Moderators: Sharp, X-Trade, Pepperpotty, karmathanever
I removed on factory R3 design the passive DA low pass filter, why? Because it use a 18 nf capacitor with a series resistance of 680 ohms and such values overloads the analog DAC output signal swing, so I use instead an active LPF solution. Basically on any DA converter the output impedance is a bit high (or weak in output current) to drive a passive LPF without lost or attenuation of audible audio signal (due the high capacitive load), also the series 680 ohms resistor and 18 nf capacitor attenuates partially the audible spectrum on high frequencies (aprox 7,5khz and above). The only advantage of a passive LPF is the reduction of production costs due the reduction of electronics components. Researching datasheets for high quality designs DAC manufacturers recommends the inclusion of an active LPF, the small brown PCB board that I added on the circuitry changes. The board performs not only the DA aliases filtering process, also it acts as an audio driver that adapts impedances between stages. At the same time I added an active EQ network on final lineout stage to increase the lows and highs frequencies, following approximately Fletcher-Munson human ear contour theory (is not an exact match of course) and to compensate the LPF roll off attenuation.Re-Member wrote:If you compare analog to digital, then yes, there's always going to be a difference. All digital synthesizers are prone to aliasing and everything you've described here is the anti-aliasing filter designed within to avoid most of that noise. While it's great that you can circuit bend the R3 to remove that filter, removing the filter is opening the door for aliasing to occur. I have the JP-8000 and while it delivers a much thicker tone, the aliasing noise pops up a lot.
The R3 has a narrower sweet spot, but it's workable. As I said, there's an EQ section to help balance out the sound better. Also in the effects section, there's an additional 2-band EQ which actually pumps out more low end than the main EQ. The Korg patches are really not a good example to compare to real analog because most of them are layered with reverb or delay effects and have shoddy EQ settings.
But honestly, a synth really doesn't need a bunch of low-end unless you are playing in a live situation through an amplifier. The thicker the sound, the more difficult it is to mix in a recording. If I were to take that sample you posted to be mixed with other instruments and then professionally mastered to disc, half of that low-end would end up being rolled off. It's nice to have, yes, but in a mix there is such a thing as having too much.
Yes, they sound great, but the point here is that the R3 is already capable of sounding great without needing to modify it. This thread was about someone asking for programming tips, but basically you're just telling everyone that unless they go out perform the exact circuit bend you did, the R3 isn't capable of sounding analog. Personally, I don't think it's a good idea to suggest such a thing since it puts people at risk of opening up their synths, damaging something and voiding their warranty.alfredokiwi wrote:To summarize the audio quality difference after circuit improvements speaks from themselves.