Jump to content


  • Content Count

  • Joined

  • Last visited

Everything posted by PeterSt

  1. You do know that a subwoofer in the room (that is, with a real LF driver inside of it) just plays along with the music, right ? (IIRC some LF designs even depend on this phenomenon) Not that I attribute real value to it regarding the tests you refer to, Paul, but I would be careful with that one.
  2. Last one from my hand (I think): ultimate transparency (= neutrality). From day one in this business I couldn't stand two tracks sounding similar (let alone two different albums from the same band, let alone even more albums from different artists). N.b.: This relates highly to mentioned "buzzing" because that is a clear flavor. Everyone using our gear knows what I am talking about. And the result ? ... Try to find even one smallest expression of customers of how things sound; only the very first DACs could encourage for telling something about the sound, but already the first incarnation with USB interface would show nothing about that. A 6 moons review ? you can look it up. Nothing about "how it sounds". There just is no sound when doing things a kinda right in this world. This counts for each of our products because I just don't want them to make sound as such. So the very key element of the "better" system is: it should not put its own flavor over your music. Not your system and also not its underlying elements (no tweaks to solve issues).
  3. My measure "forever" is about standing waves. For years already none can be detected in any corner or position in the room (I let auditioners always test this) and as soon as something re-appears regarding this, I know something is amiss. Btw, this is not only about low frequency standing waves, as "any" higher frequency is even more annoying (this buzzes). Notice that this won't really be something which people recognize; you are just used to it. Unless of course, you encountered the transition from buzzing" sound to none, regarding this. I hear ya. Room-treatment; a real fine system can do with way less room-treatment than you thought ... (in the end it deadens).
  4. That you really need to wait for applause NOT to emerge at the end of the track, to finally know it was a studio recording. Believe it or not. The echos I agree fully with. I have recordings which so extremely fill the room with sound, that it's too crazy to be true (and I know it wasn't so before).
  5. And this is exactly why this is b.s. and why it is a good thing that Marce is no longer "with us". This is totally unrelated to audio. Think of this please; About everything complies to specs for audio. And for that reason everything sounds the same. Right or wrong ? But that's Marce all right. Welcome to this objective board stuff.
  6. Ehm ... maybe you forget to use recording software ? Thus, what you record must also be output. Regarding you using Audacity, it looks to be fine, but still saying ... - Input must be captured by the proper device; - output must be output by the proper device.; both to be set in Audacity or in the subsequent programS. In Windows 7 an "loop back" soft feature existed - not sure whether that still exists. I suppose this could have helped to set up your situation. The ping sounding only testifies of the system itself also mangling in (that requiring that mentioned common sampling rate). In itself this can be solved by switching off Windows Sounds (I suppose you can find that). Not sure it is important in this case (but for recorded SQ reasons it can be).
  7. Paul^2, you can try your own measurements and tests ... And since that was about a digital application, it's a fairly strong test if that's projected on to analogue applications (like discussed in here). Want an analogue cable now ? 🙃 PS:
  8. Not sure at all whether this is going to help, but it could be related to no two applications being allowed to take exclusive control to either the output device or input device or both. Go to Settings and choose sound: In there, choose Sound Control Panel: Choose your output device (and later, your microphone) and click Properties: Go to the Advanced tab and then UNcheck the "exclusive mode" box: and see whether this helps. For best audio performance, playback software - but also recording software, may take exclusive access to the device it uses. But only one at the time can do this; with playback and microphone active at the same time, both may require that exclusive access while the sound system of Windows does not allow this. In addition I can tell that depending on circumstances not even one device can be granted exclusive access; Later, when you have it running, for best SQ performance you can try to set back step by step (one by one) to the default. If the one doesn't work out, revert and try the other. This latter can also be related to the sampling rate ("Format") you see in the last screen above; no two different bit depths and rates can be active at the same time. When you got the grasp of this somewhat, envision that it is not only your microphone which is processed on the "input" side, but this is also your windows sounds like for email notifications, YouTube and everything which can normally play sound in parallel. The system needs to convert to a common sampling rate for this and with any of the devices in "exclusive mode" this can't happen. That one source (granted exclusive mode access to the device) will play but the others remain silent. So it's my estimate that somewhere here lies your problem. Peter
  9. Chris, you told me to stay on topic (together with Alex), while I have only been busy fighting his being off topic. Apparently you regard me being offtopic as well. Highly motivating that is. Not.
  10. Kal, can you give me a couple of clues here ? Apparently I am talking in riddles to you guys. Thanks.
  11. It is to deafman's ears, as we say it over here. I never related this to audio phenomena. Today I do. It seems that these days it is the higher task to tease away people. Yes I am pissed (at).
  12. What one also must realize is how a system subject to digital VC is always at full gain. You have high sensitive speakers and a somewhat noisy poweramp ? then forget it because you will hear the noise. The clear upside: Oh, do you hear noise ? well, better do something about that (new poweramp) because if anything is bad to good SQ is it noise. Ah, we knew that. But did we also know that it is always there, and that turning down your analog VC really is the same as putting your head in the sand ? (SNR (relation) really doesn't change). Generally spoken, when you use a 1KW Spectral poweramp of which you normally open the VC only to 00:45 (quarter-to) may exhibit crazy noise when set wide open. This only means that the relation between your poweramp and the sensitivity of the speaker is not right at all. So Alex, there you have your subject ... set to 00:45 at your free will (because else too loud), but the SQ will be very poor because of a too bad SNR. There's no way out. Oh, there sure is: just accept poor SQ and that it is fine with you (whoever). Nothing wrong with it for most. But nothing wrong with good base ideas either. Unless they are no good of course. 🙃
  13. Alex, that is the original base. For me this does not necessarily mean that the bespoke "match" will be there, but this is why I point it out explicitly ("prerequisite" and such). Maybe that is how people would like to see it. But it cannot exist. It is too fragile (too many constraints to do it really well). Point is also: with the digital attenuation you can clearly/plainly see what happens (once the basic knowledge is there). Not so with the analog VC. Not at all. You thus can also easily see what it requires. Can't or won't meet that ? then no (real) dice.
  14. Well ... I recall someone with ever back (very) expensive gear that requires setting up, learning a bit (much), a healthy dose of enthusiasm etc. and ... ? What ? haha I also recall myself to solder wires on to my own DAC with the ever recurring promise to eye-measure something called a Lush cable, compare that with other cables and give some kind of yest unknown verdict. This was before purchasing that (brand new) gear. Nothing helps with me. I am at a total loss. 😁 The wires still stick out of my DAC. Maybe nice antennas for something (I never heard a downside from them). I see more people around with similar promises and spending a lot of time. Somehow measurement requires too much time to push in between jobs. Anyway, I always think: once upon a time I will find that week to work on it. Then economy started to grow and I had no time for anything any more. Today there's talk of a virus. Maybe that brings time again. But yes, of course I am eager. Aren't you ?
  15. Yes Alex, you are right. No Alex, you are wrong. You are always wrong in my personal book because once you have this up and running in the fashion I project, there is no single way to ever listen to vinyl and all you mention because it then plainly hurts. You are always right because you want to be and go off-topic for it. What's wrong with this is that people tend to first create their own subjects, in order to next be right on it. Not a good thing to do.
  16. Here is another one: An analog attenuator "squeezes" (squashes) the sound. Per the means of shifting down the whole lot of a 16 bit range (using the upper 16 bit part of 24 bits at first), by one bit (still 16 bits in use but not using the most significant bit any more) nothing gets squeezed. You could say that the dynamic range stays the same. This is not easy to prove without a scope, and influences come from more angles, but for fun try to think I could be right. This is also how an analog attenuator could lose or sustain low level detail - the way the attenuation is done (which are a few) may squeeze more or less. With digital it can be done the wrong way too, but if you envision my mentioned "shifting down" of the whole lot, nothing can be lost. It would be true though that the quality of the Least Significant bit degrades, but now suddenly so many real practice attributes play a role, that it becomes too difficult to explain in a few words (but think what upsampling implies and that with all modern applications digital filtering is a requirement (which implies upsampling)). Etc.
  17. But Alex, this doesn't make any sense to me. If others need to attenuate too much, it only implies that they have over-weighted amplifiers. Answering your question additionally does not make sense to me, because I put it forward as a prerequisite to have a matching amplifier(s) in the first place.
  18. I have a question, but I hardly dare to ask it ... This seems the xth time that I observe a discussion about what is and what is not suitable for this "Objective-Fi" part of the forum. As if we speak different languages; Chris, do you recall a quite long discussion about the phenomenon "musicality" and how dozens of posts later it appeared that you have a quite different meaning for that compared to the rest of the world ? we ended with a consensus that "in American" this bears a different meaning. Remember ? Debates like this remind me of that. In about each post DBR100 refers to MIT measuring their cables. All right, he calls it "testing" but in cable realm this would be measuring. Now, the sheer fact that nor MIT and especially DBR100 does not have testing (measurement) results, does not make MIT create cables by subjective courses. And thus, while he refers to that all the time, you put that in the subjectivity corner. I can't see why. I have been in the vdHul factory and esepcially knowing Mr vdHul personally, I can guarantee you that as a physician he is not toying around with hoaxes. That we might not grasp what he is doing, is an other matter. Maybe to draw the attention away from this back and forth (and indeed strange) subject, here's a text which possibly also is not allowed for this part of the forum: You (yesterday) told yourself about Gotham Cable must be good enough for your XLR application because of star-quad etc. topology, x shielding and possibly more. In response I told you something like "but it comes without measurement results". I told you so because it bothers me from Gotham (and especially that cable). Still you chose it, still you even dare to state it has to be good and still ... I would do too. So some characteristics tell you (and me) that this cable has to be good. Maybe my virtual math is different from yours, but anyway I selected that cable after a longer search for "best specs" and XLR. Did you tell about listening results ? no. Could I listen during my research for a good (good ?!?) cable ? no. Now the worst part: I have the measurement gear for this. Bought to test our own produced cables so people could have some data instead of "listening results" only, which are subjective indeed. I think it is two years ago now that I bought this gear (worth a small car). It is still unopened in its original packing. I should have this gear and use this gear to support my own claims of what I think is good about a cable. This includes the specs of cables like a bandwidth of 5MHz over 100m with 3dB roll off. That cable too (not being the XLR) was purchased pure on specs. No subjectivity anywhere and bought by the spool at 305m. Start producing the audio cable from it, not even listen to it first. Don't even use that gear to check the claims of the producer (I trust this world of formal datasheets). Have I lost you by now ? probably. But I don't see the difference with the MIT references, never mind it looks a bit suspicious. At least I myself don't see any subjectivity in it. My own work the same. One thing of course: it is me who decides which specs of a cable ought to be good for the audio application. Same like your star-quad and the Gotham. That could be subjective. If I don't make any sense anywhere, just tell me, so I myself understand better how this objective part of the forum can be fed with objective contribution. I am fine with any comment.
  19. Sure you can, as it doesn't work that way. Alex, as we know, almost everything we change ends up to be "colouration" because the transparency fades and characteristic sound enters. The easiest example could be an interlink or speaker cable. Different types imply a very different sound, and it does not express in THD that I know of (or could measure myself). But it implies impedance response differences and much is about that, with the notice that these responses hammer on frequency ranges and that it thus is all about consistency in the sound that bothers hence causes the colouration. If you don't have that as flat as possible to begin with, it can't be corrected afterwards (theoretically with DSP, but never mind). It never harms to start out the good (best) way and one has to start somewhere. Now coincidentally the digital volume control is so much right at the source ... you can't go deeper into that (although there surely *is* more beyond, like the software influences themselves and the PC used). So I feel that doesn't count. And obviously I started out with software alone or a better "transport" so to speak (back in 2005/2006). DACs came about a couple of years later. Speakers and such again half a decade later. And all of the other stuff. But all was based upon the source (the software), and slowly making small next steps (like the DAC being a logical next step). The references always remained - read: I never applied changes backwards in the path in order to let excel a new product further down the line. Really never - trust me. What could happen though is that new "influencers" were detected along the timeline, such as the PC being of vast importance. Or how I finally saw the light (after literally refusing for several years) in USB cables making a difference. So I tell you: none of these elements are required to let excel an other. Cables can be used stand-alone and do their audible job. The PC can. The speaker can. The software can. The DAC can; Still it would be true that replacing each of these elements with a "lesser" part, will be audible immediately. But that doesn't prevent the other elements of doing their job in a positive sense. Making this whole chain around "digital" a well performing chain is quite a task and hopefully this shows in this little VC subject alone (it really is such a small part of it !). But saying "oh well I don't have Peter's ultra high efficiency speakers so why bother" won't be the good approach. I know you are not really saying that, but still ... just saying ... Regards, Peter
  20. Anthony, what a superb post. For others, Anthony may have one of the most exciting systems on the globe. Already the effort he has put in it ... wow. But Anthony, you are right. When you have too many other sources, the input selector is a must, and with that there will be analogue sources just the same and now you need the analogue volume control. I agree from one of my last escapades that the Muses chip is the better one, and maybe you recall me working on that headphone amp (which ended up as gain stage in the G3 incarnation of your DAC) and how I obtained a "consistently working" version from China somewhere. Well, after it stayed out (shipping) for 3 months, at arrival it did not work. I think it was there where I gave up the headphone amp idea, which also would have been "a best" pre-amp (input selector with VC - no gain). Maybe I should continue a project like that. I have one cabinet for it as well - haha. Btw, I wonder ... @manisandher at some stage told he found a quite transparent pre-amp. I don't recall he shared that publically and I can't find it back in my emails. Mani, could that still be something to go for ?
  21. I did not realize there had to be a message. But if I were to hand one: Using any means of analogue attenuation, makes music sound like a dead bird. All is relative of course, but in my (say thought-over system regarding this) situation this always is so. This counts for the relatively cheap Placette, mentioned in the other thread, but also for 60K pre-amps and so many in between. What they have in common is unlinearity already at the noise level (no signal playing) or how that noise response changes when a signal plays. On an another side note, after the first version of our DAC was ready, I worked for 6 more months on a decent attenuation means (any means, cost no object on components, like USD 1200 or whatever on VC working with light), before I gave up and started shipping without attenuation. As I said in the other thread, only voltage controlled D/A chips would do what I want.
  22. A typo. But I can't edit it any more. THD.
  23. PS: I corrected some unclarities in the last two posts (I left the remainder be ;-).
  24. The very first thing I did in this audio life, was creating a digital volume control (in the software) which was lossless at its step settings. Back at the time this was a first, later this was copied by others. This is totally unrelated to 64 bits etc., but how the unattenuated digital code transfers to an attenuated code of any of the offered attenuation steps, and how they can be transferred back to the original (the "lossless" thing). This came forward from me doing the very least as possible to the signal throughout (which later went through the NOS1 DAC as well with similar thoughts - but at first working with NOS DACs from other brands). This also meant: no smart rounding whatsoever, which rounding always implies unlinearity (which in the end implies IMD distortion) but no necessity to round instead. Think like 4/2 = 2 and 2*2 = 100% 4 again. This in itself is audible already (no matter the ultra low levels where this seemingly plays) and it springs from the era of "bit perfectness". So that too implied a similar distortion and people could hear it (between their ears maybe, but alas). Digital attenuation bears another important prerequisite and that is the headroom to do it. For example, if you start out with a 16 bit file and you attenuate it 1 bit (this is how it works - you downscale the whole code one bit lower, so to speak), you need 17 bits of headroom to do it. Attenuate 2 bits, and you'll need 18 bits. Attenuate 8 bits and you need ... 24 bits. Aha. Nice. Because back at the time the 24 bit DAC's really started to emerge. So : When you attenuate 8 bits - which equals to 48dB, you lose nothing. No signal degradation (which is the first people may see as a downside). But when you have a 1000W amplifier into a 100dB sensitve speaker, you have a problem because you will undoubtedly require more than 48dB of attenuation, and you will start to lose bits. Mind you, this may not be audible per se, but you will be doing it wrongly. And so it requires a kind of match with the power amp because remember, we ditched the analogue volume control. In the end this is also related to the voltage output of the DAC, because when that is (way) too high, the problem is similar - you need to attenuate relatively more. Lastly for these "series", once you use digital attenuation, you can also use digital volume normalization (like ReplayGain in Foobar). Thus, it is now totally harmless to scale all to a certain level per track (or album), assumed you have the headroom for that on the upper (!) side. Huh ? Yes, so when you play your softest music, you should be able to play at -0dBFS (the FS is from "full digital scale") so it is audible to your liking, with the notice that boosting to e.g. +4.5dBFS usually is allowed for music which plays so soft. So this is a quite special part of the digital attenuator, just because it can do it, but it must know whether it is allowed (the digital music file is always limited for its digital headroom to use, and when we'd boost by 6dB, it requires only half or less to be used in-file). If you can't hear your softest music loud enough, your amplifier has insufficient gain (or read: you will actually need a pre-amp). So the poweramp should just match the range you play your (the) music in, and this in the end also related to the efficiency of (new !) speakers. So digital attenuation is a bit more difficult to apply well, but when that has been done, it is very worth while. Happy Easter !
  • Create New...