Korg Forums

Dan and Jeff, thanks for repsonses to my audio track available time question

I am trying to record minidisc tracks to the Oasys for live use (bass left/click for drums right). I will try the method Jeff suggested if the one I'm using is a no-goer.

My preferred way of recording minidisc tracks to the Oasys is via SPDIF from a Sony unit.

First go didn't work properly as on play back there was a slight "buzz" on bass guitar notes (clock set to internal).

Second attempt I changed the clock to SPDIF and everything recorded great, but the Oasys played back nearly 2 semi tones lower than normal. In fact every time I choose SPDIF as the clock rate the Oasys plays lower.

Any idea why this is? Effectively makes SPDIF clock rate unusable.....although i suppose I could globally re-tune the Oasys to match the recordings

I hope I am doing something wrong

Any help greatly appreciated.

Regards,

Ray.

You will definitely need to clock the Oasys to the MiniDisk player. The way you have done it with setting the Oasys clock source to the SP/Dif in is probably correct.

Your problem of samples sounding a lower pitch is most likely because one device is running at 44Khz the other at 48Khz. I don't have my Oasys in front of me, but it should be possible to set the Oasys's sample rate to match what is coming out of your minidisk recorder.

The Oasys will only work correctly if the clock that it is syncing to is 48kHz (or 96kHz IIRC). If you clock it at 44.1kHz it will appear to work, however everything will be at the incorrect pitch as you've observed (if you hit a key whilst the Oasys is synced this way you'll it plays back at the wrong pitch). I doubt that you can change the sample rate on your MD player to 48Khz so I suspect you'll need to use analog connectivity to record your audio from the MD to the Oasys. What the D/A -> A/D conversion will do to the audio is probably not as bad as what ATRAC has already done to it

Daz.

Gents,

Thanks for the prompt replies.

What I find strange is that the Oasys pitch lowers if I choose SPDIF as the clock rate even without the MD player attached.

So on that basis, SPDIF is unusable, unless I either don't play/record the keyboard along with anything recorded, or re-tune the keyboard to the new SPDIF pitch

All is not lost. I will try Jeff's suggestion for loading .wav's (although a bit fiddly as I will have to save the files as .wav's first). Failing that I will do audio, although I was trying to eliminate any noise/hum.

Thanks again for taking the time to reply

Regards,

Ray.

Ray wrote: What I find strange is that the Oasys pitch lowers if I choose SPDIF as the clock rate even without the MD player attached.

If the clock is set to S/PDIF, but no S/PDIF input is connected, the system (like most other systems that use or am aware of) will not work properly. On the OASYS, you should see a "CLOCK ERROR" message at the top of the screen.

Ray wrote:All is not lost. I will try Jeff's suggestion for loading .wav's (although a bit fiddly as I will have to save the files as .wav's first).

You'll also need to sample-rate convert them, if they were recorded at 44.1.

Best regards,

Dan

danatkorg wrote:

Ray wrote: What I find strange is that the Oasys pitch lowers if I choose SPDIF as the clock rate even without the MD player attached.

If the clock is set to S/PDIF, but no S/PDIF input is connected, the system (like most other systems that use or am aware of) will not work properly. On the OASYS, you should see a "CLOCK ERROR" message at the top of the screen.

Ray wrote:All is not lost. I will try Jeff's suggestion for loading .wav's (although a bit fiddly as I will have to save the files as .wav's first).

You'll also need to sample-rate convert them, if they were recorded at 44.1.

Best regards,

Dan

Dan,

You are right. I do see a clock error message. I'm just confused that with SPDIF clock chosen and no connection from MD player, the Oasys detunes as if the MD was conected.

Our drummer thinks he knows how to sample rate convert either from Cool Edit Pro or using an old Korg D8 bewteen the MD and the Oasys

So we'll give that a go first with fingers crossed.

Many thanks.

Ray.

Ray wrote:Our drummer thinks he knows how to sample rate convert either from Cool Edit Pro or using an old Korg D8 bewteen the MD and the Oasys

Alternately, from the OASYS internal or external USB disk:

*DISK mode

*UTILITY tab

*Select RATE CONVERT, from the upper right dropdown menu.


If you are saving back to the same folder, you need to change the name of the WAV, even if it is just a single character.

Mike Conway wrote:

Ray wrote:Our drummer thinks he knows how to sample rate convert either from Cool Edit Pro or using an old Korg D8 bewteen the MD and the Oasys

Alternately, from the OASYS internal or external USB disk:

*DISK mode

*UTILITY tab

*Select RATE CONVERT, from the upper right dropdown menu.


If you are saving back to the same folder, you need to change the name of the WAV, even if it is just a single character.

Mike,

Interesting.....

What if I recorded stereo tracks using SPDIF from a MD player? Could I then convert it to 48 kHz sample rate so that the Oasys plays at the correct pitch? I might not be explaining myself very well here, but I think you understand what I am trying to achieve.

Thanks,

Ray.

Ray wrote: You are right. I do see a clock error message. I'm just confused that with SPDIF clock chosen and no connection from MD player, the Oasys detunes as if the MD was conected.

If it actually sounds the same, that implies to me that the MD wasn't connected properly. More likely, the sound is close but not identical...

Ray wrote:Our drummer thinks he knows how to sample rate convert either from Cool Edit Pro or using an old Korg D8 bewteen the MD and the Oasys

I don't think the D8 will do sample-rate conversion in realtime. As Mike notes, the OASYS can also convert the sample rate of an imported WAV file.

Best regards,

Dan

Tell us again why you don't want to use analog? Are you actually having some quality problem, or are you afraid you might? As Daz mentioned, you have already had a quality loss due to the Minidisk compression algorithm, so I find it unlikely that you would really be losing anything material by going analog. This would be a lot easier than what you're trying to do.

danatkorg wrote:

Ray wrote: You are right. I do see a clock error message. I'm just confused that with SPDIF clock chosen and no connection from MD player, the Oasys detunes as if the MD was conected.

If it actually sounds the same, that implies to me that the MD wasn't connected properly. More likely, the sound is close but not identical...

Dan,

I mean that with no connection to the MD player (and SPDIF clock chosen from the Oasys Global page) a clock error meaage is observed and the Oasys is de-tuned.

With the MD connected there was no clock error message, but the recording was fine - if I then play it back with SPDIF clock rate the Oasys is de-tuned - if I choose Internal clock, the recording plays back too fast.

I am going to give the convert option a try this evening.

Thanks all....I'll remember you all on my first album

Ray.

Ray wrote: I mean that with no connection to the MD player (and SPDIF clock chosen from the Oasys Global page) a clock error meaage is observed and the Oasys is de-tuned.

Yes, that's expected.

Ray wrote:With the MD connected there was no clock error message, but the recording was fine - if I then play it back with SPDIF clock rate the Oasys is de-tuned - if I choose Internal clock, the recording plays back too fast.

Yes, this would happen if the MD was operating at 44.1kHz.

Best regards,

Dan

tcornishmn wrote:Tell us again why you don't want to use analog? Are you actually having some quality problem, or are you afraid you might? As Daz mentioned, you have already had a quality loss due to the Minidisk compression algorithm, so I find it unlikely that you would really be losing anything material by going analog. This would be a lot easier than what you're trying to do.

I can hear some hiss (after normalization) using analog line out from the MD player. As this is being amplified to 7.5 kW live, I want to make sure that hiss/hum is minimised. You can understand i don't want to record several hours of material to find that there is too much noise for live use. I am probably being over-cautious to be honest.....

To be fair, I'm sure I could live with it, but I just want to make absolutely sure there is as little noise recorded as possible. I have already recorded some analog and (to be fair) it sounds great through headphones - but as I say, there is a small amount if hiss.

The recordings are only a stop-gap and eventually the tracks will be midi with a few samples added here and there, so the quality will be very high.

Regards,

Ray.

danatkorg wrote:

Ray wrote: I mean that with no connection to the MD player (and SPDIF clock chosen from the Oasys Global page) a clock error meaage is observed and the Oasys is de-tuned.

Yes, that's expected.

Ray wrote:With the MD connected there was no clock error message, but the recording was fine - if I then play it back with SPDIF clock rate the Oasys is de-tuned - if I choose Internal clock, the recording plays back too fast.

Yes, this would happen if the MD was operating at 44.1kHz.

Best regards,

Dan

Dan,

It is clearly my misunderstanding of how SPDIF works. It just seems that the Oasys SPDIF clock rate setting seems to be expecting/using 44.1 kHz rate with nothing connected. Maybe the penny will drop eventually...........

As I say, I will try to convert the MD recording from 44.1 to 48 kHz tonight and hopefully things will fall in to place.

Thanks,

Ray.

Ray wrote: It is clearly my misunderstanding of how SPDIF works. It just seems that the Oasys SPDIF clock rate setting seems to be expecting/using 44.1 kHz rate with nothing connected.

If there's no clock coming in, the OASYS clock's PLL will drop to its minimum frequency, which I'd guess is somewhat lower than 44.1. It's not going to be exactly 44.1.

No clock = bad. Clocks are easy to think about and manage. See the description of the System Clock parameter in the OASYS Parameter Guide for a little more info.

For an overview, here's something I wrote a while ago for the OASYS PCI:

* * *

A few words on word clock

Whenever two or more audio devices are connected together digitally, they are
sending and receiving thousands of individual bits of data every second.The bits
are sent out continuously, one after another, at a very steady pace. This is similar to
two jugglers passing balls between each other, while at the same time keeping up
their own steady juggling patterns.

When one device sends out a bit of data (like the ball being thrown to the other
juggler), the other needs to be ready to receive it. If the data is being sent even just
slightly faster than it can be received, bits will be lost, causing errors in the audio
(like the ball being dropped).

Similarly, if the data is being sent even slightly slower than the receiver expects it
to be, then the receiver will occasionally be left without real data at its input (like a
juggler grabbing at the air). In digital audio, these errors show up as loud pops
and clicks, as well as lower-level noise.

The rate at which those bits are sent and received is controlled by the word clock,
which “ticks” for every bit of data. Each device has its own word clock, so that it
can work on its own–such as a single ADAT, or a single OASYS PCI.

When you send digital audio data between two or more devices, however, their
word clocks need to be synchronized together, so that devices send and receive
each bit at the exact same moment. One of the devices–usually the sender–will
provide the master clock; the others must be set to ignore their internal clocks, and
instead slave to the master clock.

This clock signal is normally carried along with the digital audio data, so that
when connecting two devices together in a simple configuration–such as
connecting OASYS PCI to a digital mixer, or dubbing between two DAT
machines–you don’t need to make any other connections. You will, however, need
to make sure that only one device is providing the master clock, and that all the
other devices are set so that they slave their clocks to the master (or to another
slave device).

Sometimes, it may also necessary to send the word clock separately from the
digital audio data. For instance, you may be using a dedicated word clock
generator, or a SMPTE-digital audio synchronizer. In this case, as before, you need
to make sure that all devices are set to slave to the master clock.

It is best to set up a studio to use a single clock, and to stay with that clock at all
times (or as much as possible). This will help to ensure that all timing and pitch
remains the same.

* * *

- Dan