barrows,
Do you know of anyone who as tried the B-II with ExaDevices exa U2I interface? It is a direct to I2S interface and seems to have the right types of things going for it in terms of asynchronous USB interface ... ASIO drivers etc...
Too bad the driver is windows only as I'd like something that worked for Linux...but will run windows if I must.
I've heard a bunch of different stuff about USB driver modes and playback software. So far the PC playback world doesn't seem settled when it comes to highend audio playback.
Anyone on the thread have other suggestions for USB interface into the B-II...hopefully direct to I2S?
Best Regards,
Theo
Do you know of anyone who as tried the B-II with ExaDevices exa U2I interface? It is a direct to I2S interface and seems to have the right types of things going for it in terms of asynchronous USB interface ... ASIO drivers etc...
Too bad the driver is windows only as I'd like something that worked for Linux...but will run windows if I must.
I've heard a bunch of different stuff about USB driver modes and playback software. So far the PC playback world doesn't seem settled when it comes to highend audio playback.
Anyone on the thread have other suggestions for USB interface into the B-II...hopefully direct to I2S?
Best Regards,
Theo
I...
do not know anyone using the exa. Check the exu thread for more info.
Personally, I prefer an interface which can use the native drivers with mac. In the commercial audio world this means Gordon Rankin's Streamlength code running on the XMOS chip, dCS, etc. Like I said, I am getting superb performance with the Wavelink, and I expect TPA's XMOS solution will be excellent as well (and go USB-I2S).
"I've heard a bunch of different stuff about USB driver modes and playback software. So far the PC playback world doesn't seem settled when it comes to highend audio playback."
I would not expect this to settle down anytime soon! I use the Pure Music playback software, and it continues to get updates, which keep increasing the sonic performance. Computer tweaks, and playback software optomization all seem to make differences in sound. As does the hardware implementation of the USB interface, and its associated operating code.
Note that commercial DACs are also featuring interesting USB features: DSD over USB for Playback Designs, and 24/384 data rates for Antelope Audio, both with asynchronous USB data transfer.
I do not think we are going to see much in the way of settling down in computer audio, as the pace of computer development is in command here-on the horizon is Thunderbolt... In any case, well implemented asynchronous USB audio has proven to me to outperform traditional transports in level comparisons, when all the details are attended to.
do not know anyone using the exa. Check the exu thread for more info.
Personally, I prefer an interface which can use the native drivers with mac. In the commercial audio world this means Gordon Rankin's Streamlength code running on the XMOS chip, dCS, etc. Like I said, I am getting superb performance with the Wavelink, and I expect TPA's XMOS solution will be excellent as well (and go USB-I2S).
"I've heard a bunch of different stuff about USB driver modes and playback software. So far the PC playback world doesn't seem settled when it comes to highend audio playback."
I would not expect this to settle down anytime soon! I use the Pure Music playback software, and it continues to get updates, which keep increasing the sonic performance. Computer tweaks, and playback software optomization all seem to make differences in sound. As does the hardware implementation of the USB interface, and its associated operating code.
Note that commercial DACs are also featuring interesting USB features: DSD over USB for Playback Designs, and 24/384 data rates for Antelope Audio, both with asynchronous USB data transfer.
I do not think we are going to see much in the way of settling down in computer audio, as the pace of computer development is in command here-on the horizon is Thunderbolt... In any case, well implemented asynchronous USB audio has proven to me to outperform traditional transports in level comparisons, when all the details are attended to.
Thanks for your comments Barrows and what you have said makes good sense. I'm trying to debate if I should wait for the TWP solution...which may be a long wait or if I should jump for the Exa solution.
Using default drivers is a nice feature ... if Wavelength made a USB -> I2S I'd go that route. However the solution you use I'm sure is nice but $900 bucks for a USB -> SPDIF converter is just a tad hard for me to swallow though I do understand all the development time that was put into the USB implementation.
Anyway choices choices...
Some of what is said about PC audio just doesn't seem to make good technical sense to me ... for example the updates and tweaks to the audio playback software you use I would not think would make any real difference since at the end of the day the Data must be handed to the driver anyway...does it make a huge difference as to what manner it gets passed off to the driver ?
I'm sure there is a lot at play here which I have not considered.
Using default drivers is a nice feature ... if Wavelength made a USB -> I2S I'd go that route. However the solution you use I'm sure is nice but $900 bucks for a USB -> SPDIF converter is just a tad hard for me to swallow though I do understand all the development time that was put into the USB implementation.
Anyway choices choices...
Some of what is said about PC audio just doesn't seem to make good technical sense to me ... for example the updates and tweaks to the audio playback software you use I would not think would make any real difference since at the end of the day the Data must be handed to the driver anyway...does it make a huge difference as to what manner it gets passed off to the driver ?
I'm sure there is a lot at play here which I have not considered.
With the Buffalo II I have tested extensively both a jkeny modded HiFace and the also very nice KingRex UC192 (32bit). Both features I2S out.
The exa is nice but the lack of drivers for the Mac OSX is a no-go for me.
Looking forward to see what TP comes up with!
Cheers,
Nic
TJ...
I tried a bunch of other USB-SPDIF interfaces before the Wavelink, as I really did not want to spend the $. The bottom line was that none of the (more affordable solutions) other units could beat out my CD transport, and the Wavelink handily did. I new that ultimately I was going to put a USB-I2S DIY solution inside my DAC, but wanted to have excellent computer audio (with 24/192) now! I figured I could sell the Wavelink when I get a USB-I2S interface that does the job.
BTW, for something made in small qualtities, in the US, by a manufacturer who really cares, the pricing of the Wavelink is fair. Gordon tests each unit for jitter levels, to be absolutely sure of quality. That aside, I understand that many people do not want to spend that kind of money (I didn't).
I understand your reluctance to believe how computer tweaks and software changes could make a difference to sound-especially considering that all approaches when done right remain bit perfect. The fact is, that when you get a really good DAC going (B-II/Legato) into a good amp and speakers, you will easily hear the differences. On Mac, the difference between using Pure Music and iTunes is not subtle at all. There are a lot of possible explanations right now, but no one seems to be really sure of what is going on. The software developers are just trying to develop the simplest playback software, which results in the least activity in CPU and drive access-my experience is that every positive step they make in this direction improves performance. This may be RF generation related. The latest version of Pure Music (and Audirvana) includes "integer playback" and this reduces processing by another step-avoiding floating point conversions entirely. The same applies to computer tweaks to reduce background processes-this is why many folks prefer mac, as OSX runs many less processes in the background compared to stock windows OS. I suspect that the best computer audio approach could be a purpose built computer, running a custom developed linux OS, made for only audio playback. Once audio companies get this figured out, I suspect that high end audio servers will go this route-but to get it right, a company will need both computer software geniuses, hardware geniuses, and audio geniuses.
I am guessing, but I do not think the TPA Async USB board is too far off now, also see Lorien's thread over at "digital source" here. He has an XMOS based USB-I2S/SPDIF board which looks pretty good, and I think he is ready to sell them now.
I tried a bunch of other USB-SPDIF interfaces before the Wavelink, as I really did not want to spend the $. The bottom line was that none of the (more affordable solutions) other units could beat out my CD transport, and the Wavelink handily did. I new that ultimately I was going to put a USB-I2S DIY solution inside my DAC, but wanted to have excellent computer audio (with 24/192) now! I figured I could sell the Wavelink when I get a USB-I2S interface that does the job.
BTW, for something made in small qualtities, in the US, by a manufacturer who really cares, the pricing of the Wavelink is fair. Gordon tests each unit for jitter levels, to be absolutely sure of quality. That aside, I understand that many people do not want to spend that kind of money (I didn't).
I understand your reluctance to believe how computer tweaks and software changes could make a difference to sound-especially considering that all approaches when done right remain bit perfect. The fact is, that when you get a really good DAC going (B-II/Legato) into a good amp and speakers, you will easily hear the differences. On Mac, the difference between using Pure Music and iTunes is not subtle at all. There are a lot of possible explanations right now, but no one seems to be really sure of what is going on. The software developers are just trying to develop the simplest playback software, which results in the least activity in CPU and drive access-my experience is that every positive step they make in this direction improves performance. This may be RF generation related. The latest version of Pure Music (and Audirvana) includes "integer playback" and this reduces processing by another step-avoiding floating point conversions entirely. The same applies to computer tweaks to reduce background processes-this is why many folks prefer mac, as OSX runs many less processes in the background compared to stock windows OS. I suspect that the best computer audio approach could be a purpose built computer, running a custom developed linux OS, made for only audio playback. Once audio companies get this figured out, I suspect that high end audio servers will go this route-but to get it right, a company will need both computer software geniuses, hardware geniuses, and audio geniuses.
I am guessing, but I do not think the TPA Async USB board is too far off now, also see Lorien's thread over at "digital source" here. He has an XMOS based USB-I2S/SPDIF board which looks pretty good, and I think he is ready to sell them now.
Just because someone uses an XMOS chipset doesn't mean the final result will be as good as the Wavelink. It's the small details in design and layout, clock quality etc which probably make the difference between a great solution and a mediocre one.
That said, it's probably a matter of time before there are some real good budget products.
Barrows,
Wow...even the number of running processes affect sound. That is truly hard to believe..I guess once I've got my whole system up and running with the amps I'm building I will play around with home theater PCs.
Sorry to say I'm not having a lot of luck repairing my Trident for one of my B-II channels but Russ is helping me. So hope to have it fixed up soon. I've got the current TPA USB receiver right now.
Other than that the system will consist of an Integra DHC-80.2 PrePro....balanced in from the B-II. A pair of Tannoy TD8's and a TDC....Tannoy Rev sig 4's for surround. 2 x Exodus Maelstrom-X 18" Subs power by HPA-NXv1200 amps...then 5 x Randy Sloan's Totem Pole 200watt solid state mono-blocs.
So hope it all sounds pretty sweet...I'm hoping and praying TPA makes some more Tridents soon.
-TJ
Wow...even the number of running processes affect sound. That is truly hard to believe..I guess once I've got my whole system up and running with the amps I'm building I will play around with home theater PCs.
Sorry to say I'm not having a lot of luck repairing my Trident for one of my B-II channels but Russ is helping me. So hope to have it fixed up soon. I've got the current TPA USB receiver right now.
Other than that the system will consist of an Integra DHC-80.2 PrePro....balanced in from the B-II. A pair of Tannoy TD8's and a TDC....Tannoy Rev sig 4's for surround. 2 x Exodus Maelstrom-X 18" Subs power by HPA-NXv1200 amps...then 5 x Randy Sloan's Totem Pole 200watt solid state mono-blocs.
So hope it all sounds pretty sweet...I'm hoping and praying TPA makes some more Tridents soon.
-TJ
I have been a hardware skeptic by nature, but many of my biases have been dashed. If you go over to computer audio asylum and read some of 'the art of building computer transports' pdf and thread, your head might start to swim. Explaining some of the described phenomena was over my head, but as this article has now been around for a while and I have much more experience, many of my biases about what can and cannot matter are out the window. For example, how is it that the SQ of one of my computers - a PC - is profoundly influenced merely by the brand of motherboard? Why does ASIO latency significantly alter coherence? I could go on but suffice it to say that it seems that there are no inconsequential variables - either in hardware or software.
Good luck with the tridents. Russ helped me a lot with mine (thread on TPA support forum).
Frank
labjr:
"Just because someone uses an XMOS chipset doesn't mean the final result will be as good as the Wavelink. It's the small details in design and layout, clock quality etc which probably make the difference between a great solution and a mediocre one."
Of course! Maybe you should read Lorien's thread, as he is definitely paying close attention to the details-this is why I recommended his project as having good potential. High quality clocks, careful power supply design with many separtely regulated supplies, very high quality de-coupling caps, option for USB power or external power, etc...
I do not think we will really see "budget" solutions with the XMOS processor, (on the price order of the Hiface for example), because getting the best out of it requires multiple clean power supplies, additional ICs for USB, and SPDIF, and highly accurate clock circuits, as well as a well designed board (many iterations to get it right) and fine tuning (or re-writing) of the code as well. On the other hand, I expect very good DIY boards (like Lorien's and TPA's) to probably come in close to $250 USD or so. Of course this does not include power supply, case, wiring, etc.
"Just because someone uses an XMOS chipset doesn't mean the final result will be as good as the Wavelink. It's the small details in design and layout, clock quality etc which probably make the difference between a great solution and a mediocre one."
Of course! Maybe you should read Lorien's thread, as he is definitely paying close attention to the details-this is why I recommended his project as having good potential. High quality clocks, careful power supply design with many separtely regulated supplies, very high quality de-coupling caps, option for USB power or external power, etc...
I do not think we will really see "budget" solutions with the XMOS processor, (on the price order of the Hiface for example), because getting the best out of it requires multiple clean power supplies, additional ICs for USB, and SPDIF, and highly accurate clock circuits, as well as a well designed board (many iterations to get it right) and fine tuning (or re-writing) of the code as well. On the other hand, I expect very good DIY boards (like Lorien's and TPA's) to probably come in close to $250 USD or so. Of course this does not include power supply, case, wiring, etc.
Buffalo 2 : why the shortage?
I don't want to be critical or rude in this post, I just would like to understand something.
The question is "why the shortage of Buffalo 2 DACs?"
There seems to be a lot of potential buyers, myself included, that would happily part with money to buy the B2, but it's always "sold out"
Why?
Is there a shortage of a critical part?
Is it just that you can't make them fast enough?
If the capacity to manufacture the boards is more than Twisted Pear can handle, can't you outsource? It doesn't look like a very complex board by today's standards. There should be plenty of manufacturers happy to produce it.
I could understand it if it were newly released, but now it's about to be replaced.
I don't really want a more flexible Buffalo 3. I would just like a Buffalo 2. It's perfect for my needs.
And even when the B3 is released, I have no idea whether there will be the same kind of shortage. Will I still be waiting for a B3 when you are releasing the B4?
I love the design, want to buy, but like everyone, I have a limited lifespan.
In today's climate, many businesses would be very grateful to have more customers than they can handle, and they would find a way to deliver the goods.
Can you comment please on this, and tell us why there is this shortage, and whether the same sort of thing is likely to happen with B3 (and if not, why not..what have you fixed?)
Also, why not keep the B2 alongside the B3? I would be surprised if the B3 was the same price as the B2. So it seems there would be a place for both.
Please don't make me go and buy an inferior DAC from someone else...
I don't want to be critical or rude in this post, I just would like to understand something.
The question is "why the shortage of Buffalo 2 DACs?"
There seems to be a lot of potential buyers, myself included, that would happily part with money to buy the B2, but it's always "sold out"
Why?
Is there a shortage of a critical part?
Is it just that you can't make them fast enough?
If the capacity to manufacture the boards is more than Twisted Pear can handle, can't you outsource? It doesn't look like a very complex board by today's standards. There should be plenty of manufacturers happy to produce it.
I could understand it if it were newly released, but now it's about to be replaced.
I don't really want a more flexible Buffalo 3. I would just like a Buffalo 2. It's perfect for my needs.
And even when the B3 is released, I have no idea whether there will be the same kind of shortage. Will I still be waiting for a B3 when you are releasing the B4?
I love the design, want to buy, but like everyone, I have a limited lifespan.
In today's climate, many businesses would be very grateful to have more customers than they can handle, and they would find a way to deliver the goods.
Can you comment please on this, and tell us why there is this shortage, and whether the same sort of thing is likely to happen with B3 (and if not, why not..what have you fixed?)
Also, why not keep the B2 alongside the B3? I would be surprised if the B3 was the same price as the B2. So it seems there would be a place for both.
Please don't make me go and buy an inferior DAC from someone else...
I'm a customer like you and it took me a while before I got my BII but I'd like to comment anyway.
Afaik TP-audio is a hobby-business i.e. not the primary source of income for the owners.
As such everything is done in their spare time.
To me it's smart not to bite off a bigger bite than you can chew.
Even though waiting can be a hassel I support Russ an Brian in how they run things.
It's better to wait a little and get the personal touch.
Keep up the good work.
Afaik TP-audio is a hobby-business i.e. not the primary source of income for the owners.
As such everything is done in their spare time.
To me it's smart not to bite off a bigger bite than you can chew.
Even though waiting can be a hassel I support Russ an Brian in how they run things.
It's better to wait a little and get the personal touch.
Keep up the good work.
I guess that makes sense. But their products look too professional for just a hobby business. But as for the supply of B2 - I would have thought that they could get someone to make the boards, etc.
I've seen hobby businesses have shortages due to success, but once they realize their potential sales, they soon buy in more, so the shortages are temporary.
So I guess the hobby business could just be half of the problem. I don't know. That's why I asked
I've seen hobby businesses have shortages due to success, but once they realize their potential sales, they soon buy in more, so the shortages are temporary.
So I guess the hobby business could just be half of the problem. I don't know. That's why I asked
There are a few factors at work...
1) We are a hobby business. I build, pack and ship kits/modules in the morning before work, in the evening after work (mostly after the kids go to bed), and also on the weekends (basically all of my spare time). I know Russ spends a similar amount of time doing R&D.
2) We do have the Buffalo modules built for us due to the volume. We are somewhat limited in how many we can build at a time, mostly due to the enormous up-front cost of parts, and lead times for getting XOs built. This is only a part of the delay. Also, every Buffalo module needs to hand tested and packed up, along with a hand-tested AVCC module. B-III will add Tridents to that.
3) The typical Buffalo order includes an output stage kit and a couple power supply kits. This means that for (on average) every Buffalo module, I need to built three kits, which is very time consuming. This is the primary reason we do batches of pre-orders.
4) Even after all of that, simply packing up and shipping the orders takes a lot more time than you would think. Something like 80% of orders are international, so that's a lot of customs forms for be filled out, etc.
5) Add into all of this, customer support, new product development and testing, maintaining the web site, etc., and it all adds up. You may note that some of the product web pages are a little thin on content... this is because by the end of the day I am just too tired to think of stuff to put on there.
Anyway, enough of my yammering on. We are always improving the way we do things to make them more efficient, and our ability to deliver more product is always growing, but the number of orders is also growing, so it doesn't always show in the results.
1) We are a hobby business. I build, pack and ship kits/modules in the morning before work, in the evening after work (mostly after the kids go to bed), and also on the weekends (basically all of my spare time). I know Russ spends a similar amount of time doing R&D.
2) We do have the Buffalo modules built for us due to the volume. We are somewhat limited in how many we can build at a time, mostly due to the enormous up-front cost of parts, and lead times for getting XOs built. This is only a part of the delay. Also, every Buffalo module needs to hand tested and packed up, along with a hand-tested AVCC module. B-III will add Tridents to that.
3) The typical Buffalo order includes an output stage kit and a couple power supply kits. This means that for (on average) every Buffalo module, I need to built three kits, which is very time consuming. This is the primary reason we do batches of pre-orders.
4) Even after all of that, simply packing up and shipping the orders takes a lot more time than you would think. Something like 80% of orders are international, so that's a lot of customs forms for be filled out, etc.
5) Add into all of this, customer support, new product development and testing, maintaining the web site, etc., and it all adds up. You may note that some of the product web pages are a little thin on content... this is because by the end of the day I am just too tired to think of stuff to put on there
.Anyway, enough of my yammering on. We are always improving the way we do things to make them more efficient, and our ability to deliver more product is always growing, but the number of orders is also growing, so it doesn't always show in the results.
http://www.diyaudio.com/forums/digi...chronous-usb-i2s-interface-2.html#post2585504
I have mixed feelings about that project, as in essence it's just a rehash of the XMOS reference. Sure, it's a good project for people in need of a Windows driver but it's not innovative like the project Russ is working on.
Perhaps you would care to elaborate on how innovative that project is. I would really like to know.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.