Ted Smith said If you are using a preamp it's probably not the best course of action to first attenuate the signal by about a factor of 100 and then to reamplify the signal with the preamp (and then attenuate it again with the preamp's volume control.) Any preamp will be able to handle the DS at 100 without the DS's 20dB attenuator engaged, then you'll only have one attenuator (the preamp's volume control.) Of course you can do what ever you think sounds the best, but almost always getting changing "attenuation -> amplification -> attenuation -> amplification" to just ""attenuation -> amplification" will sound better.Ted, sorry still struggling to get on board here. You advised using the 20dB attenuator as a means to reduce noise at the cost of lower maximal output. I understood this to mean for any given SPL you will have less noise using the 20 dB attenuator. Is there a cost to accuracy? If not, why not employ the 20dB attenuator's cleaner signal in all circumstances, including when a preamp is in the chain, if your system produces desired SPL?
If you use the attenuator to lower the signal and then use a preamp to amplify it back you haven’t achieved anything. The attenuator will lower the noise and then the preamp will amplify the noise again. What’s worse, a lower level signal will be more subject to interference, noise, etc. so you probably loose something by sending lower level signals thru your interconnects instead of line level signals.
As rules of thumb:
If you have a preamp, you are probably best off using it for volume control and having the DS’s volume set to 100 without using the 20dB attenuator.
If you don’t have a preamp and you typically listen with the DS volume at 60 or higher you are probably best off not using the DS’s 20dB attenuator.
If you don’t have a preamp and you typically listen with the DS volume at 60 or lower you should probably use the DS’s 20dB attenuator and a higher DS volume setting.
1 Like
OK, making sense now. 
General rule: Signal/noise ratio is enhanced with increasing volume settings on the DS. Accordingly, the 20dB attenuator is a tool to consider whenever volume settings less than 20dB of max, i.e. 60 or lower, are required.
Thanks Ted, makes sense re 1 bit. FYI I tried another interconnect with very low capacitance (Morrow with Eichmann Silver Bullets) to no avail as I thought it might be some impedance mismatch so I reverted back to the HFC CT1-U.
I learning to settle with the low level setting, putting the volume to the max which is still enjoyable but wished it would be at least +6db louder. If I want to be immersed in an opera piece, into a concert, I switch to the high setting, the noise annoys me in the quiet passages.
Again, I have had many DACs or preamps in the same system and never had noise like that, except once when one of the line stage tubes was going bad and that was more like a hum, this is hiss, sounds different. Before I had dedicated lines, sure some hum was part of the game.
Can the 20db attenuator (low/high setting) be changed by a software patch so that the high setting is lower, say 14db, yielding a lower noise floor linearly? That will be so great.
No software revision is possible.; the 20dB attenuation is a relay which shunts to ground through some nice resistors.
I understand that the noise is too high for some systems. As it turns out, almost everyone likes the sound better if we let more noise thru, that gets more extension in the high frequencies giving a more open sound, etc. The ear/brain is great at ignoring uncorrelated noise. The level of noise that we design into the hardware is tradeoff from the steepness of the output filter. We don’t want too much ultrasonic noise getting thru, but everyone wants more ultrasonic signal. Some systems out there are linear enough that using no analog output filter in the DS and the natural rolloffs in the preamp, amp, speakers, and ear are good enough - but they are rare and perhaps those people don’t have dogs or cats that wouldn’t like the extra ultrasonic noise.
We’ve already done the software accommodations we could think of: notice that we allow the volume to go to 106 when the 20dB attenuator is engaged - that difference was enough for some customers, but we can’t really go higher without raising distortion.
Other things that people in your situation have done:
Using the balanced output will gain 6dB.
An inline transformer (Purely passive, say, inline in the interconnects - you could also use the DS’s balanced outputs and a 1-1 isolation transformer into your unbalanced amp inputs.)
Using the high range (no 20dB attenuator in the DS) and an inline attenuator - you can buy reasonable quality attenuators or just put a quality resistor from hot to ground either at the source or the destination of your interconnects - 15 Ohms is -20dB, 33.7 Ohms is -14dB, 62.4 Ohms is -10dB, 136 Ohms is -6dB
Use a preamp
Some have had their amp’s input sensitivity raised
Use the mute button (or engage the attenuator) when you aren’t playing music (not something I’d think of as a complete solution.)
Ted Smith said . . . we allow the volume to go to 106 when the 20dB attenuator is engaged . . .I forgot about this. A nice extra 3dB.
Just reviving this thread as I seem to have an issue with low output.
It seems to me that the DAC has low gain. I cranked my preamp volume up to past 12 o ‘clock on most music (with other sources, 9 o’ clock is plenty loud and going beyond is deafening) and found it more likeable – it still hasn’t got that midrange richness or bloom in spades but the added volume seems to have filled that out some and there are more saturated tone colours.
Does anyone else find the low gain to be an issue? I am using the balanced outputs with an additional 6 dB of gain, have the volume output set to 100 and under Settings it is on the high volume one. I fiddled around with the filter button on the remote as I read somewhere that is an attenuator of sorts but with it engaged and volume output set to 100, it just went even softer. I would have to turn my preamp up to crazy levels unless I am doing something wrong. The other thing I found was now increasing the preamp volume control did very little incrementally when before with other sources just a small increase would boost the volume quite a bit.
Is this normal?
Thanks
The output level of the DS is about 3dB lower than most CD players (most are 2 VRMS single ended and the DS is 1.4VRMS single ended.)
Turning your preamp up by three dB shouldn’t be a problem in most systems.
The 20dB attenuator is designed for systems where the DS’s output is too high.
Louder almost always sounds better so it’s important when A/B’ing components that the output levels be matched. Playing too quietly will cause the midrange to be too soft causing the exact symptoms you cite.
Thanks for the reply Ted.
So would 3 dB be the volume loss even if I am using a balanced connection and would this equate roughly to going from 9 o ‘clock to 12 o’ clock on the preamp volume control?
I guess there may be nothing wrong with doing that but psychologically it feels like it’s running out of juice since I have never played back at such high levels before.
Still would like more midrange richness and bloom (perhaps in a future firmware update though 
)
I also realized the reduced headroom with the DS, but mainly in my case this leads to the fact that I can’t play it satisfying directly without preamp as I initially hoped.
With preamp, due to the huge differences of digital files in overall level, I indeed also partly have to turn up volume 1/4 more than before an often play from 12 o’clock up. But I use single ended, so again 6dB less afaik. Would like more headroom, but it’s not a real problem for a preamp at the end usually. Sometimes I recognize it in dynamics, rather not richness.
yacheah said So would 3 dB be the volume loss even if I am using a balanced connection and would this equate roughly to going from 9 o 'clock to 12 o' clock on the preamp volume control?I guess there may be nothing wrong with doing that but psychologically it feels like it’s running out of juice since I have never played back at such high levels before.
You aren’t alone feeling funny about using more range in your preamp than you are used to, but if you never use it what good is having it?
Some preamps do a have a sweet spot where they sound the best and many people claim that it’s at 9 o’clock, but in truth (even if there is a sweet spot for the preamp where it’s not attenuating and not adding gain) there’s no technical reason that it should be at any particular point on the dial and hence will differ from one preamp to another… Also there’s nothing technically that says that all preamps sound the best with no attenuation or with no gain - it’s not like they have three separate signal paths, all components are a series of compromises.
In the end you should just turn things up until they sound good (unless that level is high enough to be tiring or cause ringing in the ears the next day, etc.
Hi Ted
woild like to to know what happen the DS output is set to low and I’ve set the volume level to 106 at the DAC? Does this means that the DAC is amplifying the signal by 3db? Does that add any noise to the original signal?
The DS’s volume control is digital, so amplifying isn’t really a good description. In the DS there’s nothing special about a volume of 100 (see http://www.psaudio.com/forum/directstream-all-about-it/directstream-dac-output-level/#p50608 earlier in this thread.)
Mike, Ted answered you about the “Filter” button. It’s puzzling to me that your output is so low when mine is sufficient to make things loud enough for me from all sources (even if one is lower than another might like) and that is using the DS as a preamp, my tubed preamp not in the chain. But as Ted also noted, you do have enough gain on the preamp and that’s what it’s for. . . .
Ted Smith said The DS's volume control is digital, so amplifying isn't really a good description. In the DS there's nothing special about a volume of 100 (see http://www.psaudio.com/forum/directstream-all-about-it/directstream-dac-output-level/#p50608 earlier in this thread.)At 100 there will be no modification so in a way bit perfect?
tkiat saidTed Smith said The DS's volume control is digital, so amplifying isn't really a good description. In the DS there's nothing special about a volume of 100 (see http://www.psaudio.com/forum/directstream-all-about-it/directstream-dac-output-level/#p50608 earlier in this thread.)At 100 there will be no modification so in a way bit perfect?
The post cited tried to explain that there’s nothing special about a volume of 100… Let me try a different way:
Its true that we want bit perfect transportation of information - we don’t like disk drives or networks that change our data at random.
When doing math results are bit perfect only when we have infinite precision, which is impossible in computers. All upsampling involves digital filtering and (in general) digital filtering involves approximated coefficients.
The bottom line is there’s no path thru any modern DAC that’s bit perfect. Instead the key is to not loose any accuracy along the way that will affect what we can hear - i.e. don’t cheap out and just use 16 bit math, or just 24 bit math or just 32… we need to use whatever precision that’s required…
The art in the DS is to use much more accurate math when filtering, doing volume settings, doing sigma delta modulation, etc. than most DAC chips implement. Also the DS does most of it’s math at a higher sample rate than most digital processors. In addition to more bits of precision on the low end (more guard bits) than most DACs use the DS also has head room on the top for calculated values that are greater than the unity volume level so that there’s no clipping of transients thru the filtering process. The DS has 6dB extra headroom thru the digital path and 3dB of extra headroom in the analog path, both of which add ease to the DS’s output with a more dynamic sound.
A volume setting of 100 won’t sound any better than one of 101 or 99 - any differences are in the response of the following analog path in the rest of the system to the differing levels of sound.
Beautiful explanation, Ted.
lonson said Mike, Ted answered you about the "Filter" button. It's puzzling to me that your output is so low when mine is sufficient to make things loud enough for me from all sources (even if one is lower than another might like) and that is using the DS as a preamp, my tubed preamp not in the chain. But as Ted also noted, you do have enough gain on the preamp and that's what it's for. . . .Lon, do you still play CDs? How do you listen to the brighter RVG's through the DS? I played Freddie Hubbard's "Breaking Point" earlier - the air, soundstaging and low noise make the rhythm section and piano sound good but when it comes to the crescendo of horns it can be jarring and verge on earbleed unless you have Harbeth speakers which have a midrange bump to compensate for it like davidl from the Torreys thread did.
Ditto for any Jackie McLean with his sour tone.
Kinda was a throwback to my Krell owning days - on good recordings yeah, on lesser ones which is probably the majority of what I have meh. I am exaggerating - the DS is more rounded but still you get the drift.
It’s overall a nice DAC. Turning up the volume on my preamp filled out the mids somewhat on Cannonball’s “Somethin’ Else” and Dexter Gordon’s “Doin’ Allright”. In fact I preferred these on the DS than the Cary DAC I compared it with even though the latter had a richer midrange. The DS’ sweet, grainless, more extended highs won out on these recordings as well as on Jennifer Warnes’ “Famous Blue Raincoat”. The DS midrange veers towards open and clear which is nice in its own way.
It just needs like Jazznut says a richer midrange. Not too much as to be thick and slow sounding but just enough in the upper mids to keep piercing brass at bay, vocal sibilance away and give more voice to the sax.
I’m gonna try some more aggressive Trane tomorrow to see how it fares and whether to keep the DS or not.
I am wondering if I can tweak it with cables but think that’s likely to be a vicious cycle. Also, is there an older firmware that has pretty much most of what Torreys is but with a richer midrange? Alternatively, does anyone know of a DAC that sounds like what I describe - that would be the bees knees…
yacheah saidHow do you listen to the brighter RVG’s through the DS?
Do those RVG Blue Note CD’s sound great on any system? I don’t like the transfers. It’s too bad, I have quite a few…some of my favorite music.