24-bit/96Khz?

[David Evans]

Is it possible to change the bit and sample rate with the MZ-NH1? If so, could you please tell me how I can do it. The reason why I'm asking this is because I need to record high quality sounds for my major project. I have already used the MZ-NH1 with an AKG C1000S and Sony's ECM-MS957 stereo microphone for recording ambience and sounds etc. Thank you.

[greenmachine]

It's a consumer product that supports 44.1/16 only. However, because of the high quality ADC, quality is more than adequate and mainly limited by external gear (microphones, -placement).

[Qwakrz]

The NH1 does include a sample rate converter. I have fed it so far with 44.1khz and 48khz samples and it has recorded at 44.1khz with both of them, changing the samplerate on the fly.

I doubt, however, that it will be able to cope with changing from 96khz or 24bit.

[David Evans]

Thank you for your prompt replies and sorry for not replieing sooner. Out of curiosity, how do you access the sample rate converter? Thank you.

[Qwakrz]

There is no way to access the sample rate converter, you just feed the signal into the optical in. There is no configuration and the unit will convert 32, 44.1 and 48khz signals into 44.1khz for recording onto the disc.

[Prof.OND]

Qwakrz (or anyone else)

you say the signal is converted, does this mean that for any recordings up to 48 khz there is a cutoff of frequency at 44,1 khz.

And does this mean in case of 32 khz recording fed to the MD that noise will be added to about 44,1 khz or not?

Nobody's entirely wrong in considering this a freak question. You might even complement me.

[dex Otaku]

The standard rates of 32kHz, 44.1kHz, and 48kHz are supported. It's possible that any rate in between can be converted automatically as well, they generally do clain 32-48kHz functionality.

Even 44.1kHz streams are resampled.

MD and HiMD record at 44.1kHz and use 16-bit converters. There are no other options with either format.

Rather, portables use 16-bit converters. [Possible 1-bit delta/sigma PWM with conversion to 16-bit PCM, even, since these A/D converters are so much cheaper]

[e1ghtyf1ve]

The standard rates of 32kHz, 44.1kHz, and 48kHz are supported. It's possible that any rate in between can be converted automatically as well, they generally do clain 32-48kHz functionality.

Even 44.1kHz streams are resampled.

MD and HiMD record at 44.1kHz and use 16-bit converters. There are no other options with either format.

Rather, portables use 16-bit converters. [Possible 1-bit delta/sigma PWM with conversion to 16-bit PCM, even, since these A/D converters are so much cheaper]

Some decks have been using 24-bit converters for some time.

24-bits are pretty cheap now, you know!

[dex Otaku]

Some decks have been using 24-bit converters for some time.

24-bits are pretty cheap now, you know!

Decks, yes. Portables, no.

Go take a look at prices for A/D converters .. last time I looked around at this [about 12 months ago] 1-bit A/D converters [as are used in almost all consumer devices including 24-bit sound cards] for about $1.50USD .. true 24-bit linear PCM converters were somewhere around the $60USD mark for inexpensive ones. Both were in quantities of 10,000 if I remember correctly.

Also, I'm willing to bet that the other 24-bit portables out there right now [such as the newer PMD model from Marantz, and the M-Audio Microtrack 24/96] also use 1-bit converters.

From what I've seen, linear converters are a rarity outside of professional equipment.

That doesn't mean that 1-bit converters suck, though.

[mmilovan]

What are the differences between linear and 1-bit A/D converters?

From what I've seen, linear converters are a rarity outside of professional equipment.

That doesn't mean that 1-bit converters suck, though.

[e1ghtyf1ve]

Decks, yes. Portables, no.

Go take a look at prices for A/D converters .. last time I looked around at this [about 12 months ago] 1-bit A/D converters [as are used in almost all consumer devices including 24-bit sound cards] for about $1.50USD .. true 24-bit linear PCM converters were somewhere around the $60USD mark for inexpensive ones. Both were in quantities of 10,000 if I remember correctly.

Also, I'm willing to bet that the other 24-bit portables out there right now [such as the newer PMD model from Marantz, and the M-Audio Microtrack 24/96] also use 1-bit converters.

From what I've seen, linear converters are a rarity outside of professional equipment.

That doesn't mean that 1-bit converters suck, though.

I agree. Based on my experience with lots of equipment, I wouldn't be surprised if the portable HiMD units sported 1-bit converters. Their analog performance is incredible, in spite of the ridiculously low price tag. The PMD-671 I assume you are referring to (~$1000US) almost certainly uses 24-bit linear converters. It sounds amazing to my ears. The MT performs at best as a 16-bit recorder, and not a very good one at that (I compared it side-by-side to my PCM-M1). Does it use a 1-bit converter? Probably, but the analog stages are so poor does it really matter?

Cheers

[dex Otaku]

I agree. Based on my experience with lots of equipment, I wouldn't be surprised if the portable HiMD units sported 1-bit converters. Their analog performance is incredible, in spite of the ridiculously low price tag. The PMD-671 I assume you are referring to (~$1000US) almost certainly uses 24-bit linear converters. It sounds amazing to my ears. The MT performs at best as a 16-bit recorder, and not a very good one at that (I compared it side-by-side to my PCM-M1). Does it use a 1-bit converter? Probably, but the analog stages are so poor does it really matter?

Likely not.

You're the first person I've read anything by about any of the Marantz PMD series that had anything positive to say about its analogue stage. I keep reading that their mic preamps make Hi/MD portables' preamps look super-high quality by comparison.

[Prof.OND]

Coorect me if I'm wrong but i think that

one bit converters measure the frequency of a sample making an average of the output 0/1 wich is created through a delta sigma modulator at least the Nyquist rate (twice the sample frequency), wich then is put together to make a linear 16 bit PCM datastream.

I think the difference with other converters is that they use only the frequencies that are fed to the convertor.

Again I don't really understand, but that's what I've made from it.

[e1ghtyf1ve]

Likely not.

You're the first person I've read anything by about any of the Marantz PMD series that had anything positive to say about its analogue stage. I keep reading that their mic preamps make Hi/MD portables' preamps look super-high quality by comparison.

Perhaps the less than stellar PMD-660 tainted the 671's reputation somehow? Even so, I would take the 16-bit 660 over the MT any day. Those extra US$100 nets an enormous boost in quality of workmanship and reliability. I simply switch to higher output mics.

Yes, in my most humble subjective opinion, of all the sub-$1000 portable/pocket recorders available today, the HiMD analog input stages are the best, and virtually on a par with those of my PCM-M1. We're on the same wavelength here. To me, the resulting subjective sound quality is king.

If you want true high performance portable 24-bit recording, may I suggest the following:

Sound Devices 722 Portable Audio Recorder

You get what you pay for, though. With HiMD, sometimes you get a whole lot more.

Cheers

[e1ghtyf1ve]

Coorect me if I'm wrong but i think that

one bit converters measure the frequency of a sample making an average of the output 0/1 wich is created through a delta sigma modulator at least the Nyquist rate (twice the sample frequency), wich then is put together to make a linear 16 bit PCM datastream.

I think the difference with other converters is that they use only the frequencies that are fed to the convertor.

Again I don't really understand, but that's what I've made from it.

Yes, I believe you are on the right track. We are actually talking about pulse density modulation here, not frequencies per se. The 1-bit converters are inherently linear and as dex Otaku pointed out, relatively economical to manufacture. This is why most high resolution converters (such as 24-bit) use the 1-bit approach as it is so difficult to make segmented or ladder versions. Pulse width modulation is used more and more for hifi audio.

Most MD decks from 1999 onwards use 24-bit resolution converters (probably 1-bit DA/ADC hybrids) and dither down to 16-bits for PCM/ATRAC format storage. This is how S/N can be around 100dB or better; dithering and noise shaping.

I've not seen any evidence (somebody point me to it if I missed it) that HiMD portables don't use the same or similar hybrid converter scheme as the MD decks. Why wouldn't they, as it's cheap, proven technology. Maybe that's why they sound so good.

Cheers

Edited December 14, 2005 by e1ghtyf1ve

[Prof.OND]

One little supplement to my post:

As I'm looking into sigma delta conversion it seems to me that it has more difficulties in creating a digital copy of high analog frequencies.

e.g.: a sine wave at hight frequency seems to be more difficult to measure correctly then low frequencies.

Has anybody else come to the same conclusion?

PS: Not meaning that the quality of MD converter isn't good enough.

[e1ghtyf1ve]

One little supplement to my post:

As I'm looking into sigma delta conversion it seems to me that it has more difficulties in creating a digital copy of high analog frequencies.

e.g.: a sine wave at hight frequency seems to be more difficult to measure correctly then low frequencies.

Has anybody else come to the same conclusion?

PS: Not meaning that the quality of MD converter isn't good enough.

I don't know if I would use the word "difficulties" but yes you're correct. As you get into higher and higher frequencies, depending on your sample rate, "difficult" starts to become "impossible."

OT: Speaking of frequencies, did you know that mp3 has no scale factor band above about 16khz? Maybe that's why my mp3 players perform so poorly at high frequencies... hmmm Then again, mp3 ain't no hi-fi format.

Cheers

[DerTapir]

To answer all questions :

In the Sony MZ-NH1 the IC 301 is used for A/D Conversion and Mic Amp.

The used Chip for that is the -> AK 5356VN-L <-

AK 5356VN-L :

Its a 20 BIT A/D converter with output for 20 BIT OR 16 BIT (i think)

In my opinion, Sony uses only the 16 Bit part to stay compatible with the other

ATRAC systems on the market !!!

Here is the link to the data-sheet !!!

AK 5356VN-L ADC

PS: Don´t ask for details, i am no technican

Sorry i forgot:

Sampling rate 44,1 Khz !!!

[dex Otaku]

This is really interesting. Thank you for the boot to the head, DerTapir!

I don't know why I never checked my NH700 service manual for this. The NH700 shares the same ADC as the NH1.

As ATRAC itself was designed for encoding up to 20-bit sources, it's fairly likely that the internal encoder of these portables is capable of encoding ATRAC/3/plus directly from the 20-bit output of the ADC. If this is true, it's possible that the dynamic range of ATRAC/3/plus recordings could in theory be higher than that of PCM recordings.

I have no way to confirm this. Perhaps if the encoder supports 20-bit data, someone could try sending a 20-bit stream via SP/DIF to their recorder via the optical in and see what happens in various recording modes.

It would appear that our recorders have an actual linear converter in them, finally smashing any theories I had about cheap 1-bit converters.

Thanks again, DerTapir.

Also, the RH10 has the same ADC. Without checking all of the service manuals, I'm assuming the same chip is used in all current models.

[e1ghtyf1ve]

Bingo, mystery solved! You have my thanks as well, DerTapir! Based on the performance, I would guess that the signal is dithered, not truncated, to 16 bits.

Well done!

[Jammin72]

The Sony SBM-1 was like that in this regard. You had the ability to change the configuration of the chip (well those with fine soldering skills) to up it to 24bit Capture. The SBM Used Noise shaping anyway so you got a larger chuck of data for it to play with. This did increase the resolution of the unit but the larger difference in my mind was the changes that Doug made to the Analog Stages of the unit.

[e1ghtyf1ve]

The Sony SBM-1 was like that in this regard. You had the ability to change the configuration of the chip (well those with fine soldering skills) to up it to 24bit Capture. The SBM Used Noise shaping anyway so you got a larger chuck of data for it to play with. This did increase the resolution of the unit but the larger difference in my mind was the changes that Doug made to the Analog Stages of the unit.

I didn't know that one could do that. Neat. The SBM-1 was used to create some fine recordings, that's for sure. It's a great complement to my PCM-M1.