It should be noted that USB 2.0 doesn't have the bandwidth to support 384/32/8-ch, so already you'd be looking at a USB 3.0 solution.
Interesting... I haven't run the number yet myself, so I am not sure, but just curious to see your figures. Can you explain?
I likely wouldn't support sample rates that high for multi channel anyway as I don't know of any good reason to do so. There is nothing (that I know of) encoded for that.
Who said their wouldn't ever be 8 channel USB? I was just saying the first USB module will not be. 🙂
If by DSD128fs you mean bi-phase encoded DSD, then yes. The best thing for you to do here is read the datasheet. 🙂 It will support exactly what the ES9018 supports no more no less. 🙂
I meant a double resolution DSD format where audio is sampled at 128 times the sample frequency giving a data rate of 5.6448Mbit/s.
Here are what I found:
http://www.merging.com/uploads/assets//Merging_pdfs/dxd_Resolution_v3.5.pdf
http://www.sonicstudio.com/pdf/dsd/DSDIFF_1.5_Spec.pdf
I likely wouldn't support sample rates that high for multi channel anyway as I don't know of any good reason to do so. There is nothing (that I know of) encoded for that.
The only reason is to make hardware capable for the best audio 😀 It will appear after definitely. There will be a coming of BD-Audio which specs shouldn't be inferior than 8ch 32/384.
This is like one engineer from Dolby Laboratories explained a delay in developing the BD-Audio format - there is no equipment yet capable of playing such kind of audio.
Last edited:
My statement is the same. That kind of question you may need to be answered by ESS. I don't have such a source to test at the moment. So the best I could do is refer you to the datasheet.
The only reason is to make hardware capable for the best audio
In that case (absolute performance), you would likely want to use the Buffalo in Stereo or even Mono mode (four or eight modules), rather than 8-channel mode. It will, of course, work in 8-channel mode, but you get better performance with summed DAC channels.
In that case (absolute performance), you would likely want to use the Buffalo in Stereo or even Mono mode (four or eight modules), rather than 8-channel mode. It will, of course, work in 8-channel mode, but you get better performance with summed DAC channels.
With regards to DNR and THD+N values I agree, but those are insignificant compared to such 8ch solution.
(four or eight modules regretfully are not for my budget..)
Now with USB 3.0 implementation to consumer market and BIII release 🙂 we may have an ability to enjoy pure 32/384 8ch multidimensional music.
Last edited:
In that case (absolute performance), you would likely want to use the Buffalo in Stereo or even Mono mode (four or eight modules), rather than 8-channel mode. It will, of course, work in 8-channel mode, but you get better performance with summed DAC channels.
The high quality of power supply and general layout of the buffalo III may Reduce the differences between summed and lonely dac channel I think. Does anyone made tthe test between the two?
I likely wouldn't support sample rates that high for multi channel anyway as I don't know of any good reason to do so.
Here is another reason to make it 🙂
http://www.via.com.tw/en/resources/pressroom/pressrelease.jsp?press_release_no=5527
Regards.
Last edited by a moderator:
1. I'm curious what about building a single multichannel I/V stage if this is in future plans?
You use 4 stereo I/V stages for 8 channel output.
2. How to configure BIII to voltage-output mode with required rms level? What exactly depends on output stage used, as it's stated in BIII descriptoon at TPA website? In voltage-output mode can I use BIII outputs directly as SE outputs?
The DAC performs best as a current source. Use low input impedance I/V stages whenever possible. You can use BIII outputs directly, but it will hurt performance: . THD+N -108dB vs. -120dB.
4. Can I feed a digital out from my sound card directly into BIII inputs assuming that sound card digital out is TTL SPDIF?
In BIII SPDIF inputs is not like described in datasheet of ES9018? Does it means SPDIF3? BIII in this way will always have only 2ch input and output (not full 8ch) at a time due to a lack of decoder chip?
You can feed the DAC PCM or DSD for multi-channel output. S/PDIF is stereo.
6. Are there pros and cons of using LCDPS to power BIII than Placid?
Yes: the Placid obviously only has a single output, while the LCDPS offers 2. When one wants to separate the DAC board from other items like TOSLINK modules, it's easier to use the LCDPS. On the other hand, the Placid offers a more stable output. What's best? The user decides.
If you need to power 2 DAC boards the most cost effective way is to use a single LCDPS. Like Russ stated: the performance of the local power regulators will determine the noise levels.
You use 4 stereo I/V stages for 8 channel output.
2. How to configure BIII to voltage-output mode with required rms level? What exactly depends on output stage used, as it's stated in BIII descriptoon at TPA website? In voltage-output mode can I use BIII outputs directly as SE outputs?
The DAC performs best as a current source. Use low input impedance I/V stages whenever possible. You can use BIII outputs directly, but it will hurt performance: . THD+N -108dB vs. -120dB.
4. Can I feed a digital out from my sound card directly into BIII inputs assuming that sound card digital out is TTL SPDIF?
In BIII SPDIF inputs is not like described in datasheet of ES9018? Does it means SPDIF3? BIII in this way will always have only 2ch input and output (not full 8ch) at a time due to a lack of decoder chip?
You can feed the DAC PCM or DSD for multi-channel output. S/PDIF is stereo.
6. Are there pros and cons of using LCDPS to power BIII than Placid?
Yes: the Placid obviously only has a single output, while the LCDPS offers 2. When one wants to separate the DAC board from other items like TOSLINK modules, it's easier to use the LCDPS. On the other hand, the Placid offers a more stable output. What's best? The user decides.
If you need to power 2 DAC boards the most cost effective way is to use a single LCDPS. Like Russ stated: the performance of the local power regulators will determine the noise levels.
That looks pretty cool. I would love to work something around that chip.
Thanks for your replies guys. The only thing I need is to wait for your 8ch 32/384 usb transport 😀
Just a friendly note to posters. We don't allow promotion of third part products on this forum. So if you mention one then please don't be surprised if your post is deleted or edited.
The reason is that it can seem to people that we endorse those products when we may or may not,
The reason is that it can seem to people that we endorse those products when we may or may not,
The device I told is a perfect partner for the Buffalo III for multichannel as you don't have such a device to sell. This product will push many people who want to do multichannel to buy Buffalo III so I don't understant why there is a total blackout concerning this i²s product.
The device I told is a perfect partner for the Buffalo III for multichannel as you don't have such a device to sell. This product will push many people who want to do multichannel to buy Buffalo III so I don't understant why there is a total blackout concerning this i²s product.
In fact BIII supports 32/1.536MHz 8ch 😛 Would be nice to have such transport 😀
Last edited:
Haha. 🙂 I am not sure I would go that far. 😀
But in some time you will have to.. 🙂
Interesting... I haven't run the number yet myself, so I am not sure, but just curious to see your figures. Can you explain?
I likely wouldn't support sample rates that high for multi channel anyway as I don't know of any good reason to do so. There is nothing (that I know of) encoded for that.
My source actually is George from exaDevices. I asked him why the exaU2I couldn't support 8-ch/384 kHz/32 bit and he told me that USB 2.0 doesn't have support for that kind of bandwidth. I have done no measurements on my own; I'm taking his advice at face value and assuming that he has looked into this.
Also true, I can't think of a good reason for supporting something that high! Except maybe BD-audio, when it's developed.
I suspect George from axa was wrong about bandwidth limitation. 480Mb/s is more than enough for 32/384 8ch. Perhaps for 32/1536 8ch usb 3.0 data bandwidth would be required.
a reason for 8ch high res transport
crossovers done in software. if you do your equalisation in software it is best to add a few bits headroom so no information is lost, so 24bit music quickly becomes 32bit. if you then group and sum the output from your EQ filters in bands you have your crossover. each band would want to have the sample rate capability of the original file. so in my ideal case of feeding 4-way stereo speakers from 192kHz/24bit 2ch files the transport would have to be 8ch 192kHz/32bit.
what I can't quite get my head around is how to sync the clocks to make this work with 4 stereo or 8 mono Buf's. Or do you trust the I2S and keep the local clocks?
I likely wouldn't support sample rates that high for multi channel anyway as I don't know of any good reason to do so. There is nothing (that I know of) encoded for that.
crossovers done in software. if you do your equalisation in software it is best to add a few bits headroom so no information is lost, so 24bit music quickly becomes 32bit. if you then group and sum the output from your EQ filters in bands you have your crossover. each band would want to have the sample rate capability of the original file. so in my ideal case of feeding 4-way stereo speakers from 192kHz/24bit 2ch files the transport would have to be 8ch 192kHz/32bit.
what I can't quite get my head around is how to sync the clocks to make this work with 4 stereo or 8 mono Buf's. Or do you trust the I2S and keep the local clocks?
That looks pretty cool. I would love to work something around that chip.
please, please, please, release the 2ch XMOS board before you start playing with a new toy....
- Status
- Not open for further replies.
- Home
- More Vendors...
- Twisted Pear
- Buffalo III - flexibility without compromise.