I don't have any high resolution music. All of my music is on regular CD. The music I like is hard to find anyway, and never popular enough for high resolution mastering. I like to keep all my CDs as "master copies", trying to keep them clean and pristine. I ripped them with cdparanoia. The only high resolution copy I found of my CDs is Neil Young's After the Goldrush album in HDCD, but I have my doubts about HDCD.
i'm not talking about just high res, even with 44.1 at 256-320kb mp3 I would fill up the largest sd card several times over, let alone the mix of 320kb, flac/alac and wav + hires I have. so you either dont have much music, dont use uncompressed, or both. high res would fill up an SD in 10-15 albums, so I wasnt even going there.
Configuring TPA Buffalo III input wiring as Buffalo II compatible
SDTrans384 is designed to output I2S and DSD signals on the same pin set based on the source audio file type, PCM WAV or DSD DFF.
In the case of PCM, on output pins; 1 LRCK, 3 SDAT, 5 DCLK of CN8; I2S signals appear. These can be wired to D1, D2, DCK of TPA Buffalo II, respectively.
In the case of DSD, Left-channel DSD Data on 1 LRCK pin, Right-channel DSD Data on 3 SDAT pin and DSD Clock on 5 DCLK pin of CN8. This signal assignment is compatible to the DSD input signal assignment of Buffalo II without any wiring change.
Therefore, we recommend the use of "Buffalo II compatible input wiring configuration" even on your Buffalo III as far as you use not multi-channel but stereo input.
The "Buffalo II compatible input wiring configuration" on Buffalo III means;
1. Shunt both JDSD1 and JDSD2 jumper positions
(DATA1-DATA5 and DATA2-DATA6 interconnections)
2. Wire D4 and D6 throughholes at the card edge
(DATA2-DATA4-DATA6 interconnection)
3. Set SW2-8 ON
(Input remapping of DAC1-DAC3, DAC5-DAC7, DAC2-DAC4, DAC6-DAC8 enabled)
You can listen both PCM WAV and DSD DFF audio files seamlessly on both Buffalo II and Buffalo III.
SDTrans384 is designed to output I2S and DSD signals on the same pin set based on the source audio file type, PCM WAV or DSD DFF.
In the case of PCM, on output pins; 1 LRCK, 3 SDAT, 5 DCLK of CN8; I2S signals appear. These can be wired to D1, D2, DCK of TPA Buffalo II, respectively.
In the case of DSD, Left-channel DSD Data on 1 LRCK pin, Right-channel DSD Data on 3 SDAT pin and DSD Clock on 5 DCLK pin of CN8. This signal assignment is compatible to the DSD input signal assignment of Buffalo II without any wiring change.
Therefore, we recommend the use of "Buffalo II compatible input wiring configuration" even on your Buffalo III as far as you use not multi-channel but stereo input.
The "Buffalo II compatible input wiring configuration" on Buffalo III means;
1. Shunt both JDSD1 and JDSD2 jumper positions
(DATA1-DATA5 and DATA2-DATA6 interconnections)
2. Wire D4 and D6 throughholes at the card edge
(DATA2-DATA4-DATA6 interconnection)
3. Set SW2-8 ON
(Input remapping of DAC1-DAC3, DAC5-DAC7, DAC2-DAC4, DAC6-DAC8 enabled)
You can listen both PCM WAV and DSD DFF audio files seamlessly on both Buffalo II and Buffalo III.
seen the reviews, its really a pretty crappy layout, even mods cannot really save it IMO. so I would find it hard to buy. I also have no need for it when fifo decouples the i2s stream so well its much more convenient to use my mac. the sdtrans is in another league as far as layout goes compared to the QLS, but you do pay for it
For the prototypes using the Cirrus Logic CS8406 for S/PDIF transmission, was the device in hardware or software mode? Reason I ask is that if the player is playing a wav file with a 16 bit sample length and a 44.1kHz sample rate, does the I2S standard allow for this or will the player add 8 zeros to the 16 bit sample to make it a 24 bit sample. I read that I2S uses only 24 bit samples.
Thank you.
Thank you.
Hi, emuman100,
I assume your questions were directed to SDTrans.
On SDTrans384, CS8406 is set in software mode and controlled from MCU via I2C.
The FPGA outputs serial data and clock signals toward both CS8406 and I2S pin header simultaneously. For the I2S lines, the FPGA always outputs 32bit SDAT/channel, padding 0 or 1 trailer in the case of 16/24 bit. For the CS8406 line, the same method is used.
Therefore, the FPGA never outputs 16bit SDAT/channel.
The 0 or 1 padding method is like this. When a negative PCM value is sent, padded with 1 (+3.3V) and when a positive PCM value is sent, padded with 0(0V).
Though I don't think this method is mathematically correct, one early SDTrans user requested this style and we simply adopted it.
Bunpei
I assume your questions were directed to SDTrans.
On SDTrans384, CS8406 is set in software mode and controlled from MCU via I2C.
The FPGA outputs serial data and clock signals toward both CS8406 and I2S pin header simultaneously. For the I2S lines, the FPGA always outputs 32bit SDAT/channel, padding 0 or 1 trailer in the case of 16/24 bit. For the CS8406 line, the same method is used.
Therefore, the FPGA never outputs 16bit SDAT/channel.
The 0 or 1 padding method is like this. When a negative PCM value is sent, padded with 1 (+3.3V) and when a positive PCM value is sent, padded with 0(0V).
Though I don't think this method is mathematically correct, one early SDTrans user requested this style and we simply adopted it.
Bunpei
Connecting I2S/DSD signals from SDTrans to Buffalo III
Once, a Buffalo III is configured in a "Buffalo II compatible style", the necessary wiring are;
SDTrans384(CN8)___Buffalo III/Buffalo II
1 LRCK------------D1
3 SDAT------------D2
5 DCLK------------DCK
8 GND-------------GND(any)
(in the case of synchronous master clocking)
7 MCLK------------Clock (when Crystek Oscillator is removed)
I usually use short wires of 2.5 - 5 cm length.
Once, a Buffalo III is configured in a "Buffalo II compatible style", the necessary wiring are;
SDTrans384(CN8)___Buffalo III/Buffalo II
1 LRCK------------D1
3 SDAT------------D2
5 DCLK------------DCK
8 GND-------------GND(any)
(in the case of synchronous master clocking)
7 MCLK------------Clock (when Crystek Oscillator is removed)
I usually use short wires of 2.5 - 5 cm length.
I can see some people here mention the Aune S1 digital which is basically a Linux based music server based on a Core chip, and an AKM4390 DAC chip that goes up to 192kHz/32bit. This also has a built in AC power supply.
I have not heard the Aune S1, so I cannot comment on its sound. I simply want to state that what the SDTrans384 is, and the way it is used, is very different from the computer based Music Server Aune S1.
The SDTrans384 is not a Music Server, and it does not contain any computer with an operating system. The SDTrans384 is a very "single minded" battery operated device for the purest available way to play a WAV format music file.
As of today it is capable of playing multi-bit files up to 384kHz/32bit and DSD files up to 48kHz x 256 = 12.3MHz. There is (to my knowledge) no other player/transport in the world with this capability.
Of course the SDTrans384 must be connected via I2S over HDMI to a similarly capable DAC, and the only DAC-chip I know of (which is actually shipping) capable of such playback is the ESS ES9018, so the choice is among DACs that contain this chip along with capable design to extract the maximum from this chip. The actual max oversampling capability of the 9018 is actually 48kHz x 512 = 24.576 MHz.
Chiaki and Bunpei has also prototyped a synchronous clock-mode version of their SDTrans384 that has an additional PCB attached on top of the normal PCB, and this is pair matched with a remote PCB that gets mounted within the DAC. The two PCBs are connected via I2S over HDMI which also carries the synchronized clock signal to the SDTrans384 from the two custom made NDK oscillator "clocks" of 90.3168MHz and 98.304MHz mounted on the remote board inside the DAC. The clock signal is actually divided by a actor of 4 when received at the SDTrans384 for its native frequencies of 22.5792MHz and 24.576MHz. However, after testing both the asynchronous (separate clock oscillators for SDTrans and DAC) versus synchronous clock oscillators (a single set that supplies the clock signal simultaneously to both units) there is no dounb that the latter sounds significantly better.
As an owner of this system (installed as a system with a FIDELIX CAPRICE DAC) I am of the opinion that this is capable of the finest digital playback I have heard as of today, cost no object. This is also the pinnacle of Chiaki's & Bunpei's SDTrans DIY project.
It is also worth mentioning that the sound quality achievable is not dependent on being limited to the highest upsampling and bit rates. I do personally not recommend to upsample files from their original sample & bit rates, but rather recommend that they are played as they have been obtained. The best is of course when one is lucky enough to receive an original high bit & high sample studio recording that can be played back as intended. The same is true for original DSDIFF files.
There is no other digital playback system (to my knowledge) that is as capable as the SDTRans384 connected via I2S over HDMI to an ESS ES9018 DAC in synchronized clock mode.
Regarding sound; I willingly admit that I have not heard every digital playback system available in this world. Howsever the SDTrans384 based system is the absolutely best among those that I have heard.
Even though what I have written here is a summary of things already said many times previously in this thread; I though it was worthwhile to clarify what the SDTrans384 is and what it isn't.
I have not heard the Aune S1, so I cannot comment on its sound. I simply want to state that what the SDTrans384 is, and the way it is used, is very different from the computer based Music Server Aune S1.
The SDTrans384 is not a Music Server, and it does not contain any computer with an operating system. The SDTrans384 is a very "single minded" battery operated device for the purest available way to play a WAV format music file.
As of today it is capable of playing multi-bit files up to 384kHz/32bit and DSD files up to 48kHz x 256 = 12.3MHz. There is (to my knowledge) no other player/transport in the world with this capability.
Of course the SDTrans384 must be connected via I2S over HDMI to a similarly capable DAC, and the only DAC-chip I know of (which is actually shipping) capable of such playback is the ESS ES9018, so the choice is among DACs that contain this chip along with capable design to extract the maximum from this chip. The actual max oversampling capability of the 9018 is actually 48kHz x 512 = 24.576 MHz.
Chiaki and Bunpei has also prototyped a synchronous clock-mode version of their SDTrans384 that has an additional PCB attached on top of the normal PCB, and this is pair matched with a remote PCB that gets mounted within the DAC. The two PCBs are connected via I2S over HDMI which also carries the synchronized clock signal to the SDTrans384 from the two custom made NDK oscillator "clocks" of 90.3168MHz and 98.304MHz mounted on the remote board inside the DAC. The clock signal is actually divided by a actor of 4 when received at the SDTrans384 for its native frequencies of 22.5792MHz and 24.576MHz. However, after testing both the asynchronous (separate clock oscillators for SDTrans and DAC) versus synchronous clock oscillators (a single set that supplies the clock signal simultaneously to both units) there is no dounb that the latter sounds significantly better.
As an owner of this system (installed as a system with a FIDELIX CAPRICE DAC) I am of the opinion that this is capable of the finest digital playback I have heard as of today, cost no object. This is also the pinnacle of Chiaki's & Bunpei's SDTrans DIY project.
It is also worth mentioning that the sound quality achievable is not dependent on being limited to the highest upsampling and bit rates. I do personally not recommend to upsample files from their original sample & bit rates, but rather recommend that they are played as they have been obtained. The best is of course when one is lucky enough to receive an original high bit & high sample studio recording that can be played back as intended. The same is true for original DSDIFF files.
There is no other digital playback system (to my knowledge) that is as capable as the SDTRans384 connected via I2S over HDMI to an ESS ES9018 DAC in synchronized clock mode.
Regarding sound; I willingly admit that I have not heard every digital playback system available in this world. Howsever the SDTrans384 based system is the absolutely best among those that I have heard.
Even though what I have written here is a summary of things already said many times previously in this thread; I though it was worthwhile to clarify what the SDTrans384 is and what it isn't.
Last edited:
Hi elecon,
As one of the early SDTrans users I can confirm that this is one of the best sounding player devices there is.
And the only player device with the LSB extension algorithm.
When I tested the early SDTrans384 model and compared to my modified SDTrans192 units there was at that point of time no improvements that could justify to buy the newer version.
As I have heard there have been improvements from the first SDTrans384 release, but I have not been able to have any units for test.
There have been a while since I last played music with SDTrans due to I have several player systems that sounds better than my old modified SDTrans192,
and I can play the SD cards made for SDTrans, from USB disks, SATA disks, CD (ripped first to memory), NAS servers etc.
And I can play WAV, FLAC, AIFF, DXD, DSD and other lossy formats,
and use my iPad or iPhone as remote control with cover pictures and meta tagging also for WAV and AIFF.
The only thing I cannot play yet is the still non-existent DSDXXX formats.
As one of the early SDTrans users I can confirm that this is one of the best sounding player devices there is.
And the only player device with the LSB extension algorithm.
When I tested the early SDTrans384 model and compared to my modified SDTrans192 units there was at that point of time no improvements that could justify to buy the newer version.
As I have heard there have been improvements from the first SDTrans384 release, but I have not been able to have any units for test.
There have been a while since I last played music with SDTrans due to I have several player systems that sounds better than my old modified SDTrans192,
and I can play the SD cards made for SDTrans, from USB disks, SATA disks, CD (ripped first to memory), NAS servers etc.
And I can play WAV, FLAC, AIFF, DXD, DSD and other lossy formats,
and use my iPad or iPhone as remote control with cover pictures and meta tagging also for WAV and AIFF.
The only thing I cannot play yet is the still non-existent DSDXXX formats.
Hi RayCtech
Nice to hear from you
It is true that the SDTrans has come a long way and that there has been many incremental improvements.
To me the biggest breakthrough has been the add-on kit to achieve synchronized clocking of the SDTrans384 and an ESS ES9018 based DAC.
Unfortunately this system has not been made available for purchase partly because it is very time consuming to make, and also partly because the raw cost is very high. (It also requires modification of the ESS ES9018 based DAC which is used together with the SDTrans384.)
I was lucky to have the opportunity to buy a prototype which cost me (on top of the SDTrans) almost twice as much as the SDTrans384 itself.
The question is if there is a high end market even within DIY audio?
******
I am sure that you have alternative digital player systems that also sound very good. Like I stated in my original post, it is not possible for one person to hear everything which is available. All I can say is that what I hear from the system I have is truly remarkable. Living in Japan and also traveling to audio shows etc around the world, I have a chance to listen to a lot of equipment, some of it frighteningly expensive. I still stand by my statement that the digital system I have right now is the best that I personally have experienced. I get a smile on my face every time I turn the system on, and life is very, very sweet
Nice to hear from you
It is true that the SDTrans has come a long way and that there has been many incremental improvements.
To me the biggest breakthrough has been the add-on kit to achieve synchronized clocking of the SDTrans384 and an ESS ES9018 based DAC.
Unfortunately this system has not been made available for purchase partly because it is very time consuming to make, and also partly because the raw cost is very high. (It also requires modification of the ESS ES9018 based DAC which is used together with the SDTrans384.)
I was lucky to have the opportunity to buy a prototype which cost me (on top of the SDTrans) almost twice as much as the SDTrans384 itself.
The question is if there is a high end market even within DIY audio?
******
I am sure that you have alternative digital player systems that also sound very good. Like I stated in my original post, it is not possible for one person to hear everything which is available. All I can say is that what I hear from the system I have is truly remarkable. Living in Japan and also traveling to audio shows etc around the world, I have a chance to listen to a lot of equipment, some of it frighteningly expensive. I still stand by my statement that the digital system I have right now is the best that I personally have experienced. I get a smile on my face every time I turn the system on, and life is very, very sweet
Last edited:
Thank you for your reply Bunpei. If a 44.1kHz 16 bit wave file is played on the player, do you pad the sample with 0's to make it 32 bits in length?
If the MSB of the 16 bit is 0 then 0s are padded else 1s are padded.
This is the way that RayCtech requested.
How to obtain DSD256 sources
I am quite addicted in listening DSD256 sources these days.
DSD256 sources can be prepared in the following procedures.
At least six SDTrans users in Japan are practicing the method.
1. Downloading an original DSD256 recorded audio file of 1-bit Consortium
You can find one sampler music of DSD256 on 1-bit Consortium web page.
Its format is WSD. We can convert WSD to DSDIFF by using Korg AudioGate program.
2. Converting PCM 44.1, 88.2, 176.4, 352.8 KHz/16bit, 24bit PCM sources into DSD256(11.2/12.3 MHz)
A. Upsampling or downsampling original sources to 176.4 or 192 kHz/24 bit PCM WAV files by using Korg AudioGate
B. Rewriting sampling frequency data in WAV headers to those of 44.1 KHz
C. Applying Sony DSD Direct program for converting PCM WAV to DSD DSF
D. Changing the DSD DSF format into DSDIFF DFF format by using Korg AudioGate
E. Rewriting sampling frequency data in DFF headers to those of DSD256(11.2 MHz or 12.3 MHz)
3. Converting DSD64(2.8 MHz) or DSD128(5.6 MHz) sources into DSD256(11.2 MHz or 12.3 MHz)
A. Converting DSD64 or DSD128 DFF sources to 176.4 kHz/24bit PCM WAV by using Korg AudioGate
B. The same procedures described in 2.
DSD256(11.2/12.3 MHz) sources can be also played on the combination of exaU2I+ES9018 DAC other than SDTrans384+ES9018 DAC.
In the case of exaU2I+ES9018, you can't enjoy seamless play of PCM and DSD. You must change wiring when you change sources from PCM to DSD, DSD to PCM. You may suffer a big noise between DSD tracks.
I am quite addicted in listening DSD256 sources these days.
DSD256 sources can be prepared in the following procedures.
At least six SDTrans users in Japan are practicing the method.
1. Downloading an original DSD256 recorded audio file of 1-bit Consortium
You can find one sampler music of DSD256 on 1-bit Consortium web page.
Its format is WSD. We can convert WSD to DSDIFF by using Korg AudioGate program.
2. Converting PCM 44.1, 88.2, 176.4, 352.8 KHz/16bit, 24bit PCM sources into DSD256(11.2/12.3 MHz)
A. Upsampling or downsampling original sources to 176.4 or 192 kHz/24 bit PCM WAV files by using Korg AudioGate
B. Rewriting sampling frequency data in WAV headers to those of 44.1 KHz
C. Applying Sony DSD Direct program for converting PCM WAV to DSD DSF
D. Changing the DSD DSF format into DSDIFF DFF format by using Korg AudioGate
E. Rewriting sampling frequency data in DFF headers to those of DSD256(11.2 MHz or 12.3 MHz)
3. Converting DSD64(2.8 MHz) or DSD128(5.6 MHz) sources into DSD256(11.2 MHz or 12.3 MHz)
A. Converting DSD64 or DSD128 DFF sources to 176.4 kHz/24bit PCM WAV by using Korg AudioGate
B. The same procedures described in 2.
DSD256(11.2/12.3 MHz) sources can be also played on the combination of exaU2I+ES9018 DAC other than SDTrans384+ES9018 DAC.
In the case of exaU2I+ES9018, you can't enjoy seamless play of PCM and DSD. You must change wiring when you change sources from PCM to DSD, DSD to PCM. You may suffer a big noise between DSD tracks.
...and at least one person in Sweden
I have several SACD:s and these I would really like to rip the SACD-music from. Especially my SACDs from the Swedish OPUS3 music company. The only way I have seen is to use a Playstation PS3 - has anyone seen any other method?
/S
I have several SACD:s and these I would really like to rip the SACD-music from. Especially my SACDs from the Swedish OPUS3 music company. The only way I have seen is to use a Playstation PS3 - has anyone seen any other method?
/S
Hi Everyone.
I purchased a new SDTrans3.0 board a month ago from Bunpei.
It now seems that I will not have a use for it. I have too much stuff.
I would like to sell it.
It has not been used. Has only been out of it's packing once.
Please PM me if you wish to buy.
Bunpei. If you know somebody who wished to purchase but missed out. Please contact me, and we can sort something out.
Cheers,
Dave
I purchased a new SDTrans3.0 board a month ago from Bunpei.
It now seems that I will not have a use for it. I have too much stuff.
I would like to sell it.
It has not been used. Has only been out of it's packing once.
Please PM me if you wish to buy.
Bunpei. If you know somebody who wished to purchase but missed out. Please contact me, and we can sort something out.
Cheers,
Dave
Congratulations to Bunpei & Chiaki 50/500
I just realised that this thread has reached 50 pages and 500 posts which is a milestone for the SDTrans project by Bunpei & Chiaki.
I happen to live in Japan and got to know these two very nice gentlemen in person after contacting them via this thread generated by Mr. Bunpei Matoba who takes care of all international communication. Mr Chiaki Nakajima does all the nuts and bolts with the SDTrans, including making personally making each one in his limited spare time.
Despite having full time jobs (which in Japan means long working hours (including overtime) and a long commute), these gentlemen have tirelessly kept on developing and promoting the SDTrans player as a true DIY effort for the benefit of the DIY-community..
As an enthusiastic user of the SDTrans I have been following it through its development with steadily growing capabilities. At times it would feel that the main point was to break records (like in pursuing a Guinness Book award) with ever higher bit and sampling rates (now going all the way up to DSD at 12.3MHz I think.)
However, the point really is that, especially after incorporating synchronised clock mode where the modification of a ESS ES9018 based DAC gets a sub-board with custom made 90/98MHz NDK clocks and the clock signal gets sent vie the HDMI cable over to the SDTrans where the clock signal gets decided to the 22/24MHz values employed by the SDTrans, the sound of any music file starting with "CD level" 44.1kHz/16bit sounds stunning. It is my opinion that the synchronised clock mode SDTrans together with an ESS ES9018 based DAC sounds on par (or better) than any top level digital system regardless of price (I have heard many candidates with daunting price tags). It makes me very happy every time I listen to music via such an SDTrans based system, despite that my ultimate preference for listening to music is still the analog vinyl record on a top notch turntable system.
It is my opinion that music files as much as possible should be played at the native rates they were recorded at: e.g. 48kHz/16bit, 96kHz/24bit, 192kHz/24bit, 352.8kHz/24bit (or 32bit) all depending on what is available.
The SDTrans makes it possible to play just about any file converted to WAV at its native rate. (And sometimes you can have fun using various conversion software to upsample etc.)
Well, I just wanted to congratulate with a truly excellent project worthy of praise on the excellent site DIYaudio.com.
Cheers
I just realised that this thread has reached 50 pages and 500 posts which is a milestone for the SDTrans project by Bunpei & Chiaki.
I happen to live in Japan and got to know these two very nice gentlemen in person after contacting them via this thread generated by Mr. Bunpei Matoba who takes care of all international communication. Mr Chiaki Nakajima does all the nuts and bolts with the SDTrans, including making personally making each one in his limited spare time.
Despite having full time jobs (which in Japan means long working hours (including overtime) and a long commute), these gentlemen have tirelessly kept on developing and promoting the SDTrans player as a true DIY effort for the benefit of the DIY-community..
As an enthusiastic user of the SDTrans I have been following it through its development with steadily growing capabilities. At times it would feel that the main point was to break records (like in pursuing a Guinness Book award) with ever higher bit and sampling rates (now going all the way up to DSD at 12.3MHz I think.)
However, the point really is that, especially after incorporating synchronised clock mode where the modification of a ESS ES9018 based DAC gets a sub-board with custom made 90/98MHz NDK clocks and the clock signal gets sent vie the HDMI cable over to the SDTrans where the clock signal gets decided to the 22/24MHz values employed by the SDTrans, the sound of any music file starting with "CD level" 44.1kHz/16bit sounds stunning. It is my opinion that the synchronised clock mode SDTrans together with an ESS ES9018 based DAC sounds on par (or better) than any top level digital system regardless of price (I have heard many candidates with daunting price tags). It makes me very happy every time I listen to music via such an SDTrans based system, despite that my ultimate preference for listening to music is still the analog vinyl record on a top notch turntable system.
It is my opinion that music files as much as possible should be played at the native rates they were recorded at: e.g. 48kHz/16bit, 96kHz/24bit, 192kHz/24bit, 352.8kHz/24bit (or 32bit) all depending on what is available.
The SDTrans makes it possible to play just about any file converted to WAV at its native rate. (And sometimes you can have fun using various conversion software to upsample etc.)
Well, I just wanted to congratulate with a truly excellent project worthy of praise on the excellent site DIYaudio.com.
Cheers
Last edited: