Output voltage and signal to noise ratio

[trivium]

I bought a DAC that is only rated at 1.1 v output from the RCA instead of 2V output. Other than having to turn the volume up more on my integrated amp does the lower voltage affect the signal to noise ratio since I have to turn the amp up more?

 

[Lambda]

does the lower voltage affect the signal to noise ratio since I have to turn the amp up more?

Not nervelessly maybe your DAC has also very low noise so signal to noise stays the same.

 

[win]

All else being equal, yes it diminishes SNR, and is the reason that pro audio line level runs so much higher voltage.

 

[trivium]

Not nervelessly maybe your DAC has also very low noise so signal to noise stays the same.

Gotcha, well it appears to be a very silent USB DAC so turning the amp up where it is now doesn’t appear to change the background hiss compared to turning it down. There is a slight noticeable hiss that is normal when I put my ear right up to the speaker that doesn’t appear to change whether the volume is low or high on the amp. I know with a power amp typically most would have it about 3/4 volume or so and use the pre amp for the volume. I’m basically doing the same thing except the DAC volume control as the pre amp and the integrated amp as the power amp.

 

[trivium]

All else being equal, yes it diminishes SNR, and is the reason that pro audio line level runs so much higher voltage.

I thought it was higher voltage to offset voltage drop at long distances, typical of professional audio.

dumb question here though, could I not increase the gain in windows to compensate for the low output voltage of my DAC?

 

[win]

If it's boosting it that could cause clipping.

Honestly if it's not noisy and outputs enough volume it shouldn't be a problem. I use an rca wireless transmitter that only outputs 1.1v and haven't had any issues with that from an output perspective.

 

[trivium]

If it's boosting it that could cause clipping.

Honestly if it's not noisy and outputs enough volume it shouldn't be a problem. I use an rca wireless transmitter that only outputs 1.1v and haven't had any issues with that from an output perspective.

Hmm your probably right. I have a calibrated microphone though so couldn’t I measure the background noise/hiss with the volume on the amp at different settings? Is signal to noise ratio not the same thing?

 

[win]

I think that puts more variables, like your amp and especially your speaker sensitivity into play. What DAC is this anyway? We like to pontificate here, but realistically you're probably well within the full-transparency region of distortion with something modern.

 

[trivium]

I think that puts more variables, like your amp and especially your speaker sensitivity into play. What DAC is this anyway? We like to pontificate here, but realistically you're probably well within the full-transparency region of distortion with something modern.

Amp is a marrantz pm6006 and I’m using a Xonar u7 USB sound card (CS4398 chip). I was using the integrated dac (Also CS4398) before but I found it was noisy at high volumes with nothing playing. Likely because dispite the EMI shielding its placement inside the chassis is right next to two huge silver power caps. The reason I switched to the xonar though is to separate the sub and mains so that I can apply an 80hz crossover with equalizer APO and manage the sub separately. The downside is I have to use my music app in windows to control the volume since I usually keep windows volume maxed. Also the reduced voltage output is another downside, it’s more than loud enough just hoping I didn’t sacrifice signal to noise ratio.

 

[AnalogSteph]

Sounds like SNR is rather better than before.

You may want to generate some rather low-level test tones (I'm thinking like -60 dBFS) and compare both setups that way.

Setting your main volume in the player app seems cumbersome when you've got both a soundcard and an integrated amp with volume dials on them, just saying. As long as you're steering clear of 0 dBFS to some degree, in-player volume control and Windows volume control should be pretty much equivalent.
https://www.audiosciencereview.com/...nding-the-windows-audio-quality-debate.19438/
(I tend to encourage the use of ReplayGain and similar if available. I tend to use a rather conservative pre-gain of -3.3 dB, and YouTube and other players are generally set to 50%.)

If you want to align full power output on the Marantz with full output on the U7, that should take about -14 dB... about 1 or 2 o'clock on the volume dial, depending on the pot. You should not generally need to go any higher than that. On paper (I did some calculations based on actual PM-6006 circuitry, with Source Direct assumed on) having a 1 Vrms source instead of a 2 Vrms one costs you about 3 dB of SNR simply due to the change in volume pot setting, although absolute noise level at this point still is fairly moderate due to the amplifier's moderate gain (about 200 µV out, or -83 dB ref. 2.83 V, or -87 dB ref. 5 W / 4 ohm).

I thought it was higher voltage to offset voltage drop at long distances, typical of professional audio.

Goodness, no. Do you have any idea how long a cable would have to be before a noteworthy level drop would occur?

Let's say we have an output impedance of ~0 and and input impedance of 10 kOhms. Let's also assume we are using twisted pair cable made from 28AWG (~0.08mm²) copper wiring. That has a resistance of 220 mOhm / m, and it would take a cable 9.4 km in length even for 3 dB of attenuation. At 150 pF/m, that's ~1.4 µF of cable capacitance, and if characteristic impedance is ~110 ohms that's ~17 mH of inductance (~2.1 kOhms by 20 kHz). Strictly speaking, the cable wouldn't even be correctly represented by a lumped element model any more, given that one only is sufficiently accurate for lengths less than about 1/10 the wavelength, which is ~14 km at 20 kHz in vacuum and rather smaller in materials. In any case, we are talking completely unrealistic dimensions.

Pro levels are higher in order to overcome the commonly higher noise of balanced input stages (at times exceeding -100 dBu), and for generally increased robustness.

 

[trivium]

Sounds like SNR is rather better than before.

You may want to generate some rather low-level test tones (I'm thinking like -60 dBFS) and compare both setups that way.

Setting your main volume in the player app seems cumbersome when you've got both a soundcard and an integrated amp with volume dials on them, just saying. As long as you're steering clear of 0 dBFS to some degree, in-player volume control and Windows volume control should be pretty much equivalent.
https://www.audiosciencereview.com/...nding-the-windows-audio-quality-debate.19438/
(I tend to encourage the use of ReplayGain and similar if available. I tend to use a rather conservative pre-gain of -3.3 dB, and YouTube and other players are generally set to 50%.)

If you want to align full power output on the Marantz with full output on the U7, that should take about -14 dB... about 1 or 2 o'clock on the volume dial, depending on the pot. You should not generally need to go any higher than that. On paper (I did some calculations based on actual PM-6006 circuitry, with Source Direct assumed on) having a 1 Vrms source instead of a 2 Vrms one costs you about 3 dB of SNR simply due to the change in volume pot setting, although absolute noise level at this point still is fairly moderate due to the amplifier's moderate gain (about 200 µV out, or -83 dB ref. 2.83 V, or -87 dB ref. 5 W / 4 ohm).

Goodness, no. Do you have any idea how long a cable would have to be before a noteworthy level drop would occur?

Let's say we have an output impedance of ~0 and and input impedance of 10 kOhms. Let's also assume we are using twisted pair cable made from 28AWG (~0.08mm²) copper wiring. That has a resistance of 220 mOhm / m, and it would take a cable 9.4 km in length even for 3 dB of attenuation. At 150 pF/m, that's ~1.4 µF of cable capacitance, and if characteristic impedance is ~110 ohms that's ~17 mH of inductance (~2.1 kOhms by 20 kHz). Strictly speaking, the cable wouldn't even be correctly represented by a lumped element model any more, given that one only is sufficiently accurate for lengths less than about 1/10 the wavelength, which is ~14 km at 20 kHz in vacuum and rather smaller in materials. In any case, we are talking completely unrealistic dimensions.

Pro levels are higher in order to overcome the commonly higher noise of balanced input stages (at times exceeding -100 dBu), and for generally increased robustness.

"Setting your main volume in the player app seems cumbersome when you've got both a soundcard and an integrated amp with volume dials on them, just saying. As long as you're steering clear of 0 dBFS to some degree, in-player volume control and Windows volume control should be pretty much equivalent."
The only reason i do is because i can control the player volume with my phone via tidal or plex. I cant really use the volume pot on the amp anymore because i have the subwoofer now plugged into the xonar's subwoofer output for separate bass management in windows. When you mentioned 1 or 2 oclock on the volume dial you meant the amp volume pot around 65% ish?

Alternatively i have a Sanskrit S10 DAC which has a 2.2V output which is even better, however the downside is i would lose the bass management since the sub would have to use the high level inputs through the amp. I could also go back to what i was doing before which was running a shelf filter of -8DB or so at 80hz than just turning up the subwoofer gain to compensate, it basically gets me a very similar result of a 80hz crossover on the mains.

All in all, you are saying that i will only lose about 3DB of signal to noise ratio by having 1V instead of 2V? That level difference is probably inaudible especially considering a quiet room is 30DB or so as a baseline, no?

 

[trivium]

Alright problem solved, i moved the Sanskrit s10 into the listening room, i wasn't using the full 2.2V output anyways as it was clipping my headphone amp at full volume. I lost bass management of course but using the xonar with 80hz crossover on the mains vs running full signal with a -10DB shelf filter at 80hz and turning up the sub up to compensate sounds identical. In fact, I measured the frequency response since im running REQ anyways and there is no difference, furthermore the difference in distortion according to REQ is negligible and inaudible in all cases.

Had i known this was how i was going to set things up i would have just went with a power amp instead of an integrated since i dont even need a volume control. Dually so because the Sanskrit even has its own volume control separate from windows.