Can you give any insights as to the volume difference in the new 2021 presets? (The volume is significantly lower when using the LCD-X 2021 preset, than say the LCD-4 preset that you were previously recommending be used until the updated presets for the 2021 revisions were released.
Also the “More Info” link in the preset is currently a 404 error
Did the volume issue come up at all during beta testing?
Not as far as I know. But I am now tempted to try some LCD-Xs!
Thanks for reporting, I will have to double check
Can confirm about the volume difference. Noticing at my end as well.
I can confirm that for the low latency presets the gain went down by 6dB. I realize that we were overly conservative on the gain in general. You can safely disable headroom and then add 11 or 12dB gain like below (Procedural EQ->volume). For the next release I will make sure the gain normalization is done better.
I also strongly recommend trying the Customize->Audeze linear, there will be a latency but imaging and clarity improves.
Should I apply both settings you mentioned (Procedural EQ and Audeze Linear) for LCD-X 2021 model? I did notice the volume reduction.
It seems the volume issue is only impacting the Audeze low-latency filter (which is the default). So if you prefer that filter then the Procedural EQ is needed, another option is to just switch to the Audeze Linear filter which doesn’t have the same issue. I’m listening to the linear filter now and its quite nice.
Thanks, I’ll try Audeze Linear (without Procedural EQ).
Even with Linear, you can gain a lot of headroom with the procedural EQ, all it is doing is increasing the gain and it will not affect SQ except, your DAC or Amp will now have more headroom and will help with dynamics.
Sorry to butt in 
There is a thread on here somewhere about the linear filter affecting MQA playback - not doing the bit depth restore before passing the signal to the endpoint.
I can confirm this is the case for me as well.
Low latency works as expected and no issues with MQA.
To my knowledge both linear and low latency filters are processed at 64bits by Roon, then requantized to whatever the bit depth of the DAC. I don’t see bit depth being an issue, but latency could. Frankly I am surprised any DSP processing is even possible with MQA.
Dunno what to say. I’m really bad at tech I love my Audeze and can’t tell a difference anyways - bit depth restore or no
Just want to point it out given it’s been discussed on here.
My external USB DAC is a Gustard A18 which is a full MQA decoder.
Signal path with low-latency:
Signal path with linear:
OK so it looks like this is impacting the chain if you are doing decoding or full hardware rendering on the DAC itself. (My Gustard X16 does this but I’m circumventing it since I am using HQ Player to upsample, I have Roon do the first unfold (decode only) and HQP take its from there). In your case I would switch back to the Low Latency Filter and add the Procedural EQ to boost the volume.
@KMan can hopefully shed some light on the technical aspects of why this may be happening to the Linear filter. Here is a link to the support thread with more details.
Thanks for this, I understand better, but this ball is in Roon’s court, not much I can do.
Hi - It seems the Linear filter is taxing the CPU and processing speeds are extremely slow <1x for DSD256 playback which makes it unusable. Changing it to low latency improves significantly. Any thoughts on if this is expected. I have Roon on an i7 NUC so more than capable of handling DSD256 with DSP applied.
This signal path is with low latency filter. Processing speed is 3x.
This one is with Linear filter and the processing speed drops to 0.9x.
I do prefer the sound of the linear filter but its making listening to DSD128 or 256 impossible.
