Topic: Korg MR1, WTF??? SOS

[MattH]

the dsd files are really representing something of quality, yes.
its just the conversion that seems to be a little tricky.

try going DSD>CD direct.  i'm loving what i hear out of mine

For me, I only care about DSD and conversion to CD. I have no interest in 24 bit anything from this point forward. I don't think CD will sound any better or worse than any decent 24 bit recorder conversion to CD.

[it-goes-to-eleven]

For me, I only care about DSD and conversion to CD. I have no interest in 24 bit anything from this point forward. I don't think CD will sound any better or worse than any decent 24 bit recorder conversion to CD.

How are you going to play the DSD?

A lot of us don't convert our 24 bit recordings to redbook for our playback. We play them at 24 bits...

The data suggest that the Korg DSD > redbook conversion is likely flawed.  But even the guys who did those tests don't seem to care enough to test 16/44.

[MattH]

For me, I only care about DSD and conversion to CD. I have no interest in 24 bit anything from this point forward. I don't think CD will sound any better or worse than any decent 24 bit recorder conversion to CD.

How are you going to play the DSD?

A lot of us don't convert our 24 bit recordings to redbook for our playback. We play them at 24 bits...

The data suggest that the Korg DSD > redbook conversion is likely flawed.  But even the guys who did those tests don't seem to care enough to test 16/44.

Will be playing DSD on the MR-1000. I also have Discwelder Bronze (1000-edit) to convert DSD to redbook flawlessly.

[mandoman]

I believe it can record in native PCM.  Anyone tried that at 24/96 or 24/192?

I second the request to try recording 24/192 native on the Korg and see how
the recording comes out. Do you still see the noise?

[it-goes-to-eleven]

I believe it can record in native PCM.  Anyone tried that at 24/96 or 24/192?

I second the request to try recording 24/192 native on the Korg and see how
the recording comes out. Do you still see the noise?

The answer is yes:


Korg MR-1. 15 kHz -20dB tone. Line in, PCM 192k recording. More than 40 dB of noise increase!

From here:

http://www.unipv.it/webcib/res_techtest_uk.html

[MattH]

I believe it can record in native PCM.  Anyone tried that at 24/96 or 24/192?

I second the request to try recording 24/192 native on the Korg and see how
the recording comes out. Do you still see the noise?

The answer is yes:


Korg MR-1. 15 kHz -20dB tone. Line in, PCM 192k recording. More than 40 dB of noise increase!

From here:

http://www.unipv.it/webcib/res_techtest_uk.html

The issue only appears to be with 24/192 whether recorded natively or when DSD is downconverted to 24/192 with AudioGate.

Personally, I could care less about 24/192 if that's the only issue with the new Korg DSD products.

[WiFiJeff]

You know, I have been concentrating on the right side of these two graphs (DSD to 192. 192 direct on the MR-1), because of the ugly things it does to the pcm wave in an editor.   Am I correct that the left side shows a noise floor well under -132 dB below 20 kH?  Does this mean we have real 22 bit performance here? 

Jeff

[it-goes-to-eleven]

Personally, I could care less about 24/192 if that's the only issue with the new Korg DSD products.

The issue also occurs at 24/96. Safe to assume it also occurs with 16/44 conversion.
Korg either released the product knowing about these issues or their testing was very weak. I'd guess they knew. Since there is no good way to listen to the DSD, the PCM is absolutely critical to listening. This is a major gaffe on their part.

You know, I have been concentrating on the right side of these two graphs (DSD to 192. 192 direct on the MR-1), because of the ugly things it does to the pcm wave in an editor.   Am I correct that the left side shows a noise floor well under -132 dB below 20 kH?  Does this mean we have real 22 bit performance here?

No, wrong type of test for that.  The 722 results are quite a bit lower and closer to -140dB in that freq range.. We know that just isn't the case.

[SClassical]

You know from what has been said here it seem as if this recorder should be kept as a DSD recorder and any files made with it should be achieved for future resampling using a better program.... Forget about using it as a PCM recorder...I think Korg concentrated in developing a DSD recorder only (not a player or a PCM recorder - even though it can do all that). We should relying on our "ears" more and judge it by the sound quality. What is disappointing for me now is the AudioGate program it comes with it not how the recording sound. I used this recorder 3 times and the first 2 times were disspointing (my fault) but the last time I used it (last night) the result was mind blowing. Tonight I will record a voice recital...we'll see how that goes :p

[ghellquist]

Just a guess, may be wrong: Going tape out from the 722 is not really giving the MR1 the full benefits. The tape out is first AD-converted in the 722 and then DA-converted (tell me if I am wrong on this one). And if you go into the MR1 it is AD converted once more.

Gunnar

[WiFiJeff]

I don't think the 722 tape-out goes through any A/D or D/A conversion.  In any case, the problems were evident on a Korg tape from DPA4060s through an MMA6000, with no digital stuff at all upfront (and I can't imagine the Korg preamp is better than an MMA6000, or the 722 pre for that matter).

Jeff

[live2496]

The person who had prepared the graphs had also done some checking on this. He asked someone who is involved with DSD conversion about it. His opinion was that the noise likely had to do with noise-shaping.

[WiFiJeff]

I talked with a tech at DiskWelder about their DSD conversion software (the DiskWelder Bronze 1000 with DSD conversion is available only through Tascam, to get DSD conversion from Diskwelder you have to upgrade to Chrome II, which is $3000).  They have two settings, roughly accurate and fast.  The accurate one takes around 20 hours to process a DVD-A worth of files.  This is clearly NOT what we have in AudioGate, and probably not what is implemented in their hardware either.

Jeff

[Nick's Picks]

speculation
all of it, at this point.  We'll know more soon, i'm sure.

[Nick's Picks]

The person who had prepared the graphs had also done some checking on this. He asked someone who is involved with DSD conversion about it. His opinion was that the noise likely had to do with noise-shaping.

in speaking w/doug oade this morning, he thought the same thing.