acalex said:
What's the difference? Thanks!
OK. I'll try to keep it brief and simple as much as possible.
every DAC has to have an analog output stage to amplify weak analog sygnal post D/A conversion into something a line level preamp can wok with (usually 2V RMS). plethora of DACs (and all budget ones) use off-the-shelf industrial opams to do the job. opamps are solid state (transistor) based devices. transistors amplify different frequencies in not to much linear way, i.e. some frequencies are amplified more than others. this is not acceptable in audio where linear freq response is always sought. so one day someone came up with an idea how to solve this problem using negative feedback. negative feedback attenuates frequencies that are amplified too much. it should be noted here that the worse quality transformers used the less linear amplifying characteristics are.
how it works. original sygnal passes through amplifying stage where it's amplified. then it's most likely too much amplified so it needs to be attenuated therefore a part of the amplified signal, but out of phase to the original sygnal, is fed back into the amplifying stage to do the job. this solution is a brilliant idea if we deal with stable sygnals. however, musical sygnal is faaaar from being stable. therefore when the fed back sygnal gets into amplifying stage it tries to regulate something that it should not because new sygnal is now completely different from what it originaly was. all this of course happens very quickly but still, due to the nature of musical sygnal, feedback system is always lagging behind. in one word it distorts the output sygnal in that it makes it ever so slightly different from the original.
that doesn't happen when no feedback is used. what you get in return is better low level musical clues retrieval hence better ambience retrialval, better low level dynamics, truer timbre of instruments and voices and sounds, much bigger depth of sounstage (it has a lot to do with ambience retrieval), more stable and precise positioning of images in soundstage.
if you've got a feedbackless DAC (or other devices) you can be sure of one thing, that high quality, having more linear characteristics components were used in output stage because you can't go without feedback with basic components.
there's another big advantage of ditching feedback. if you use feedback you get a bonus in greatly reduced THD figures. every active electrical component (i.e. transistor) generates harmonics next to what originally was asked to amplify. usually it's second order harmonics and is quite large % of original sygnal. feedback reduces greatly this 2nd order harmonics but it introduces many other, low level harmonics (3rd, 4th, 5th, 6th, 7th...) in return. this is the main reason solid state devices sound "shrill" and "bright" compared to tube gear, where you usually get a lot of 2nd order harmonics and nothing else. human ears don't like high order harmonics, even low level.