@MarcelvdG, If your ears tell you clipping is the worst problem, then by all means go after it first. If only your instruments tell you...well, the mastering engineer and client decided they wanted clipped.
The other thing I keep getting reminded of is not to design based on compensating for bad recordings. Use good recordings with simple micing that capture room sound. Try to get it sounding real. The bad recordings are a distraction. For example, maybe try Janis Ian's "Breaking Silence." Lots of dynamics and space in it. Sounds like real instruments in a real space.
The other thing I keep getting reminded of is not to design based on compensating for bad recordings. Use good recordings with simple micing that capture room sound. Try to get it sounding real. The bad recordings are a distraction. For example, maybe try Janis Ian's "Breaking Silence." Lots of dynamics and space in it. Sounds like real instruments in a real space.
I have not seen any proof that B excludes A. And why would it? No reason why well performing system could not also sound good.
D90 measures clearly better than DAC-3 so this contradicts your "System A" vs. System B" comment. Another thing is that both D90 and D90SE seem to have problems with clocks and Vref as the noise skirts are far from perfect.
Don't see a contradiction as a practical matter. There is no requirement that a better sounding device measure worse, only that it be chosen on the basis of better sound rather than on better measurements.
Also, agree about D90 noise skirts and clocking. Still a good sounding DAC for the price IMHO, but not the best AK4499 could potentially do. Makes a good backup dac.
Design goals are not the problem, philosophical or otherwise. The targeting/selecting of specifications/metrics toward the objective of delivering the design goals, however, can be a problem when it comes to music reproduction systems. I suspect that we would most all agree, that the ultimate goal of a home HiFi system should be to obtain reproduction which sounds like whatever the original live captured event sounded like, had we been there to hear it. With this set as the ultimate design goal, as engineers, we then logically proceed to specify a signal transfer path which is as perfect as possible. Although, it should be highlighted that audio engineering specifications do not need to exceed the resolving limits of the human hearing system. Specifying to exceed those limits is overkill, but such is the nature of engineering 'progress'.
The trouble seems to come about from the subjective listening experience via a surprising number of superbly specified systems. An experience which often runs counter to what those superb specifications would seem to dictate. At least, regarding typically provided performance specifications. Said another way, the trouble seems to come from objectively poor measured performing systems, despite subjectively sounding excellent. My own logic is that, since physical devices behave according to physical laws, physical performance specifications are required to engineer the physical signal path. The issue is, that it seems, there is significantly more going on regarding human ear/brain perception (particularly, via stereo listening) than is accounted for in engineering only for the physical signal path. Once a holistic view of the reproduction chain, including the ear/brain behavior, is fully and properly taken in to account (of which, I'm insufficiently knowledgeable) there will necessarily be physical parameters which enable that. Parameters which can serve to identify those engineering specifications which more truly/effectively describe the subjective listening experience.
To be clear, I do not see this as a problem with measurement instrument resolution, because test instrument resolutions already easily surpasses the resolving ability of human hearing. Logically, this is, instead, an issue of whether the right parameters are being measured, within the context of their dynamic behavior, and within an interrelated interpretation. Finally, how to effectively communicate those relevant parameters, which might easily be quite complex in nature, to human eyes.
Last edited:
MarcelvdG said
Do you use digital volume control or is the volume always at 100%? How does your DAC fare on this test? 7040483 with the digital volume at maximum?
The digital volume is always at maximum, I use the (analog) remote control of my Prima Luna Dialogue Two amp. I do the same in my other sound equipment, in the same one, my noble Technics SPLG300, which uses the revered -in those years and for some- MASH chip. It also has a volume control - digital - and I don't use that either. I have read that they introduce distortion, because they do not act on the analog volume potentiometer, but instead do the conversion on the DAC. I am not an expert in digital technologies, but it seemed credible to me. So I always, always control the volume analogically from whatever amp it is, with or without RC. Why you ask ? Help me understand and maybe I'll change my habits.
Anyway, in case you didn't understand, my previous post was in an ironic sense, this discussion is the same as in the 70's when the different brands competed in the distortion rates of the amplifiers, first it was the harmonic distortion, then that is not it was enough and a tremendous discovery had to be measured and published, the intermodulation distortion! But a vintage Dynaco ST 70 tube kept rolling any SS amp from those years on the floor in SQ! (except for my NAD 3020, you know, there are always exceptions, and to my credit... 😉 )
And nobody discusses or measures the high distortion of the speakers? (read the above link I provided) Does your high distortion play a role in this match? I know audiophiles who invest fortunes in electronic gear and listen with a pair of vintage 2-way bookshelf´s...
I think that the difference in SQ between Mash and ES9038PRO, after so many years, is easy to perceive -for trained ears- but I also think that it does not add much to the final quality of a current system to make this type of comparisons between DACs. s current, the cow no longer gives more milk....
But, to each his own, I think the OP's statement about the Saber chip is very biased, too much. Perhaps he was on purpose looking for controversy? You could also start a new discussion here by hijacking the thread:
Why doesn't any DAC offer SQ similar to a good vinyl record played on a good MC cartridge? With my Denon DL 103 cartridge there are no wholesale details, nor is there too much dynamic range -in that aspect the limit is the support, the vinyl, and its necessary Riaa EQ-, but hey, that sound is musical, very real, without details that do not affect the general enjoyment of the work! Does anyone at a live concert, even if located a few meters from the musicians, hear all those details? I'm talking about music without amplification, obviously. Did the violinist drop the sheet music on the floor and have to put it back on the music stand? Who cares to hear that? It reaches how many DB? Well, not me, but it may be that some people want to listen to that and also the flatulence that musicians expel from time to time....if they like it, that's it.....😳
Last edited:
Do you know which artist did that?
Trying to spread FUD again. 🙄
Which test method revealed such small differences between those DACs? That's an extraordinary claim. It would be only fair to have it accompanied by extraordinary evidence. Please share it with us.
I just wanted to know if your test conditions were comparable to those of the OP. They are regarding this point: no digital volume control in either case.
Digital volume control has its advantage and its disadvantage. The disadvantage is that the noise floor at low volumes can be higher than with a volume control further down the chain, that is, than with an analogue volume control. That noise floor includes any low-level artefacts of the DAC (for example, some sigma-delta DACs suffer from idle tones that get frequency-modulated by the signal).
The advantage is that as long as the volume is at least a few dB below maximum, you automatically have headroom to prevent clipping on intersample overshoots. This is only true if the signal passes through the digital volume control before getting to the first digital filter.
I didn't understand that, thanks for clarifying.
I certainly can't always distinguish the individual voices in a choir or individual instruments in an orchestra when listening to a live performance without amplification.
I use some of my DACs also as signal generators when making measurements which makes low distortion and low noise worthy goals. Also IMO reaching the performance specified in datasheet is an indication of successful implementation. These goals do not preclude good sound.
It would be nice to have the inclusion of the phase noise figures in spec sheets, in addition to all other specifications that we already have, like distortion and noise.
No proof on my part, just opinions.
Your information about the digital volume control is interesting. I'll put the Oppo a few DB before the maximum, thanks.
And with regard to the identification of the voices, it is true, in some cases it is impossible, the more members there are in a choir, the more complicated things are, but the whole is enjoyed. It is the idea of the choirs, to achieve vocal simultaneity and sound amplitude, magnificence.
Uhm, perhaps. But ime some digital volume controls are great and others suck at any volume.
Looking at the posted DR results for that not particularly impressive. https://dr.loudness-war.info/album/view/62602 Whilst not a conclusive metric we don't have much else to go on and I'd file that as average+ for dynamics.
😊That gives hope to digital lovers, I am currently satisfied with my means of reproduction, I highlight vinyl because I am a war veteran in this hobby and I have hundreds of LPs, but I am not a vintage fundamentalist either. Streaming has allowed me to discover a lot of new music for me.
The only support that I definitively abandon is the cassettes because I ran out of deck and I'm afraid to buy used ones. But I would like to have a Nakamichi ZX7, or an Akai 570 D in good condition to enjoy many cassettes that I recorded with musical works that I never listened to again. Not everything is on the web.
P.E. a recording of Mark Levinson Recording that I don't remember the name, but it's out there, somewhere in my cave, I'll see if I can find it, maybe someone can help me .....
I had a chance to buy one in excellent condition but there were not enough funds in the bank......
https://audio-database.com/NAKAMICHI/player/zx-7-e.html
😢
Because digital audio is too good as sound reproducing device to downgrade and mimic vinyl on MC cartridge.
Its utility is not limited to its measured dynamics. It has good imaging, well recorded room reverb tails, was recorded in a way intended to sound correct on a system optimized to sound like real music, etc. The name of the song suggests the sudden and dramatic, yet only occasional, variations in dynamics. In order to hear the low level room sound decays, the volume level needs to be turned up. The few 'breakings of silence' are then powerful to human perception. Its a piece more useful for human use in system evaluation than for impressive measurement numbers of one parameter. I wouldn't have recommended it if it were not of good utility, at least once learned and familiar to a attentive listener. If no interest, that's fine. I would still recommend familiarization with it to others who may have interest.
- Status
- Not open for further replies.