You are doing this the wrong way... You are not making measuring equipment. You need to use your ears, to decide what is best. In the end, it's your ears that has to live with result of your effort! And if your ears cannot hear the difference, why spend time on optimizing??
If you want to know how audio gear sounds when R&D is based on measurements, just buy some stabdard $199 DVD player. These normally have impressing measurements. But sounds like "Biiippp".
I guess the BOM cost is somehow higher than the Buffalo. My best guess is, that BOM for a Buffalo is way under $150 (I really do not see why it must be priced at $599).
Since the BOM for our DAC is not yet finish (I did not have the time to update it yet), I cannot say the excact number. But no doubt it will be less than a $599 Buffalo
. And compared to the Buffalo, this is NOT just for a PCB with some screw terminals to connect regulated power supply, input and output.This is for a main DAC PCB including high quality RCA connectors and TOSLINK. All regulators are onboard. Also included is a transformer PCB including 2 UI core transformers.
All you need extra to finish the DAC, is a chassis power socket, power switch and some wire between the 2 PCB's.
If you really are thinking about building the DAC, I can supply you the PCB's without components at a nice price
.Also, if you wat the whole package, I will be willing to ship you PCB with all components mounted and tested. Just send me a message, and then we will find out.
I have seen something like that with >10k$ audio equipment and maybe this is what your DAC4½ is actually destined for (commercial Danish made Hi-Fi can be pretty good and expensive). Or what do you actually have in mind with it?
As it is way beyond my competence to do prototyping, I greatly appreciate and rely on the effort and intellect that many people put into the design and realization of diy circuits. Especially, if kit versions are also made available to hobby diyers for a reasonable price.
I'm sure that Buffalo is not perfect and that the mods that some of us are doing is not going to make it so. Nevertheless, its for real, sounding great (IMO) and will do just fine until something significantly better comes up.
Cheers,
Nic
We fully agree... And that's really my point.
And that really makes the Buffalo an mainstream DAC like most others. This one just costs $599
lol
definitely needs to be the official name
that is what characterizes the dac the best no?
Last edited:
I assume your comparisons were subjective only. I have no experience with the Buffalo or the DAC 1. But I have to wonder how the Benchmark unit with its impeccable digital performance and excellent overall measurements falls so short. Are there any comparable measurements on the Buffalo unit? I also wonder if you used an original DAC 1 for your comparisons or if you had a later unit with upgraded analog stages like the DAC 1 HDR.
I have owned a Theta Gen Va balanced DAC for about 15 years. I have never felt the need to replace it. I've compared it extensively to the Pass D1 and the Logitech Transporter, among others. I have found the differences between these DACS to be quite subtle. Tomorrow I'll have a DAC 1 HDR on loan, which I will again compare extensively to the Theta. I'll report my findings when I'm done.
The Buffalo 32 is intriguing in that it offers a lot of technological bang for the buck. But I have to admit to being very skeptical about getting huge improvements to Red Book digital beyond what I already have.
Subjective only.
We were all curious how it'd fare against a DAC1 which is considered a good contender in it's (commercial) price range.
As I said, the owner the "test" system stopped the music halfway through his favorite track. It was Sunday when we did the comparison. On Monday he put an add for the Benchmark and on Tuesday it was sold.
5/5 listeners were impressed by the Buffalo and/or disappointed by the DAC1 (depending on their expectations before the shootout)..
We are now looking for a Weiss Minerva to put against a 32S with shunt regs.
I'm a big fan of Daniel Weiss and his products (I've met him in person and he sure knows what he's talking about) so I think it's gonna be a tough challenge for the Buffalo.
It was a stock DAC1 (not USB).
There are plenty of measurements on the Buffalo thread but I don't have the exact post # atm.
It's a huge thread but I guess with the new forum platform it'll be easy to find them.
So was I, until that damned day some friends brought a Buffalo (from the first batch ever sold) at my place.
BTW. When I replaced the 3.3V reg for the clock with a PH I, by mistake, turned on the DAC without having powered up the PH. I had a moment of panic as the DAC actually still outputs music - muddy and unacceptable, but still music. Powering up the PH fixed everything and brought it back to baseline or above. I know that the Sabre chip in the B32 is likely to configured for the use with an external clock and me not powering this cannot be anything but detrimental.
Nevertheless, it makes me speculate that maybe the quality of the clock on the B32 is one of the key reasons for this DAC's great sonics.
Nic
Subjectively, the buffalo 32 sounds very good. It only took a couple of seconds into the first song to recognize it sounds significantly better than the built-in DAC in my Marantz cd-6000 KI. Where the Marantz sometimes can sound a tad harsh and in-your-face with violins/crash-cymbals the b32 just sounds relaxed and displays those sounds with a better depth in the sound-stage, and not harsh at all. I think the (slight) harshness is what also makes the cd6000 ki sometimes sound up-front and hence obscuring the soundstage.
Wider and deeper soundstage, better micro dynamics, effortless with complex sounds. Relaxed without being laid-back. Maybe that is what others describe as the B32 being transparent or not adding anything of itself to the sound.
Some weeks before I completed my B32 some friends came over to compare the DAC of the Marantz with the Benchmark (non USB, earlier analog circuitry). The Benchmark just barely "won" over the Marantz and it took some careful listening to selected recordings to hear the differences. We were expecting a much bigger improvement than what we heard to the big disappointment for the Benchmark owner. All of us thought the benchmark was better, but not worth the investment.
I still have to do an A-B comparison between the Benchmark and the B32, but if the improvement going from Marantz to B32 was a lot bigger than going from Marantz to Benchmark, I would be surprised if the same bunch of (audiophile) friends would find the B32 anything but the best yet, overall.
Upgrading my Marantz cd6000 ki with the B32 is a no-brainer price/performance-wise IMHO.
Last, but not least, I find the guys at TPA to be knowledgable, honest, professional and straightforward to do business with .
Wider and deeper soundstage, better micro dynamics, effortless with complex sounds. Relaxed without being laid-back. Maybe that is what others describe as the B32 being transparent or not adding anything of itself to the sound.
Some weeks before I completed my B32 some friends came over to compare the DAC of the Marantz with the Benchmark (non USB, earlier analog circuitry). The Benchmark just barely "won" over the Marantz and it took some careful listening to selected recordings to hear the differences. We were expecting a much bigger improvement than what we heard to the big disappointment for the Benchmark owner. All of us thought the benchmark was better, but not worth the investment.
I still have to do an A-B comparison between the Benchmark and the B32, but if the improvement going from Marantz to B32 was a lot bigger than going from Marantz to Benchmark, I would be surprised if the same bunch of (audiophile) friends would find the B32 anything but the best yet, overall.
Upgrading my Marantz cd6000 ki with the B32 is a no-brainer price/performance-wise IMHO.
Last, but not least, I find the guys at TPA to be knowledgable, honest, professional and straightforward to do business with
.As promised, posting again with more listening impressions and results of mods.
The DAC did sound closed in and stuffy as I last reported. I had not cut the traces to lower output impedance as recommended in the user guide. The explanation for why to cut them was not thorough enough to explain why it must be done, it only says that if you will use it to drive headphones then don't cut. If use amps and cables, then cut to lower ZOut. I have tube preamp 250kOhm Zin, so I didn't care what the DAC outputZ was, and i thought I might use it as headphone DAC at some point so I didn't cut the traces.
Anyway, an engineer friend was over listening, comparing amps and we plugged the DAC into an amp that was hot. This started a soft but audible squeal a oscillation. My friend explained that the leads that the manual suggests to cut should be cut to increase stability at ultrasonic freqs. He went into a long explanation which made sense at the time, but in a couple days I just remember the answer: cut them.
After I made this change, and added some nice nexgen output RCA jacks I had another opportunity to compare it to our reference DAC, the Oracle DAC1000. The stuffyness and close in feeling of the Buffalo were gone. The Oracle still has a slight edge in ultra resolution, but it is not worth the 10x price premium over the Buffalo. The audible differences were only noticable on audiophile type recordings of steel string guitars and such where micro tonal and dynamics nuances are extreme. On regular music and older but high recordings there is no audible difference. This is playing into LFD NCSE amp and Harbeth 40.1 via all JPS SC3 and Aluminata wire, and very well acoustically treated room with bass traps, diffusion on walls and ceiling. Awesome system, revealling, uncolored, neutral. This comparison happened Saturday night, with about a dozen experienced audiophiles in attendance. Some of the opinions are expressed on audionervosa.com
The next day, a friend and I compared Buffalo32 against Audio by Van Alstine T8 DAC (6N1P) using another system, with Usher Be20, MR RM200 tube amp, Cary SLP98 pre and all JPS SC3 wiring. The T8 is very pretty sounding, and incredible value for a moderate priced tube DAC. Buffalo was clearly superior in everyway, including soundstage detail and depth, typically tubes will excel in doing space. T8 was fuzzier and warmer. This acytually helped on some tracks like symphony recordings which can be very dry and analytical. The tubes thickens the tone a little, but so much other detail is lost compared to Buffalo. The owner of the AVA T8DAC wants me to build him a Buffalo now, which I am happy to do!
We have 2 Buffalo32 DACs here in town. The other one is owned by the owner of the Oracle. That one is receiving mods in effort to determine the absolute potential of the chip. The engineer doing the mods noted to me how great the PCB layout is. He is man of few words so his praise is out of the ordinary, especially for someone like Russ whom he doesn't even know. He swapped 4562 and voltage reference chips for quieter chips, I can't remember which. 1632 stayed. He also modified the PS with NichiconKZ caps, higher performance regulators and rectifier diodes. DAC power and I/O was wired with solid core JPS speaker wire soldered direct to PCBs instead of the screw terminals provided in the kit. Input is switchcraft XLR jack for AES/EBU input, via transformer. Analog outputs WBT nexgen. Mine is wired with Mouser tinned solid copper hookup, screw terminal headers and stock circuits right from Brian's kit. In direct ABC comparison, A was my stock Buff32, B was modded Buff32, C was Oracle, there was only barely audible difference between A and B, very slightly more open and clear on some tracks, identical on most others. But that could easily be the JPS wire, or slightly quieter PS or the new opamps. But there was no aha! moment or big enough difference to warrant repeating those mods again on a new build. Both the Buffalo DACs seemed to really improve a lot after some warm up, which was strange. The modded one was brand new, never played. But my stock has 100+ hours on it, but still sounded much better after 30 minutes. Before warm up it was much easier to tell the difference between it and the Oracle. We also compared it to Exemplar Audio DAC that night, but it was brand new and was trounced by all. We'll try that one again after it breaks in.
I see all the armchair engineers redesigning Buffalo32 for sport, I suspect few have heard it in a system capable of fully revealing its performance potential. I suspect fewer still would waste their own company's billable hours on a project that is so unnecessary. Even if there is more performance still to be found from Buffalo32, as stock this is still one of the best commercially available DACs available. For <$1000 it is a great opportunity for real audiophiles to improve their system by a big step. Especially for people using cheap CDP or network music servers as digital transport, Sabre32 eliminates jitter and makes all transports sound the same, from Sonos ZP-80 (bad!) to Lynx22 (great!) Just make sure you cut those output traces per the instructions!! It makes a huge difference.
I paid full price for my stuff and am not acquainted with Russ or Brian.
Rich
The DAC did sound closed in and stuffy as I last reported. I had not cut the traces to lower output impedance as recommended in the user guide. The explanation for why to cut them was not thorough enough to explain why it must be done, it only says that if you will use it to drive headphones then don't cut. If use amps and cables, then cut to lower ZOut. I have tube preamp 250kOhm Zin, so I didn't care what the DAC outputZ was, and i thought I might use it as headphone DAC at some point so I didn't cut the traces.
Anyway, an engineer friend was over listening, comparing amps and we plugged the DAC into an amp that was hot. This started a soft but audible squeal a oscillation. My friend explained that the leads that the manual suggests to cut should be cut to increase stability at ultrasonic freqs. He went into a long explanation which made sense at the time, but in a couple days I just remember the answer: cut them.
After I made this change, and added some nice nexgen output RCA jacks I had another opportunity to compare it to our reference DAC, the Oracle DAC1000. The stuffyness and close in feeling of the Buffalo were gone. The Oracle still has a slight edge in ultra resolution, but it is not worth the 10x price premium over the Buffalo. The audible differences were only noticable on audiophile type recordings of steel string guitars and such where micro tonal and dynamics nuances are extreme. On regular music and older but high recordings there is no audible difference. This is playing into LFD NCSE amp and Harbeth 40.1 via all JPS SC3 and Aluminata wire, and very well acoustically treated room with bass traps, diffusion on walls and ceiling. Awesome system, revealling, uncolored, neutral. This comparison happened Saturday night, with about a dozen experienced audiophiles in attendance. Some of the opinions are expressed on audionervosa.com
The next day, a friend and I compared Buffalo32 against Audio by Van Alstine T8 DAC (6N1P) using another system, with Usher Be20, MR RM200 tube amp, Cary SLP98 pre and all JPS SC3 wiring. The T8 is very pretty sounding, and incredible value for a moderate priced tube DAC. Buffalo was clearly superior in everyway, including soundstage detail and depth, typically tubes will excel in doing space. T8 was fuzzier and warmer. This acytually helped on some tracks like symphony recordings which can be very dry and analytical. The tubes thickens the tone a little, but so much other detail is lost compared to Buffalo. The owner of the AVA T8DAC wants me to build him a Buffalo now, which I am happy to do!
We have 2 Buffalo32 DACs here in town. The other one is owned by the owner of the Oracle. That one is receiving mods in effort to determine the absolute potential of the chip. The engineer doing the mods noted to me how great the PCB layout is. He is man of few words so his praise is out of the ordinary, especially for someone like Russ whom he doesn't even know. He swapped 4562 and voltage reference chips for quieter chips, I can't remember which. 1632 stayed. He also modified the PS with NichiconKZ caps, higher performance regulators and rectifier diodes. DAC power and I/O was wired with solid core JPS speaker wire soldered direct to PCBs instead of the screw terminals provided in the kit. Input is switchcraft XLR jack for AES/EBU input, via transformer. Analog outputs WBT nexgen. Mine is wired with Mouser tinned solid copper hookup, screw terminal headers and stock circuits right from Brian's kit. In direct ABC comparison, A was my stock Buff32, B was modded Buff32, C was Oracle, there was only barely audible difference between A and B, very slightly more open and clear on some tracks, identical on most others. But that could easily be the JPS wire, or slightly quieter PS or the new opamps. But there was no aha! moment or big enough difference to warrant repeating those mods again on a new build. Both the Buffalo DACs seemed to really improve a lot after some warm up, which was strange. The modded one was brand new, never played. But my stock has 100+ hours on it, but still sounded much better after 30 minutes. Before warm up it was much easier to tell the difference between it and the Oracle. We also compared it to Exemplar Audio DAC that night, but it was brand new and was trounced by all. We'll try that one again after it breaks in.
I see all the armchair engineers redesigning Buffalo32 for sport, I suspect few have heard it in a system capable of fully revealing its performance potential. I suspect fewer still would waste their own company's billable hours on a project that is so unnecessary. Even if there is more performance still to be found from Buffalo32, as stock this is still one of the best commercially available DACs available. For <$1000 it is a great opportunity for real audiophiles to improve their system by a big step. Especially for people using cheap CDP or network music servers as digital transport, Sabre32 eliminates jitter and makes all transports sound the same, from Sonos ZP-80 (bad!) to Lynx22 (great!) Just make sure you cut those output traces per the instructions!! It makes a huge difference.
I paid full price for my stuff and am not acquainted with Russ or Brian.
Rich
Richidoo writes some funny stuff.
Taken from the above, the conclusion must be, that a Buffalo DACseems to very good. But the Oracle is just that bit better.
The Buffalo features the ESS Sabre 32 bit DAC chip followed by an op-amp output stage and I/V, the Oracle uses Crystal Semi 4397 or 4398 in conjunction with an ASRC , which at that time only can be Crystal, AD, BB or one of these loaded into a DSP, either as standard or with minor modifications, the DAC chip is then followed by a discrete analog stage.
It seems like a 6½ years old Crystal DAC does pretty well afterall.
Or the discrete analog stage is somewhat better than the opamps.
To me there is no doubt what so ever, these and a lot of other op-amps really sucks. We tried most of them out, with absolutely no go at all. The whole idea with linearising an amplifier with NFB is IMHO wrong. Every stage throws away 100 dB of signal level and reamplifies afterwards, on top of that they are weak and sounds just precisely like that - even the strongest of them.
I ´m sure the Oracle will sound a lot more different from the Buffalo if it was given the chance, as it features a very strong output stage (Oracle says someplace 25 Watts which of course is both impossible and sales **), but it should be pretty powerfull afterall.
Taken from the above, the conclusion must be, that a Buffalo DACseems to very good. But the Oracle is just that bit better.
The Buffalo features the ESS Sabre 32 bit DAC chip followed by an op-amp output stage and I/V, the Oracle uses Crystal Semi 4397 or 4398 in conjunction with an ASRC , which at that time only can be Crystal, AD, BB or one of these loaded into a DSP, either as standard or with minor modifications, the DAC chip is then followed by a discrete analog stage.
It seems like a 6½ years old Crystal DAC does pretty well afterall.
Or the discrete analog stage is somewhat better than the opamps.
To me there is no doubt what so ever, these and a lot of other op-amps really sucks. We tried most of them out, with absolutely no go at all. The whole idea with linearising an amplifier with NFB is IMHO wrong. Every stage throws away 100 dB of signal level and reamplifies afterwards, on top of that they are weak and sounds just precisely like that - even the strongest of them.
I ´m sure the Oracle will sound a lot more different from the Buffalo if it was given the chance, as it features a very strong output stage (Oracle says someplace 25 Watts which of course is both impossible and sales **), but it should be pretty powerfull afterall.
Kurt, it's pretty obvious you wish Sabre32 was a drop in replacement for the DAC chip you're currently using but please get over it!
There are other ways you can promote your DAC and design decisions and a good start is to make it avalaible for people to listen and evalute it..
TP guys didn't go around bashing other people's job to get the Buffalo name heard.. They just released a product and let it build a name by itself...
There are other ways you can promote your DAC and design decisions and a good start is to make it avalaible for people to listen and evalute it..
TP guys didn't go around bashing other people's job to get the Buffalo name heard.. They just released a product and let it build a name by itself...
- First of all, our DAC can be listened to in the same way the Buffalo can, You buy the PCB and mount it yourself or you do buy a completed one, or you do your own layout and build it up from scratch.
- Second, I truely think the Sabre DAC is worthwile listening to, or at least I hope so.
- Third, I still think the implementation of the Sabre DAC in the Buffalo design is both boring, and completely foreseeable after these few months of lifetime for the chip itself, and in addition it is simply not very ambitious. If the ESS chip is really good, it deffinately has not been shown yet.
- Fourth: How can anyone leave it to even DIY users to choose their own PSU, and think the result will be good? I don´t get it. The PSU is next to the analog design, the most vital part of a DAC.
- We did not yet design any analog stage and PSU for this chip, hence we are not trying to market anything by bashing someone else.
- I am an enthusiast, who regardles of how large the obstacles may be, are trying to get as close as possible to the recorded event - reproduction, and I´ve tried to many times to achieve this by following application notes, as well as using obscenely specced operational amplifiers. It simply does not work, because no one bothers to listen to it, but relies on specs. If really did work, you´d never find any hifi component sounding different from the eval board.
- The last but not least interesting thing of it all is: Have your guess about which, in this case particularly interesting person, asked for the complete schematics of our discrete DAC and our discarded discrete common base I/V converter. Tehere is no prise but try anyway.
- If someone missed it until now I´ll repeat - operational amplifiers sucks, and their only upsides are fast and cheap design, followed by a not completely mandatory catastrofic result.
The absolutely last thing to be said about anything is: There is no rule without an exception, but I think it would be somewhat incidentally if that would comply with the B thing.
Last edited:
KvK.....
THis is simply the wrong approach.
You are doing your product a lot of marketing harm by continuing - having been told sooo many times - to post on this thread in the manner which you have used. You have your own thread...stay on that as far as any mention of your DAC is concerned.
You have been shamelessly knocking a generally well accepted product which is in competition to yours, your blatant advertising (by negating the Buffalo product) is not welcome. It is not your business to publicize any shortcomings that product may - in YOUR opinion - have. IF you have any genuine wish to help the promoters, Twisted Pair, then mail them direct.
I have to be honest and say that your behavior has been so distasteful that I am very surprised that a Moderator has not yet taken action against you.
THis is simply the wrong approach.
You are doing your product a lot of marketing harm by continuing - having been told sooo many times - to post on this thread in the manner which you have used. You have your own thread...stay on that as far as any mention of your DAC is concerned.
You have been shamelessly knocking a generally well accepted product which is in competition to yours, your blatant advertising (by negating the Buffalo product) is not welcome. It is not your business to publicize any shortcomings that product may - in YOUR opinion - have. IF you have any genuine wish to help the promoters, Twisted Pair, then mail them direct.
I have to be honest and say that your behavior has been so distasteful that I am very surprised that a Moderator has not yet taken action against you.
Last edited:
KvK.....
THis is simply the wrong approach.
You are doing your product a lot of marketing harm by continuing - having been told sooo many times - to post on this thread in the manner which you have used. You have your own thread...stay on that as far as any mention of your DAC is concerned.
You have been shamelessly knocking a generally well accepted product which is in competition to yours, your blatant advertising (by negating the Buffalo product) is not welcome. It is not your business to publicize any shortcomings that product may - in YOUR opinion - have. IF you have any genuine wish to help the promoters, Twisted Pair, then mail them direct.
I have to be honest and say that your behavior has been so distasteful that I am very surprised that a Moderator has not yet taken action against you.
The funny thing is, that we´ve got no product at all, TPA has.
And the thread is named "how good is the Buffalo DAC".
My answer to that is: "Just almost as good as any other standard product based on application notes" No more no less.
You forgot to make your guess about who wanted the schematics in complete, to give you a tiny clue, his first name starts with a D
- Status
- Not open for further replies.