cfmsp Posted January 26, 2010 Author Share Posted January 26, 2010 "I might add you handle your self very nicely amongst this whirlwind of questions!" Thanks Jeff, most of the credit should probably go to those who provided constructive criticism, such as yourself. so, thanks for that as well, clay Link to comment
labjr Posted January 26, 2010 Share Posted January 26, 2010 "You can not have a low-jitter clock if it is powered by a SMPS" Metric Halo Seems to do a good job. So do the HRT Music Streamers. Link to comment
Audio_ELF Posted January 26, 2010 Share Posted January 26, 2010 ar-t said... "You can not have a low-jitter clock if it is powered by a SMPS and regulated with a '7805." I think there are a few manufacturers who would take offence at such an absolute statement... Eloise Edit: Quote editied to clarify the statement I was referring to. Eloise --- ...in my opinion / experience... While I agree "Everything may matter" working out what actually affects the sound is a trickier thing. And I agree "Trust your ears" but equally don't allow them to fool you - trust them with a bit of skepticism. keep your mind open... But mind your brain doesn't fall out. Link to comment
ar-t Posted January 26, 2010 Share Posted January 26, 2010 They can take all the exception they want. So, you guys are trying to tell me that you have a low-jitter device, that has noisy supplies inside? No, I don't buy that. So, when I measure some other gear, and I can see a change in the noise sidebands of the SPDIF signal by changing the wall wart to a quiet linear supply, that supplies don't make a difference? http://ar-t.co/ Link to comment
machinehead Posted January 26, 2010 Share Posted January 26, 2010 I think you are right, some folks work hard at making switching PSU's work very well. I think it might be like most things, in that if you really work hard at the details, you can make just about anything work well. \"It would be a mistake to demonize any particular philosophy. To do so forces people into entrenched positions and encourages the adoption of unhelpful defensive reactions, thus missing the opportunity for constructive dialog\"[br] - Martin Colloms - stereophile.com Link to comment
Audio_ELF Posted January 26, 2010 Share Posted January 26, 2010 In my case, I wasn't saying that power supply doesn't make a difference ... only there are several companies (such as Chord and Linn and I'm sure others) who use SMPS to great effect in both amplifers and digital technology such as DACs and Digital streamers. Eloise Eloise --- ...in my opinion / experience... While I agree "Everything may matter" working out what actually affects the sound is a trickier thing. And I agree "Trust your ears" but equally don't allow them to fool you - trust them with a bit of skepticism. keep your mind open... But mind your brain doesn't fall out. Link to comment
cfmsp Posted January 26, 2010 Author Share Posted January 26, 2010 re jeff's comments: http://en.wikipedia.org/wiki/Switched-mode_power_supply Here's a link that describes some of the shortcomings of SMPS that would need to be overcome. Re; Labjr's post, I don't know the actual jitter specs of the Metric Halo ULN-8, but by some accounts, the jitter is low enough to be considered a world class performer, sound quality-wise, despite their use of SMPS. Presumably they've addressed the shortcomings to a high degree. OTOH, I noticed a difference in sound quality when using a linear power supply with my ULN-2. There's been debate here as to whether the improvement was likely due to the ULN-2 itself or whether, as was posited, the improvement was only due to no longer dumping noise back into the AC circuit and degrading the sound quality of the other components. Given that I had not yet improved the quality of my AC, I"m inclined to believe the latter. clay Link to comment
labjr Posted January 26, 2010 Share Posted January 26, 2010 It would be cool if MH made a 2 channel "Audiophile" version of the ULN-8 with a linear power supply. Then we could compare the differences if a difference can actually be heard. Link to comment
glt Posted January 26, 2010 Share Posted January 26, 2010 High End companies differentiate their products by implementing proprietary digital filters. The ESS DAC does not have an interface for a separate digital filter. You could program your own, but people have not gotten that far. The Wolfson DAC has its own set of digital filters, The BB PCM chips have been designed to interface with an external digital filter. I believe those companies that have already invested in proprietary digital filters will continue to use that as a differentiating factor. Hence the selection of a particular DAC chip. www.hifiduino.wordpress.com Link to comment
ryw Posted January 29, 2010 Share Posted January 29, 2010 ESS Sabre 32 DAC can interface to external digital filter. It has a by-pass mode if the designer does not want to use the internal filter. ryw Link to comment
PeterSt Posted January 29, 2010 Share Posted January 29, 2010 "Besides, I think the DAC chip is only about 10% of the overall sound quality. The analog circuitry is probably 50%, the power supply is probably about 30%, the digital filter is about 10% and the DAC chip is probably around 10%. YMMV." It doesn't work like this because all is relative. The DAC chip architecture is 95%. The analog circuitry is 90%. The power supply is 96%. The digital filter is 90%. The DAC chip is 20% or less. The proper combination hence synergy is 100%. Now what ? This happens when you work on the individual elements as mentioned. And yes, even the last one is an element (and the most difficult one). It may not be for everyone to have all those aspects under your control, and it sure will be hard to interpret when you don't work on DACs in the first place. But if you would be able to change a DAC for these individual elements after one of them has been applied, I'd dare to mention the relative figures as I just did. For example : Once the analog side of the DAC has been squeezed to the best you can manage - and sound improved because of it by unheard differences compared to what you started off with, applying a well thought power supply already for the digital part only creates an improvement which is even more than what the analog stage implied. Once the chip architecture has been chosen wrongly, there's not much more to do on the filtering part. But do that the other way around, and the filtering creates the sound (which I indeed created myself). So, of course the above list could be turned into a normal total of 100% which would give something like this : The DAC chip architecture is 95% -> 25% The analog circuitry is 90% -> 22% The power supply is 96% -> 26% The digital filter is 90% -> 22% The DAC chip is 20% or less -> 5% The proper combination hence synergy is 100% -> hmm ... But it doesn't tell much now. The problem is, we need a commen reference, and we don't have that. An example to emphasize the importance of this (start shooting) : the ESS Sabre doesn't cut it for me. At all. It may sound nice, it may exhibit detail, you (not me) may be able to listen to it for hours and days in a row, but it will NOT show instruments. Because its architecture is as it is, it will not allow for an analogue stage as I know what can be done with it normally. Because of the same architecture, there is no way I would know of to create a filtering that would even come close to what can be achieved with "the other" architecture. Because the base is as it is (see before two sentences) no PSU will help improving it. So, what is worth "my" figure on the PSU now ? Despite the all so good (measuring) figures on the chip, what are they worth if ... well ... the measuring applied says nothing (not to me). This by itself is caused by the filtering again, and the "science" about that (which is just wrong). If the filtering is not working, the architecture is not working (think about this). This shouldn't be a discussion about the ESS Sabre being right or wrong, and to emphasize that within itself let me add that I think it is the best sounding chip within its ... architecture. No problem, because 95% of people is listening to such an architecture. Within that you can always pick your best sound. For me that is the ESS Sabre (in a well done implementation). Funnily enough, the other 5% is listening to the right (says me) architecture, but with the wrong implementation. Always. This starts with the filtering, and that implementation allowing for no, wrong, bad or whatever means of filtering, just because you can do anything (which sure is not the case with e.g. the Sabre). So, to end with something maybe not known to everyone, that other arhitecture easily shows 30% THD+N for certain frequencies and most people love it. Because of this, nobody is ever really trying to improve by means of PSUs or anything, *or* it is theory only (hey, what to measure when the base is 30% HD !). Could I have been more vague ? I guess not. But the relative figures - nicely adding up to 100% - say nothing. They count for Hong Kong DACs just the same but it sounds sh*t. Peter Lush^3-e Lush^2 Blaxius^2.5 Ethernet^3 HDMI^2 XLR^2 XXHighEnd (developer) Phasure NOS1 24/768 Async USB DAC (manufacturer) Phasure Mach III Audio PC with Linear PSU (manufacturer) Orelino & Orelo MKII Speakers (designer/supplier) Link to comment
cfmsp Posted January 29, 2010 Author Share Posted January 29, 2010 Peter, thanks, as always, for your reply. I read (and deciphered) this post faster than normal, so your ability to communicate in non-native language is improving, and for that I am grateful as I enjoy your posts, but don't always know for sure that I understand what you're trying to say. so, it sounds like, as with much of technology, once we start combining (too?) many elements into a single 'chip', we (well DAC designers such as yourself) lose the ability to maximize capabilities of the other aspects of DAC design - all in the name of technology advancement. Thanks for pointing to the DAC chip architecture (esp. it's shortcomings) as having the real impact on the sound quality, i.e. way more than the 5-10% of a DAC chip currently. clay Link to comment
Wavelength Posted January 29, 2010 Share Posted January 29, 2010 Gang, Listen to Pat (art) about this stuff. He is the master at oscillator and jitter reduction. The problem with switching supplies and oscillators is really something these companies need to take a look at. The ground noise caused by the switching element really kills the low jitter aspects of any oscillator. Ok here is an example that I just ran. I have a power supply for my oscillators that Pat and I worked on. His is different from mine but both of them are very low noise in like the less than 50nVrms range. Used with my 22.5792 oscillator (44.1, 88.2, 176.4) and the input to this very low noise supply yields about 1.4-1.7ps of jitter on the output using a linear 9V wall wart regulated supply. Now it I change that linear 9V to a good 9V switching supply that jitter goes to 12.2ps-15ps. If I put a secondary filter on the supply that only drops it to 11.8-14.6ps. Ok so all jitter specifications are based on the word clock. The reason is that this will be the highest value and is a direct result of the master clock which is in this case 22.5792Mhz. For each division of this clock the MINIMUM amount of added jitter is 3x. So if we drop the master clock down to 44.1K then that is a division of 9 times: 11.2896Mhz 5.6448 2.8224 1.4112 705.6KHz 352.8 176.4 88.2 44.1 Ok so 9*3 = 27 so 20*log(27) = 28.6272752831797 So the minimum jitter that the original WCLK could have is 1.4*28.6272752831797=40.0781853964516 The minimum under the switching supply would be 11.8*28.6272752831797 = 337.801848341521 This does not even account for the added jitter each of these division would have due to the same ground and supply noise issues that the switching supply would create. So really the switching supply would be at least 2x that value and the noise for the linear would probably put it in the 50ps area. Sorry for all the math, too much coffee this morning. ~~~~~~~~ Ok so the clocks are only one aspect. Think about it this way... Pat and I spent six months developing the power supply for the oscillators. Pat has spent a lifetime developing his oscillator. These only account for like 15% of the design. Get an appreciation for people like the Charlies of the world. Heck I have a QB9 here and there are no... I mean no 3 terminal regulators in the design at all. There is easily 20 or so power supplies on the analog and digital filter board. There is more discrete transistors on this board than I would dare to count. Not one opamp was used in this total design.... think about it, open your units up and see if the same can be said. Thanks Gordon J. Gordon Rankin Wavelength Audio http://www.usbdacs.com/ http://www.wavelengthaudio.com/ http://www.guitar-engines.com/ Link to comment
Lars Posted January 29, 2010 Share Posted January 29, 2010 "the ESS Sabre doesn't cut it for me. At all. It may sound nice, it may exhibit detail, you (not me) may be able to listen to it for hours and days in a row, but it will NOT show instruments." I think you haven't heard a well designed ESS Sabre based DAC. My real world experience is obviously different than yours. Wavelength Silver Crimson/Denominator USB DAC, Levinson 32/33H, Synergistic Research Cables and AC cables, Shunyata Hydra V-Ray II with King Cobra CX cable, Wilson Sasha WP speakers with Wilson Watch Dog Sub. Basis Debut V Vacuum turntable/ Grahm Phantom/Koetsu Jade Platinum. MacBook Pro 17\" 2.3GHz Quad Core i7, 8GB RAM, Pure Music, Decibel, Fidelia, AudioQuest Diamond USB Cable. Link to comment
PeterSt Posted January 29, 2010 Share Posted January 29, 2010 My real world experience is obviously different than yours. Hi Lars - Why didn't you leave it to just that ? IOW, what do you want for an answer ... that I think you never heard a good sounding DAC ? It's useless. So consider I didn't say that. Peter Lush^3-e Lush^2 Blaxius^2.5 Ethernet^3 HDMI^2 XLR^2 XXHighEnd (developer) Phasure NOS1 24/768 Async USB DAC (manufacturer) Phasure Mach III Audio PC with Linear PSU (manufacturer) Orelino & Orelo MKII Speakers (designer/supplier) Link to comment
Lars Posted January 29, 2010 Share Posted January 29, 2010 "Hi Lars - Why didn't you leave it to just that ?" Wavelength Audio displayed their Crimson DAC based on the Denominator module (ESS Sabre 32) at the last RMAF. I think it is one of the best DACs I have yet heard. Peter, what are the ESS based DACs you have heard to arrive at your conclusions? Thanks, Steve Wavelength Silver Crimson/Denominator USB DAC, Levinson 32/33H, Synergistic Research Cables and AC cables, Shunyata Hydra V-Ray II with King Cobra CX cable, Wilson Sasha WP speakers with Wilson Watch Dog Sub. Basis Debut V Vacuum turntable/ Grahm Phantom/Koetsu Jade Platinum. MacBook Pro 17\" 2.3GHz Quad Core i7, 8GB RAM, Pure Music, Decibel, Fidelia, AudioQuest Diamond USB Cable. Link to comment
PeterSt Posted January 29, 2010 Share Posted January 29, 2010 I have just been typing for an hour or so, but when finished and reading it through I realized that I might be offending here and there. Explainable perhaps, but I don't like it much. We can leave it like this but no complaints please, or ask for it again and I'll post it with the message at the end that I hope not to offend. Hmm ... (stupid) Peter Lush^3-e Lush^2 Blaxius^2.5 Ethernet^3 HDMI^2 XLR^2 XXHighEnd (developer) Phasure NOS1 24/768 Async USB DAC (manufacturer) Phasure Mach III Audio PC with Linear PSU (manufacturer) Orelino & Orelo MKII Speakers (designer/supplier) Link to comment
machinehead Posted January 29, 2010 Share Posted January 29, 2010 VERY good info, thanks Gordon. Jeff \"It would be a mistake to demonize any particular philosophy. To do so forces people into entrenched positions and encourages the adoption of unhelpful defensive reactions, thus missing the opportunity for constructive dialog\"[br] - Martin Colloms - stereophile.com Link to comment
barrows Posted January 29, 2010 Share Posted January 29, 2010 That Peter has stated in other threads that he does not believe that any sigma delta DAC can sound good. So by nature he is not going to like ESS, TI (except perhaps the old school 1704), AD, Crystal, Wolfson, AKM (although they claim their sigma delta chips do not sound like sigma delta chips) etc. DAC chips. I would presume (and yes, presumption can be wrong) that by extension he is also not a believer in DSD, as it shares in common the extreme noise shaping used in all sigma delta DACs. SO/ROON/HQPe: DSD 512-Sonore opticalModuleDeluxe-Signature Rendu optical with Well Tempered Clock--DIY DSC-2 DAC with SC Pure Clock--DIY Purifi Amplifier-Focus Audio FS888 speakers-JL E 112 sub-Nordost Tyr USB, DIY EventHorizon AC cables, Iconoclast XLR & speaker cables, Synergistic Purple Fuses, Spacetime system clarifiers. ISOAcoustics Oreas footers. SONORE computer audio | opticalRendu | ultraRendu | microRendu | Signature Rendu SE | Accessories | Software | Link to comment
Telstar Posted January 29, 2010 Share Posted January 29, 2010 What barrows said apply for me too AKM dacs are indeed a strange breed and it is probably the main reason why the MH ULN8 sounds good (havent heard it yet, but I have listened to other AKM based units). Link to comment
The Computer Audiophile Posted January 29, 2010 Share Posted January 29, 2010 Peter - Based on our conversations I believe you have a vested interest involved here. Please disclose this interest or do not comment further about DAC performance. Founder of Audiophile Style | My Audio Systems Link to comment
cfmsp Posted January 29, 2010 Author Share Posted January 29, 2010 "VERY good info, thanks Gordon." I agree with Jeff. Labjr, yes it would be great if Metric Halo made versions of their DACs which used more typical (read: normal for audiophile gear) AC so that we can hear for ourselves. The more I try to improve my understanding of AC and it's impact, the more I start to wonder if SMPS isn't the limitation of the ULN-2, and perhaps even the ULN-8. cheers, clay Link to comment
Telstar Posted January 29, 2010 Share Posted January 29, 2010 Chris, I guess that Peter's interest is no more no less like Gordon or Pat (art) here, with the only difference that his electronics are not on the market yet. Link to comment
The Computer Audiophile Posted January 29, 2010 Share Posted January 29, 2010 I agree 100%. Gordon & Pat make their interests well known whereas Peter has yet to say anything about his. Founder of Audiophile Style | My Audio Systems Link to comment
ar-t Posted January 29, 2010 Share Posted January 29, 2010 I dunno.........he has a link in his sig to a forum of his own. No idea what it is about. Maybe someone would like to give us a 25 words or less description. http://ar-t.co/ Link to comment
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now