Spot, a slight quibble about MiniDisc

riredale wrote on 11/25/2005, 10:01 PM

I remember a few weeks ago there was a thread on this Board containing a link to your site. The topic on the site was about using an alternative to record audio for HDV, and as I recall you stated that MiniDisc would not be a viable method because of a severe frequency cutoff around 10KHz.

I've used little cheap Sharp MD-MT15 Minidisc recorders quite a bit for my surround-sound field recording, and after reading the article, I shrugged my shoulders and thought, "What the heck; for surround sound even 10K is probably good enough."

But curiosity got the better of me tonight, and I recorded a white noise sample onto Minidisc (using one of the Sharp units) and then back off into Vegas. I threw the Noise Reduction filter on the timeline, not to use the noise reduction effect, but rather to use it as a quick spectrum analyzer. To my surprise, this is what I got. To my eyes, the Minidisc recording looks absolutely flat to about 17.5KHz, then drops like a rock.

Of course, Minidisc uses a lossy data-reduction method, which in some cases might somehow interfere with certain kinds of post-processing. Nonetheless, a 17.5KHz reach is probably good enough for a whole lot of practical uses, and is far better than the 10KHz cutoff you experienced. Incidentally, I typically buy these little Sharp Minidisc units on eBay for about $70.

I'm not as well-versed with Sound Forge, but I put the clip in SF6 a little while ago and the spectrum analyzer there shows the same thing.

If this data is true, you might want to have someone run tests on several Minidisc recorders in order to see why there is such a large discrepancy.

Comments

Spot|DSE wrote on 11/25/2005, 11:13 PM

We only tested two MD's, both older. We didn't test the new Sony HDMD we bought because it was out on a shoot. I realize this is what NR shows, but what does the spectrum analyzer show? If it's this flat, then there is a whopping difference between the Aiwa and Sharp. The Sharp was better by just a hair, so that's what we used for the post-test. The actual cut on ours was around 11.5 if I remember right.
We also didn't use white noise for this, as the test header indicates, we used a music sample to also test the dynamic.
Thanks for the input, and would you please send me a copy of what you got with the newer Sharp on a spectrum analyzer?

riredale wrote on 11/26/2005, 9:30 AM

Will do; however, just in the past few minutes I Googled "minidisc frequency response" and saw all sorts of interesting stuff. Even the ATRACS 3.5 standard (~1995) was supposed to show flat response out to about 18KHz. The newer ATRACS 4 standard (~1999+) supposedly pushes that out to 20KHz.

From what many are saying, a decent Minidisc recorder can be functionally identical to DAT.

john-beale wrote on 11/26/2005, 10:37 AM

Here are two different MiniDisc units I tested with RightMark Audio Analyzer:

http://beale.best.vwh.net/measure/audio/MD-recorders.htm

the frequency response plot looks really rough, but note the scale is 0.5 dB per division, and that's because this analyser test used white noise. There is some quantization error in the Atrac under those conditions, but it's hard to detect by ear. With a swept-sine test the data should look much better.

I'm not an "audiophile" and my own hearing drops off at 16 kHz but I have never heard any problems with my HHB recorder (note, this is a "pro" MD, over $1k). I have heard problems with MP3 playback.