Here is a question for Frank:
I was checking Dr. Oehrlich's site. He did this input receiver mod but for 192kHz and you decided to keep 96KHz. It seems that main difference is in selection of clock, apparently for 192 cannot be lower than 28MHz. Dr. Oerlich also has great kit for Pioneer universal players, that allows full hi res digital output, SACD, DVDA... Any particular reason that you feel that 192 mod would not work? I am thinking in getting this Pioneer player and kit in order to get digital out for all digital formats. As you know I have exchanged DACs that are capable of working in 192Khz. Do you think that it is possible of passing through that kind of signal if your mod is adjusted? If it is possible what would you suggest to do? I do not know if other DSP chips would be affected with this change, but it seems that Dr. Oehrlich mod is working at that rate. ( I am not really clear on that if he is using it like that or not)
Thank you
AR2
I was checking Dr. Oehrlich's site. He did this input receiver mod but for 192kHz and you decided to keep 96KHz. It seems that main difference is in selection of clock, apparently for 192 cannot be lower than 28MHz. Dr. Oerlich also has great kit for Pioneer universal players, that allows full hi res digital output, SACD, DVDA... Any particular reason that you feel that 192 mod would not work? I am thinking in getting this Pioneer player and kit in order to get digital out for all digital formats. As you know I have exchanged DACs that are capable of working in 192Khz. Do you think that it is possible of passing through that kind of signal if your mod is adjusted? If it is possible what would you suggest to do? I do not know if other DSP chips would be affected with this change, but it seems that Dr. Oehrlich mod is working at that rate. ( I am not really clear on that if he is using it like that or not)
Thank you
AR2
Hello AR2,
The SRC needs a minimum clock input which is sample rate times 138. Using the on-board low-jitter 24.576 MHz clock this results in a maximum sample rate of 178 kHz. For 192 kHz you would need a clock frequency between 26.5 and 30 MHz and therefore a second clock oscillator which isn't synchronous to the other devices (PLL, DSP, DAC).
I optimized the mod for max sonical quality not max input sample rate. Because the DCX operates always with an internal rate of 96 kHz SRC would have to down sample your 192 kHz source. I'm not sure whether it wouldn't be preferable to use a 192 kHz crossover than. I think most of us have data sources with 44.1 kHz?
So it is possible to support a 192 kHz input rate but only with an additional clock oscillator which could be stacked on the existing one as e.g. a DIP08 size and 2 short wires.
Frank
The SRC needs a minimum clock input which is sample rate times 138. Using the on-board low-jitter 24.576 MHz clock this results in a maximum sample rate of 178 kHz. For 192 kHz you would need a clock frequency between 26.5 and 30 MHz and therefore a second clock oscillator which isn't synchronous to the other devices (PLL, DSP, DAC).
I optimized the mod for max sonical quality not max input sample rate. Because the DCX operates always with an internal rate of 96 kHz SRC would have to down sample your 192 kHz source. I'm not sure whether it wouldn't be preferable to use a 192 kHz crossover than. I think most of us have data sources with 44.1 kHz?
So it is possible to support a 192 kHz input rate but only with an additional clock oscillator which could be stacked on the existing one as e.g. a DIP08 size and 2 short wires.
Frank
oettle said:
Sure you did! It does miracles.
I have good DVD/SACD/DVDA player and even with just analog output only, certain SACD discs are sounding really good. Dr. Oerlich's kit for Pioneer players is very interesting to me. I believe these new formats, SACD and DVDA haven't been accepted fully because in the first place there were two of them and in the second place it allows us to use only analog portion. So we are stuck with player's installed DACs. Besides the fact which makes me very angry, that I cannot fully utilize what I paid for, I was using an analog input in DCX in order to play my SACD. Just that eliminate the reason to purchase higher priced discs. That is why - tapping into full potential of these great formats makes me very excited about.
If my understanding of your explanation is correct, even with additional clock upgrade, 192 format would need to be down sampled in order to be used. Than in that case I would agree with you that it is not worthy of the effort. I do agree with you that majority of users are using 44.1 KHz, which include myself 90% of the time, but once you hear "Kind of Blue" on SACD you would want to have more if not all of that.
Thank you
AR2
AR2 said:
in case of abandoning entire idea of using digi out from your universal player , and continuing of using analog outs , you have just one solution ....... bettering analog outs of universal player ;
you can either look at Lukasz Lampizator site for ideas ( even if his writing style is just little less chaotic than mine
) ........ or you can look for few other solutions ........... besides that - depending primary of fact how your main system will look in the future , you can also think about bettering analog inputs in your crummy DCX
edit:
I forgot usual smiley :
Zen Mod said:
in case of abandoning entire idea of using digi out from your universal player , and continuing of using analog outs ,
No, no no nooo.
I know what you mean, but I like capability of outboard DAC. Than, having active crossover... its added bonus. Besides, with all upgrades this "crummy DCX" sounds like real grown up DAC! Don't you hate that - they want you to buy discs and they do not let you use them in full potential. Than you have to monkey around and try to find how to do something what you suppose to do in a first place.
Hi AR2,
I myself are able to hear frequencies up to 17-18 kHz. So I'm pretty happy with max 20 kHz (44.1 kHz sample rate). The DAC AK4396 has a max dynamic range of 120 dB and a typ THD of 96 dB. That's an eqivalent of 20- and 16-bit data resolution and I'm not aware of much better DACs if at all.
So that's the reason why SACD and DVD-Audio are more or less dead which support 24-bit resolution at sample rates of 44.1, 48, 88.2, 96, 176.4 and 192 kHz.
The IIR filters build with the DSP inside the DCX work with 96 kHz only independet of input data rate (That's why you need a SRC.).
You would have to modifiy firmware (filters) and would need more DSP performance to switch to 192 kHz.
As you already mentioned it's most probably not worth the effort.
Frank
I myself are able to hear frequencies up to 17-18 kHz. So I'm pretty happy with max 20 kHz (44.1 kHz sample rate). The DAC AK4396 has a max dynamic range of 120 dB and a typ THD of 96 dB. That's an eqivalent of 20- and 16-bit data resolution and I'm not aware of much better DACs if at all.
So that's the reason why SACD and DVD-Audio are more or less dead which support 24-bit resolution at sample rates of 44.1, 48, 88.2, 96, 176.4 and 192 kHz.
The IIR filters build with the DSP inside the DCX work with 96 kHz only independet of input data rate (That's why you need a SRC.).
You would have to modifiy firmware (filters) and would need more DSP performance to switch to 192 kHz.
As you already mentioned it's most probably not worth the effort.
Frank
Hey guys,
you bought my second small lot (12 pcs) within 2 days. So 'miracle mod' is completely sold out and I myself do not plan to produce further ones because I do not have enough time to make business from it.
But don't worry. If everything works out well Ward Maas from Philgram Audio is able to deliver further 'miracle mods' soon. So please contact him: http://www.pilghamaudio.com/index.asp?pgid=7
Frank
you bought my second small lot (12 pcs) within 2 days. So 'miracle mod' is completely sold out and I myself do not plan to produce further ones because I do not have enough time to make business from it.
But don't worry. If everything works out well Ward Maas from Philgram Audio is able to deliver further 'miracle mods' soon. So please contact him: http://www.pilghamaudio.com/index.asp?pgid=7
Frank
I do not know if anyone is aware, but I came across this:
http://switch-box.com/3x_S_PDIF_Digital_Output.html
Digital out board for DCX! I know that this question came many times in the past. So here it is. I do not know anything more than what is listed on the web,but decided to share it with you, as I am sure many people were interested in this. Cool.
AR2
http://switch-box.com/3x_S_PDIF_Digital_Output.html
Digital out board for DCX! I know that this question came many times in the past. So here it is. I do not know anything more than what is listed on the web,but decided to share it with you, as I am sure many people were interested in this. Cool.
AR2
tms0425 said:
That is my board linked to above.
To be able to feed the I2S lines directly to another piece of equipment you would need to satisfy several conditions.
1) the other device would need to be able to run at the same master clock rate and therefor the same master clock ratio (MC/Fs)
2) You could slave the second device to run off the DCX2496 MC, bit clock and L/R clocks. If any devices are configured as masters in the second device you are either out of luck or need to be able to reconfigure them as slaves to the DCX2496.
3) The device you are connecting to accepts 96/24
4) The data format of the device is expecting data in the same format as the DCX2496. (I2S, Left justified, right justified...etc...etc)
Or you can go with S/PDIF which avoids everything but #3 and an external SRC can be used if needed in that case via S/PDIF.
Shawn
oettle said:
Frank,
There is no SRC on my board, it outputs at the native rate of whatever it is connected to. In the DCX2496 that is 96/24. As far as 'jitter' that is the catch all that everyone says about anything digital. You can have jitter in the DCX2496 feeding the internal DACs too. Jitter is only an issue at the DAC/DSP itself, if the clocks are clean at the DAC/DSP then what happened earlier is irrelevant.
The point of the board is for those that want to stay in the digital domain longer. Either to connect to another digital processor or to feed something like a Panasonic XR amp digitally. In doing that you avoid the additional D/A, analog output of the Behringer and A/Ds of the Panasonic completely.
Shawn
You know Frank, various people - various needs. This is the only solution that I have seen available for someone wanting to try different DACs. There were many people asking this question - how to take digital signal out. As far as I remember there was even one thread that someone was making attempt to take the digital out.
Tom, that is very good idea in taking I2S out. I do not know if anyone have done that, it would be interesting to see. I do not know what is needed in order to send that signal out. Maybe Frank could suggest something since maybe that approach would avoid consequences that he pointed out in S/PDIF mod.
The 3x S/PDIF out board is interesting because it is buffered and allows to still have analog and digital out. In another word the mod would not change or disable analogue out of AKM DACs.
Tom, that is very good idea in taking I2S out. I do not know if anyone have done that, it would be interesting to see. I do not know what is needed in order to send that signal out. Maybe Frank could suggest something since maybe that approach would avoid consequences that he pointed out in S/PDIF mod.
The 3x S/PDIF out board is interesting because it is buffered and allows to still have analog and digital out. In another word the mod would not change or disable analogue out of AKM DACs.
AR2 said:
Running the I2S lines directly does not automatically avoid jitter. If there is jitter inherent in the Behringer it would be on the I2S lines as well and it would be injected directly into the connected device.
The inductance of the cables carrying the signals would also come into play. The signals are all of course square waves and for the MC would be in the mHz range. As you likely know a square wave is a sine wave with many odd order harmonics. If the inductance of the cables rolls off any of those harmonics it alters the shape of the square wave. If that gets altered enough to change the crossing point of the equipment on the receiving end that is jitter... instability between the clocks. Likewise noise picked up (or carried from the source) on the lines could effect clocking and so on.
Since you are directly injecting the I2S into the second device there is nothing you can do to try and reduce the jitter.
Over S/PDIF you can put a good S/PDIF receiver on the other end to try and reduce/eliminate jitter. If it is clean to the DACs what happened over the transport is irrelevant. For example look at the jitter resistance of the Benchmark DAC1. That feeds all data through a SRC and of course has a PLL on its receiver too and it is very resistant to jitter on the transport.
Shawn
As I already mentioned if it's a must having digital outs because there is e.g. another processor the 6-channel digital out would be better than a DAC and an ADC in a row. Question would be why is there an additional processor at all?
But regarding quality it's best to have only volume control and amps directly connected to the internal DACs of the DCX.
You are right there is no PLL and SRC on the 6-channel digital output board but using S/PDIF you would need both at the receiver side. S/PDIF is NO lossless transport because SRC is not lossless and you ADD jitter even if reclocking!!! That’s different e.g. with a non isochrone USB interface.
For testing different DACs you could connect them directly to the existing AK4393/6s without anything between (but with additional power supply for the new DAC chips).
You can't use I2S for transport longer than a few centimetres. That's because there is a 12MHz master clock and a 1ps jitter L/Rclock both with 5V TTL level. 1ps is an equivalent of about 1000 GHz! So regarding jitter each single millimetre you can reduce length of wire between clock oscillator and DACs is an improvement.
Frank
But regarding quality it's best to have only volume control and amps directly connected to the internal DACs of the DCX.
You are right there is no PLL and SRC on the 6-channel digital output board but using S/PDIF you would need both at the receiver side. S/PDIF is NO lossless transport because SRC is not lossless and you ADD jitter even if reclocking!!! That’s different e.g. with a non isochrone USB interface.
For testing different DACs you could connect them directly to the existing AK4393/6s without anything between (but with additional power supply for the new DAC chips).
You can't use I2S for transport longer than a few centimetres. That's because there is a 12MHz master clock and a 1ps jitter L/Rclock both with 5V TTL level. 1ps is an equivalent of about 1000 GHz! So regarding jitter each single millimetre you can reduce length of wire between clock oscillator and DACs is an improvement.
Frank