@wesayo: So in fact you have exactly what krivium describes and what I am planning.
I have 4 normal modes
-wife in house (not loud)
-wife in bed (quiet)
-wife out (loud)
- Movie night (need 10db boost as TV digital out is low level for some reason).
Once miniDSP is in place all fine adjustment will be there with course adjustment on an 8 channel pre-amp type thing.
I have 4 normal modes
-wife in house (not loud)
-wife in bed (quiet)
-wife out (loud)
- Movie night (need 10db boost as TV digital out is low level for some reason).
Once miniDSP is in place all fine adjustment will be there with course adjustment on an 8 channel pre-amp type thing.
An interesting and really simple read about digital volume control: The Case for Digital Volume Control | It’s only audio.
In short, analog noise should always dominate over digital noise; and in case of real world electronics, it does (-144 dB digital noise vs. -120 dB analog noise in case of 24 bits and best analog components). "Bits and resolution lost" is, IMO, a misunderstanding (which I had for years), because bits and resolution is just signal-to-noise ratio.
Analog volume control loses bits and resolution, too, since in the end it always reduces SNR ratio (there is no absolutely silent amplifier output stage or absolutely silent room). So, according to my understanding, changing volume in digital domain is perfectly good, and even preferrable (no issues with channel balance, for example). Of course, digital signal should always be properly dithered.
Oh, and there's that lovely bit:
In short, analog noise should always dominate over digital noise; and in case of real world electronics, it does (-144 dB digital noise vs. -120 dB analog noise in case of 24 bits and best analog components). "Bits and resolution lost" is, IMO, a misunderstanding (which I had for years), because bits and resolution is just signal-to-noise ratio.
Analog volume control loses bits and resolution, too, since in the end it always reduces SNR ratio (there is no absolutely silent amplifier output stage or absolutely silent room). So, according to my understanding, changing volume in digital domain is perfectly good, and even preferrable (no issues with channel balance, for example). Of course, digital signal should always be properly dithered.
Oh, and there's that lovely bit:
according to my experience, using digital volume is disastrous. it kills the dynamic, low level details, all in all very bad.
can you tell me how to try to have a good digital volume and how to properly dithered. I have only tried volume control from jriver, foobar, itune, ect
You've got it! 😀
And only one of those setting I ever rate of being of good enough quality! My movies run from JRiver as well so they'd be as loud as we can stand it... I have not bothered yet to include signals from the outside, as that would need an AD conversion. I sometimes miss having my radio though... Internet radio just doesn't cut it.
Thanks everybody for the feedback and discussion. I'm going to do a little experimenting to see whether I can hear the difference between different methods of volume control.
I'm running JRiver MC21 out Toslink, to a Schiit Audio Modi 2 Uber DAC, to a Schiit Audio Asgard 2 class A headphone amp, to a set of Hifiman HE400i planar magnetic headphones.
Would everyone agree this setup should be resolving enough to hear any differences?
I've been keeping the digital out set to 100 percent and controlling volume through the headphone amp. I'll try to do some back to back listening while controlling volume both ways. And yes, I understand enough to level match as well as I can between the two.
Is there any particular music or frequency range where I should specifically focus?
I have thousands of FLAC files but they are all 16/41. Should that be a high enough bitrate to make differences obvious or should I find some higher bitrate files?
-Chris
I'm running JRiver MC21 out Toslink, to a Schiit Audio Modi 2 Uber DAC, to a Schiit Audio Asgard 2 class A headphone amp, to a set of Hifiman HE400i planar magnetic headphones.
Would everyone agree this setup should be resolving enough to hear any differences?
I've been keeping the digital out set to 100 percent and controlling volume through the headphone amp. I'll try to do some back to back listening while controlling volume both ways. And yes, I understand enough to level match as well as I can between the two.
Is there any particular music or frequency range where I should specifically focus?
I have thousands of FLAC files but they are all 16/41. Should that be a high enough bitrate to make differences obvious or should I find some higher bitrate files?
-Chris
iScream, I'll give you one pointer: make it that volume control in JRiver is at 100% at the (maximum) preferred listening level. Attenuate from there digitally. Do not drop all of it digitally, that would be throwing away bits like you don't care! Do you get what I'm trying to say here? Gain structure is important and needs to be considered.
disable the jriver volume control entirely.
even when set at 100% gain in jriver, compared to disabelling the VC, theres a big sq difference. go to options and disable internal volume control
even when set at 100% gain in jriver, compared to disabelling the VC, theres a big sq difference. go to options and disable internal volume control
I think we all agree about that and this is a 'con' for multi amp active system versus passive solution. Well , it's important in there too but usually gives much less headache!
If not a con this is something you must think of if you want an effective system.
Last edited:
I think I get the idea of having a max realistic volume where I will listen. I would want to be at 100 percent digitally when listening at that volume. That leaves the most "disposable" bits before signal above the noise floor starts to be removed.
Close enough?
I'm trying to figure out the best way to make it happen though. Other than a high and low gain switch on the headphone amp I don't have an easy way to attenuate the analog signal. Well, other than the volume control on the DAC.
I'm thinking that starting with the digital at 100 percent and choosing a specific volume level through the headphone amp is the way to go. I can easily mark that spot on the volume knob as a a reference and either turn the knob down from there or reduce the volume in JRiver with the knob set to that point.
It's not blind AB testing and I'm already biases toward believing I won't hear a difference but I'm only trying to prove this to myself.
I'm pretty far from having golden ears anyway. I'm over 40 and I abused my ears when I was younger so I couldn't hear over 16K the last time I checked. And the last time I checked was pretty long ago.
Maybe I have brass ears, or copper at best.
-Chris
Close enough?
I'm trying to figure out the best way to make it happen though. Other than a high and low gain switch on the headphone amp I don't have an easy way to attenuate the analog signal. Well, other than the volume control on the DAC.
I'm thinking that starting with the digital at 100 percent and choosing a specific volume level through the headphone amp is the way to go. I can easily mark that spot on the volume knob as a a reference and either turn the knob down from there or reduce the volume in JRiver with the knob set to that point.
It's not blind AB testing and I'm already biases toward believing I won't hear a difference but I'm only trying to prove this to myself.
I'm pretty far from having golden ears anyway. I'm over 40 and I abused my ears when I was younger so I couldn't hear over 16K the last time I checked. And the last time I checked was pretty long ago.
Maybe I have brass ears, or copper at best.
-Chris
I don't know. Never heard any of the signal path you own. But usually given this is good quality gear you should be able to hear differences if any. Maybe you 'll hear nothing... sometimes it happen, even with musicians! 🙂 I've encountered people which doesn't notice how mp3 of poor quality change the acoustic cymbal of a drum set into TR808 synthesized one! 😀
All that to say that in some case you'll have to learn what to hear or practice your hearing to recognize some problems.
I'm not meaning anyone can discern some problems just that sometimes a learning curve is in action.
About particular music to use for tests it is usually easier to spot issues using acoustic music with not to heavily processed music (not overly compressed,eqed,...). Some Jazz recording are usually good example but if it's not your kind of music please find something you like! Because we don't want you to become one of this audiofool which stop listening to music to be some kind of 'medical examiner' of sound system! 😛
If you want i can make some crude downsampling, bit reduction or other horror to some of your audio file to have a clue of what to hear for. 😀
I can hear the scratchy, grating cymbals in low bitrate MP3 files very well. Really bugs me. 256 bit is about the lowest I can tolerate.
The system I built in my car with Scanspeak mids and tweeters is especially unforgiving of low bitrate sources. I think the analog to digital to get into my DSP makes it worse. I will hardly let my wife connect her iPhone to my car Bluetooth because she has so many 128 bit songs on there.
I listen to a lot of blues and fair amount of jazz so I've got a nice selection, including some live stuff where the microphones were carefully placed. Even have some well recorded classical with high dynamic range.
-Chris
Hi,
I'm afraid I can't give any "rule of thumb". I personally use Foobar and it's volume control seems to be transparent to my ears - but I've got no "golden ears". If it helps I use Asus Xonar DSX ASIO output set at 32 bit and volume control in precise 3 dB steps (just a personal preference). Graphical interface is done with WSH Panel Mod, though it only controls the default volume control (and does it logarithmically, unlike the default UI slider). If you're interested, I can share my GUI script; it only does buttons, volume, track name, rating and playback time.
I usually keep volume above -30 dB, plus another 10 dB digital attenuation in my DCX2496, so 16-bit files are basically played "as is" - 16-bit files with 24-bit stream gives 8 "disposable" bits, which is around 48 dB.
As I mentioned that I have DCX2496, I should state my opinions about active.
I have heard great passive speakers and bad active ones; IMO, one is not necessarily or automatically better than the other. It's all about the implementation.
Passive is better for the current state (really more like status quo) of the industry - tweakers are pleased, but it's yet not too complicated for an average music lover to play with different amps, speakers, and, of course, cables! Nothing can really go wrong, most of today's amplifiers even survive shorted speaker outputs.
Active needs to be an integrated system (like B&O Beolab series), not good for tweakers. Or you need to have rather good technical knowledge and ideally acoustical measurement setup (you need to be a serious tweaker). Get things wrong, and not only will it sound horrible; you can easily blow your 400€ tweeters! There is some middle ground; I believe Dali Megalines have an active "black box" crossover and require two power amps, but it's just easier and cheaper to make a passive commercial system.
What appeals to me about active is that I have the flexibility and adaptability. Whatever speaker I choose to build next I no longer need to buy expensive passive crossover components. I no longer need an expensive amp that's great in bass, mid and high frequencies, I use cheap amplifiers of which one is good at bass and the other at mids and highs. I still use passive x-overs between mids and tweeters but the components needed for a 5 kHz x-over are small and cheap.
Impossible for me to do that. I need the headroom for my processing as there is boost involved, another no no we usually learn here. 😀
But I do wonder how you use JRiver, what drivers and what settings and what equipment. As this is a strange thing you claim to hear. Have you ever done a loop back measurement of your DAC? To see if it performs as it should? Very enlightening to see what can go wrong if you haven't properly setup the equipment or drivers. By loop back I mean go from DAC out to mic input with a cable and measure the whole chain (including JRiver) with something like REW. That will show you what the setup is doing.
But don't blow up your mic input, you need to attenuate somewhere... Do not measure the amp this way without some serious resistors. Measure before the amp.
Last edited:
From my experience, first things to suffer are low level informations, informations about place where the recording is done (reverb cues), noise (in the recording) which become more 'electronic' than white or pink noise kind, stereo width and depth which suffer, sort of non pleasing 'grain' which appears on the whole picture (if i can say this way), space between instruments blurred, position of instrument in pan not clearly defined... When severely plagued you have that kind of 'edgy' sound.
I'm surprised. I would have think your line array won't need much more of a boost as it is probably relatively high sensibility and you re less subject to critical distance. What is your amp Ronald? And how much boost do you need overall?
About the nono theorically if you are using floating point bit depth there's no reason you can't boost. With fixed and low bit rate yes it's a nono but floating point allow boost no problem (as long as you dont clip inside treatments/plug ins).
OT: you should have something in your mailbox.
Last edited:
Pretty much mirrors my experience with mp3s.
Incidentally a couple of times I've come across where somebody posted sound files of what is removed when encoding them.
Sounded very much like listening to a reverb return on full wet.
Got it, thanks...
I've got no less than 16 dB of boost maximum at ~30 Hz. I play these lines from top to bottom, they do bass too. From 160 to 950 Hz I actually cut more than 10 dB. Above 10 KHz I need some boost too, but nowhere near as much as at the low end. All these numbers are relative and you need to keep in mind that on average it will be ~96 dB 1 meter, 1 watt sensitivity. It only drops 3 dB per doubling distance so most of the music is coasting along on less than a watt. So each of the 25 drivers is only getting a very small part of that watt to handle.
The processing is indeed floating point and as said, gain structure is definitely monitored within this setup. Clipping is avoided while maximizing the gain. It's a balancing act (lol).
My amp is a Pioneer A-757 Mark II. Would love to try different amps, just lack the funds right now. I've got a little sister amp, the A-447 to do ambient duty.
That doesn't sound like a description of my setup, judging a review of my system by someone less biased than me: http://www.diyaudio.com/forums/full-range/242171-making-two-towers-25-driver-full-range-line-array-169.html#post4589772
Last edited:
Typical digital artefact. When properly done and using high sampling rate/bit depth and using linear format digital is a great for capturing and reproducing transients in a maner analog tape could never do, but once problems occurs this destroy this advantages.
Mp3 principle is the most destructive you could think of: using masking effect (high volume low end signal mask all the even multiples higher up in frequency) and dynamic level comparison in order to 'keep' the information suposedly needed for accurate reproduction... all this driven by an algorythm... Well use a chainsaw with your message and you'll have the same results.
No surprise all info taken away are info about space!
Charles another very interesting thing to do is to perform MS matrixing and set mono signal to zero db (keep only the side information) then dematrixing to L/R. Very informative on which effects used and the informations included in 'stereo' mixes... 🙂
That's a serious amount of boost. From what you describe of your eq you have probably applied an iso compensation eq (kind of loudness adapted to your listening level). I don't know if this is intentional but it should be very confortable sound to listen to.
I've already read the review, as read your thread. Nearly cried when saw the box broked! 🙂 But at the end i don't know if i don't prefer the box how they are finished with epoxy and black paint! Anyway in a shop you could sell them for a serious amount of money!
I never said your system wasn't great sounding or flawed. But i think that if you have the opportunity to use another set of converters and a more powerfull amp one day(Given the eq on bass this could push the whole system to another level improving headroom and probably give a sense of easy drive imho!) maybe the need of a pad could appear. Anywway i would like to listen to the result as it is now!
Your amp look almost like one i used for more than 10 years with my monitors, but it was from Technics not Pioneer. Could be twins! What is the global impedance of each tower?
Last edited:
- Status
- Not open for further replies.