Hello Oehlrich,
I am confused about this latest board. I am currently using your first pcb, running into my Benchmark DAC1. What advantage is there for me to use this new board? I don't want to limit the output to 96 kHz, as my DAC will handle 192, and I have heard bad things about upsampling lower rates in non-even ratios (eg. 44.1 -> 96). I would like to improve/replace the clock in my player, is this what this new board is for? Can it be made to output 192kHz?
Take care,
Doug
I am confused about this latest board. I am currently using your first pcb, running into my Benchmark DAC1. What advantage is there for me to use this new board? I don't want to limit the output to 96 kHz, as my DAC will handle 192, and I have heard bad things about upsampling lower rates in non-even ratios (eg. 44.1 -> 96). I would like to improve/replace the clock in my player, is this what this new board is for? Can it be made to output 192kHz?
Take care,
Doug
The Peasant said:
Hi Doug, operating the DAC-1 with the V-1 kit there will be no improvement for you using the new kit!
The new standard kit works the same as the V-1 Kit but the extended kit has some new features.
What the new kit can do is up- or downsampling the signal from the palyer to a constant 96 kHz if it is jumpered to do that. The clock to generate this 96 kHz output may either be delivered by an onboard XTAL like the XO chip from Tentlabs or by an external XO2. This will provide an output signal with dramatically reduced jitter.
Operating a DAC delivering superclock (256 x fs) this superclock could be used to synchronize the kit’s output. So the S/PDIF output would be in sync to this superclock and to the D/A. Jitter is avoided totally in this setup because the output of the kit (being in sync with the DAC) would just be reclocked by the DAC so you only will have the jitter of the DAC clock.
Using the DAC 1 there is no need to up- or downsample the player’s output because this would be done inside the DAC-1! The DAC-1 uses the same AD1896 chip as the kit does to up- or downsample the incoming signal to a constant output samplerate (I guess about a bit above 100kHz). Just open your DAC-1 and you will see the AD1896 and an XTAL providing the constant clock signal for the D/A converters.
Regarding your comment that upsampling from 44.1 to whatever uneven frequency will degrade quality. I have to say that this statement is definitely wrong. Listen to a CD with your DAC-1 and you will listen to an upsampled signal and this would have a very good quality which is widely accepted ;-))
Operating the DAC-1 the standard kit either V-1 or V-2 is the perfect choice. Because of the reclocking inside the DAC-1 you don’t have to care about the jitter of the player. Improving it will give you NO improvement at the DAC-1’s output.
Charly
Re: Re: Harman Kardon DVD-47
The CXD2753R in the DV656 produces 6ch DSD. I would imagine the 989 does the same. Question is what to do with it.
fmak said:
The CXD2753R in the DV656 produces 6ch DSD. I would imagine the 989 does the same. Question is what to do with it.
Re: Re: Re: Harman Kardon DVD-47
The question is:
Does the chip combo process DSD data from decoder thru PCM1738 in native, not PCM format? And how to get this out in proceessible form.
No doubt the 989 can output up to 192k in PCM format.
rfbrw said:
The question is:
Does the chip combo process DSD data from decoder thru PCM1738 in native, not PCM format? And how to get this out in proceessible form.
No doubt the 989 can output up to 192k in PCM format.
Re: Re: Re: Harman Kardon DVD-47
As you say later, PCM 1738 takes DSD data from the decoder direct. Question is:
What does it do with this data?
I can email you the info when I have the scanner in a few weeks'
time.
Fred
ps: is your new kit ready?
oehlrich said:
As you say later, PCM 1738 takes DSD data from the decoder direct. Question is:
What does it do with this data?
I can email you the info when I have the scanner in a few weeks'
time.
Fred
ps: is your new kit ready?
Re: Re: Re: Re: Harman Kardon DVD-47
Can't comment on the '989 but the '656 uses the DSD interface of the PCM1738. No conversion to PCM takes place. Other than sending it to a SM5816 or performing a Meitner style conversion, can't see much use for raw DSD data.
fmak said:
Can't comment on the '989 but the '656 uses the DSD interface of the PCM1738. No conversion to PCM takes place. Other than sending it to a SM5816 or performing a Meitner style conversion, can't see much use for raw DSD data.
New V 2 Kits available now!!
Hello out there,
I had announced it some time ago and now the V2 boards are here in my hands waiting for you!
Today I updated my website (http://freerider.dyndns.org) providing all details and all new photos you should need. Like at V1 there will be two kits (Standard and Extended) using the same PCB.
There are several new features which I hope you would enjoy:
- Limit output rate for DVD-A 192 to 96kHz
- Provide a constant output rate of 96kHz no matter what kind of media is played
- Low jitter operation at constant output rate if tentlabs XO DIY oscillator (24.5760MHz) is used
- Input for synchronisation to an external (256 x fs) superclock (e.g. from Apogee Big Ben)
- Seamless operation with Tentlabs XO-3 for best jitter performance
I had to increase the price a bit because the V1 board was a bit cheaper than the new one and I last time forgot that PayPal steels some money of your transfer. For your convenience I added a tiny cooler for the 7805 power regulator.
Have fun
Charly
Hello out there,
I had announced it some time ago and now the V2 boards are here in my hands waiting for you!
Today I updated my website (http://freerider.dyndns.org) providing all details and all new photos you should need. Like at V1 there will be two kits (Standard and Extended) using the same PCB.
There are several new features which I hope you would enjoy:
- Limit output rate for DVD-A 192 to 96kHz
- Provide a constant output rate of 96kHz no matter what kind of media is played
- Low jitter operation at constant output rate if tentlabs XO DIY oscillator (24.5760MHz) is used
- Input for synchronisation to an external (256 x fs) superclock (e.g. from Apogee Big Ben)
- Seamless operation with Tentlabs XO-3 for best jitter performance
I had to increase the price a bit because the V1 board was a bit cheaper than the new one and I last time forgot that PayPal steels some money of your transfer. For your convenience I added a tiny cooler for the 7805 power regulator.
Have fun
Charly
Re: Re: Re: Re: Re: Harman Kardon DVD-47
~G
one cool use of raw DSD data is in far superior DSD to PCM processing than any of the current ASICs can do.. and yes, 656A/757A indeed sends unprocessed DSD right to the DACs and I'm hoping to see much more players do the same..rfbrw said:
~G
Re: Re: Re: Re: Re: Re: Harman Kardon DVD-47
And what is this marvel that will provide superior DSD to PCM conversion?
I don't think that was in dispute.
Glassman said:
And what is this marvel that will provide superior DSD to PCM conversion?
I don't think that was in dispute.
rfbrw,
Without giving too much away, 95% of digital filters are designed incorrectly – this combined with the fact commercial designs are heavily compromised to reduce silicon area, it was not to hard to develop a better filter – once we understood which ‘parameters’ where important.
We have spent a great deal of time understanding these parameters and there effect on audio quality. Its difficult to talk much about our filters without helping the competition, rest assured the DSD to PCM filters Glassman mentioned is the best on the market (at least to our current knowledge) – and yes listening has confirmed there is a GREAT deal that can be done to improve ‘off the shelf’ filters.
John
Without giving too much away, 95% of digital filters are designed incorrectly – this combined with the fact commercial designs are heavily compromised to reduce silicon area, it was not to hard to develop a better filter – once we understood which ‘parameters’ where important.
We have spent a great deal of time understanding these parameters and there effect on audio quality. Its difficult to talk much about our filters without helping the competition, rest assured the DSD to PCM filters Glassman mentioned is the best on the market (at least to our current knowledge) – and yes listening has confirmed there is a GREAT deal that can be done to improve ‘off the shelf’ filters.
John
JohnW said:
The limitation of OTS filters are not exactly unknown, subject as they are to the whims of the great god, computational efficiency. That being said I don't think it is fair to say designed incorrectly. I'm sure the designers would loved to have created an exotic filter with infinitesssimal passband ripple and down by 150dB at 22K but I don't think too many would be prepared to pay for a pair of AD TigerSHARC's or a million gate fpga.
Here's hoping it is all within the reach of us mere mortals.
rfbrw,
I choose my words carefully when I stated that 95% of digital filters are designed incorrectly - it has nothing to do silicon area, MIPS, tap length etc.... but a very much more basic parameter.
Only PMD100 came close... but then not close enough... forget the PMD200...
As to price, a FPGA capable of a 1024 Tap digital filter is cheaper then NPC or others charge for there Digital filters.... no need for DSP, these are VERY inefficient for DSD to PCM conversion....
John
I choose my words carefully when I stated that 95% of digital filters are designed incorrectly - it has nothing to do silicon area, MIPS, tap length etc.... but a very much more basic parameter.
Only PMD100 came close... but then not close enough... forget the PMD200...
As to price, a FPGA capable of a 1024 Tap digital filter is cheaper then NPC or others charge for there Digital filters.... no need for DSP, these are VERY inefficient for DSD to PCM conversion....
John
JohnW said:
This is afterall audio and every designer is convinced that they have found the Holy Grail. I am sure Wadia,dCS,Krell,Theta and a number of others with proprietary filters are just as convinced as you are as to the validity of their approach. The proof will be in the eating. Then again it is possible for two to hear the same and disagree absolutely as to its merits.
Now FPGA's are cheap and large. When the SM5842 and the PMD100 where new, the XC4000 and the FLEX10K were the FPGA's of the hour and in quantities the smallest devices were a lot more expensive than an OTS filter. The larger devices even more so , starting in the mid £100's. Specify a higher speed grade and the sky's the limit. Stick a single 16 bit parallel multiplier in a XC4003, the next smallest in the family, and that's around 75% of the device gone. We used to pay £30 in quantity for these things.
rfbrw,
You should take a look at the new Lattice ECP2 range, in Asia pricing on a decent size device is less then a meal for 2 from McDonalds!
John
You should take a look at the new Lattice ECP2 range, in Asia pricing on a decent size device is less then a meal for 2 from McDonalds!
John
JohnW said:
Thanks, I look every now and then but to be honest a kind of comfort zone has set in. I've always used Xilinx so X it is.
that 5% JohnW reserved is meant for various custom filters (not nearly all of them though), the rest is ordinary ASICs on the market where the engineers are really seeking the holy grail, which is the least sillicon area for given filter performance in terms of passband ripple and stopband attenuation.. but these parameters are about as important for the sound as are THD measurements for amplifiers - good for spec sheets, useless as a measure of sound goodness.. that said, our filters are orders of magnitude better than any ASIC on the market even in these specs, but thats a secondary thing..
Well, in the absence of a device calibrated in sound goodness, I'll stick to the traditional parameters used in guaging the efficacy of a digital filter.
understanding the very basic concept of antialiasing filters is plenty enough to judge the current filter offerings.. there's no black magic in it, it's all about meeting the requirements, which you can read in each and every textbook on this topic.. first study the basic concepts, then see what ordinary digital filters do and hopefully you'll understand.. we do..
- Status
- Not open for further replies.