Just in case we are having a language problem, may I ask if you know what the word "troubleshoot" means? Google says the translation to Spanish would be, "solucionar problemas."
However, that doesn't seem exactly correct. "solve problems" is not the same as "troubleshoot." Maybe more like, "prueba para encontrar el problema."
Yeah I want to make my tailored filter specífic for me (as I do with my pasive LPF) to use in the FPGA
I know because I make my own 3 way loudspeakers xover, the 4th order LPF for noDac & lately change the filters for Chipamp....
Also noted playing with Daphile + i7 laptop differences of SQ using differents DSD upsampling to PCM, not always the high upsampling sounds better...because the power of the laptop processor make an important role in SQ, could be if you use maximum power of laptop to make for example DSD256 the laptop is running near the límit of the processor so not working in optimal conditions also surely PSU working at límit or near also not optimal.
Thanks for those measurements. Difficult to compare SDM order performance as it's 4th order DSD256 and 6th order DSD128.
Did not know about those LVDS outputs on the FPGA, interesting! You probably already know but just to be on the safe side, there are ample I2S-to-HDMI kits available to do it today without using the FPGA LVDS outputs. AliExpress has a couple of perfectly functional ones, though they don't reclock. IanCanada's is better but in my opinion overpriced. When I'm done designing my own reclocking HDMI output module to go with PCM2DSD I'll post it here.
@merlin el mago kindly may I ask to keep posts together? Makes reading and notifications easier.
I will try.
Sometimes I forgot something & I write a new post, sorry for the inconvenience.
Sometimes I forgot something & I write a new post, sorry for the inconvenience.
Are there any plans to make a version with a bigger FPGA? I presume that less compromises would need to be made then? Though at a higher cost of course...
PJotr25 and @olo111 have given the diy community a whole lot of DSD converter performance for free. Its been a lot of work for them. What if they did something like Iancanada does and sold diy boards with a pre-programmed and more powerful FPGA included? Then commercial dac companies couldn't steal their IP so easily without at least paying a little something for it. Would that be a fair way to compensate the two gentlemen for what would be a lot more hard work on their part?
That's not the point Mark. Everybody agree to give big thanks to both persons.
My vote for a more powerfull FPGA.
Also the possibility to make own filters if someone could tech will be great to learn following the DIY spirit.
Also the possibility to make own filters if someone could tech will be great to learn following the DIY spirit.
I do not know what computational needs the conversion algorithms you use have, but I have my eye on this architecture:
https://beagleboard.org/ai-64
It has a ton of floating point / DSP capabilities. And also quite some generic ARM performance. So it has the potential to be a great platform for network based audio, upsampling, EQ/crossover multi channel processing and hopefully also PCM to DSD conversion...
https://beagleboard.org/ai-64
It has a ton of floating point / DSP capabilities. And also quite some generic ARM performance. So it has the potential to be a great platform for network based audio, upsampling, EQ/crossover multi channel processing and hopefully also PCM to DSD conversion...
Some people may have no idea of the difficulty of some this. Optimizing dither involved a great deal of experimental work, since there is no good theory for dithering a one-bit modulator. Just a heads up. I will leave it at that.