Without ever listening to the Buffalo, I need to say, that I really hope it will kick som serious DAC1-***. The DAC1 may be the most over-rated DAC on the marked this moment. It really sounds nothing but boring.
About the Buffalo: I could have the impression, that it is made primarily to achieve good specifications. When you spend some years on R&D of a DAC, you will find that specifications isn't the way to State Of The Art. It's much more a matter of implementation of the DAC.
PSU, regulation, decoupling, analog stage, PCB layout.... These are the things that will make a great DAC.
During our 4½ years of R&D on our DAC, we have testet a lot of DAC chips. The difference is quite small, as long as you go for relative new chips (Old NOS types realy does sound "old". Like it or not).
Working on the above areas on the other hand, has improved performance extremely.
Since I have not listened to the ESS chip used in the Buffalo, I cannot say how it performs. Again my guess is, that specifications has been more important than listening tests during the R&D at ESS.
Looking at the Buffalo design, I need to say, that this is not the way I would design a high performance DAC.
- The design is seperated on different PCB's, You need a seperate SPDIF receiver, connected with wires. This will create more jitter than a well designed PCB havin the SPDIF receiver onboard.
- The analog stage is 100% based on op-amps. This will most likely lead to a sound just like any other DAC based on op-amps.
One of the things we discovered was, that the negative impression "Digital Sound", has nothing to do with the signal being digital. It is caused by the use of op-amps, and replacing the analog stage with a discrete design, normally removes this "digital sound" completely.
You may have a look at our DIY DAC: http://www.diyaudio.com/forums/showthread.php?t=138230
This is the product of 4½ years of R&D and listening tests.
About the Buffalo: I could have the impression, that it is made primarily to achieve good specifications. When you spend some years on R&D of a DAC, you will find that specifications isn't the way to State Of The Art. It's much more a matter of implementation of the DAC.
PSU, regulation, decoupling, analog stage, PCB layout.... These are the things that will make a great DAC.
During our 4½ years of R&D on our DAC, we have testet a lot of DAC chips. The difference is quite small, as long as you go for relative new chips (Old NOS types realy does sound "old". Like it or not).
Working on the above areas on the other hand, has improved performance extremely.
Since I have not listened to the ESS chip used in the Buffalo, I cannot say how it performs. Again my guess is, that specifications has been more important than listening tests during the R&D at ESS.
Looking at the Buffalo design, I need to say, that this is not the way I would design a high performance DAC.
- The design is seperated on different PCB's, You need a seperate SPDIF receiver, connected with wires. This will create more jitter than a well designed PCB havin the SPDIF receiver onboard.
- The analog stage is 100% based on op-amps. This will most likely lead to a sound just like any other DAC based on op-amps.
One of the things we discovered was, that the negative impression "Digital Sound", has nothing to do with the signal being digital. It is caused by the use of op-amps, and replacing the analog stage with a discrete design, normally removes this "digital sound" completely.
You may have a look at our DIY DAC: http://www.diyaudio.com/forums/showthread.php?t=138230
This is the product of 4½ years of R&D and listening tests.
There is no comparison between DAC1 and Buffalo.
The Buffalo24 with IVYs which we compared to DAC1 was in a different performance category altogether.
We were gonna do extensive testing but it proved to be pointless half-way through the first track we compared them on..
Buffalo32 takes that performance a notch higher.
The Buffalo24 with IVYs which we compared to DAC1 was in a different performance category altogether.
We were gonna do extensive testing but it proved to be pointless half-way through the first track we compared them on..
Buffalo32 takes that performance a notch higher.
on "modern" DACs like the CS4398:
the "no negative feedback" and discrete vs op amp arguments are completely silly when you look at what’s inside one of these delta-sigma DACs
“noise shaping” is a consequence of the Negative Feedback Loop Gain in delta-sigma circuits, you can read off the loop gain (technically the “desensitivity”) as the inverse of the noise amplitude – in fact they have some of the consistently highest feedback factors over the audio range of any audio circuitry
we're looking at over 1/2 dozen op amps internal to the 4398, implementing >120 dB of feedback to achieve the noise shaping - delta-sigma converters have the highest amount of negative feedback of any audio chip
and the op amps internal to the chip are implemented in a compromise semiconductor process with the fact that they also have >100K digital gate equivalents on the same chip
it seems to me seems very foolish to argue that well implemented, properly selected topologies using highly optimized modern analog process op amps external to the DAC chip is somehow going to compromise the DAC output
additionally the feedback inside the DAC chip is largely implemented switched C elements - are you saying this 5 V CMOS mixed signal system is "better" at processing audio than say bulk foil resistors around a composite jfet or bipolar op amp +/-15 V circuit implemented in a recent complementary isolated process?
except for low noise front ends (MC phono pre) and power amp outputs discrete semis are losing ground, in the last 20yrs we're talking Billions of dollars and tens of thousands of man years improving op amps for DSL and other high speed signal processing as well as precision industrial/instrumentation apps
on the discrete front for audio apps we're only losing the best devices as they're discontinued
Last edited:
Perhaps you should listen to it before commenting.
Incorrect. the original Buffalo had the I/V stage on a separate board to allow people to use other output stages. The current Buffalo (32S) integrates the I/V stages into the same module for lower inter-stage RCL.
The Buffalo has NEVER required a separate module for digital input (S/PDIF, I2S and DSD).
This is a pretty silly thing to say as a blanket statement. Also, the Buffalo uses a fully differential analog output stage which is not typical of op-amp output stages.
I have the first generation Buffalo and would love to try the new Sabre chip. However, I somewhat agree with the opamp analogy in that I have never heard n opam output stage that sounds as good as a discrete one. I have neer heard the Buffalo 32S and probably never will because of the price. Mine in running through a pair of Lundahl LL1674 transformers for the I/V resulting in a massive difference in SQ over voltage out mode. I have compared it to a few good players and can say it is in a league of it's own. I have a Pass D1 I/V stage that I would like to populate and try to compare for giggles.
The guys at Twisted Pear have done great work and I would love to hear from someone who has heard the new one and the old one with transformer output and can compare the two. I guess I could order the 32s and sell the one I don't like as much.
The guys at Twisted Pear have done great work and I would love to hear from someone who has heard the new one and the old one with transformer output and can compare the two. I guess I could order the 32s and sell the one I don't like as much.
I know it may seem silly, arguing against the op-amp. But that's the difference between the mainstream DAC and the one we made. We really did not care how the DAC was build. We just wanted top performance. That's why we spend 4½ years of R&D and listening to different implementations.
Along the way we did discover some intersting things:
- When we use an op-amp based analog stage, most DAC chips sounds the same. When we use our non feedback discrete version, we start hearing the difference. The non feedback design really reveals the diffrence between the chips.
- If we take our non feedback design, and follow it with an op-amp, it sounds just like any other op-amp based DAC.
- When we use an op-amp based analog stage, the sound will go towards the term "digital sound" in it's negative way (Remember the late 80's/early 90's discussion on digital sound). When changing nothing but the analog stage, to our non feedback discrete design, this "digital sound" disappears = "Digital sound" isn't a digital thing.... It's basically caused by a bad analog stage
.All of this has shown us, that going for the absolute best performance, op-amp really is a No-Go!
If you aim for lower performance, op-amps may be OK. The advantage is, that it will take less than a day of work to complete the schematics, and you never really fail completely. However... You also newer really reach the top! If you wanna know what I mean, just listen to some ML. That's what op-amps sound like.
Did you try every op amp and every possible implementation? Perhaps it was your implementation that was lacking?
If you weren't trying to sell your own design in commenting on ours, you might sound a bit more credible.
I mean, really...
Sorry... I was looking at the current version, with an 100% op-amp based analog stage. Also it looks like they use ceramic capacitors in the analog filtering. Unless the DAC is made by someone with a serious hearing problem, this can only mean one thing: No listening tests have been made to select components.
Ceramic capacitors are great for decoupling digital supply-pins, and that's about it. In the analog domain, ceramic capacitors is No-Go if you want top performance.
I actually tested this once, when I changed 2 capacitors in the analog stage in our DAC, before I left it at KvK's place. When we powered it up, KvK asked me within the first seconds, if something was changed, since the sound had really changed to the worse. At that point, he hd not listened to it for 2 days. And still the 2 ceramic capacitors cut deep inside the ear.
The only positive thing may be, that the op-amps may not reveal that much of the weakness in ceramics
The old version with an off-board analog stage may seem OK. But then you have 3 seperate boards:
1) SPDIF receiver
2) DAC
3) Analog stage
All these connected with Wires and screw-terminals.... Please explain to me, how this can be an ideal solution on a DAC where you need:
- Low jitter in high speed digital signals.
- Ultra low noise (-120dB)
I guess you need some serious WooDoo to see why this is ideal....
Well... When have you tried every op-amp and every implementation?? Guess that's not possible!
But... I guess I have tested at least 50-75 different op-amp. Many of them are well known "Audio" op-amps. The funny thing is, that even though they have very different specifications, they generelly sound very much the same. I could try even more, but since they sound so much the same, I guess it will make no difference.
About implementation...: Well... Most of the DAC's I have seen, use a almost 100% copy of some app-note in the analog stage. I have even seen a CD-player, where the designer had spend time on telling how he had choosen every single part in the analog stage. By the way he told about it, it must have taken years.... Anyway... It was a 100% copy of a Crystal Semi App-Note
. That's how High End are made.... The implentation in the Buffalo (if I'm right about the ceramic capacitors in the analog filter), really do not impress me. I guess it will take a few hours to draw the schematics for such a DAC.
And if you use ceramic capacitors, you certainly have not spend time on listening tests.
You may be right.... Someone claimed that I was trying to sell my DAC in here... Guess what.... We uploaded the schematics = You can make it yourself.
And it's not ment as an insulting. I just need to clear out, that the world is floaded by DAC's based on op-amps. We really need something a bit more "well done", if we are looking for the absolute best.
If you are aiming a little lower, you may go for the op-amp...
Your question is a tough one. If you framed it slightly differently:
Is the buffalo a worthwhile build and worth the cost?
Then the answer really has to be yes. If you are asking is the buffalo better than your existing DAC, then we can't answer - and most likely you can't either until you hear both built side by side. However, I'm sure that if you built the buffalo and didn't like it, the resale would be swift.
I'm not criticising, just saying its a hard one to answer.
*****************************************
To Hurtig:
It does seem a bit unfair to kind of run down the buffalo when you haven't even heard it. Not saying that there isn't logic in what you are saying. However, I do happen to believe that a large amount of time has gone into each buffalo version released so far. In fairness, it would have been very easy for TPA to sit and just continue with the older ESS chip.
**************************************************
To declare my own interest: I do have a buffalo (the older version), and I built a second one for a friend, both with counterpoint discrete output stages. Were the discrete counterpoint IV stages better than the opamp based IVYs? Well yes, we both thought so. I have no great loyalty or dedication to TPA, other than enjoying the buffalo. I do think they offer good help and service. If I thought the buffalo sounded poor, then I would be saying so.
However, I, and my friend, have a fair bit of listening experience under our belts and can say that the buffalo compares excellently with nearly anything thrown at it. The only piece of kit that really clearly surpassed it (although my listening to really high end gear is very limited) was another friends DCS system. Just can't recall exactly the names of the 3 (or was it 4) boxes that made up the DCS system he has, but IIRC, the transport was a from the previous generation, but the DAC, clock and PS ( I think) was the more recent generation.
Putting it another way: you need to have a bloody good DAC built to exceed the sonics of the buffalo. Hurtig: If you have that, fair play to you, well done. More people here should be reading, building and listening to your DAC.
Fran
PS: khundude, do you have a parts list for the pass d1 to allow it to be used with the buffalo. IIRC, it is made for a lower current output and there was something that needed changing. Do you know what that was?
Hurtig, have you looked at the ESS data sheet and the TP website?
The ESS chip has a built-in SPDIF to I2S converter, so there are no wires, not even PCB traces, to cause jitter here. If I remember correctly, the chip also has a built-in ASRC, so the DAC clock is not derived from a PLL.
And let me assure you that the TP people are well aware that top performance can only be achieved with a carefully optimized, tight layout. This is why they have come up with a second board, which includes the IV and output stage (yes, based on opamps). And since everything is dc-coupled (in both versions), there are no caps in the signal path, ceramic or otherwise.
For those who don't like opamps, they offer the Counterpoint, a fully discete IV and output module.
In view of all of this, and since you have never even listened to the Buffalo, I don't really understand your criticism.
Kurt
The ESS chip has a built-in SPDIF to I2S converter, so there are no wires, not even PCB traces, to cause jitter here. If I remember correctly, the chip also has a built-in ASRC, so the DAC clock is not derived from a PLL.
And let me assure you that the TP people are well aware that top performance can only be achieved with a carefully optimized, tight layout. This is why they have come up with a second board, which includes the IV and output stage (yes, based on opamps). And since everything is dc-coupled (in both versions), there are no caps in the signal path, ceramic or otherwise.
For those who don't like opamps, they offer the Counterpoint, a fully discete IV and output module.
In view of all of this, and since you have never even listened to the Buffalo, I don't really understand your criticism.
Kurt
Not sure how it compares to any other $1000 commercial DAC but you have our results when put head-to-head versus DAC1 (which is supposed to be one of the best in its price range).
Did I mention the owner of that DAC1 unit sold it two days later?
Hey guys,
This is the bottom line for me. Brian and I listen to what we build as well as what others produce.
If we thought we weren't doing something special we would not bother. We both have very real very demanding careers outside of audio and spend a hell of a lot a time we could be doing other things doing our best to help people enjoy music at the same level (or better) as we do.
So here is the thing. Could I do zero feedback? Sure no sweat. I have. Does it sound nice? Yep. Better? Wow. That's up to you. In my book probably not.
So what is better? Not a chance in hell I could answer that in a way that will please all camps.
I like what I have done. I did not come easy or fast.
Cheers!
Russ
This is the bottom line for me. Brian and I listen to what we build as well as what others produce.
If we thought we weren't doing something special we would not bother. We both have very real very demanding careers outside of audio and spend a hell of a lot a time we could be doing other things doing our best to help people enjoy music at the same level (or better) as we do.
So here is the thing. Could I do zero feedback? Sure no sweat. I have. Does it sound nice? Yep. Better? Wow. That's up to you. In my book probably not.
So what is better? Not a chance in hell I could answer that in a way that will please all camps.
I like what I have done. I did not come easy or fast.
Cheers!
Russ
- Status
- Not open for further replies.