Do DAC chips have their own sound?

[Martinh]

This may be a dumb question but I'll ask it anyway.

Do DAC chips determine the resulting sound character or is it all in the implementation?

The reason I ask is that I just bought a Marantz CD/DVD player and I prefer the sound that it's CS4398 DAC makes compared to my dacmagic. It sounds very analogue and weighty compared to the dacmagic. The Marantz has a discrete output stage, rather than op amps.

If I were to buy a CS4398 based dac from somewhere, would it have that smooth, weighty sound?

Cheers,

Martin

[DSJR]

My experience is that it's really the power supplies and the audio output that really dictates the performance.

The old much loved (by "we") Philips TDA1541 chipset was quite complex in its requirements I think, the best giving great sonics, but when I look inside my CD player I cringe at the loads of components inside - can't really be good...

The QED Digit Opto I bought had a classic "Bitstream" sound on receipt, nice and warm, but a bit clogged. The Positron power supply ran hot and I found that two of the internal caps were leaking. Replacing these and later doing the twin-supply mod made a stupid difference sonically I found. First, the Positron supply ran totally cool under the reduced load and I'm sure the now humble 5532 op-amp output improved with its own dedicated supply. Now, the twin-supply Digit sounds totally different, having an ease lacking in the standard version, the ripe bass being all but banished (made even better I thought with the horrible cheap-n-nasty output caps replaced with Muse type ones).

These days, with everything done for you in the chipsets, it's difficult (thank goodness) for manufacturers to ruin the performance. These chips are a very few quid too, so why a DAC needs to cost much more than £500 from a dealer, unless the casework is lovingly crafted at huge cost, totally beats me.

That's part of my experience anyway..

[AlfaGTV]

I think Dave has a point there, you would'nt just get along with merely the DAC chip! :scratch:
And it's more up to the conditions the manufacturer of the whole unit provides to the DAC chip. Such as power,clocks, buffer stages etc. etc.
Given good working conditions even the oooold Tda-1541 can sound wonderful. But so can more modern DACs also, and i do think it's easier for a constructor today to achieve a good sounding DAC than it was 20 years ago.

Tbh the "big" audio differences with digital sources has mostly passed me by...
I recently hooked up an old Nakamichi CD-4 from mid 90's and it sounds just fine to me! Even when compared to my Marantz SA15S1 from a couple of years ago.

But, as always, YMMV

BR Mike

[dave2010]

Isn't that a bit simplistic? There are surely quite a number of different designs for DACs. Some "simply" convert the PCM data into an equivalent current flow. This is then converted to a voltage source in some designs, and this is often achieved using op amps.

I'd have to think a bit more about this, and look at books and articles on DAC design. Simply attributing the performance to one or more op amps without considering the circuits used seems hit and miss to me, though of course if some people make discoveries by experimentations that's OK, but it would be a lot better if experimentation could be driven by theories and understanding.

[AlexM]

Power supplies have an influence, as do clock design and implementation, I/V conversion, analogue stages etc.

In terms of the Dac, I think that the combination of Dac and digital filter does have a signature, more filter than Dac, I think. In many chips, the Dac is integrated with an upsampling digital filter, so you are talking about the sonic signature of the comination. Having said that, I don't think that you can make any real assumptions about the sound of a DAC or CD player just on the basis of the chips it is using. Some would have you believe that any competently designed DAC that measures similarly will sound the same... I'm not in that camp, but I do believe that other components have a greater variation in terms of SQ.

Cheers,
A.

[DSJR]

I have it on good authority that TODAY'S DAC chips do most of it internally, needing very little externally at all, including filtering. Stan has already shown that a ten year old design doesn't really need the output op-amp if a short length of cable is to be used into a sensitive enough line stage (I wonder how a cap-less direct output from the chip itself if one of his DAC's would be into a vintage Quad or Naim with high sensitivity and copious amounts of band-limiting?).

These days, I'm absolutely convinced that adding shedloads of electronics to a DAC can only make things worse and add severely to the cost. I do appreciate the power supplies and casework cost, but a well trained engineer should be able to make something that doesn't cost loads of dosh, surely? Thing is, we audio-nuts are pre-conditioned to spend ludicrous amounts of money for so-called improvements which can sometimes be bad design in the first place (I'm thinking of all the "improvements" made to the expensive ARC Reference preamps - weren't they "right" in the first place?????)

[Reid Malenfant]

At the end of the day no matter what does whatever, you are listening to a power supply. The final factor in how anything is going to sound will be governed by the quality of the PSU. Different DACs may well sound slightly different given the implementation & what has been mentioned by other members, but the final arbiter on quality is what is feeding the circuit power - the PSU.

If you think otherwise then you need to do a bit of reading & thinking ;)

[dave2010]

Again, surely it depends ... Results will not be good with a poor power supply, but if we assume that the supply is good, then other factors can clearly affect the sound. The power supply should be stable, and not induce obvious periodic signals into the audio (trivial example - mains hum), and it should also be as stable as possible under load. With a current summing DAC the power supply should be capable of rapidly supplying current under load, and the voltage levels should not change significantly. I suspect that with many power supplies that this is not a problem, BUT the degree of stability required to deliver even 16 bit resolution has to be better than 0.002% under possibly quite varying degrees of load. If we imagine that 24 bit resolution is something to aim for (which may be quesionable), then the required power supply stability should be around 0.00001% under all load conditions.

It could well be that few power supplies are able to operate at such absolute levels of stability, but that most do not induce perceived noise disturbances. This is likely to depend on the frequency spectrum of any noise introduced by power supply variations under loads.

It's perhaps likely that power supply induced noise due to load effects is going to be more obvious with high quality DACs. It goes without saying that power supplies should also isolate the equipment being driven from externally generated noise, such as mains borne, or radiated noise.

[Martinh]

Thanks chaps,

It looks like my search for the ideal DAC isn't going to be an easy one then :(

I guess that the output that I'm getting from my Marantz player wouldn't measure very well - a distinct emphasis on the bass and a rolled off top end. Hence the "analogue-ish" sound.

Maybe I should try something tweakable, like a Caiman? Can you change the sonic balance with a few component changes?

Or maybe look at something else like different speakers?

What do you think?

Martin

[bobbasrah]

Can you change the sonic balance with a few component changes?

Or maybe look at something else like different speakers?

What do you think?

Martin

There are many devices available ranging from commercial to simple DIY Martin, 20 bucks to mega....:eek:
Have you googled your player to see if there are mods or tweaks? Worth a look as at the very least you can improvements from what you already have....;)

Just my 2c

[dave2010]

There are many devices available ranging from commercial to simple DIY Martin, 20 bucks to mega....:eek:
Have you googled your player to see if there are mods or tweaks? Worth a look as at the very least you can improvements from what you already have....;)

Just my 2cIf you go fully digital, with digital rips, you could implement digital filters, both to alter the sound from the recording, and perhaps also to modify your room sound. If you're good at programming this might cost very little, otherwise the time and effort will cost you the equivalent of a small fortune. Not really a very serious proposition for most people.

The thing about digital filters is that once you've workd out the filters you want, they will not change, though of course the analogue systems around the rest of your kit can change due to heat, interference etc. It is possible to devise very precise digital filters. The big problem is knowing what you want to adjust and/or correct.

Hamming's book has just about all you need to get started -
http://www.amazon.com/Digital-Filters-Dover-Mechanical-Engineering/dp/048665088X

[dave2010]

This presentation on DA circuit design may be helpful to some - http://www.ecse.rpi.edu/homepages/ieee/presentations/RPI_DAC_Lecture_Oct_08.pdf