You are not logged in. Please login or register.
RME User Forum → ADI-2 & 2/4 Pro series, ADI-2 DAC series → [ADI-2 DAC FS] High frequency decay with slow filter @44.1 kHz
Hello!
There are a lot of diagrams showing the (early starting) decay with slow filter @44.1 kHz. I couldn't find any diagram @48 kHz. Is it because it would look the same or is it because the problem is already negligible? Maybe there are any and I haven't found them?
Thanks!
… I couldn't find any diagram @48 kHz. Is it because it would look the same…
Yes, it looks the same for any given sample rate, just shifted up in frequency proportionally to the sample rate relation, ca. 9% for 48 kHz.
ADI-2 Pro’s manual page 86 and ADI-2/4 Pro‘s SE manual page 90 shows the effect for AD-conversion.
There is one exception for the ADI-2 Pro, intended for the possible usecase as measurement frontend:
At 384 kHz RME activates an internal compensation EQ in the DA path that linearizes the fixed Slow Filter’s frequency response up to ca. 120 kHz, see manuals page 81.
ADI-2/4 Pro SE doesn’t need this compensation, as opposed to ADI-2 Pro filter selection is possible at sample rates up to 768 kHz sample rate (manual pages 85, 86).
Thanks Kai,
you mean something like it is described here probably for a TI chip at figure 4.
The decays at 20 kHz would be (for the TI chip mentioned above):
Sampling frequency: 48.0 kHz at 20 kHz (0.42) -> ~ -1.8 dB
Sampling frequency: 44.1 kHz at 20 kHz (0.45) -> ~ -3.0 dB
Decay for the TI chip starts at
Sampling frequency: 48.0 kHz at ~ 14 kHz
Sampling frequency: 44.1 kHz at ~ 13 kHz
The ESS / RME-Device seams to start earlier (10 kHz) and looses 2 dB more at 20 kHz / @44.1 kHz SF.
But the above diagram doesn't meet your 9% rule - or did I misunderstood something? Is it filter specific?
…
Decay for the TI chip starts at
Sampling frequency: 48.0 kHz at ~ 14 kHz
Sampling frequency: 44.1 kHz at ~ 13 kHz…
But the above diagram doesn't meet your 9% rule - or did I misunderstood something? Is it filter specific?
I don‘t know the TI chip, but the above figures follow the sample rate relationship:
48/44.1 ≈ 14/13, about equal.
Imagine the frequency response curve shifted to the right for higher sample rate frequency.
Figure 4 for the TI chip isn‘t even scaled in absolute frequency, but relative to sample rate:
I found the Specs for ES9028Q2M. On page 34 they say, that for slow roll off filter at 20 kHz -3 dB are lost @44.1 kHz (for 48 kHz fs, the -3 dB limit is reached at ~ 21,8 kHz). I wonder where the difference comes from (RME says ~ -4.8 dB - ESS says 3.0 dB - strange).
I found the Specs for ES9028Q2M. On page 34 they say, that for slow roll off filter at 20 kHz -3 dB are lost @44.1 kHz (for 48 kHz fs, the -3 dB limit is reached at ~ 21,8 kHz). I wonder where the difference comes from (RME says ~ -4.8 dB - ESS says 3.0 dB - strange).
I don‘t have an ADI-2 DAC to do the reality check, but my ADI-2/4 Pro SE (ESS), Slow Filter, 20 kHz, delivers measured by me:
44.1 kHz sample rate -3.3 dB,
48 kHz sample rate -1.2 dB.
Maybe @MC hasn’t yet updated the ESS curves in the DAC manual, as the figures fit to the AKM.
In ADI-2/4 Pro SE’s manual the -3.3 dB @44.1 kHz is correctly displayed in the published curve
Thanks Kai for doing own measurements!
Your measurements sound reasonable. This would match my audio experience, too - I can't hear any difference between sharp and slow @48 kHz regarding high frequency decay - especially as the decay starts at ~ 14,8 kHz according ESS specs. I don't hear anything over 14 kHz anyway ... .
Indeed these measurements were not updated. I might do that at a later time. They still show what you need to know - early treble loss when using this filter option.
They still show what you need to know - early treble loss when using this filter option.
It is important to know exactly how high the high-frequency roll-off is and when it starts - both at 44.1 kHz and at 48 kHz fs, in order to be able to assess the extent to which you can or want to live with it.
MC wrote:They still show what you need to know - early treble loss when using this filter option.
It is important to know exactly how high the high-frequency roll-off is and when it starts - both at 44.1 kHz and at 48 kHz fs, in order to be able to assess the extent to which you can or want to live with it.
At the end you will decide with your ears. It is no big issue to not have the exact curve.
MC wrote:They still show what you need to know - early treble loss when using this filter option.
It is important to know exactly how high the high-frequency roll-off is and when it starts - both at 44.1 kHz and at 48 kHz fs, in order to be able to assess the extent to which you can or want to live with it.
Look into the ADI-2/4 Pro SE manual page 91 to find it.
Thanks for your hint, Kai!
To summarize, one can say that a higher sampling rate reduces the decay of treble. Wouldn't oversampling achieve exactly the same result (no treble decay under 20 kHz)? I'm probably misunderstanding something ... .
RME User Forum → ADI-2 & 2/4 Pro series, ADI-2 DAC series → [ADI-2 DAC FS] High frequency decay with slow filter @44.1 kHz
Powered by PunBB, supported by Informer Technologies, Inc.