Posted by ghellquist

In digital recorders it is the same thing really. When sampling is done at exactly the same speed in both machines they will stay in relative positional sync forever once recording is started. This can only be achieved reliably with "sample clocking", usually done through a word clock connection. The only thing left to cater for is exact starting time of the recording, easily compensated for by sliding samples in post production. A clap to synch to is an easy answer.

Word clock allows samples from several units to stay in synch down to a very low jitter (less than sample time). This will allow for a stable stereo picture even if you record the left and right channel on two different machines. This of course goes for surround stability as well.

SMPTE is generally not supported by todays crop of boxes, so we can quickly rule that one out. It is more or less a thing of the past with ADAT-s or (shudder) reel-to-reel.

Several of the boxes today support Time Code, used mostly on film recordings. The implementation of time code is done by putting the start time of the recording into a special field in the .wav file (the format used is called bwav). Time code here does not compensate for diffent sample speeds so units will slowly drift out of synch.

Slow drift between the clock in two machines can to a certain degreen be compensated for in software by resampling. The function could be called time stretch or something like that.

The 7xx boxes allows you to connect one cable between two boxes which transfers both word clock and start/stop commands between them. It is not perfect but it brings home the bacon. In contrast the MR1000, to my knowledge, has no way of synching two machines.
Gunnar

[ghellquist]

Sorry matyas for any misunderstandings.

This discussion started, as far as I understand it, with how to synchronize two field recorders. The idea was to make long recordings, two hours, on two machines and then "mix" these into a coherent surround sound scape. The machine suggested for use was the MR1000.

I do not personally find the MR1000 to be the most appropriate box to use in that application. One reason is that it does not have word-clock synchronization. This means that the two machines will slowly drift out of synch. Soundwise this means that the surround sound scape will have phasing effects, or after a while even hearable echo effects.

I wanted to point out tha SMPTE is not a solution is this situation. It is not available on the current crop of field recorders for one thing. It of course has its uses when synching separate analog or digital decks but the asked for application is not a prime example. Word clock does not have any transport control. Sorry if that could be implied from my answer.

Transport control is not used if you run fully inside a DAW-program. Here the files are imported fully and exists totally "inside the computer". No external units are synched using transport control in this scenario. Synchronizing of two files needs to be done only once, when the files are placed in the time line. (Synching start points is what time code is used for in this kind of work setup).

The problem when clock frequences differs between two machines is that the number of samples will differ over, say, an hour. You can see this in that one hour of "real-time-recording" will not show up as exactly one hour in the time line in the program.

I find basically three solutions:
1 - assume (or better, verify) that the magnitude of the problem is small enough in your application
2 - synchronize the word clocks in the field when the recording is done
3 - resample.

Solution no 3 is quite feasible in most software programs today. You simply align the start points and then "stretch" one or the other sound file to get the same length. You will get small sound artifacts, often small enough not to be heard. The result can turn out quite useful most of the time, but there are no guarantees that the drift between the two machines will be exactly the same over time. The stretching will handle the static difference in frequency between the two clocks, but will not necessarily handle if the clocks over time are not stable.

Of course, an even better solution would be to use a multi-channel machine, but if needed I suggest solution 2 instead.

Once again, the whole point is that SMPTE is not an option as it is not available. Even if it was available it would not necessarily solve the issue. Word clock is available and is the sure solution.

Gunnar

[ghellquist]

Originally Posted by Ruphus

To my ears DSD recordings made with the Korg stay different from PCM recordings yet after conversion.
Especially HF contents seem to remain much more smooth and pleasent.
No sizzle with DSD / PCM converts, as far as I could tell.

Ruphus

Not my experience. I have tested a MR1000 against a 722 in a field recording of a symphony orchestra. Using the built-in preamps you could hear a difference. Using an external preamp with Y-splitter I cannot hear any difference. Test was done at 96kHz/5.8MHz and conversion at highest quality in Audiogate.

I suggest a careful AB-test -- record the source using an Y-splitted preamp into two recorders. This, to me, is the only way to really verify.

Gunnar

[tsvisser]

Can someone tell me if the analog outputs on this unit, when recording are zero latency... that is the signal path is:

mic pre > analog output

or is there some latency?

mic pre > A/D stage > D/A stage > analog output

[lucey]

If there is latency it's extremely brief. I have nothing for you otherwise.

Originally Posted by ghellquist

Not my experience. I have tested a MR1000 against a 722 in a field recording of a symphony orchestra. Using the built-in preamps you could hear a difference. Using an external preamp with Y-splitter I cannot hear any difference. Test was done at 96kHz/5.8MHz and conversion at highest quality in Audiogate.

I suggest a careful AB-test -- record the source using an Y-splitted preamp into two recorders. This, to me, is the only way to really verify.

Gunnar

The MR1000 needs the AD mod to really shine IMO. I ran an analog 2" 16 track mix to each of my mastering converters and to the Korg at 5.6, then dropped it to 24.44.1. PCM recorded at 24/44.1 Fixed level, printing one after the other in rapid succession.

I would not want to use the Korg instead of the Pacific Microsonics for mastering, but I wouldn't want to use the Pacific for mixing. The Korg is even handed with no smear and no hot spots ... although not euphonic in anyway, as is the HDCD converter.

[Shadow_7]

I did run my Korgs RCA outs to my full duplex soundcard and didn't notice any latency. Not that I've tried multi-tracking to it yet. But I've completely disconnected my DMP3 and am about to put it on craigslist, since I'm not actively using it, and can use the Korg instead, should the need arise. I still need a couple XLR to TRS cables to use use my current gear in the manner that I wish. And some shorter XLR to XLR cables.

As far as recording at PCM or recording at DSD on the Korg then using audiogate to convert to PCM of the same specs. There is a noticeable difference in that instance. For my Avenson STO-2s that difference is best described as clarity of the spoken word. Others might describe it as transient detail. But since I'm in essence using the same mics on the same source on the same recorder, that jargon might not be applicable. It basically gives my SDCs a hint of LDC detail. It's still an Omni SDC that needs proximity for best results. But I can put a little more distance in there, that I couldn't otherwise do, and still end up with good results. In my opinion.

[JonathanEspinosa]

Hello everybody

Does anybody know when playing mp3s from the KORG why the PEAK leds flashes ?

I know that indicates clipping when recording using either the MIC inputs or LINE inputs.

>Thanks!

[pesciolino]

What is the AD mod?

What mods for the MR1000 exist? (I'm assuming for the pres...)

Originally Posted by lucey

The MR1000 needs the AD mod to really shine IMO.

[lucey]

* The led is flashing as your mp3 is clipping.

* The AD Mod is some ICs in the analog path, and caps. Busman Audio.

Originally Posted by Lynn Fuston

This reminds me of a product looking for a customer.

Tell me why I'm wrong.

2.5 years later I guess this one has gone pretty well for them Lynn!

[Marik]

Originally Posted by lucey

The MR1000 needs the AD mod to really shine IMO.

I agree. I don't think that would be hard to accomplish. The only problem might be getting into bunch of SMD stuff...

Best, M

[lucey]

Originally Posted by Marik

I agree. I don't think that would be hard to accomplish. The only problem might be getting into bunch of SMD stuff...

Best, M

Yes it's a simple mod ... but yes it's surface mount critters! I went back and forth with Busman to get a sound I liked, which is the mod he offers, but there are many options that are all more open on top and cleaner on the bottom than the old TL072 or similar that's in there.

[Ogrady]

Newbee here,
have followed this thread on the Korg DSD product for years,

Hope this is the place to announce this;

Wavelab 6, (and now a few other programs as well) is now 384khz, 24 bit capable. The realization, (I think, am I right?) that all that's left to edit our DSD files recorded with the Korg is to have the next release of AudioGate convert from DSD 2.8/5.6 to DXD (384khz) for editing and back to DSD for playback thru the Korg thus creating for the first time an affordable editing solution for those not gifted with 10's of thousands of $$ for a Pyramix or Sonoma solution has my attention big time! This potential is very exciting to me!

I thought this thread to be an excellent place to get the word out in hopes those at Korg are 'listening'. Are you there Korg? Please make this possibility a reality ASAP!

Sean

[tsvisser]

Originally Posted by Ogrady

Newbee here,
have followed this thread on the Korg DSD product for years,

Hope this is the place to announce this;

Wavelab 6, (and now a few other programs as well) is now 384khz, 24 bit capable. The realization, (I think, am I right?) that all that's left to edit our DSD files recorded with the Korg is to have the next release of AudioGate convert from DSD 2.8/5.6 to DXD (384khz) for editing and back to DSD for playback thru the Korg thus creating for the first time an affordable editing solution for those not gifted with 10's of thousands of $$ for a Pyramix or Sonoma solution has my attention big time! This potential is very exciting to me!

I thought this thread to be an excellent place to get the word out in hopes those at Korg are 'listening'. Are you there Korg? Please make this possibility a reality ASAP!

Sean

out of curiosity, what are the other programs that now support 384Khz/DXD?

[Ogrady]

A quick web search came up with Cakewalk SONAR 8 and another Steinberg program, HALion 3.

[CharlesCola]

Hi, can anyone tell me at which point the rec level pots are unity gain on the mr1000? I would like to know where on the scale the unit is neither attenuating or adding gain....

[lucey]

Originally Posted by colnbishop

Hi, can anyone tell me at which point the rec level pots are unity gain on the mr1000? I would like to know where on the scale the unit is neither attenuating or adding gain....

Just insert it and listen for it as you switch in and out. 12:45 on ours, but it has the AD mod.

[CharlesCola]

thanks

[zorf]

Any thoughts on the latest ( June/2009) audiogate software update?

[bobkeiser@ieee.o]

Well, I'm getting a little tired of having to plug the MR-1000 into the MacBook Pro to authenticate every time I want to use AudioGate 2.0...but otherwise it's a pleasure. Upgrades include batch-editing options; ability to burn discs--including DSD discs--directly from the software; choice of Triangular Probability Distribution Function (TPDF) and KORG AQUA dither (haven't compared yet).

Does anyone have experience with getting the authentication to behave? TIA!

[lucey]

Originally Posted by bobkeiser@ieee.o

Does anyone have experience with getting the authentication to behave? TIA!

Not an issue for PC users. For once, something nice about a PC.

[macula]

Originally Posted by bobkeiser@ieee.o

Well, I'm getting a little tired of having to plug the MR-1000 into the MacBook Pro to authenticate every time I want to use AudioGate 2.0...but otherwise it's a pleasure.

Does anyone have experience with getting the authentication to behave? TIA!

On my Macbook I only had to authenticate ver. 2.0 once, and it's been permanent ever since.

[reynaud]

Originally Posted by bobkeiser@ieee.o

Well, I'm getting a little tired of having to plug the MR-1000 into the MacBook Pro to authenticate every time I want to use AudioGate 2.0...

Korg has released an update to AudioGate (2.0.1) that fixes this issue.

cheers,
Reynaud