Jump to content
IGNORED

HQ Player


Recommended Posts

On 1/12/2022 at 6:16 AM, Miska said:

 

Yes. Dealing with settings and such is web browser access + HQPlayer Client. So just like Desktop, but instead of VNC you use browser.

 

Version changes is that you just flash new version with Etcher.

 

 

Anyone knows where to find the latest version of Etcher for macOS? The download link at https://www.balena.io/etcher/ for the macOS DMG leads to a '404' at Github (the download subdirectory to which the former link refers doesn't exist anymore), and the download link at https://github.com/balena-io/etcher leads back to the broken link at https://www.balena.io/etcher/ - a nice vicious circle 🙄 When searching the web for Etcher, I get across all sorts of download sites, but I rather get my DMGs from a reliable source, and hence my question. TIA!

Link to comment
  • 1 month later...

Last weekend, I added a Fitlet2 with SFP+ optical port as a HQPlayer NAA to my home network and, necessarily, an additional switch. I can only confirm what many others have said before on this forum, and that is that adding an NAA to the audio setup and isolating it by means of fibre results in a significant boost in SQ. While I am hence very much pleased with the result, there’s one aspect that makes me feel like I have painted myself to some extent into a corner, and this has to with routing the sound signal of internet video streams through my DAC to my loudspeakers.

 

As an illustration, please see the schematic drawing of my home network below. Prior to adding the Fitlet2 to my network, my Mac Mini M1 and the RME ADI-2 DAC were directly connected by means of an USB cable. Not only did the Mac Mini act as a HQPlayer server (which it still does), I also used it to play internet video streams made available by Dutch Public Broadcasting and watch these at my television, while at the same time routing the sound signal via my DAC and the power amps to my loudspeakers. This way, I was able to enjoy a much better SQ than the speakers of my Loewe television have to offer. However, with now the Fitlet2 as a NAA in my audio chain, this routine no longer works; as I already feared, the sound signal is no longer fed to the DAC. The atrocious speakers of the Mac Mini not being an alternative, I at first thought I was condemned to keep switching out the USB cable that feeds the DAC between either the Mac Mini or the NAA, dependent on whether I want to listen to music or to watch an internet video stream. Since this wasn’t something I wanted to keep doing (the USB ports of the Fitlet2 are really tight, and I don’t want to ruin them in the longer run), I decided to experiment with a trial version of Rogue Amoeba’s Loopback. This software appeared to be astonishingly easy to configure, and by selecting the Safari ‘Pass-Thru’ as input device in HQPlayer, within minutes I was able to do what I wanted to achieve; relaying the sound signal of internet video streams through HQPlayer and the NAA to my DAC, and from there, to my loudspeakers.

 

However, one big caveat was immediately apparent too, and this is the noticeable latency of the audio signal; image and sound are out of sync by some 1-2 seconds, and this is way too much. I have tried to solve this problem by making processing in HQPlayer Desktop as light as possible (PCM, no filtering, no dithering, various sample rates), but whatever I have done so far, nothing has helped and the problem remains. @Miska: is this latency something that can be solved or reduced by applying the right settings in HQPlayer, or is it inherent to what I am doing? If the latter is true, I’ll be giving up on Rogue Amoeba’s Loopback or whatever other virtual audio cable. After all, I’d still be able to reproduce the sound trough the speakers of the television itself by choosing this as a setting in macOS, but being able to send the signal to my loudspeaker through my DAC, as I did previously, would certainly be considered an added bonus. If you could give me any advise as to what to do, I'd be delighted.

 

image001.thumb.png.25e4b7c4a1393311e414669bd63d576f.png

 

Link to comment

@El Guapo, @Miska Thank you very much for your replies. I have experimented over the weekend with the settings suggested by the both of you, by combining 'short buffer' with various filters and dithering options. Unfortunately, nothing of what I've tried has solved the problem and hence, a significant lip sync issue remains. Since so far I haven't been using VLC or any other video player, but merely played the video stream in Safari, it might be worthwhile to try VLC and see to what extent it allows me to delay the video play to get it back in sync with the audio signal. But indeed, I don't specifically need HQPlayer DSP for the video audio (although it would have absolutely been a nice bonus if it worked), and therefore there's still also the option of bypassing the HQPlayer and the Fitlet2. In a variation on Miska's idea of adding an Apple TV to my setup, I am now looking at the possibility of directly connecting one of the Mac Mini's USB ports to the vacant S/PDIF coax port of the DAC. If only I could find a decent USB-S/PDIF converter that is less pricey than the Halide Design S/PDIF Bridge and more trustworthy than the numerous anonymous converters that are all over the Internet (which invariably need yet again another LPS), I guess I'd be fine. But this is a question I better post in another thread.

Link to comment
  • 1 year later...

@Miska Last week I became aware of a problem which curiously is only there when I am playing DSD files. It consists of a continuous crackling noise, which is noticeable during quiet passages:

 

 

 

Once you know it’s there, you tend to focus on it, and this is quite distracting. This crackling noise is only discernible when music plays, but also when it is paused – when a file has ended playing, the effect is gone, and the same can be said when HQPlayer Desktop is stopped; my audio system itself is completely silent without an active source.

 

This audio artefact also isn’t there when playing FLAC files. Since these make up the vast majority of my music library and I am hence playing DSD files much less often, I can’t say with certainty that the problem didn’t already occur with HQPlayer Desktop 4. However, the crackling is so obvious that I can't imagine not hearing it before – my suspicion is that it only started manifesting itself after installing HQPlayer 5, which I did a couple of days after this  new version became available. First suspecting that something might be the matter with cable connections or hardware (I have a couple of cats responsible for running the household), it took me a while to figure out that the problem lies in the file format, or, probably more correct, in the settings that are annex with DSD. I have, however, so far failed at finding a combination of settings that solves the problem, and this is why I am here to solicit any thoughts or ideas that may help me to eliminate this crackling noise.

 

The (two-channel) DSD files I am referring to are DSD256, which is the same format in which the recordings were made, and these were mostly downloaded from nativedsd.com. The settings with which I first encountered the problem are as follows, but like said, fiddling with other combinations of filters, modulators, integrators and converters hasn’t helped:

 

Schermafbeelding2023-06-05om13_45_00.thumb.png.e2e0e496af1bcb8c6a7e0f2566c3604c.png

 

Schermafbeelding2023-06-05om13_45_23.thumb.png.f71e44418f1e61ff9f0c3b3463187296.png

 

Schermafbeelding2023-06-05om13_45_42.thumb.png.bddadf7df14dd3c37b8df7879b74b09d.png

 

Schermafbeelding2023-06-05om13_46_02.thumb.png.8120bd01dce8b63a1ae59a2b49ccc2bd.png

 

Schermafbeelding2023-06-05om13_46_24.thumb.png.c09546b246f8274c7f8fad08dec59d9c.png

 

For further information, please see the attached log file:

 

HQPlayer5Desktop.log

 

My network layout hasn't changed since more than a year, but this how its part relevant for the current topic looks like:

 

Thuisnetwerk-audio05062023.thumb.png.9b3d2cc4705a04ce85664694020e5c3a.png

 

The Apple Mac Mini M1 has 16 GB RAM.

 

Like written, any suggestions that help me troubleshooting the issue and eliminate this artefact, will be greatly appreciated.

 

Cheers, Simon

Link to comment

@cpcat @Fredc @Schafheide and @Miska: thank you very much for replying to my message, and for your observations and suggestions which have helped me to solve (or should I say circumvent?) the issue.

 

Changing the SDM output from 48x256 to 44.1x256 did the trick, as this eliminated the crackling noise I described above. Only deactivating the 48k DSD option under the 'Outputs' tab isn't enough (or not even necessary) to solve the problem, changing to a 44.1k output bitrate under SDM tab is what appears essential to get rid of the crackling. Ticking 'Adaptive rate' under 'Outputs' (instead of 'greying' it), without changing the other settings, has the same effect.

 

@Schafheide: I guess like you, I updated my Mac Mini to the latest incarnation of macOS Ventura shortly after it became available. Since both this computer and my Mac Studio are running under this OS, and I don't have a spare Mac laying around which still runs under a prior version of Ventura, I am not able to test whether the current OS is indeed a contributing factor to this problem. It would be interesting to hear if other forum members who have their M1 Macs still running under and older version of Ventura are indeed unable to reproduce this crackling noise when playing DSD files at an output bit rate of 48k.

 

@Miska: changing the modulator to ASDM7EC-light (or any other, for that matter) without touching the settings mentioned above doesn't make a difference, in this case the crackling is still there.

 

7 hours ago, Miska said:

I just tested on my M1 Mini with 16 GB and it works fine as long as I have matrix processing disabled and Adaptive rate enabled. Same settings as you have. Source is DSF file from my "NAS" (Ubuntu 22.04 based SMB server) over gigabit Ethernet to local (USB) ADI-2 Pro output (AKM chip). Output to T+A HA 200 with T+A Solitaire P headphones over balanced. When I pause and turn the HA 200 volume to maximum it is still completely silent.

 

My ADI-2 DAC fs has the AKM chip too (AK4493). Since your settings are the same as mine, and the network layout you describe basically seems the same as what I depicted in the diagram in my original posting, I am at a bit of a loss as to what exactly is the cause of the problem that I encountered - especially so, since I found an older posting of yours from 2019, in which you ascertain that 'ADI-2 supports DSD also at 48k base rates just fine'. How come then, that I need to switch to 44.1k for a base rate, as suggested by @cpcat and @Fredc, to get eliminate the problem?

 

 

Link to comment

@cpcat No, my ADI-2 DAC features an AKM chip (AK4493). As to your suggestion to drop down to 48x192 or lower, this I didn't try yet, but I will do so and see whether this leads to noticeable changes in SQ. Changing to 44.1x256 still sounds fine though, and I am pleased that this has solved the issue.

Link to comment

@Miska

 

9 minutes ago, Miska said:

 

Because your source is 44.1k multiple (DSD) and you output 48k multiple because the Adaptive rate is set to grayed. This overloads your CPU since it is much heavier to process 44.1k x256 -> 48k x256, than processing 44.1k x256 -> 44.1k x256.

 

 

Thank you for explaining, all this is a matter of (a lack of) computing power then. In this case I owe you an apology for suspecting in my first posting that the problem only started manifesting itself after installing HQPlayer 5. If a lack of computing power is the root cause, this must also have been the case when I still used HQPlayer 4.

 

Are you able to reproduce this crackling sound effect, when playing 44.1k multiple DSD at your Mac Mini M1 and outputting 48k multiple with Adaptive rate set to greyed?

Link to comment
On 6/6/2023 at 6:32 PM, cpcat said:

@Quokka_61 if you want to test offline created 48k 256 DSD then go below and download and play - this will tell you if the issue is related to cpu loading or not.  Play it with “direct sdm” checked so it is not processed by hqplayer.  Thanks to @bogi for this -who just posted below btw.

 

@cpcat Thank you for this suggestion, and @bogi for supplying these test files and for your additional advice. After I first had to get rid of the error warning shown in the first screen shot below, it appeared that my RME ADI-2 DAC fs didn't have a problem being fed 48k x 64 DSD, 48k x 128 DSD or 48k x 256 DSD via DirectSDM, for which see the second screen shot. The music played just fine, without the crackling sound effect that led to my first post on this subject. @Miska must hence be right that a Mac Mini M1 16GB is too light in resources to handle converting 44.1k multiple (DSD) input to 48k multiple (DSD) output. Out of curiosity I may try and see what happens when I install HQPlayer Desktop 5 on my Mac Studio M1 Max 64 GB, but in the mean time I am fine toggling off the 48k DSD option under the outputs tab - the sound artefact is gone then, and I don't hear any change in SQ or whatever.

 

Schermafbeelding2023-06-07om14_23_40.thumb.png.d3ad99366a4d49622ea17a202068fe92.png

 

Schermafbeelding2023-06-08om13_17_55.thumb.png.541f9ac87c21bb1cb9e2f9454f9ddcc0.png

Link to comment
On 6/8/2023 at 8:01 PM, Miska said:

You can keep the 48k DSD checked. Just make sure you have "Adaptive rate" fully checked (not greyed or blank), and that your rate limit is 48k x256. This should work fine.

 

@Miska Thanks for this tip. Things work like a charm this way, and like written before, being limited by my Mac Mini M1 by missing the option to convert 44.1k multiple (DSD) input to 48k multiple (DSD) output is not a problem.

Link to comment
  • 5 months later...
  • 5 months later...
8 hours ago, bogi said:

I'm not familiar with Mac computers, just generally: Make sure you have only one network interface enabled on Mac Mini. Otherwise multicast packets used for NAA discovery may end on some other interface.

It's my experience there's no need to fiddle with this on a Mac. When an ethernet port is active and the NAA can be found by HQPlayer, things work without the need to mess around with network settings.

Link to comment
13 hours ago, Miska said:

I have created corrections for the ADI-2 Pro's I have. But those have the AK4490 chip. While the ADI-2 DAC fs rev B has AK4493 chip which is rather different. And then the rev C model has ESS chip which is again different. This creates a bit of issue as it would need three different correction profiles. And I have only one of the applicable models at the moment.

 

@Miska Thank you for your reply. I understand this is complicated, given that there are two versions of the ADI-2 DAC fs, one of which is apparently no longer for sale. I would consider lending my rev B with AK4493 chip out to you for development purposes, were it not that this is my only DAC. Hopefully someone living closer to you will be able to send you one on a loan, but otherwise I guess there isn't much you can do.

Link to comment
7 hours ago, copy_of_a said:

Me personally in any case I only use Album Replay Gain.

Do you happen to use „Yate“?

If so I could send you (or point you to) a script („Action“ in Yate) that copies the Album Replay Gain values to each track of the respective album.

You can batch process your library with Yate over night (that‘s at least what I once did).

Doesn‘t work if the source is online (streaming), of course.

So an option to select the desired mode within HQP would be really nice.

 

@copy_of_a I am a fellow user of Yate, and I'd hence be much interested in receiving the action script you are referring to. I am experiencing pretty profound differences in volume levels between albums, despite activating the Replay Gain option while ripping CD's to my NAS with dBpoweramp, and activating Adaptive Gain mode in HQPlayer Desktop. I am therefore looking forward to see what your action script can do for me. TIA!

Link to comment

@copy_of_a  One quick question to get started. If I am understanding you correctly, I am supposed to run two actions in Yate on a selection of tracks, and such in the following order;

  • Replay Gain Album
  • Album Replay Gain to Track

The latter action I have downloaded using the URL you provided. However, as regards the first one, I have checked the resource page of 2ManyRobots (https://2manyrobots.com/resources/), but no such action is to be found. If you're referring to the fact that the metadata of a file should comprise values for ReplayGain Album Gain and ReplayGain Album Peak to start with, as shown the picture below, I am fine. Otherwise, I'd much appreciate if you could provide me with a link for Yates Replay Gain Album action. 

 

 

 

Schermafbeelding2024-05-10om14_52_42.thumb.png.917c86275a8089d97166a4b26ee0e27b.png

 

Link to comment

@copy_of_a Many thanks for providing your own script! Like I wrote, for those Flac files I ripped myself and the metadata of which contains information on 'ReplayGain Album Gain' and 'Album Peak', Yate's own 'Album Replay Gain to Track' action will probably do. However, I have downloaded quite a few high-res Flac or DSD albums from vendors like Presto Music, NativeDSD and Qobuz, and especially for these I am curious to see if your script helps to lessen the differences in volume levels. These are pretty remarkable; given the volume setting of my DAC / pre-amp, which isn't touched, for some albums a volume setting in HQPlayer of - 18 dB is sufficient, while for others it needs to be increased to - 8dB or -7dB. So, I am going to experiment with your script this weekend, and hopefully it helps to smooth things out a bit!

Link to comment
On 5/10/2024 at 10:24 PM, copy_of_a said:

In general Album ReplayGain works really well for me!

As far as DSD goes ReplayGain doesn't work with *.dsf or other DSD-files - at least Yate can't analyse DSD files.

Don't know if there's a metadata editor that can decode DSD files to analyse gain levels?

 

You could manually enter the respective values in Yate - I just have no idea whether or not HQPlayer does accept these entries with DSD files. Actually I wouldn't be surprised if it would do so 🙂

To check the RMS level of DSD files you could for instance use Xivero MusicScope (freeware since the company went bust), but that would require to analyse each file separately so that's a lot of manual work.

 

@copy_of_a Thank you again for your help. As Xivero MusicScope doesn't seem to be a Mac solution, I'll start running the action scripts on my Flac files, which constitute the majority of my music library anyway and among which the most profound differences in volume levels occur. I guess adjusting volume levels of DSD files requires some further digging into what solutions there are for macOS, but for now, Flac files are my priority.  

Link to comment
3 hours ago, creativepart said:

Thanks for that. So, I just changed mine from "default" to 16 pipelines. This did lower the app's RAM usage from 4.16gb to 1.42gb. So, not as low as yours. Are there other changes to that Advanced tab that I should use?

 

@creativepart You may want to grey (not check) the Multicore DSP selection box in the Advanced settings tab, for which see the manual. Like you, I have HQPlayer Desktop running on a M1 Mac Mini (16GB memory), and when playing music, HQPlayer uses only 470 MB of its memory - pretty remote from the memory usage you quoted.

Link to comment
5 minutes ago, creativepart said:

Thanks... but that made memory go from 1.42gb to 1.46gb. That small rise is inconsequential but it didn't go down.

 

What files (Flac, DSD) do you play, and what filters and ditherers / modulators do you use? Some filters are much more taxing on memory than others, so this may also be a difference...

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...