If you can make it to the next PNWAS meeting we can play them through my DAC, and you can tell me if I'm delusional or not.
Will do!
DSD Battle Royale!
- Thread starter Bruce B
- Start date
Will do!
So Bruce,
when you going to provide the quad rate DSD
Orb
It's there Orb... did you listen to it? Unfortunately unless folks have an exaSound or Horus DAC, you won't be able to hear it.
I'm sure the manufacturers are now working on a DSD512 DAC as we speak!! More is better, right??
I'm sure the manufacturers are now working on a DSD512 DAC as we speak!! More is better, right??
Yep, up to the point that the logic elements can handle it. It's a settling time issue. Andreas, even using discrete ECL logic circuit elements, found that 128fs was about the upper limit for data reliability. That's of course implementing with discrete logic. Integrated circuit logic, with its much lower inter element inductance and stray capacitance of course can operate faster than discrete logic. Witness the ARDA AT1201 and ESS9016/18 chips. But now it's no longer a garage shop venture. Gotta be able to sell a million of em before a foundry like TI or AD gets interested.
Hi Bruce!
Last edited:
I'm sure the manufacturers are now working on a DSD512 DAC as we speak!! More is better, right??
Only until the FCC starts regulating them based on radio transmission broadcast regulation...
One reason I think Merging with the Horus will corner this segment.
And then to really play it back ideally you also need a similar architecture; for now the Horus converters.
Who are the ones that are truly 1-bit DSD designed DAC architectures; Aside from the primary core studio focused companies with ADs/DAWs/DACs, Playback and possibly only one or two others (such as Meitner/EMM Labs using FPGA) that do not utilise OEM DAC chips.
Why I think there will be technical arguments about the need for quad DSD with pros/cons.
Still, interesting some of the big producers rave about quad DSD (using Merging products).
Cheers
Orb
I agree, for the professional market certainly. But while Merging has a complete DAW solution, other ADC's and DAC's however contain the same chips. Ayer for one uses the ARDA AT1201 A/D Converter chip, and eXa E20 and E28 use the ESS 9018 configured to run at up to 12.288MHz sampling rate. And Merging itself is not far away from both the hardware and software support for the Emotion server/player.
As I mentioned above, the discrete component circuit DAC's have a difficult time running much faster than 5.6MHz due to parasitic inductance and capacitance of the layout.
I think Charles did post 6 files, 3 in wav and 3 in dsf. At CompAudiophile, but said he would not leave thenm on for long.
I did manage to snag 'em.
http://www.computeraudiophile.com/f8-general-forum/needle-drop-comparisons-18486/
I did manage to snag 'em.
http://www.computeraudiophile.com/f8-general-forum/needle-drop-comparisons-18486/
I listened last night to the three sample rates of the Bruce's Horus samples. I could not, with this source content, hear a difference. However, I did prefer the 64fs on an emotional level. It just felt more comfortable.
I listened using two SoundLab M-1PX full range electrostatic speakers in near field, Parasound JC-1 amplifiers, my design 1db step passive attenuator, Merging Horus, and Pyramix Mastering pack software.
The first consideration was the three clips are at different levels, with the 128fs being loudest. The 64fs clip is -1.2dB lower, and the 256fs clip -3.2dB lower. I corrected the differences with the analog attenuator.
A great feature of the Horus and Pyramix is the ability to have several projects mounted simultaneously, and even though they're at different sampling rates, and switch between them with only about a one second switching time.
I mounted the three samples, and repeatedly listened to the lead-in voice and first guitar stroke(s). For me they were identical. Repeated longer listening to whole clips (the 256fs is twice as long as the 64fs and 28fs) is what gave me the "more comfortable" feeling of the 64fs clip.
I don't mean to imply in the above that there is no difference within the three DSD sampling rates. Just not for me with this material.
Thanks Bruce for so generously sharing these clips!
I listened using two SoundLab M-1PX full range electrostatic speakers in near field, Parasound JC-1 amplifiers, my design 1db step passive attenuator, Merging Horus, and Pyramix Mastering pack software.
The first consideration was the three clips are at different levels, with the 128fs being loudest. The 64fs clip is -1.2dB lower, and the 256fs clip -3.2dB lower. I corrected the differences with the analog attenuator.
A great feature of the Horus and Pyramix is the ability to have several projects mounted simultaneously, and even though they're at different sampling rates, and switch between them with only about a one second switching time.
I mounted the three samples, and repeatedly listened to the lead-in voice and first guitar stroke(s). For me they were identical. Repeated longer listening to whole clips (the 256fs is twice as long as the 64fs and 28fs) is what gave me the "more comfortable" feeling of the 64fs clip.
I don't mean to imply in the above that there is no difference within the three DSD sampling rates. Just not for me with this material.
Thanks Bruce for so generously sharing these clips!
Last edited:
The DSD samples are mono. Guess he's having problems with the code or something.
I couldn't get it to work before I had to send them back. Something just wasn't right.
My preference in descending order: Horus, Grimm, EMM.
Listened to the files through the MR-2000s using Beyerdynamic DT770 phones.
Thank you, Bruce
Looking for information on Playback Designs handling of signal I found a recent interview with Andreas Koch at the-ear site. http://www.the-ear.net/features/andreas-koch-playback-designs He summarizes the handling of PCM in PD units as (quoting from the interview)
" In our devices we use the 44.1 kHz sample frequency as a basis, convert the digital signal into 16FS PCM, make a DSD stream of it in 5.6 MHz, multiply the sample rate by another factor of 8 and finally, at a sample frequency of around 45 MHz, we decode the signal into analogue sound. "
Can some one elucidate me why this intermediate 5.6MHz DSD is needed? Why not direct decoding to what looks like DSD 43 MHz?
BTW, if you transcode your PCM to 5.6 MHz DSD and play it using the PB you get similar sonic results as playing the PCM file?
" In our devices we use the 44.1 kHz sample frequency as a basis, convert the digital signal into 16FS PCM, make a DSD stream of it in 5.6 MHz, multiply the sample rate by another factor of 8 and finally, at a sample frequency of around 45 MHz, we decode the signal into analogue sound. "
Can some one elucidate me why this intermediate 5.6MHz DSD is needed? Why not direct decoding to what looks like DSD 43 MHz?
BTW, if you transcode your PCM to 5.6 MHz DSD and play it using the PB you get similar sonic results as playing the PCM file?
Similar threads
Staff online
-
treitz3Super Moderator
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Co-Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |