Using MD deck as DAC

[bluecrab]

After seeing quite a few threads here and there about using an MD deck as a standalone DAC, I decided to try it myself, especially when my JA333ES deck began emitting a loud, click-like sound upon turn-on and disc-read attempt. That can't be good.

I had the JA333ES pretty much set up to accept optical in, as it is connected to a digital switch, itself connected to an SCMS stripper. I set the JA333ES to receive optical out from an MXD-D400 CD/MD deck. As is known from experience and from discussion here, the 'D400 is great for copying from CD to MD, but falls somewhat short in the DAC department.

The output from the 'D400 sounds noticeably better when run through the JA333ES's DAC. The sound is louder, fuller, and there is more low-end. It was easy for me to switch between the 'D400 straight-up and the 'D400 played through the JA333ES's DAC - just a button-push on my amp. That was how I tested, using various CDs and MDs run through both DACs.

It's simple to do this. You just connect the digital source to the target MD Deck, press record on it, and start playing the source. No need to insert an MD. The display shows "-DA", indicating that the unit's DAC is in use, I would think. The DAC in the JA333ES is much better than the one in the 'D400 and the resulting SQ is likewise better.

I would expect that the JA20ES and the JB940 also have DACs superior to that of the MXD-D400. What it all means is that I will be able to actually play CDs and MDs on the MXD-D400 and get good SQ out of it. I wish I had thought to try this months, if not years ago.

Anyone else hijacking a good MD deck's DAC?

[sfbp]

No (as discussed I use my Onkyo Receiver as an excellent D->A for the '400).

However I do the opposite. None of the receivers put OUT a digital signal that you can record (or if they did the rights would be restricted). However capitalizing on the A->D circuits in a good deck works wonders, because as described elsewhere, MD A->D actually goes through 24-bit processing before producing a nice optical signal, which explains why MD decks record conventional analog signals so brilliantly.

All this is not to say that a portable is hopeless on line input... but those analog (metal) jacks have a horrible tendency to be very noisy. I pipe the optical OUT from the deck to the input of a HiMD recorder, and as they say "Bob's your Uncle" (and Fanny's your Aunt).

So your deck says "- DA", mine says "AD - DA", with the second part being irrelevant (no, it's a hippopotamus).

Stephen

[KJ_Palmer]

My MDS-JE640 is digitally fed from 2 sources, an Inoi 500Gb Media player which only has a coaxial output, and the optical TOSLink from an LG DVD/USB deck. The JE640 in turn feeds its analogue RCA outputs into a Denon Receiver. Don't mind saying it's quite a nice setup for me, most audio-visual bases being covered, including MD of course.

I could use my JB980 for the same job, and perhaps get a slightly better result from its DAC, but the JE640 seems fine, and is slightly less obtrusive. Maybe I'll switch over to the JB980 one day, but for now it's happily ensconced in my MD-CD only system.

[Chickenvenus]

Does anyone have a dedicated stand alone DAC attached to an MD Deck? I have one I use with a iMac, and have been thinking of also using it with a MD Deck.

[sfbp]

The only disadvantage of standalone DAC is that by the time the signal has been converted to optical, you already lost its 24-bitness. So certain MD decks will BEAT an external DAC. But the 640 + external DAC (as noted higher up) beats the 640's own D->A.

I'm curious, what was the mod to your 640 that you show in your signature?

[Chickenvenus]

Hi sfpb,

My 640 came out of a sound studio and has had 2 modifications.

1. Fitted with a rack mounting kit.

2. of more interest, fitted with a pair each (L & R) of TRS Jacks, these are fitted to the rear panel right hand side.

Other than that it's OOTB, and runs well.

I'm going to run an digital optical out to a Cambridge Audio 'DAC Magic' for upsampling to 24bit/192Khz.. That 'should' improve sound quality.

[sfbp]

fitted with a pair each (L & R) of TRS Jacks, these are fitted to the rear panel right hand side.

Other than that it's OOTB, and runs well.

I'm going to run an digital optical out to a Cambridge Audio 'DAC Magic' for upsampling to 24bit/192Khz.. That 'should' improve sound quality.

Interesting! Nowadays one can probably pick up an E-11 or E-12 for the price of doing a mod like that. But you do end up with XLR and optical on the same unit which is otherwise unobtainable.

Seeing as MD is inherently 24-bits, for the upsampling is it better to try and get a mod closer to the disk than at the optical out? Seems to me that by the time you are sampled at 44.1 you've already chucked out the higher resolution. If you are technically minded maybe it would be good to look at the decks which have 24-bit out, to see how to do this. That is, the JA555ES, the JB940, the JA33ES or the JA333ES (the JA30ES has optional 20-bit out only). Perhaps staring at the service manuals would give you a clue about the design. But I'm not sure what the advantage of upsampling something that just got downsampled, if you see what I mean.

Stephen

[PhilippeC]

Naim has done a DAC that bring music from optical source to another level. This DAC maybe would a real difference. www.naim-audio.com

Web : The secret is in the respect of the timing clock, «Zero S/PDiF Jitter Design» (my turn to become technical but I don't understand it myself).

Bring your deck to a Naim Audio reseller...

[sfbp]

I'm going to run an digital optical out to a Cambridge Audio 'DAC Magic' for upsampling to 24bit/192Khz.. That 'should' improve sound quality.

This upsampling business confuses me a bit. Seems likely that when I feed my TOSlink to the input of my (Onkyo) AV Receiver which is 192kHz capable, presumably it upsamples to that before converting to analog. Does anyone know if that is the case?

[PhilippeC]

This upsampling business confuses me a bit. Seems likely that when I feed my TOSlink to the input of my (Onkyo) AV Receiver which is 192kHz capable, presumably it upsamples to that before converting to analog. Does anyone know if that is the case?

The real good question...

[Chickenvenus]

This upsampling business confuses me a bit. Seems likely that when I feed my TOSlink to the input of my (Onkyo) AV Receiver which is 192kHz capable, presumably it upsamples to that before converting to analog. Does anyone know if that is the case?

I don't know the specifics of your AV Receiver, but normally if an AV is capable of 192kHz, it means that it is capable of converting 192kHz digital to analogue. I'd say then it's not 'upsampling' anything.

[Chickenvenus]

Interesting! Nowadays one can probably pick up an E-11 or E-12 for the price of doing a mod like that. But you do end up with XLR and optical on the same unit which is otherwise unobtainable.

Seeing as MD is inherently 24-bits, for the upsampling is it better to try and get a mod closer to the disk than at the optical out? Seems to me that by the time you are sampled at 44.1 you've already chucked out the higher resolution. If you are technically minded maybe it would be good to look at the decks which have 24-bit out, to see how to do this. That is, the JA555ES, the JB940, the JA33ES or the JA333ES (the JA30ES has optional 20-bit out only). Perhaps staring at the service manuals would give you a clue about the design. But I'm not sure what the advantage of upsampling something that just got downsampled, if you see what I mean.

Stephen

I'd agree that if the optical output of an MD deck is 44.1/16bit, whereas the MD itself is 44.1/24bit, then you are loosing some bit depth from the MD, to then upsample it again in an external DAC to 192/24bit sounds like a less than ideal arrangement. I guess the thing also is that a lot of MD recordings are going to be copies of CD's, so the original material was only ever 44.1/16bit, recorded and presumably "upsampled" itself during recording to 24bit. The other thing to ponder, is the extra 8bits of depth highly significant when compared to the huge loss of bit rate when comparing a CD to an MD (CD = 1,411.2 kbit/s vs MD (Stereo Play) = 292kbit/s).

Having said that, the JE640 does sound good outputted to the external DAC, then again, the JE640 sounds pretty good to me on it's own.

[sfbp]

Where the bit depth comes in is that now you can compress the CD sound more accurately owing to the wider dynamic range. This is actually the trick that enables MD to mimic a CD at much higher bitrate - by concentrating the bits where they are needed for actual sounds. Sony coined the term "psychoacoustic" and it stuck, being in wide use now. So repeated compressions and expansions are going to lead to rounding errors, essentially. Make sense?

[sfbp]

I don't know the specifics of your AV Receiver, but normally if an AV is capable of 192kHz, it means that it is capable of converting 192kHz digital to analogue. I'd say then it's not 'upsampling' anything.

You've lost me. If the input is Fs 44.1 kHz then converting internally to 192 is upsampling, no?

[Chickenvenus]

You've lost me. If the input is Fs 44.1 kHz then converting internally to 192 is upsampling, no?

Do you have the model number?

Where the bit depth comes in is that now you can compress the CD sound more accurately owing to the wider dynamic range. This is actually the trick that enables MD to mimic a CD at much higher bitrate - by concentrating the bits where they are needed for actual sounds. Sony coined the term "psychoacoustic" and it stuck, being in wide use now. So repeated compressions and expansions are going to lead to rounding errors, essentially. Make sense?

As clever as it might be, and as much as I am a MD fan, I doubt a MD will sound better than the original source CD.

I do agree upsampling and downsampling, followed by upsampling would have a negative impact on the data.

[sfbp]

Actually there's a fair bit of evidence on this point (CD sounding better than original), both logical, and also anecdotal. More to the point if you record an analogue source to MD you may well get something better than CD quality, and eminently suitable for mastering a CD. This is precisely how I got into MD.

My Onkyo TX-SR605 has a TMS320DA708 as the main DSP chip - it seems that most processing is done by this chip. Tell us what you know.

[Chickenvenus]

Actually there's a fair bit of evidence on this point (CD sounding better than original), both logical, and also anecdotal. More to the point if you record an analogue source to MD you may well get something better than CD quality, and eminently suitable for mastering a CD. This is precisely how I got into MD.

My Onkyo TX-SR605 has a TMS320DA708 as the main DSP chip - it seems that most processing is done by this chip. Tell us what you know.

• I'm fine with people having alternate views to me, always healthy to explore ideas together.
  
• For me an MD copy (in SP or less) cannot have the same 'quality' as the CD source it originated from. Sony talk about 'CD like', and 'close to CD quality', I note. I accept the sound on an MD may be 'different, and 'different' can be said by individuals to 'sound better', as that's always subjective.
  
• Copying an analogue source (like an LP) onto an MD is a different matter. I think i'll pass on commenting on that one.
  
• I had a bit of a look at the information available online regarding the Onkyo TX-SR605. I'm unable to satisfy myself that it is upsampling audio to 192kHz. Very happy to pointed in the right direction on this though.
  

Thanks

Interesting topic.

[sfbp]

For me an MD copy (in SP or less) cannot have the same 'quality' as the CD source it originated from. Sony talk about 'CD like', and 'close to CD quality',

Doesn't that depend a bit on how the CD was read? I recall all the stuff about "oversampling" which I assume is routine now - but was only in high end players (I had one) 20 years ago. So if there existed a way of getting better sound off a CD (say 20 bits) wouldn't that actually be more permanently captured by 24-bit MD?

I am groping here, since like you and everyone else I don't see how to get something out of nothing. However I keep seeing all this ABCD stuff in the circuits (and some CD's have EF as well). Is that not the 4 (or 6) extra bits? Just my general ignorance, would be happy to have someone fill in the gaps here.

Interesting addendum: the TMS320DA708 chip in my Onkyo has a datasheet here. In the section entitled "Industry Standard Software"

The PA software solution includes

- Decoders

Dolby® Digital, Dolby Digital

EX; DTS® Digital Surround,

DTS-ES Discrete 6.1, DTS-ES

Matrix 6.1, DTS 96/24; MPEG-2

AAC multichannel, MPEG-4

AAC stereo; Windows Media®

Audio version 9, WMA9 Pro,

MP3, HDCD, ATRAC3plus;

DSD-to-PCM conversion

- Encoders

Stereo: MP3, Windows Media

Audio version 8, ATRAC3plus,

MPEG-2 AAC; Multichannel:

Dolby® Digital 5.1 creator

Are y'all still awake???? I see the magic phrase ATRAC3plus in there. What does this mean in the context of such a datasheet? Does this mean that in some way receivers with this chip are specially suited to ATRAC3+? I have no clue but it seems an odd thing to be claiming.

[jim.hoggarth]

Stephen - the ABCD and EF stuff. I assume you mean the A-F signals coming from the laser unit? All optical disc lasers need ABCD in order to focus on the data. An algorithm such like A-C + B-D generates an error signal for the focus coil. The theory is that any slight change in focus causes an imbalance in the signal received by these four photo-diodes and can be used to keep the laser's target lens in focus with the track despite any up/down movement of the disc relative to the sled carrying the optical pickup.

The E and F photo-diodes are used to generate a lateral error signal for perfect tracking. These diodes are not associated with the data stream, a separate set of photo-diodes read this, I and J. I understand there are other systems which use two I and J diode pairs for both signal and tracking too. Plus, you can also find another photo-diode used as a control for the laser output, to stabilise the light output from that. Phew! I am sure there are some errors in all that BS I have just typed, so apologies now.

Incidentally, when you get faults on the optical pickup it is often these photo-diodes that have failed, rather than the focus or tracking coils, or indeed the laser itself.

Jim

[Chickenvenus]

Hi,

• I see this list of features is for an audio software that utilizes the TMS320DA708 chip. My reading here is that if you want to use this software/chip combination to encode into ATRAC3plus, it's available. I'm unsure that this has a special significance when the chip is part of an AV.
  

I'm sure we are way off topic though here.

[Chickenvenus]

Killed this one off didn't i...