Page not found | Stereophile.com
It can be speculated that the reason OS mode sounded dark and muddy in HoloAudio was that the designer tried to hide the harshness through the filters, much like Soren did in 1.19/1.20 linear filters. I haven't gone back to 1.19 in a while but 1.20 had zero harshness yet hightly distorted (by mistake most likely...) with an enormous bass emphasis. Not sure why their NOS sounds good though, maybe hardware filtering?
I really do like the idea of "linear compensation" of some sort, ideally in software, but maybe it's ignoring the "elephant in the room".
I wonder what dac1541 sounds like with HQPlayer. The giant power supply circuits in HoloAudio seems distasteful - does it really matter?
P.S. Maybe the PSU is audible... but still there is something to love about the dac1541 simplicity and the giant screw you to the rest of the HiFi industry.
It can be speculated that the reason OS mode sounded dark and muddy in HoloAudio was that the designer tried to hide the harshness through the filters, much like Soren did in 1.19/1.20 linear filters. I haven't gone back to 1.19 in a while but 1.20 had zero harshness yet hightly distorted (by mistake most likely...) with an enormous bass emphasis. Not sure why their NOS sounds good though, maybe hardware filtering?
I really do like the idea of "linear compensation" of some sort, ideally in software, but maybe it's ignoring the "elephant in the room".
I wonder what dac1541 sounds like with HQPlayer. The giant power supply circuits in HoloAudio seems distasteful - does it really matter?
P.S. Maybe the PSU is audible... but still there is something to love about the dac1541 simplicity and the giant screw you to the rest of the HiFi industry.
Last edited:
The developer "Bogi" created also another solution that seems more robust:
Manage HQPlayer playlist from Foobar2000
The latest script and manual pdf can be found in this post.
I bet what dac1541 really needs is a bit more vref capacitance... completely unfounded claim of course
The point is not about whether it is beneficial. Ask yourself how many ppl in HiFi is willing to use SMPS in their high-end products and is it really justified? I would never want something like a Audiogd product in my room.
Last edited:
That's 2015 before the other post right? I'll check after a much-delayed lunch.. Also zero stability problem on my end so far. Works flawlessly except for the sampling rate change issue. Lots of development efforts needed to get that to function automatically it seems.
Btw just out of curiosity, why not use a shunt regulator then?... Constant and low impedance over all frequencies. I was also bothered by the fancy capacitors they advertised in their level3 version, but I'm rather heavily biased in that regard I have to admit... There may be something to the expensive caps i just don't know
But his latest SendToHQP v1.04 script is later: 20 oct 2016
This is also the solution that he uses according to this post.
Last edited:
Thanks so much! I'll see if I can set it up quickly
Btw, your speakers look awesome! Must've taken a ton of efforts. Great work!
The manual is very well written, going through it now. One thing though, I'm not sure if I buy into RAM disk being better than SATA... I actually have most of my 300GB library on an SD card. Not because it sounds better, but that NVMe SSD is expensive and I don't want the hassle of upgrading to 1TB... That said, I do have 32GB ram on my laptop so even if I must use RAM disk it should still work...
Last edited:
This is so true. And should be your first order of business. ATM you are trying to cure hardware harshness with software.
Why would there be harshness in opamps?... I'm expecting less clarity than what is possible though
You really need to just go try it. This approach is groundbreaking. I give you my word
And most importantly, you need to first explain why an absolutely more accurate upsampling would degrade the sound in a way to mask problems "caused by opamp...". Let's not do this again....
Last edited:
Ummm... I had no idea how it would work in practice after I read the manual (though I was sure I could follow the steps and click the buttons...), so I watched the video. Are you sure it's more convenient....?
I admit it's a great way to get DSD files to play properly though, not that I have any in my library...Edit: It's easier than the video made it seem. You'd select tracks in foobar2000 and then one-click a shortcut to a batch command file and it puts everything in HQPlayer playlist for you. Only drawback is you'd still have to interact with the HQPlayer GUI (sorry Signalyst...). With the UPnP solution, unless you change sampling rate, it's set and forget. You can control everything, including track seek, inside foobar2000 and just leave HQPlayer running. When you have to change sampling rate, you ctrl+p in foobar, toggle one dropdown option, go to HQPlayer to hit play again because it automatically stopped upon sampling rate change, then everything is ready for you in foobar2000...
Last edited:
Thanks, I'm finishing version 2 of my speakers, 99% done
32GB is plenty! That won't be a problem for setting up a RAM disk.
I thought it should be more convenient and that it eliminates the problem of manually switching the sample rate in upnp
Last edited:
Thanks, I'm finishing version 2 of my speakers, 99% done
32GB is plenty! That won't be a problem for setting up a RAM disk.
Nice! Check my edit in the last post. I'll stick with UPnP solution since I don't need DSD playback... Also conversion would be less real-time. I'm keeping the manual in case I decide to use it at some point though.
Is there a way that you could make your speaker design into a kit? Or is it generally too difficult in diy?...
I don't see myself making a kit of my speakers, way too much work.
There's a lot of woodwork to do, I won't count the hours that I've put in it
In a new thread I'll share the crossover, the cabinet dimensions and construction.
OK, upnp is then the most seamless one.
That's fair. At this point it's more than just a pair of speaker and a hobby... Plus it's a huge cost saving, whereas in our case I would advise ppl to carefully consider the option of just getting a dac1541...
I hate to say this but TNT might've been right again. I still believe that HQPlayer is much better than SoX, but it seems closed-form-m (with 16 million taps... though not warmly recommended by the author) is even less harsh sounding. I was listening to Julliard quartet's Mendelssohn string quartet no.2 and poly-sinc-xtr causes a bit of headache with strings on the very high notes, and there were a lot of high notes.
Btw, very good news. It must've been a glitch last night with foobar, maybe I was using the wrong sampling rate over UPnP which forced sampling rate conversion on the fly in FB2K. The CPU usage now in FB2k is about 0.2% with 44.1Khz (slightly more with higher sampling rates). So no huge computation overhead with this method! Also note that my CPU clock is still quite low @1.8~2.3Ghz so really nothing to worry about at all. The cooling fans haven't even been triggered...
Last edited:
Seriously? Do you think you can perfect your DAC in a way the imperfections present in the soundfile go away? You can only control the final piece in a very long chain.
So again, on the recording / mixing / mastering side of things there were ADC and DACs involved with significantly less processing power and a need for realtime performance, so you can bet their filters don't use a million taps. Yet many suceed in sounding great. Not all of them, but for more reasons than the filters....
I've got 12 bit studio samplers from the 80s that sound really nice... It can surely be done in an FPGA in real time.
I would buy a SAR chip based ADC designed by Soeren. It could be a game changer.
I'll look into that.
Yep... It's been a really interesting hour trying different filters again, though I don't think I made much progress in getting a better sounding dac...
Our desire to get better sounds and to know how to do it is great. I don’t believe that there is anything cringeworthy about it, though it could lead us into trouble. Not knowing isn’t easy, but it’s not always justified to remain contentedly ignorant. While I don’t have too much sympathy for those who are too weak to take action, I do recognize the enjoyment that I inevitably derive from acting more vigorously. With this in mind, we may be able to prevent ourselves from giving up far too easily on our efforts.
Before I started testing filters, I was so sure that I knew the traps of audiophilia. I thought that even if the difference between a cloth hanger and the silver braided handcrafted cable is audible (which I still don’t believe by the way), it must be trivial. I thought that the only explanation was an utter weakness of mind for anyone except the most hopeless audiophile who can perfectly justify his thousand-dollar purchases based on nothing but how he felt when he wrote the check. Or perhaps it was the pleasure we get when we appear to succeed in our efforts, and the fear of being thought of as failure, of admitting that we wasted our time, that forced us to lie?
What I didn’t expect was how strongly I would believe in the differences I hear from the filters. It’s actually hard to be a true audiophile on a DIY forum. A/B tests in hardware often cost more than the mod itself. If we know it to be more or less harmless, we certainly wouldn’t take the time to remove it, and leave it at YMMV. What a perfect escape clause, by the way. Filters, however, are a different poison. No more difficult is it for us to switch a filter, than it is for the archetypical audiophile to swap his headphone cables. The circumstances that prevent us from experiencing the conviction that a true audiophile does about his investments after extensive A/B tests no longer hold true. I thought a good many audiophiles must’ve said to themselves, “It doesn’t matter if it’s true, as long as it satisfies me. This $500 silver-coated doubly-insulated extra-thick USB cable definitely sounds clearer to me than that ugly-looking one from Walmart.”, and I knew, that I felt nothing like it when I asked people to try software up-sampling, with the anticipation of great results. So, I concluded, my judgement must be somewhat reliable, especially given how sure I was about it. In hindsight, I didn’t keep in mind the possibility that our sensory perception and auditory memory are inherently imperfect. The truth, as it appears to me, is that no book knowledge will be enough to prepare you. You will believe; you will be sure, even if you know that you could be wrong. People buy exotic cables not simply because they are addicted to imprudent spending, at least not everyone. Audiophiles truly expect wonderful rewards. And it’s always easier to become an audiophile than you think.
I just did another round of tests and I now think that the filters most likely didn’t help with the harshness. One possible explanation for my failure might be provided by the observation that after listening to the same fragment a few times, I started hearing things to be softer than they are. After I tried to control for this by taking a break before two tests, the differences seem to be less than I remembered. Granted, it’s late and I’m tired, but I don’t think it makes my hearing much less or more reliable than it was yesterday. I never believed in using hearing tests only to tell which equipment is better, especially when they could be close in performance, since it’s not inconceivable that we might easily mistake distortions and coverups as perfection. “Technical gold-digging” sounds much more reliable as it only requires the listener to judge whether the improvement is audible. But given the fundamental unreliability of our hearing that may be caused by unconscious factors far beyond our personal control, it’s not as reliable as one would believe, but is made all the more dangerous because of it. I admit at one point I was asking myself if I was really hearing a difference between non-upsampled fb2k and HQP. If there’s a way to produce insanity, audiophilia should be high on the list. If something is proven to be beneficial, however trivial it might be, wouldn’t we see only people who claim to have heard an improvement and people who didn’t? Did we try to reject the null hypothesis? I did switch back to stock 1.19 filters, however, after I thought the EQHQv5 is low on bass compared to the software filters. The 1.19 linear filter was closer to the software solution, I think. Did the software filters offer an improvement in dynamic range compared to the stock filters? It is said that proper dithering increases perceived dynamics, but I worry that we don’t know for sure. For now, I’ll keep my HQPlayer setup. But I’ll look into proper ABX test solutions to make the final decision.
Does this mean that our listening test results are not worth anything? I don’t think so, nor are we able to accomplish much sometimes if we must rely on complete and objective data. But given their relatively low reliability, they probably should matter less than we often believe. Information entropy is not kind.
On a more positive note, it seems I’ve learned to fully enjoy this DAC.
P.S. Jogi, I do hope your DAC has seen more actual improvements than mine. Perhaps your system has more potential! But uncertainty gets us all sometimes, I wouldn't take it hard.
P.P.S. I found what looks like a fascinating read. I think it'll make my day :-D Some of it might still be BS though, as with any academic work
Last edited:
"In the translation of an analog electrical signal into binary code, performance is robbed of a part of its being, as digital sampling of a signal can only occur at a rate less than infinite— in the case of compact-disk technology, 44.1 thousand times per second."
It's a fun and therapeutic read, but I don't think people here believe what he says we should believe. He must have an interesting circle of friends.... In my opinion he also seriously fell short of the promises in the intro. Here's his webpage: Faculty Profile > USC Dana and David Dornsife College of Letters, Arts and Sciences . Btw, teaching professors in the US are quite a bit lower on the academic rank than regular research professors and don't usually have tenure. Not that I wouldn't be interested in his class...
"To know perfectly is to know what can and cannot be known, to know the ambiguity of particular experience."
I especially like the potential self-contradictory interpretation of the latter half. Lovely.
It's a fun and therapeutic read, but I don't think people here believe what he says we should believe. He must have an interesting circle of friends.... In my opinion he also seriously fell short of the promises in the intro. Here's his webpage: Faculty Profile > USC Dana and David Dornsife College of Letters, Arts and Sciences . Btw, teaching professors in the US are quite a bit lower on the academic rank than regular research professors and don't usually have tenure. Not that I wouldn't be interested in his class...
"To know perfectly is to know what can and cannot be known, to know the ambiguity of particular experience."
I especially like the potential self-contradictory interpretation of the latter half. Lovely.
Last edited:
ynmichael... you must first hear some for you well known record on best system that is available for you and put every single note into your mind... if you want to tweak your gear. That process is not so easy. Of course this presentation should be as real for our ears as possible (what is real in right way?). Therefore in our demo rooms we have one piano, and in another one 10pcs drum set. ok manly beacuse of a not enough space but also to try sometimes how real instrument is hearing in that room. At last show we had interesting comparrison that one famous saxophonist brought his 24bit-96khz studio master records and we compare how they sounds in compare to live play in the same room. He even play in the middle of record to hear the difference.
With headphones...no way to hear the real modifying result...i can*t and tried several times. I use both ... ears and scope mesurements and good memory.
You don*t hear differences in cables? you are lucky or not, depends.
With headphones...no way to hear the real modifying result...i can*t and tried several times. I use both ... ears and scope mesurements and good memory.
You don*t hear differences in cables? you are lucky or not, depends.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.