ESS Sabre Reference DAC (8-channel)

[Russ White]

Originally posted by rossl
Most of the inexpensive transformers do color the sound. Good sounding transformers cost a lot.

The demo board as I have it now is very transparent.

I am not talking about the cheap ones, nobody would even consider using those. I am sure what you have sounds fine, but I would not want to add the distortion and loss of linearity you get with even the best transformers.

I have worked with a lot of pro gear (I am in Nashville, and a lot of my buddies are studio engineers and artists) , and hifi stuff too. Transformers are used a lot, but not really (usually) because they sound "better", more because they allow you get around technical issues. In that role they excel.

Adding a transformer always increases distortion, and it wil do some nasty things to linear performance(to some degree depending on quality) at the top and bottom. The price is irrelevant. It may simply lessen the degree of bad effects, but its still there.

By all means, if you like transformers on the output use them, but I really don't think they are the all around "best" solution all the time. I don't mean to discourage you.

Cheers!
Russ

[rossl]

Originally posted by Russ White

Transformers are used a lot, but not really (usually) because they sound "better", more because they allow you get around technical issues. In that role they excel.

Yes, the technical issue is unwanted ground current in the wires.


I have the Behringer digital eq before the DAC. I measure the room with a microphone. Digital eq corrects for room standing waves and non-linearities in the equipment and speakers.

These are measures that most normal people wouldn't take.

[flshzug]

by what means you correct for standing waves? High pass filter ?

its real bad i get these topic notifiers , even though rossl is ignored ...

[Russ White]

There are much better ways to insure there are no unwanted ground currents than to resort to a transformer. Eliminate them by design.

Now if you are stuck with them, then you may not have a choice. In which case you can easily add them.

My opinion is If that if you must use transformers then you must, but if you don't need to, then for the sake of best performance don't.

Cheers!
Russ

[Andre Visser]

Originally posted by Russ White
My opinion is If that if you must use transformers then you must, but if you don't need to, then for the sake of best performance don't.

Cheers!
Russ

I agree, transformers are OK for power supplies.

André

[Spartacus]

The ESS output stage using AD797s is actually very good, but so is my transformer stage. Also, I've tested with and without the transformer, and can say that it's pretty transparent, but really the differences between the various approaches are very small in comparison to other stuff that can be done.

The AD797s are -120dB THD, far more linear than any transformer - but that doesn't translate to any audible advantage.

[rossl]

The audible advantage, that is the key.

The typical speakers and listening room are far more non-linear than any of our schemes for coupling that we have discussed.

There are differences in distortions. I do not find the distortion from a good transformer objectionable. Some people want to hear the distortion from tubes. I find that some op-amp circuits sound objectionable even though they may be linear.

I doubt if we all are ever going to agree, but the passive output I am listening to now sounds very good to me.

[Cappy]

Ross,

I agree. One man's neutral is another man's colored.

I seem to be particularly sensitive to transistor colorations. I've been in situations where other people call a component neutral and I hear bleached sound missing correct harmonics and a mechanical treble.

Also, the brain is good at filtering out distortion. If people have been listening to transistors for many years, for them the sound may indeed be neutral. There are probably many neuron pathways that were set up in the brain over time to handle it, even! Same goes for me filtering out tube distortion.

[jcx]

at least read the transformer line driver section of Walt's book:

http://waltjung.org/PDFs/ADI_2002_Se..._Drivers_I.pdf

6.69+, p27+

[fierce_freak]

Originally posted by Russ White
I have a circuit which both sounds and measures very neutral and incredibly low THD (measured), and it takes advantage of the very high DNR of the chip. GND loops have not been an issue at all, and the balanced (symmetrical) nature of the circuit rejects common mode noise very nicely.

Does this mean you've had a chance to get one of your boards up and running, or is this in reference to a past circuit?