With the capabilities your DAC has and the bit perfect transport that Roon does, it’s not surprising to me that you don’t hear any differences. I get and respect that a lot of people say they hear differences. I spent a good amount of time measuring a few different DACs with a variety of tones, multi-tones, etc. running through Roon/Raspi/DAC and over a few different transport mechanisms to satisfy my curiosity on the quality of the digital path from Roon to DAC output. Sometimes it’s fun to play with high sample rates but it isn’t audibly different to me. It’s one thing I don’t have to worry about.
1 Like
HQP isn’t bit perfect by design though - it can’t be if it upsamples - so this is a different topic/discussion.
1 Like
HQPlayer has three primary functions all combined into one:
-
a replacement for DAC’s digital filters and delta-sigma modulator, without restrictions in available processing power
-
a big general purpose DSP engine for running digital room/headphone correction, cross-feed, cross-over filters for multi-way speaker systems (including DSD sources, without any intermediate PCM downconversion etc)
-
a player / endpoint for playing content from local library and streaming sources, either native or for example over UPnP (HQPlayer is also a UPnP Renderer)
So essentially a software replacement/solution for all those upsampler (dCS/Esoteric/Chord) and streamer devices out there. With the difference that you can run it on a regular computer which makes it more power and and cost efficient at the same time.
Then finer details include things like DAC corrections and algorithms to correct errors introduced for example by the ADC chips used for recording.
It is just a long term project I started in 1998 when I encountered too many limitations with the available hardware solutions (I designed quite a bunch of DACs in early 90’s). There is still a long way to go in terms of future as well, my TODO list is growing faster than it is possible for me to implement.
12 Likes
A belated thank you very much to all those who took the time to reply to my original question (especially Jussi). All your replies were helpful, and instructive.
There are more useful features of hqplayer beyond the three you listed.
in my case, connecting my turntable to a ADC usb interface, set hqplayer input, then output the vinyl album to any or all four of my systems through NAA AND incorporate audiolense convolution for room correction and subsonic filter for each system…
3 Likes
I always find it interesting to see people who know they have limited hearing ability use science authority to critique folks exploring their own human potential. I don’t understand the point.
1 Like
Clean up the digital chain? There is no distortion in digital chain. Bits are ones and zeros. There is nothing between zero and one. That is why they call digital data “binary” only two states. If a bit is lost this is known via the check sum and the lost bit is re-transmitted.
3 Likes
I’d say, your post qualifies as axe grinding, since I’m definitely not critiquing anyone.
I merely responded to @Calum_Mackay’s question to aid in finding peace of mind, offering my experience in rigid testing, using proper and verifiable, readily available tools.
I started doing such decades ago, long before age related deficiencies became apparent.
On the other hand, there’s no way of verifying anyone’s alleged super-human powers, as long as they’re purely anecdotal.
4 Likes
Well, the data Jussi is showing is scientifically rigorous: there are differences in oversampling algorithms and their results. So there’s no debate about that, I assume. The question is whether it’s audible and significant.
What you’re saying is that you’ve tested the limits of human hearing, but there’s ample evidence in all studies of sensory perception that there are differences between humans. Some of that is innate, some of it is training. Most differences are relatively small, but there are always outliers and they can be rather specific. One can say that certain claims are implausible, but it’s often hard to exclude the possibility and it shouldn’t be done without proper research into the abilities if the subject that made the claim.
The question remains whether a small difference that may be audible, is significant. And in my opinion that’s a subjective qualification and can be discussed, but is hard to prove. There is some research in neuroscience that’s quite interesting in this respect and that actually shows qualitative differences in signal processing between humans.
2 Likes
It is always reassuring when someone finally reaches the stage of their audio journey where they can confidently explain that bits are ones and zeros. We were getting dangerously close to discussing what happens after those bits arrive at the DAC, so the intervention was certainly timely.
That said, the learning path is a treacherous one, particularly for those of us for whom English is not a first language. “Cleaning up the digital chain” is a perfectly valid expression.
The response would therefore be highly relevant if anyone had suggested that HQPlayer repairs corrupted bits. Since nobody did, it serves mainly as a reminder that understanding the words in a sentence and understanding the sentence itself are not always the same skill.
9 Likes
These days it’s quite easy to educate oneself on the matter and reliably find one’s individual limits by ways of eliminating interfering biases and such.
Skipping on such rigorous testing, yet offering anecdotal opinions, just keeps the circle of confusion - as Floyd E. Toole coined the term - spinning without any true value for advancing the understanding.
Beware of the hi-end audio industry feeding FOMO to keep their businesses running.
I’ll just sit back in front of my room-acoustics DSP’ed DIY-Full-Range-Line-Source speakers, fed by non hi-end-brand-named electronics to indulge in subjectively and objectively transparent 44.1kHz/16bit files and streams, as rendered by the DAC’s standard sharp filter.
1 Like
I appreciate that this is the approach that works best for you.
At the same time, I think many of us here combine room acoustic treatment and DSP and still find value in using HQPlayer instead of relying solely on a DAC’s internal filters and upsampling. Once the room and the big variables are handled, the quality of the digital filtering and modulation can still make a meaningful difference — to those who can hear the difference.
3 Likes
I think the main issue is that people are upset that their hearing is not as good as others and so must resort to these kind of arguments of using specific scientific authority to make themselves feel better.
Are you being dishonest or do you not understand the words you write? Who is axe grinding here?
Isn’t it more that people are amused that there are people who think that their hearing is so good that they can hear things that do not even exist?
1 Like
Perhaps it’s both, and both assume things about the other. I understand the critique of golden ears and over priced audio gear, but it seems those making that critique more often than not also claim to have limited hearing ability. I don’t understand the point of doing that in this thread where folks are looking for ways to make audible and significant changes to what they hear for themselves. It would be one thing to say here are some tools (assuming you find them to be valuable) that might help you hear differences, but to make it all about the subjective/objective debate and take a side seems pointless.
Everyone has limited hearing ability. Some people act as if they don’t.
3 Likes
I think it helps to take a different perspective. Take a CD and play it through a dac that works on 44.1kHz and 16 bits. No problem, fine audio.
Play that same CD in a dac that works on 48kHz and 16 bits. Now you have to oversample the source. It’s possible to do this perfectly, but only with unlimited time and processing power. You don’t have that, so you create an algorithm that’s fast, but also introduces some distortion. This distortion is in those 16 bits, the audible range.
The solution (other than buying a different dac) is to get more processing power to lower the amount of distortion and add bits to move the distortion out of the audible range. That’s what HQplayer does.
Now maybe you’re fine with the oversampling distortion in the 16 bits, but you add processing like room correction or an active crossover. Not uncommon nowadays. This processing can increase the distortion and make it audible.
Again, to solve this, use a tool like HQplayer to reduce the distortion in the oversampling step, so the total of the distortion decreases and is moved out of the audible range.
This basic scenario doesn’t require anything superhuman or a top level system. My car audio distorts when feeding it 44.1kHz audio, it upsamples and does some processing to make the sound more spacious or whatever and maybe it does more evil stuff. It’s awful. Feeding it 48kHz sources makes it a bit less awful. I reliably tested that single blind. Not enough for science, enough for me.
1 Like
Science is a practice or a method to understand what is true about the world. It’s easy to use a scientific authority or method that could be inaccurate or distorted to try to understand what is true. I’d say you are doing science and using your discernment capacity to determine whether what you hear sounds good to you. I doubt anyone on this thread knows the limits of all humans’ capacities. Why not let people who want to explore what’s possible do that and stop claiming to know what’s true about what another person can hear especially if you know your own capacity is limited?