-
Χωρίς να ξέρω αν είναι συχνό αυτό το "κολπάκι", παραθέτω την επίσημη εξήγηση της Benchmark για την επιλογή τους (τη βρήκα σε αυτό το thread του audioasylum):
Hello,
I am John Siau, Director of Engineering at Benchmark Media Systems. Your questions are good ones, and our choice of a fixed 110 kHz conversion frequency may seem strange, but there are very good performance related reasons for this topology.
The AD1853 converter has a built-in interpolation filter that can operate at a ratio of 8x, 4x, or 2x. The 8x mode is intended for 44.1 and 48 kHz applications, the 4x mode is intended for 88.2 and 96 kHz applications, and the 2x mode is intended for 176.4 and 192 kHz applications. The performance of this filter is optimized for the 4x mode of operation. Unfortunately, the interpolation filter is rather poor in the 2x interpolation mode. This problem is not unique to the AD1853, nor is it unique to D/A converters. Every A/D and D/A conversion IC that we have tested performs better at 96 kHz than at 192 kHz. In most cases THD+N, SNR, passband ripple, and stopband attenuation are all poorer at 192 kHz than at 96 kHz. Based upon these tests, I am not surprised that there is not yet any conclusive evidence that 192 kHz is better than 96 kHz. Given the current state of the art, 192 kHz should sound poorer than 96 kHz (unless the defects mentioned above are outweighed by the increase in bandwidth of a 192 kHz system).
Our tests show that the current crop of converter 192 kHz converter IC’s exhibit increased defects in the 0 to 20 kHz band in exchange for increased bandwidth (between 48 to 96 kHz). Remember this added bandwidth is useless unless frequencies between 48 kHz and 96 kHz are recorded, reproduced, and perceived by the listener.
Our tests only confirm what is published in many of the converter data sheets. A Google search on AD1853.pdf will return the data sheet for the AD1853. On page 3 the digital filter characteristics are as follows: Stopband attenuation is 20 dB better at 96 kHz than at 192 kHz, and passband ripple is +/- .0005 dB (DC – 39.95 kHz) instead of +/- .5 (DC-65.4 kHz).
Other converter data sheets tell similar stories, or fail to specify performance at 192 kHz. Let the buyer beware.
We chose 110 kHz because this is near the maximum frequency that the AD1853 can reliably operate in the 4x interpolation mode. We measured the performance of the AD1853 vs. sample frequency before choosing 110 kHz. Given a different converter IC, we could choose a different conversion frequency (our SRC topology allows this flexibility).
Another important consideration is the location of the passband of the interpolation filter. At 110 kHz with 4x interpolation, the AD1853 passband is 45.8 kHz (instead of the 39.95 kHz published on the data sheet). When combined with the filters in the SRC, the result of this (rather unconventional topology) is a D/A conversion system that is almost entirely free of images (alias effects).
John Siau
Κοντολογίς, δεν την παλεύει το φίλτρο παραπάνω!
Posting Permissions
- You may not post new threads
- You may not post replies
- You may not post attachments
- You may not edit your posts
Forum Rules