Are you using any EQ anywhere along the audio path? Or a volume setting which allows volumes greater than 100%?
That could indicate problems with intersample overs, depending on your filters this could very well be the case on recordings reaching 0dBFS
I am not aware of the Fusion Amps correcting for this, but that can be easily remedied if you are using Roon or similar. Typical guide seems to be -3dB
https://benchmarkmedia.com/blogs/application_notes/intersample-overs-in-cd-recordings
I am not aware of the Fusion Amps correcting for this, but that can be easily remedied if you are using Roon or similar. Typical guide seems to be -3dB
https://benchmarkmedia.com/blogs/application_notes/intersample-overs-in-cd-recordings
Yes, this problem occured the first time when I enabled ALSA 10 band equalizer for some corrections that I had not yet transferred into the Hypex filters. Curiously, I immediately took note of it when I first did this and where looking for a place to lower total levels. But: it went away, I had not found a place within PiCorePlayer and could listen at 100% digital output without problems. I had to write the SD again, and set upt the player another time and since then, it is always there, only 90% solves the issue, 95% is still too high.
So are your actual output level at 100% or >100% ;-)
Did you understand the reference to intersample overs?
I have not read the text, sorry, it's longer than what I can handle atm 
The picture above is not easily understandable, sorry. 66 is the zero position in pCP alsa 10 band EQ. According to this post, it is even slightly below: https://forums.slimdevices.com/show...liser-settings&p=969247&viewfull=1#post969247
With my above setting, there is some boost and I cannot rule out that the signal is higher than 100%, but as I had noted, this "clipping" also occurs with all the faders set to 66, which ought to be no boost at all.
The picture above is not easily understandable, sorry. 66 is the zero position in pCP alsa 10 band EQ. According to this post, it is even slightly below: https://forums.slimdevices.com/show...liser-settings&p=969247&viewfull=1#post969247
With my above setting, there is some boost and I cannot rule out that the signal is higher than 100%, but as I had noted, this "clipping" also occurs with all the faders set to 66, which ought to be no boost at all.
That is in short what the article is about - you need a margin, even 100% will normally cause clipping due to intersample overs, caused by the filter ripple in the DAC's.I have not read the text, sorry, it's longer than what I can handle atm
The picture above is not easily understandable, sorry. 66 is the zero position in pCP alsa 10 band EQ. According to this post, it is even slightly below: https://forums.slimdevices.com/show...liser-settings&p=969247&viewfull=1#post969247
With my above setting, there is some boost and I cannot rule out that the signal is higher than 100%, but as I had noted, this "clipping" also occurs with all the faders set to 66, which ought to be no boost at all.
So while you are at it - ensuring a margin - make sure you add some extra (3dBFS) or the problems will be back on recordings that are maxed out.
I have another request: since I now use two sources, RCA-in for watching movies from computer and S/PDIF from my music server, I need to switch between sources more frequently. Despite that the Hypex remote gives the option, I face the following problem:
I used the EQ function that is on the device panel for room EQ to my satisfaction. But when I switch between sources, the EQ function often gets disabled, sometimes this only happens on one speaker, I cannot recall whether in master or in slave. This is kind of annoying. My current solution is to use auto detect function, but I find it not as good because it always takes the first seconds of a song until the amp recognized the source.
Am I doing something wrong? And is there a way to get the EQ working, as in never being deselected, when I use multiple sources?
I used the EQ function that is on the device panel for room EQ to my satisfaction. But when I switch between sources, the EQ function often gets disabled, sometimes this only happens on one speaker, I cannot recall whether in master or in slave. This is kind of annoying. My current solution is to use auto detect function, but I find it not as good because it always takes the first seconds of a song until the amp recognized the source.
Am I doing something wrong? And is there a way to get the EQ working, as in never being deselected, when I use multiple sources?
Still in 1.46 there are other weird problems:
During REW sweeps both I and a friend have problems with drop-outs during the sweeps
I noted blinking red LED, then it got back again - he also noted that it shifted from AES to S/PDIF
Only running the shortest sweep durations and have pre-sync signal shortly before made it possible to avoid drop-outs, initially thought it was entering sleep-mode and turned that completely off - even though that should not happen that fast, and not in the middle of a sine sweep signal.
The signal level was not high, probably -20 to -25dBFS
The most annoying problem was that even though I never used the EQ and it was set flat - on completely new from factory modules. Bypassing this finally completely opened up the imaging on the LX521.4's.
The linearity was ok so it definitely seemed there was some garbage left in the modules, registers not cleared,? - expecting phase shift, or timing issue.
So make sure to bypass or overwrite (thought I already did that by having it ruler flat/untouched but active) what ever is in there on the EQ.
I did not at all expect a problem like that on a ruler flat EQ setting, so it took a long time of desperate search before I eventually tried that out.
I also note setup settings changing when connecting USB on the fly, never had any other issues with that - no noises, output gets muted, HFD filter settings can be re-written right away w.o. closing down, handy for replicating the settings on the other speaker. But then again it seems that EQ is swapped on/off. So I now go through all settings and do a close down in between of HFD not to risk any more hiccups
This has unfortunately made me distrust the modules and as the root cause for this behavior is unclear, it will take a long time before that trust is regained.
Then there is the S/PDIF, even with dedicated crimped Canare/Bomar/winchester 75ohm RCA and 75ohm Mogami/SPDIF/video cables 24bit/192K sample rate the signal is lost after module number 2 (of 4) cable length is <6m
It all works on AES, Hypex assures that the "control signal will not be affected" so I am running AES for sound and SPDIF for control. Interestingly my friend is able to get the SPDIF through on his modules - but only in a specific routing order. Noise or tolereance issues on the interface maybe?
When it works it is great, but the GUI is really quirky - and I never had any glitches like this on my miniDSP units, which include 2x4HD, DDRC24, 4x10HD and SHD.
One last thing which has been mentioned often before, it is really annoying not having even a simple volume/level read-out without having to have a pc connected to one of the modules at all times.
During REW sweeps both I and a friend have problems with drop-outs during the sweeps
I noted blinking red LED, then it got back again - he also noted that it shifted from AES to S/PDIF
Only running the shortest sweep durations and have pre-sync signal shortly before made it possible to avoid drop-outs, initially thought it was entering sleep-mode and turned that completely off - even though that should not happen that fast, and not in the middle of a sine sweep signal.
The signal level was not high, probably -20 to -25dBFS
The most annoying problem was that even though I never used the EQ and it was set flat - on completely new from factory modules. Bypassing this finally completely opened up the imaging on the LX521.4's.
The linearity was ok so it definitely seemed there was some garbage left in the modules, registers not cleared,? - expecting phase shift, or timing issue.
So make sure to bypass or overwrite (thought I already did that by having it ruler flat/untouched but active) what ever is in there on the EQ.
I did not at all expect a problem like that on a ruler flat EQ setting, so it took a long time of desperate search before I eventually tried that out.
I also note setup settings changing when connecting USB on the fly, never had any other issues with that - no noises, output gets muted, HFD filter settings can be re-written right away w.o. closing down, handy for replicating the settings on the other speaker. But then again it seems that EQ is swapped on/off. So I now go through all settings and do a close down in between of HFD not to risk any more hiccups
This has unfortunately made me distrust the modules and as the root cause for this behavior is unclear, it will take a long time before that trust is regained.
Then there is the S/PDIF, even with dedicated crimped Canare/Bomar/winchester 75ohm RCA and 75ohm Mogami/SPDIF/video cables 24bit/192K sample rate the signal is lost after module number 2 (of 4) cable length is <6m
It all works on AES, Hypex assures that the "control signal will not be affected" so I am running AES for sound and SPDIF for control. Interestingly my friend is able to get the SPDIF through on his modules - but only in a specific routing order. Noise or tolereance issues on the interface maybe?
When it works it is great, but the GUI is really quirky - and I never had any glitches like this on my miniDSP units, which include 2x4HD, DDRC24, 4x10HD and SHD.
One last thing which has been mentioned often before, it is really annoying not having even a simple volume/level read-out without having to have a pc connected to one of the modules at all times.
Last edited:
Thank you, I had not taken notice of a new release, my mistake.
This is great!
Considering FA123 for a 2way design, running BTL driving a purifi 4ohm driver. It’s a relatively low sensitivity driver that can push to 250W peak. Appreciate thoughts and possible shortcomings going btl path. I assume that tolerance driven mismatches between btl channels will contribute to loss of performance. I’m worried of major bugs/issues that might hinder bridging, especially as I plan to use the DSP for xover. Or should I shelf out the diff to 253. Appreciate inputs
I use FA123 in BTL mode with 16 Ohm drivers. The software can handle it BTL mode is easy to set with a setting in the software and following user manual wiring instructions. No bugs there. DSP works fantastic.
But the modules will be current limited if you BTL a 4 Ohm driver. The modules will see a 2 Ohm load and will not output full power, that's explained in manual of the modules. I'd suggest higher Ohm version of the driver to take full advantage of the available power.
Question:
The scenario: a previous filter was in non-btl mode, for both HF and LF drivers, the amp ‘saw’ nominal resistance 8 ohm. When changing to btl, only LF driver is seen as 4 ohm, HF still 8 ohm. This changes filter, but by how much dB SPL? 6? 10?
Does this mean it also doubles output power and if so, another question:
The scenario: a previous filter was in non-btl mode, for both HF and LF drivers, the amp ‘saw’ nominal resistance 8 ohm. When changing to btl, only LF driver is seen as 4 ohm, HF still 8 ohm. This changes filter, but by how much dB SPL? 6? 10?