|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
75.52.172.207
In Reply to: RE: Marantz SA11 Filter Settings posted by f1 fan on July 18, 2007 at 13:12:37
Hi, All,Here is some info on the three "CD" filter settings available on the SA-11:
The frequency response graphs here are from a very extensive (and technical) test of the SA-11 in the July, 2006 issue of Hi-Fi News and AV Tech Magazine.
The testing was done by Miller Audio Research; the complete test results, with measurements of filters 1-3 for both CD and SACD are available here:
http://www.milleraudioresearch.com/avtech/
(You will have to "register" on the site in order to access the results. It's free, and I haven't gotten any "spam" in the 18 mos. I've been registered).Here is a comparison of the three filter settings for "Red Book" CD:
(Note the different dB scales for the three graphs.)
Filter 1 has a subtle "softening" effect, with the frequency response being down -.5dB at 14.5kHz and only -1.33dB at 20kHz.
(This filter will take just a bit of the "edge" off bright recordings.)Filter 2 has a definite HF roll off. FR is down -3dB by 14.5kHz, and is down a whole 5dB at 20k!
Filter 3 has the least effect on frequency response, being down only about .33dB at 20k.
I took some "low tech" oscilloscope photos of the appearance of a 1kHz square wave for each of the three filter settings.
Filter 3 looks just like the square waves produced by my old Magnavox CDB-560, which had a 4x oversampling filter and a "traditional" 16-bit ladder-type DAC. The square wave has the usual symmetrical "Batman" appearance!
The other three photos show the effects of the three filter settings on a constant 20kHz sine wave (actually 19,997Hz, from the CBS CD-1 Test Disc).
Filters 1 and 2 definitely have some kind of modulation effect on the sine wave. This "wavy" effect can be seen on a sine wave sweep as well, and begins to appear at about 10kHz.
So... Filter 3 appears to "do" the least to the music, with almost no effect on frequency response and the least effect on high frequency waveforms.
I do most of my listening using filter 1. It seems to have just a bit of a "sweetening" effect on the sound, and also makes my older, early-digital-era recordings easier to take. I find I can listen to some edgy recordings on the SA-11 that I never could on my previous players. Over-bright recordings don't have me reaching for the "off" button as often either!
I'm sorry I'm not able to find out more about the SACD filter settings. I don't have access to a test SACD!I do all my listening to SACDs using filter 1. I haven't been able to hear any difference at all between the three filters on SACD.
Ken
Edits: 07/19/07 07/19/07 07/19/07 07/19/07Follow Ups:
I have done similar measurement. The 3 filter's frequency responses are compared.
The most interesting thing is that Filter One is a non-phase-linear filter. Look at the ringing -- there is more on the left side of the square waves than the right side of the square waves.
This is more like the way an analog filter behaves. With an analog filter, *all* of the ringing will be on the left side (after the transition) and *none* will be on the right side (prior to the transition).
It's interesting the the consensus of the posters seems to be that Filter One sounds the best, because for decades everybody has been hyping digital filters as "inherently superior" because they are phase-linear. But at least in this case, apparently not.
As Marantz puts it:With RBCDs:
"This is an original filter using short FIR with asymmetric preecho and postecho in the impulse response. The analog characteristics exhibit slow roll-off. This creates a sound venue with richness, smoothness, and depth consistent with the image of analog record performances."With DSD/SACD:
"This is a direct mode that does not perform any filtering on the DSD data. The original source data is reprocuded without any alteration."Since this is primarily bought as an SACD player, I wonder whether it might be the latter that actually matters: no mingling whatsover with the DSD data. The other settings apparently have an attenuating effect on some segments of the bandwidth.
I've been doing my RBCD listening with setting #3, thinking that I hear a greater overall "cleanliness" or clarity across the spectrum with it. Not a huge difference though and this decision is not very studied.
TL
almost a default. Best for both.
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: