Erik Haas

Ohh wow…. I did not know this. I may need to investigate this more since I do have a use case for this when using a input of 44.1. When I do use that source I cannot upsample to DXD.

It would be great if I can get away with a AMD 7950x and a 3060…. But I have a feeling a 4090GPU is needed and those things are still stupid expensive.

AP: Thanks for responding. What are your PC specs and are you running multi-channel convolution through your HQP/NADAC system? My MacMini is too old to process DSD but it has no issues with upsampling everything to DXD and process 6 channels of convoluted audio. I’d like to upgrade to HQP5 and could buy a newer used M1/M2 MacMini to do the same things I can do now. But struggle if spending 3/4k on a Windows PC cable of processing DSD worth it or not. I’m running some very high efficiency speakers that are very transparent to changes, which I would have to imaging DSD could be a welcomed change.

What benefit does NAA have when using a Ravenna System? I could see it on a USB system to keep the main processing PC out of the listening space.

Yeah, thats the issue…. My MacMini from 2012 isnt up to the task since I’m convolving 6 pipeline channels. Though…. I could see if I can play vinyl at 256dsd through the Merging ADC then have HQP to pass the 6 channels. I am doubtful though. Also, isnt Max rate on a mac dsd128? ( i read that on mergings site) Good news about those rumors..

I’m still running my Merging Hapi Mk2. and trying to weigh the benefits of upsampling and if it's worth investing the $$ on a computer build to do so. Would I still be able to use volume control in DSD via HQP? I heard rumors that you dont have the ability to adjust volume with DSD. Thanks

Very Noob question but: I’m currently upsampling PCM sources to DXD 352.8khz with HQPlayer. What are the benefits of upsampling to 256DSD? Thanks

Right on. How do you think performance would be with omitting a 4090 GPU for a period of time. As we know GPUs are still super expensive, would like to soften the financial blow a little bit. In my use case, I'm only processing 6 pipelines with 65k Tap FIR filters. I can up sample to 352.8 DXD with no problem with my old 2012 MacMini but would like to utilize the DSD capabilities with my Merging DAC

Happy to see you are still testing away!! I still have yet to piece together a new PC since other priorities around the house takes precedence. I see you are using 4x 16gb sticks of ram, why did you not go with 2x32GB sticks ? I thought I read something about the 7950x and the 670E motherboards slowing ram speeds when more than 2 ram sticks are added.

Thank you sir!

Looking for some clarification.. if I’m using Audiolense and pulling sweeps @ 96khz does it make sense to create filters at rates higher than this? I have been generating filters @ 352.8khz but not sure if it’s worth doing so.

Which variant is preferred, has there been any multi channel benchmarking between the two? I’ve also read these chips have a weird trait of slowing memory speeds when adding more than two sticks of memory. that being said I’m assuming 64gb ram is plenty? (2x32gb)

Merging Hapi with it's premium ADC card.. I believe the Anubis can do this as well with the same quality. I'm actually running my vinyl rig through my Hapi less a Phono stage and actually sounds better than my McIntosh Phono stage.

Hello Everyone, I'm finally getting around to upgrading my current HQPlayer machine ( 2012 Mac Mini) but still confused on the direction I should go with the build. Objective : Up sampling PCM formats (as low as 16/44.1) to 256DSD Channels to up sample : 5 pipeline channels using convolution 65k taps DAC: Merging Hapi via Ravenna From what I've been reading most everyone is using a traditional 2 channel system which is fine but this is where I get confused because I'm doing a multichannel pipeline for my active xo horn speakers. It seems like the recommendation is 2 cores per channel PLUS a GPU or 4 cores per channel. Right now this is how I am leaning: 1. AMD 7950x + 3090 (or) 4090 GPU 2. AMD 7050x 3D + 3090 (or) 4090 GPU 3. AMD 7960/70x Threadripper 4. Mac Studio M2 Ultra - this option is at the bottom of the list because I dont want to use bootcamp for Audiolense and I believe Jussi mentioned something about Macs not being able to process DSD correctly? (and merging still has yet to certify the M2 Ultra units) ( Should I consider a Intel processor though I'm not sure if e-Core processors are recommended for multi channel?) Again, any guidance would be greatly appreciated.

Nice!!! What CPU are you going to use? I’m still on the fence with a build… Right now Its between a Ryzen 7950x / 7950x3d with a 4090 GPU or a Threadripper 7970x. I’m curious to know if the Threadripper 7970x is up to the task to do multi-channel / 65ktap convolution for 6 channels upsampling from PCM to DSD with out the help of a GPU...