I have A/B tested this over several days with the same result. When I enable DSP Engine to get the -3dB headroom control the sound quality takes a step back. I disable DPS Engine and it sounds great again.
With DPS it sounds flat, no depth, veiled. Without DPS it blooms, has better attack/decay, musical and enjoyable.
ROCK>Uptone Iso Regen/LPS-1 powered>Kii Control-Kii Three speakers
I agree with this. I also prefer to listen to Roon with all DSP turned off.
However there are times when I need to use DSP so there are good reasons to keep exploring. The Parametric EQ feature especially is a powerful tool that can really help enhance your listening experience. An example of this is Internet Radio where some stations I listen to only have low res feeds that have a high frequency “hash” when playing on my main system. I now use a filter to tone down the HF - making it much better to listen over prolonged periods of time.
OK, I went back and retested several times to confirm these results:
Used my Roon library Ayre Acoustics Test CD - Pink Noise for level testing referenced all to 73dB.
With DPS Enabled, all presets off except Headroom.
As expected, with Headroom enabled (-3dB) the level measured 3dB less and with Headroom disabled it goes back to reference 73dB.
With DPS Disabled, level is 73dB and does not change with Headroom on or off since DPS is off.
So with Headroom and all other presets off, when I toggle DPS on and off I DO hear a difference in SQ. It is subtle but it is there. With DPS disabled I perceive more air around the instruments, attack/decay improved, double bass sounds woodier, cymbals have more metallic sheen, just more musical sounding to me.
My system - Roon ROCK/HDPlex LPS > ISO Regen/LPS-1 > Kii Control/Kii Three speakers powered with balanced transformer.
The OP’s system is very resolving. With the Kii speakers I am not surprised he can hear a difference. I have tried the same in my setup and I imagine it sounds more open and relaxed without DSP, but nothing as dramatic as Frank_Palm describes. I guess this is system dependent and my speakers are certainly not as transparent as the Kii Three. Yes, I adjust for the level difference.
If I switch directly while playing I would of course hear a 3dB level difference. Even 0.1dB would make the loudest sound *better’. I don’t do it this way. I play part of a track, then stop and adjust and then play the same part again. My Devialet amp has a dB scale volume display so it is easy to adjust.
No human will be able to hear a 0.1 dB difference at 20 khz, it’s just to make sure that the PEQ is enabled and in use. For me personally, I want to be able to switch and compare with no delay in between. And there is no way for an amp (or preamp) to know the level a pair of speakers will play at a specific output, so that reading is a rough estimate only (unless its calibrated with the speakers used).
I do not want to hijack this thread, but I think you misunderstand. I am not talking about 0.1 dB eq at 20kHz, but a level difference of 0.1 dB. It is well known that even such a small level difference is enough to skew a comparison. And the 3dB adjustment in Roon is a digital level adjustment. Same as the volume adjustment in the Devialet. I’d say they are both exact.
The reason I suggested the PEQ and 0.1 dB at 20 khz was that I have heard some people say that enabling the PEQ lowers the SQ, and wanted to see if the OP could hear that difference as well (I can’t).
Lowering the volume in the digital domain will reduce sound quality, that’s known. Here is how, on a simple level, I think Roon adjust the volume ( @brian can maybe confirm or explain it in more detail if needed):
Original input (16 bit for CD quality), is converted to a 64 bit double
Volume adjustment is made on the double value.
The double is converted back to 16 bit
The lowering of sound quality is partly because the resolution will be lowered (it might “remove” bits), but also because the conversion back to 16 bit will require rounding, and the rounding part is lost.
Using 24 bit sound should result in less reduction of SQ, maybe @Frank_Palm can verify?
I did not think that a -3 dB adjustment would result in noticeable lower SQ though.
This may sound strange, but it really is an honest question : what is the intended purpose of this thread ?
A notification ? Discussion, gathering opinions ? A request for improvement ? There are no questions…
If ‘discussion’ is the purpose : you are describing night and day differences. Should be then be very easy to prove for you, with repeated 10-out-of-10 scores, in A/B’s under controlled conditions. Levels matched etc.
Not the case ? Then it’s not as night-and-day as you thought. (I could discriminate an apple from an orange, at least 1000 times in a successive row…)
EDIT:Forgot one thing. When A/B-ing, be sure to turn off to to turn OFF headroom management (or 0dB). Due to that, be sure to use test tracks that do not use 0dBFS; they should have just a little headroom ‘built in’. 3dB is already way more than enough.
That’s true if you start with 16bit and end with 16bit, but that’s not what Roon does.
In Roon, if you start with 16bit and do a -3dB headroom adjustment, the process looks like this:
Notice that the stream going to the DAC is 32bits, not 16.
This is because Roon expands headroom when performing DSP to 64bits (since 64bit float is our intermediate format). Then, when it’s preparing to stream to the device, it converts to the widest format that the device supports. In this case, 32bit.
6dB = 1 bit. So a 3dB adjustment moves the audio down 1/2 of a bit. At the same time, we’ve expanded headroom to 32bits. You can mathematically reverse the adjustment without reducing the number of significant figures–so there is no information loss either. To be clear, this means you could:
Start with a 16bit signed integer signal
Convert to 64bit float
Apply a -3dB gain adjustment
Convert to 32bit signed integer
Convert back to 64bit float
Apply a +3dB gain adjustment
Convert back to 16bit signed integer
And the output signal would be bit-perfect to the original. So from an information-theoretic perspective, the gain-adjusted 32bit signal is lossless.
That said, there are some ways that lowering the volume earlier in the playback chain can result in a reduced perception of quality despite the fact that no information about the signal is lost. For example:
Psychoacoustics. Any test of a volume control that does not employ precise level-matching is flawed. Quieter statistically sounds worse
After level matching, you may discover that the amplifier (or another component) does not perform equally well in all volume ranges.
The second one is particularly applicable to volume leveling–since the adjustments are often ~10-15dB. I like volume leveling, but not all amplifiers sound as good cranked 10dB louder to compensate, especially headphone amps.
Assuming 32 bit output, that would mean that all volume adjustments in steps of 3 would be without signal quality loss, and other volume adjustments would result in a very very tiny SQ loss (an maximum error of 0.000000011%).
Just got an idea that would please the audio purists: have an option to round volume leveling to the nearest multiplicative of 3 dB, since this is in effect “bit-perfect”
Thank you Brian for that well written reply but my original intent of this thread was simply to share an observation - that turning on the DSP Engine with all presets off including headroom results in lower SQ for me, with my system in my room. With DSP on and all presets off the path shows no conversion…looks just like having the DSP off. But again…I hear a loss of SQ with the DSP switch in the enable position.
If I do use DSP with Headroom -3dB enabled the SQ really is set back. I noticed this one night so I was curious and experimented with the DSP Engine settngs and discover the above mentioned listening experience.
So in summary - DSP on = less SQ (with all dsp presets set to off)
Nothing to do with volume, just flipping the DSP switch, volume level remains the same 73dB in either position.
Turning on the DSP Engine switch does nothing whatsoever. It’s actually active in the off position–it’s a detail of the graphic design that we present it as turning something “on”.
In the off position, its behavior is a shortcut for flipping the enable switches on each of the sections in the DSP Engine to “off” if they are not already off. It does nothing extra. When “on” it has no effect at all–just the absence of forcibly turning that other stuff off.
But–if those switches are already off, then the DSP Engine switch truly does nothing at all.
All of that switch flipping happens before playback starts–once it’s running, the signal path is doing its thing, and not interacting with the switches in the user interface at all anymore.
I’m not sure what the cause is for the difference you’re hearing–but it isn’t the master switch on the DSP Engine.
Not doubting that you’ve heard something, but effects need causes that make technical sense. Before I explained this, you didn’t have the technical background in how that switch works to understand what its potential effects were–now you do. If I were you, I’d be looking for a more plausible root cause.