If you start with a TIM(30) Sine-Square test signal that is often available with modern test equipment like the SR-1, and then put it through a passive inverse RIAA network, you get essentially the same sort of rise-time test signal (fig 8). Compare 7 and 8 carefully and you will see a remarkable similarity between them. Note that they have different time scales, but with the same time scale, they would overlay virtually on top of each other.
They clip into an A/D, then reduce the level digitally. It's still standard practice not to ship CDs at 0 dBFS. Clipping has become preferred to limiting in many cases because it is believed to sound better by some of the people who have compared the two.
Last edited:
That's an interesting observation. Does that imply that the cart is also 30kHz band-limited, like the TIM30 signal?
Jan
Well you don't play vinyl so that is not suprising. As I have said my view is that a 1-2 sample clip pales into nothing compared to the catridge going into a tank slapper. Playing live you would hardly notice it, and if you were de-clicking the sofware seems to have an easier job than if the transient were 'smeared' by filtering.
It's all a bit of fun as trying to improve vinyl is after all a fruitless persuit, but keeps me amused.
Jan: Based on some more recent measurements there it would seem the cantilever may behave more like a transmission line rather than a rigid beam. Given all the different materials for a cantiliver and damping methods this could end up with all sorts of different responses to a high rise time stimulus. It has been proven that a number of modern cartridges can play CD4 records which have a 38kHz subcarrier which could be due a cantiliver resonant mode at the right place.
Having said all that, electronics that does DC to 100kHz is almost trivial next to the mechanical problems.
Having said all that, electronics that does DC to 100kHz is almost trivial next to the mechanical problems.
Last year, I bought some jitter measuring gear and also a couple of Mastering recorders as well as the ADC -1 to document the recording side of the chain;
Besides the DAC3 increasing headroom to prevent a too often common problem of intersample overs, there are more issues when there is jitter. One area that Benchmark has improved with the DAC 3 over previous models is in further jitter reduction... in the area of its affect on the anti-alias filter. "small amounts of jitter can severely degrade stop-band performance, and can render these filters useless for preventing aliasing"
THx-RNMarsh
Besides the DAC3 increasing headroom to prevent a too often common problem of intersample overs, there are more issues when there is jitter. One area that Benchmark has improved with the DAC 3 over previous models is in further jitter reduction... in the area of its affect on the anti-alias filter. "small amounts of jitter can severely degrade stop-band performance, and can render these filters useless for preventing aliasing"
THx-RNMarsh
Last edited:
Fancy buzzwords to describe how they use an SRC4392 ASRC on all their incoming audio before they feed it to the ESS DAC.
I am not sure about their newest claim, but I think it's essentially that by increasing the sample rate first with the ASRC, you are now moving the transition band to a much higher frequency. The point of which, I guess, is now allowing you to use a DAC with a digital filter that is not in the stopband by Fs/2 but not have to worry about any aliasing due to that.
They spend a lot of time on various sources of jitter... their design (?) uses their UltraLock (tm) converters. It eliminates jitter as a major variable.
"No traces of jitter-induced distortions are detectable at our measurement limits (-144 dBFS)."
I have not read their tech info on the filter mentioned. But its interesting.
THx-RNMarsh
Last edited:
But didn't benchmark claim all the same anti-jitter stuff on the DAC2? Other than the new DAC chip the rest of the digital path seems very similar? And hey if you've done the R&D nothing wrong with that. I'm still not sure it's worth the premium over a dacmagic plus, but even that is currently out my price bracket as I need multiple channels.
Riddle me this. If you are driving your speakers active rather than passive, do the DACs need to be as good if only doing part of the frequency band ?
Riddle me this. If you are driving your speakers active rather than passive, do the DACs need to be as good if only doing part of the frequency band ?
It was a bit of a loaded question. if they still have the external SRC and then use an ESS chip its a bit redundant. The ESS has an SRC at its input that can be defeated with effort and uses a much higher frequency clock that is local. One can argue that the higher frequency clock (100 MHz) will have higher jitter. Its redeemable with heroic efforts http://www.wenzel.com/wp-content/uploads/VHF-ULN.pdf but there is a valid argument that there is an absolute floor below which jitter or phase noise will have no effect on the analog output.
I think Benchmark is getting close to the marketing story departing from engineering facts. From personal experience it gets very uncomfortable trying to defend stories that are not based on reality.
I think Benchmark is getting close to the marketing story departing from engineering facts. From personal experience it gets very uncomfortable trying to defend stories that are not based on reality.
??
https://benchmarkmedia.com/products/benchmark-dac3-hgc-digital-to-analog-audio-converter
"The DAC3 builds upon Benchmark’s highly successful DAC2 product family. The DAC3 maintains the familiar DAC2 form factor and feature set, but adds the higher performance available from the new ES9028PRO D/A converter. " The 9028 pro is pin compatible with the 9018. The 9038 is ESS's premium converter. The ESS converters all have the SRC at the input.
I mean, come on now. There is no such thing as UltraLock. They run their audio through a TI SRC4392 (unclear if data originating from USB sources go through it, but SPDIF does I am pretty sure) and then an ES9018 or ES9028 depending on the generation.
It is nothing more than marketing. There are dozens of other products using the same chips. All of their claimed jitter rejection comes from the sample rate conversion process. Not that it's wrong or bad, but they are doing nothing unique here.
Last edited:
I am not sure if they use the external SRC for SPDIF sources only or also for USB, but they do still use it. I don't really know why, it may be they have some issue with how the ESS part operates.
Only if all your assumptions hold.
Why guess though? Just ask them.
I don't use the USB port yet.
What ever they did that is an upgrade to the DAC2 works very well as it sounds a lot cleaner and clearer.
They also have a clever distortion cancellation addition which cancels the 2H and 3H out thru and including the analog output stages.
And, redid the level control structure .
-RNM
Last edited:
There's no need to ask. I have seen the teardown and we know how the ESS9018 and 9028 work.
No one is claiming it's bad. It would be my first choice for a 2 channel DAC in that price range.
And I bet if you ask them, they will admit to you that the improvements are likely inaudible over the already well-made DAC2.
No one is claiming it's bad. It would be my first choice for a 2 channel DAC in that price range.
And I bet if you ask them, they will admit to you that the improvements are likely inaudible over the already well-made DAC2.
Dave Hill is the designer at Crane Song and the owner of Dave Hill Designs. He makes products for mastering and mixing, not for home theater audiophiles. Crane Song just announced a 5th generation DAC called Solaris SOLARIS
Mr. Hill seems to think jitter is a significant factor in DAC design, so it's not only Benchmark that has the idea. Hill wrote a bit about some testing he did here: http://www.cranesong.com/A_Matter_ Of_Time_The _Audibility_Of_Clock_Jitter.pdf
Eventually, Crane Song is talking about making retrofit DACs for Avocet and HEDD. If they do one for HEDD, I will probably upgrade my unit. The DAC never was as good as the ADC. The old HEDD DAC might be about the same sound quality as DACMagic+ IMHO, although HEDD DAC can probably sync to a wider range of sample rates.
Neither the HEDD or DACMagic+ are as clean and clear sounding as Benchmark DAC-1, IMO. And DAC-3 sounds better than DAC-1, again IMHO.
The wife in the kitchen type of stories being pretty old and worn, I will chance mentioning that my son has no problem hearing a difference between DAC-1 and DAC-3, although he couldn't hear any differences in PMA's 24/96 test files on DAC-1. I don't know if that helps give any idea as to some minimum amount of difference between DAC-1 and DAC-3, but it's all I have at the moment.
Mr. Hill seems to think jitter is a significant factor in DAC design, so it's not only Benchmark that has the idea. Hill wrote a bit about some testing he did here: http://www.cranesong.com/A_Matter_ Of_Time_The _Audibility_Of_Clock_Jitter.pdf
Eventually, Crane Song is talking about making retrofit DACs for Avocet and HEDD. If they do one for HEDD, I will probably upgrade my unit. The DAC never was as good as the ADC. The old HEDD DAC might be about the same sound quality as DACMagic+ IMHO, although HEDD DAC can probably sync to a wider range of sample rates.
Neither the HEDD or DACMagic+ are as clean and clear sounding as Benchmark DAC-1, IMO. And DAC-3 sounds better than DAC-1, again IMHO.
The wife in the kitchen type of stories being pretty old and worn, I will chance mentioning that my son has no problem hearing a difference between DAC-1 and DAC-3, although he couldn't hear any differences in PMA's 24/96 test files on DAC-1. I don't know if that helps give any idea as to some minimum amount of difference between DAC-1 and DAC-3, but it's all I have at the moment.
Last edited:
- Status
- Not open for further replies.