Possibly chosen to ensure it's always deliberately asynchronous
Precisely.
Not yet - right now I have a Behringer DCX2496 driving the ncores, but aiming to replace it with the DCLP - just got shipping acknowledgement.
Not sure what's so excellent about it, as 16 or 36 dB is just as good as 26, as long as the overall gain structure of your system makes sense.
Easier, yes, but not better. It is never a very good idea to over-amplify and then have to attenuate. Ideally you have just enough amplification in your whole chain to give full output with full-scale input. The nice thing about a digital source is that you know for sure what the highest possible input amplitude will be.
Where is your system volume control when using the DCX2496?
It sounds like you will now have to un-modify your Ncores for use with the DLCP. Some simple, fixed attenuators installed between your DCX2496 could have (now) easily been removed to facilitate your DLCP integration.
We don't always have the option to avoid an amplify/attenuate scenario when integrating various commercial gear. A fixed attenuator is a sonically transparent option that can optimize gain structure in most situations.
Cheers,
Dave.
Our ears works on a logarithmic way.what I mean is; if you have to turn the digital volume down to a point where it is noticeably effecting the sound quality, your gain structure is wrong, because you shouldnt be needing to turn it down that much, if you do have to, you have too much gain.
Reducing the volume (level control) in the digital world reduce the numbers of available bits. While analog volume control reduce the levels of each step, but keep their number (definition) the same. We have to remember too that our ears are able to ear signals UNDER the noise floor.
If 24 bits are preferred for audio reproduction, it is not for the signal/noise ratio (96dB is more than enough) but for definition of the lowest signals.
Last edited:
edit: sorry for snippyness, need to think, just opened eyes, but I really at this point do not see the relevance
with proper gain structure and 32bit+ volume, with 24bit dacs and proper gain structure, remind me of how any of that has any impact?
you are using the very same mechanisms used in the production process to mix that music
with proper gain structure and 32bit+ volume, with 24bit dacs and proper gain structure, remind me of how any of that has any impact?
you are using the very same mechanisms used in the production process to mix that music
Last edited:
Our ears works on a logarithmic way.
Reducing the volume (level control) in the digital world reduce the numbers of available bits. While analog volume control reduce the levels of each step, but keep their number (definition) the same. We have to remember too that our ears are able to ear signals UNDER the noise floor.
If 24 bits are preferred for audio reproduction, it is not for the signal/noise ratio (96dB is more than enough) but for definition of the lowest signals.
Isn't extra noise exactly what happens when you reduce the number of bits, thus the signal / noise ratio is lowered?
And 16 bits is enough for music, so in practice with a 24 bit dac you have 8 bits of padding you can shave away without losing anything at all. And when you are shaving off more than 8 bits then the signal is too quiet for you to notice the increased noise anyway.
This of course as some others here already have said requires that your gain structure is ok.
it also ignores the difficulties of building a multichannel high grade analogue balanced control and the effect it has, with 12 decks minimum, with lower CMRR due to resistor matching (thus higher noise and higher distortion), OR a chip based digitally controlled analogue attenuator, which adds more active elements and lowers the performance to its performance.
man this is a hard nut to crack, the audiophile stubbornness is so strong with this issue, people that normally understand much more complex mechanisms somehow manage to miss the point with digital volume.
its made even more ironic by the venue here ie. a DSP/Crossover that uses these mechanisms at its very heart, if you cant palette digital volume, perhaps you should look at other devices....
man this is a hard nut to crack, the audiophile stubbornness is so strong with this issue, people that normally understand much more complex mechanisms somehow manage to miss the point with digital volume.
its made even more ironic by the venue here ie. a DSP/Crossover that uses these mechanisms at its very heart, if you cant palette digital volume, perhaps you should look at other devices....
Last edited:
Am I understanding this correctly, if using the AES input you can only do stereo ? But if you use the analog inputs you get 8ch in ?
If you apply dither you can shave of more than 8 bits with digital volumecontrol!
Ok a part list:
Dac: AK4396
Dac intergrator: NE5532
Balanced output buffers: OPA1632
DSP: TAS3108
USB audio: PCM2704C
ADC: PCM4202
SRC: SRC4382
Ok a part list:
Dac: AK4396
Dac intergrator: NE5532
Balanced output buffers: OPA1632
DSP: TAS3108
USB audio: PCM2704C
ADC: PCM4202
SRC: SRC4382
well I guess people will first work out an I2S input mod to avoid that 2704
the opa1632 is a great choice though, love that chip!
the opa1632 is a great choice though, love that chip!
Why this poor NE5532 ? High slew rate OPAs are better here, on my experience.Dac: AK4396
Dac intergrator: NE5532
Balanced output buffers: OPA1632
Why this poor NE5532 ? High slew rate OPAs are better here, on my experience.
What is wrong with the NE5532??? Because it is an old design?????
true high slewrate is preferred with first integrator IMO too, but the 5532 isnt as bad as many make out. I would choose something else, but then we didnt design it did we?
Slew-rate too slow, omho. And not very neutral (i use to compare with strait wire) .What is wrong with the NE5532??? Because it is an old design?????
Its main interest is high current, where we don't need here, because a better driver follows.
Are-they DIP or SOIC ?
Last edited:
CorrectIt maybe possible that PGA stands for programmable gain adjustment, aka the gain jumpers on the board. If I have received mine, I will check this!
The DLCP is not only able to drive the NC400, so that's why the output stage can drive the NC400 into clipping.hope so, since that would send an ncore into clipping
Plus 1.
Can anyone with more knowledge then me comment the strange samplingrate?
DSP sampling rate Fs 93.75 kHz
ADC sampling rate Fs 93.75 kHz
Correct, but there are a few other reasons too.Possibly chosen to ensure it's always deliberately asynchronous
There is one AES input available (L+R), and two analogue inputs (L+R). The optional Input board has 4 stereo analogue balanced inputs, switched with relays.Am I understanding this correctly, if using the AES input you can only do stereo ? But if you use the analog inputs you get 8ch in ?
If you apply dither you can shave of more than 8 bits with digital volumecontrol!
Ok a part list:
Dac: AK4396
Dac intergrator: NE5532
Balanced output buffers: OPA1632
DSP: TAS3108
USB audio: PCM2704C
ADC: PCM4202
SRC: SRC4382
The NE5532 isn't used as DAC integrator! It's a voltage buffer, only used for a reference supply voltage.
it also ignores the difficulties of building a multichannel high grade analogue balanced control and the effect it has, with 12 decks minimum, with lower CMRR due to resistor matching (thus higher noise and higher distortion), OR a chip based digitally controlled analogue attenuator, which adds more active elements and lowers the performance to its performance.
man this is a hard nut to crack, the audiophile stubbornness is so strong with this issue, people that normally understand much more complex mechanisms somehow manage to miss the point with digital volume.
its made even more ironic by the venue here ie. a DSP/Crossover that uses these mechanisms at its very heart, if you cant palette digital volume, perhaps you should look at other devices....
Qusp,
to be honest, digital volume out of ESS chip is the first volume coming out of a silicone that sounds good. No other solution I ever heard was good, and that particularly stands for those PGA chips (not part of digital conversation, but still silicone). Analog was far better than any of those. Hearing digital volume out of ESS chip was revelation as well as relieve. As you are very well aware doing volume on multichannel is not a small task. As someone who was purist and kept analog volume as long as I could, I have to say, doing volume in digital domain, if done properly will do as good job as analog... almost. 😀
man this is a hard nut to crack, the audiophile stubbornness is so strong with this issue, people that normally understand much more complex mechanisms somehow manage to miss the point with digital volume.
I guess they are so fixated on the incorrect image of the wave consisting of digital "steps".
I understand better this choice, now.The NE5532 isn't used as DAC integrator! It's a voltage buffer, only used for a reference supply voltage.
The AD797 has only 20V/us slew rate and as an integrator on the output of a Dac its the gold standard for THD and Noise performance in the audio bandwidth.
Qusp I have not found a part that has higher slew rate that beats in THD and noise performance in the audio bandwidth.
The Ne5532 bang for buck is very hard to beat, no part at its price has better or equal THD and noise performance.
Arthur
Qusp I have not found a part that has higher slew rate that beats in THD and noise performance in the audio bandwidth.
The Ne5532 bang for buck is very hard to beat, no part at its price has better or equal THD and noise performance.
Arthur
Where is your system volume control when using the DCX2496?
In my source. Not ideal, but OK as a "transitional" system.
It sounds like you will now have to un-modify your Ncores for use with the DLCP. Some simple, fixed attenuators installed between your DCX2496 could have (now) easily been removed to facilitate your DLCP integration.
I agree - it would have been easier to use attenuators. Well, the mod is not too hard to undo, just fiddly.
We don't always have the option to avoid an amplify/attenuate scenario when integrating various commercial gear. A fixed attenuator is a sonically transparent option that can optimize gain structure in most situations.
I almost agree. I would say "mostly sonically transparent". You will have a less good impedance match because of the attenuator and a decreased signal-to-noise ratio compared to lowering amp gain, but mostly the difference shouldn't be audible.
- Home
- Source & Line
- Digital Line Level
- Hypex DSP module(s)