Topping dx7 pro fir filter

I think monkeyxx uses one and loves it, I think the non-pro version.

Here is one of Jim Williams posts in GS about the DX7:

www.gearslutz.com/board/showpost.php?p=13997380&postcount=114

Yes I use a Topping DX7S. Not the pro version. I'd love to be able to upgrade to the Pro.

It's my word vs. Jim Williams word LOL.

I hate the one he likes, and he hates the one I like.

That's exactly right, thank you.

The DX7s is more midrange focused, a little thicker sounding, slightly subdued top end compared to many converters. Kicks in the lows enough for me. To me it sounds more 'analog'. I wonder how much of this sound signature is related to the Burr Brown audio op amps. Which have always sounded a little darker to me when compared to other op amps.

Pro Ject Pre Box Digital S2 I hear as being very shiny and open on the top, very spacious, lots of front to back depth, almost holographic. But to me it lacked that solidity, and the low end punch that I crave so much. After a while I found it disorienting to listen to compared to what I'm used to. It was just too different for me.

VERY different sound presentation based on some of the same components. Just shows you that the "chip" is not everything, not even close.

The Topping DX7 Pro version is interesting to me because I do think the DX7S could be improved upon. Even though I use it happily.

Get a Tascam UH-7000

I just did a little shootout of the UH7000 DAC yesterday vs a Focusrite Clarett 4Pre.

The Tascam is a little behind in 2019. It sounds good but the Clarett just had a little more detail and "information."

I think the Tascam headphone amp might beat the Clarett headphone amp though, the Clarett phone amp is a little zingy, almost harsh. The UH7000 phone amp is more neutral sounding and pleasant.

My current use of the UH7000 is as a mic preamp / ADC input. I think they are stunning for that. I have two of them now.

Obviously I would use the UH7000 as a DAC/interface if I needed to. I am just comparing them in a critical way here because I just did a shootout.

I used the UH7000 as a DAC for quite a while. It was beating the silver Apollo, the Emotiva DC-1, and the Audient iD22. Then I moved to Clarett, and finally moved to the Topping DX7S that I am using currently.

I guess I'm not the kind of converter user who settles on something. I love these little boxes.

Hey, a buddy who really knows this stuff sent me an interesting comparison of Topping dx7 s vs Pro.

Read on:

So DX7s is good, and in my test sounded about equal to Apollo X DA. I really couldnt tell any difference in a level matched test, but sometimes thought the low end was a hair more focused on the Topping. The I couldnt confirm that because it would sound the same again. UA did a good job on their DA.

The Pro however has some upgrades from the DX7s and it unanimously sounds better than the X to me. Not enough that most people should worry about it, but there is a detail, smoothness, and separation I just love to hear. The output level is only 4Vrms, so the Apollo and your Apogee are way hotter, though it shouldnt matter much in the scheme of things. Plenty of places to get level when you need it.

DX7s uses 2 9038K2M, the 2 channel basic 9038. Two channels summed per side.
Pro uses the full 9038 Pro 8 channel chip. 4 Channels summed per side.

Pro upgrades the crystals to Accusonic Femtosecond clocks, from more vanilla clocks.One for 44.1, one for 48 and one for the ESS chip itslef [the DSP]. Since the ESS is asynchronous, it ignores incoming clock and jitter and is immune to any source jitter. The internal clock is your clock, and the one on the pro is superb. Its cleaner than Apollo and surely the Apogee too, though all should have jitter way below audible levels.

DX7 has 3 large filter caps, and two local regulators with heatsinks, and there isnt a hint of noise.
Pro upgrades that to 4 [overkill] filter caps on the AC, then has ultra low noise regulators, AND THEN uses the ESS 9311 chip which is designed to provide ultimate power to the 9038 chip and simplify design. That means power is double regulated, therfore ultra clean.

DX7s feeds two stages of OPA1612 op-amps [among the best]. One converts the current output of the ESS Dac to a voltage output and is the most critical stage in the design. The second is for the low pass filter. They do a great job.
Pro upgrades this to an un-named current to voltage converter. They scratched the name off and wont tell me what it is. Then that feeds an LM49720 for the filter stage [just as good as the BB and same UA and Apogee use]. Th ereason this matters is that is the first chip that the DAC sees at its output, and this is the main weakness of the ESS 9038. In mot implementations there is a mid frequency rise in distortion called the ESS hump. Whatever they are using for the I-V stge [current to voltage conversion] is correcting this in their design. Thats why they wont share what it is.

All the Toppings Ive seen have pretty dirty circuit boards, and seem like they are done by hand. Its a small problem but I cleaned mine of flux and solder spatter just to be safe. Also, they dont use thermal tape on their heatsinks. they are just screwed on dry so I put some thermal tape on mine. they really dont need to be there but they do help the chips keep stable, at the maybe the cost of a little warmup period.

Other than that the DACs feel very high end, and the design ooks amazing to me. 9038 can run in current or voltage output mode, but ESS says current mode it the best. That means that I-V stage is important to get a voltage output, and topping figured out something.

You can select filters on both of them, but I am using linear phase, fast, which is the basic and most transparent.

Of course, guys like dangerous and crane song go a step further and design their own custom filters to tune their hardware to thir liking Thats why they use such strange mixxes of parts and get stellar sound despit not showing big numbers on spec sheets. Crane Song and Benchmark also upsample everything to 211KHz so that the filters reside outside the audio band, but whether or not thats necessary is a big discussion and nobody agrees.

I will say this, there is nothing at $500 that I think touches the DX7 Pro. Its probably a game of 5-10%, but theres no denying its just a joy to listen on.

My next best recommendation would be the ADI-2 DAC. But its also double the price. Granted you get all of the DSP RME is famous for, and build quality is like a Mercedes. Still, if you dont need the DSP the Topping actually specs a little better even.

Paralleling chips reduces average noise floor. It doesn't do much for signal quality besides creating some small intermodulation and burning power. It's mostly for SNR specsmanship.

The TI ultralow regulators are the ones that I used in my AD/DA box. At the time they were the lowest noise LDO available in the currents I needed. I wonder where they got the idea..

Double regulated doesn't mean ultra clean. It means that it's just double regulated, which also means that if believe you have to do double regulation, your decoupling/layout is poor. Regulators have bandwidths like any other amplifier, because they technically are amplifiers, but with really slow slew rates so that they don't oscillate. Regulators are not designed to "reduce noise", that's what decoupling and filtering is for, and better suited to do. A "low noise" regulator means that it's not contributing much to the noise figure of the output, not that it's actively reducing noise. Overly relying on regulators to snuff noise just makes them burn more power and waste slew rate. More decoupling and better layout makes a larger difference overall.

I/V stage IS probably one of the most important pieces, but not for why most people think. I/V conversion is easy. What people think they need are super wide-bandwidth opamps here, but those opamps typically have very poor input current noise, which helps ruin SNR.. They also typically have poor CMRR and PSRR, which is also hampered by the mistaken belief that multiple stages of regulation somehow absolved the designer of needing large and careful decoupling... And then the fast opamps typically need lower feedback resistor values to stay in a lower noise region, but then burn a lot more power doing so. A regular audio opamp would be absolutely fine here as long as the slew rate is fast enough for the bandwidth.

My guess is that the I/V part with the scratched off markings are actually a low-bandwidth audio opamp like the NE5532 or something and they scratched off the markings so that folks wouldn't be unhappy to see them in a design they're buying mostly on specs and the expectation of faster=better.. I mean, why wouldn't they scratch off all the parts? Because they want you to see the parts you'd be happy to "upgrade" to, but not the parts that everyone deem inferior [even though they're perfectly fine].

Topping just used the common knowledge that I is better output because most V output DACs simply use a shunting resistor to convert I to V. You can still do that, and a lot of DAC purists believe it's better, but more folks believe the active I/V conversion maintains better linearity over frequency. You can also incorporate a stage of LPF into that I/V conversion for free as well and then use the second buffer as the second LPF stage for better alias/image performance.