Which HQP Filter are you using?

Any idea of how to check/control that it is an original hires?

For example using audio editor and check out the spectrogram. Free Audacity is enough for the purpose. Unfortunately you will only know after buying if it was real or fake…


Nice! And how do you like Sinc-L with 64 million taps? How would you compare it to Sinc-Ll with 32?

I always like sinc-Lm. This is the first time I ever tried sinc-L from 44.1. I was intrigued by the ram usage and was seriously expecting more cpu usage. I’ll have to sit down and listen now.

And now that this build is (dare I say it) :shushing_face: stable :face_with_peeking_eye:… I’ll attempt to stop messing with it.

@jussi_laako Sorry if my original post wasn’t clear enough.

What I meant is that, unless I’m missing something, a 96kHz file would need to be upsampled 128 times in order to reach the same frequency (what we would refer to as DSD256) as a 48kHz file would when upsampled 256 times (same for 44.1 vs 88.2, of course, just different frequencies). This appears to be confirmed by both the time it takes for a song to play as well as the RAM usage which is halved when I’m playing a file that is already higher-res (leaving aside the issue of whether it was actually recorded at said higher-res or if it was upsampled afterwards).

If my thinking is correct, then my two issues would be:

  1. Does using 44.1/48kHz provide the HQPlayer with more opportunities to filter out the noise (more upsampling cycles/higher conversion ratio = the noise being pushed further away from the audible frequency range)? Or, perhaps (as I’ve alluded to above) - it does not matter because the higher-res content already has the noise pushed further into the “abyss”? (Talking about conversion into DSD, though, and the noise that is created in the process.)

  2. With filters such as Sinc-L and Sinc-M(x), using higher-res files would appear to decrease the number of taps that are applied (lower conversion ratio = fewer taps). Again - perhaps it’s not that relevant to the content that is already higher-res to begin with.

I used to use Spek as it was very lightweight. I recall having issues with it before but that was a few OS’s ago, and the way I tinker with things, it was probably something I did that screwed it up. :joy::rofl::joy:

Filter length is altered to retain same characteristics regardless of conversion ratio. So for smaller ratio you need less taps to retain same response.

I don’t understand what I’m hearing…
Henrik Lindstrand - Klangland
Track 7 CPH - ARN
Qobuz 24/48 version

The difference between poly-sinc-gauss-xla and sinc-L

It’s like someone screwed up the master with one filter and it its a totally different experience with the other. Maybe I just need more coffee.

I’m using ASDM7ECv2 to DSD256x48 on an ADI-2 DAC FS

1 Like

For all Lampi users out there…

I’ve been using HQP to upsample to a Baltic3 via a dedicated M1 MacMini. Until recently, my go-to settings for HQP were gauss-long/gauss-hires-lp and ASDM5ECv2 at DSD256.

On a whim, I changed that today to use the following settings. No change to the filter settings but changed the modulator and doubled the upsampling. Great improvement in detail, soundstaging. Wish I had experimented more before now. Highly recommended.


That reminds me I need to give AMSDM7EC 512 + fs another go at DSD512 now that I have a proper NAA.

I just came across an old post of @jussi_laako on another forum where he said this psuedo multi bit is largely DAC dependent, and works especially well with Holo Audio (I have a May KTE). I wonder if the same could be said for your Lampi

1 Like

I’ve been using these same settings for this past month with my Lampi L4G5 fitted with the chipless DSD512 upgrade.
Mac mini M1 to Sotm 200 ultra naa endpoint.
Sounds wonderful.


I swear I tried AMSDM7EC 512 + fs (DSD512) on my M1 Mini awhile back and it stuttered. So I’ve been on ASDM7ECV2 (DSD256) ever since. Based on the above post I tried it again and it works flawlessly. Don’t know if this was from a recent update or from me adding an NAA to the equation. At any rate I’m thrilled.

1x - gauss-long
Nx - gauss-hires-lp

Initial impressions are AMSDM7EC 512 + fs (DSD512) > ASDM7ECV2 (DSD256) > AMSDM7 512 + fs (DSD512)

In other words the psuedo Multibit at DSD 512 (with Extended Compensation) is a sweet spot for the M1. You get DSD512 and EC with the same processing power as ASDM7ECV2 DSD256. @jussi_laako says the pseudo Multibit is largely DAC dependent as it places a larger load on the analog filter itself. This works well with DACs like Holo Audio, but is a poor implementation for ESS for example

Feel it’s got better instrument separation, placement and a slightly wider stage.

I’m certain that optimum HQP settings are totally DAC-dependent. Anyhow, this latest combination manages to dig out a noticeable increase in detail which is a very pleasant surprise.

I think I read somewhere that the latest versions of OS X for the M1 are more efficient. Maybe that has something to do with it? Ditto for later versions of HQP.

Glad you found the same to be the case with your Holo.


1 Like

AMSDM7EC 512+fs

I just saw this. Is there a discussion of the difference between this and ASDM7ECv2?

Differences I noticed at 1024 rates was either a slight pull back of mids or more v shaped presentation to me. Density/body/weight could seem slightly greater with ASMDM7ECv2 because of the presentation difference. It is more intensive though and still not possible without stuttering every 20-30 seconds in my 13900k/4090 testing using an AIO 280mm cooler in a mini itx Meshlicious case.
From what I heard, I preferred the AMSDM7EC 512+fs so it was an easy decision to return the 4090.

AMSDM7EC 512 + fs is a psudeo multi-bit modulator designed for DSD512 (or higher). Scroll up a couple of posts and I’ve linked to a post of @jussi_laako over at AS where he goes into greater detail on the differences of the two (which are largely DAC dependent)

So I tried a 3090 24GB from Asus. Sinc-L does not start at DSD512 from 44.1 kHz. Is this because 24 GB is not enough?

Did you check the HQPlayer log file for errors that could indicate what is going wrong?

It is close to the memory limits, but it could work. Depending on how much other tasks and the desktop are consuming GPU RAM.

Here is the last attempt, maybe you can identify something? Thanks!

2023/03/19 20:28:21 Play
2023/03/19 20:28:21 Offload: resampler=disabled convolution=disabled

  • 2023/03/19 20:28:21 Playback engine running
    2023/03/19 20:28:21 IntegratorM: FIR2
    2023/03/19 20:28:21 NAA output set sampling rate: 22579200 (22579200)
    2023/03/19 20:28:21 Automatic output rate: 22579200
    2023/03/19 20:28:21 Rate or blocksize change triggered
    2023/03/19 20:28:21 Rate: 44100, block size: 2352, frame size: 588
    2023/03/19 20:28:21 Block size: 2352 (sample: 2)
    2023/03/19 20:28:21 Oversampling: sinc (low)
    2023/03/19 20:28:21 Modulator: adaptive seventh order 1-bit ec v2
    2023/03/19 20:28:21 Integrator: FIR2
    ! 2023/03/19 20:28:22 clHQPlayerEngine::Execute(): clAlignedAlloc::SetSize(): lAllocSize < 0
    2023/03/19 20:28:22 Stop request (reset)
    & 2023/03/19 20:28:22 Stop…
  • 2023/03/19 20:28:22 Playback engine stopped
    & 2023/03/19 20:28:22 …stopped
    2023/03/19 20:28:22 Set volume: -8 +
    & 2023/03/19 20:29:05 Pause
    & 2023/03/19 20:29:05 Play
  • 2023/03/19 20:29:05 Playback engine running
    2023/03/19 20:29:05 IntegratorM: FIR2
    2023/03/19 20:29:05 Rate or blocksize change triggered
    2023/03/19 20:29:05 Rate: 44100, block size: 2352, frame size: 588
    2023/03/19 20:29:05 Block size: 2352 (sample: 2)
    2023/03/19 20:29:05 Oversampling: sinc (low)
    2023/03/19 20:29:05 Modulator: adaptive seventh order 1-bit ec v2
    2023/03/19 20:29:05 Integrator: FIR2
    ! 2023/03/19 20:29:06 clHQPlayerEngine::Execute(): clAlignedAlloc::SetSize(): lAllocSize < 0
    2023/03/19 20:29:06 Stop request (reset)
    & 2023/03/19 20:29:06 Stop…
  • 2023/03/19 20:29:06 Playback engine stopped
    & 2023/03/19 20:29:06 …stopped
    2023/03/19 20:29:06 Set volume: -8 +
    2023/03/19 20:29:07 Store GUI state
    & 2023/03/19 20:29:07 Playlist save: C:\Users-\AppData\Local\HQPlayer\current.m3u8
    2023/03/19 20:29:07 Shutting down…
  • 2023/03/19 20:29:07 Control ended from
  • 2023/03/19 20:29:08 Exit

You are not on latest HQPlayer version?