Hi,
So it is claimed. Associates have spend time and tested I2S in from their CDP vs. SPDIF. They report I2S is notably better.
So it would seem the Sabre does clean jitter, but either having less input jitter overall reduces the problems or the process to remove large amounts of jitter is not absolutely Transparent.
Dusty, who designed the Chip used to hang out here. maybe worth asking him.
Ciao T
So it is claimed. Associates have spend time and tested I2S in from their CDP vs. SPDIF. They report I2S is notably better.
So it would seem the Sabre does clean jitter, but either having less input jitter overall reduces the problems or the process to remove large amounts of jitter is not absolutely Transparent.
Dusty, who designed the Chip used to hang out here. maybe worth asking him.
Ciao T
I wonder what this really is ?:
MFAST™ vs. conventional PLLs
Most converters utilize PLL (Phase Locked Loop) circuits. MFAST™ has two distinct advantages. It's a high-speed asynchronous system that locks almost instantaneously to any data stream. Moreover, unlike PLLs which merely attenuate jitter, MFAST™ strips jitter out of the audio stream completely. Enabling you to enjoy pristine sonic clarity whether the incoming data stream is pure or anything but.
EMM Labs - Meitner Design - High Performance Audio
MFAST™ vs. conventional PLLs
Most converters utilize PLL (Phase Locked Loop) circuits. MFAST™ has two distinct advantages. It's a high-speed asynchronous system that locks almost instantaneously to any data stream. Moreover, unlike PLLs which merely attenuate jitter, MFAST™ strips jitter out of the audio stream completely. Enabling you to enjoy pristine sonic clarity whether the incoming data stream is pure or anything but.
EMM Labs - Meitner Design - High Performance Audio
Very interesting discussion.
A couple of ideas:
1- the TI ASRC chip has a bypass mode where it does reclocking, so two clocks to handle 44.1 and 48K material
2- S/W control is very easy nowadays with Arduino. Suggest have the option to tap into the s/w interface of the chips.
3- If chip has a sister chip with optional s/w interface, suggest choosing the one the option of s/w interface (e.g. to use digital volume control in the DAC)
4- Allow access to balanced output of DAC
Question for ThorstenL: Can the intrinsic jitter of 8804 be improved with better supplies?
A couple of ideas:
1- the TI ASRC chip has a bypass mode where it does reclocking, so two clocks to handle 44.1 and 48K material
2- S/W control is very easy nowadays with Arduino. Suggest have the option to tap into the s/w interface of the chips.
3- If chip has a sister chip with optional s/w interface, suggest choosing the one the option of s/w interface (e.g. to use digital volume control in the DAC)
4- Allow access to balanced output of DAC
Question for ThorstenL: Can the intrinsic jitter of 8804 be improved with better supplies?
Last edited:
Hi,
The SPI control version is SRC4193.
Actually, nowadays SW control is mostly mandatory to unlock the potential of the more recent chips.
Well, it has VCO on chip (analog supply pin), where noise will be reduced by the on chip PLL loop, but only so much. Making it's job easier can't harm. I only ever tried the 880X with decent supplies...
Ciao T
The SPI control version is SRC4193.
Actually, nowadays SW control is mostly mandatory to unlock the potential of the more recent chips.
Well, it has VCO on chip (analog supply pin), where noise will be reduced by the on chip PLL loop, but only so much. Making it's job easier can't harm. I only ever tried the 880X with decent supplies...
Ciao T
Thanks. How about the crystal? (I think you discussed this elsewhere also): what is better a crystal or a clock (one of those 1 psec ccxx models), and
12MHz frequency or 24-27Mhz frequency? (in h/w mode only 12 is allowed, but in s/w mode higher frequencies)
Hi,
I have not studied the 880X in such great detail. But I suspect the same as usual applies...
Crystals wrapped around a gate (colpitts oscillators) are not very good, they are super sensitive to supply/ground noise. The limit for such noise is the resistance and inductance of the actual leadouts of the chip.
This implies also that problems get worse as clock speeds increase.
So, an external clock oscillator carefully tied into the oscillator part for lowest problems and using lower clockspeeds seems suggested, all else being equal.
Ciao T
I have not studied the 880X in such great detail. But I suspect the same as usual applies...
Crystals wrapped around a gate (colpitts oscillators) are not very good, they are super sensitive to supply/ground noise. The limit for such noise is the resistance and inductance of the actual leadouts of the chip.
This implies also that problems get worse as clock speeds increase.
So, an external clock oscillator carefully tied into the oscillator part for lowest problems and using lower clockspeeds seems suggested, all else being equal.
Ciao T
Hai Kat,
Ta, I figured there was some small room to improve on what they had in mind. And it had not yet turned into the latest group buy and they offered "open sauce" AND they did invite comments.
Hence I broke my own rules about giving free advise.
I have found people do not value advise until you call yourself Consultant (note, only Consultants, Lawyers and ****** have clients) and you charge your clients at least $ 1,000 US per day for giving it. In fact, if you charge $ 1,000 US you can be incredibly rude and blase about things too, it would get you thrown off boards like this one here.
Ciao T
Ta, I figured there was some small room to improve on what they had in mind. And it had not yet turned into the latest group buy and they offered "open sauce" AND they did invite comments.
Hence I broke my own rules about giving free advise.
I have found people do not value advise until you call yourself Consultant (note, only Consultants, Lawyers and ****** have clients) and you charge your clients at least $ 1,000 US per day for giving it. In fact, if you charge $ 1,000 US you can be incredibly rude and blase about things too, it would get you thrown off boards like this one here.
Ciao T
To the bored, modulators and other bosses...
For star star star star stars sake (shall we call it "for a flying french coitus's sake"?), can we leave some robust anglosaxon vocabulary uncensored?
Being a White Hero Opposing Red Extremism (W.H.O.R.E.) or being a broad that spreads for cash used to be very proper. Let's not let political (in)correctness run away and lets preserve the purity of English and of the anglosaxon language.
Yours truely, Concerned (a Saxon)
For star star star star stars sake (shall we call it "for a flying french coitus's sake"?), can we leave some robust anglosaxon vocabulary uncensored?
Being a White Hero Opposing Red Extremism (W.H.O.R.E.) or being a broad that spreads for cash used to be very proper. Let's not let political (in)correctness run away and lets preserve the purity of English and of the anglosaxon language.
Yours truely, Concerned (a Saxon)
How about using the asrc as reclocker only? there is a mode where the sample conversion is bypassed and the signal is passed with the new clock
Hi,
Actually as advocate of "small government" and as a member of the ultra-extremist militant center (aka anarcho-libertarians) I like this one way better:
Ciao T
Actually as advocate of "small government" and as a member of the ultra-extremist militant center (aka anarcho-libertarians) I like this one way better:
An externally hosted image should be here but it was not working when we last tested it.
Ciao T
This project doesn't seem that different than mine, which was completed three years ago, or the TPA COD DAC for that matter (which was done after mine). I hope it's worth the effort...
no, in bypass mode only shift registers works (you can convert from RJ to IIS or LJ for example)
Simply by using the ears. Ceramics are really easy to point out.
At one audition, I had by mistake replaced a film decoupling with a ceramic. When we started listening, we both agreed, that something was wrong. But we did not understand why. Just something very wrong.
After an inspection we found the ceramic, changed into film... and suddenly everything was fine.
The thing is, that in theory, ceramics are the best for decoupling... But then again... In theory, an op-amp with THD at -130dB should sound extremly good 😀
There is no text book way of creating good audio gear. You must use your ears and not your eyes, since it is your ears you have to please.
Last edited:
If you have an idea of how to make the analog stage, please make a drawing. The we can try it out, and maybe implement in the design. Would be great.
- Status
- Not open for further replies.