Page 3 of 3
Posted: Thu May 05, 2005 8:35 am
by lukpac
Interesting. Maybe I should post a whole clip or something. I should also post a clip of Ryan's Xfer.
The EQ in ProTools that I use is parametric (see above), so that should be interesting.
Posted: Thu May 05, 2005 11:07 am
by Dob
That graph is much more jagged that I would've expected. For the left channel we have a swing of 8 db up from 10982 to 11197 and then back down 7.4 db to 11262. That doesn't seem right to me...almost as if that was phase related, not tonality related.
Luke, IMO an analysis like this would be more accurate if you concentrated on just a few seconds (or even a fraction of a second) of simpler audio. Acapella vocal would be very good (especially if it is of constant pitch) or isolated cymbals/drums, as you feel that is the area where the discrepancies are most audible.
Posted: Thu May 05, 2005 11:19 am
by krabapple
That's sort of hte point I was getting at with my talk of EQ bands -- probably more reasonable to focus on the adjusting based on the overall shape of the difference curve, using a normal spread of EQ bands, rather than adjusting for each individual level differences at each 20-Hz(!) increment. I would expect that gets you in the ballpark.
There's probably a way to 'smooth' the curves , but I haven't fussed with those parameters yet.
I'm should probably also read up on how those EQ matching tools work...
Posted: Thu May 05, 2005 11:25 am
by krabapple
One thing that's obvious jsut from looking at frequency analysis of *either* sample, is that the let and right channels are quite unbalanced , albeit in a frequency dependent manner -- IIRC there's a fairly consistent ~10 dB difference between left an right channels throughout about half the audible spectrum, midrange on up. It's true for both the LP and the CD samples. Maybe it's one of htose mixes where treble isntruments are concnetrated in one channel.
Another thing to note is that profiles from *mp3s* won't be identical to profiles from the original .wavs.
They may be quite different.
Posted: Sun Apr 16, 2006 10:08 pm
by lukpac
All of the recent talk about pre-emphasis got me thinking. The (seemingly) jacked up high end on the CD is a lot like that on PE CDs (played without PE, obviously). I burned a CD with PE set, and while the EQ isn't a complete match to the vinyl, the high end in particular is much closer. I can't fathom how a CD from 2005 could have PE, though, unless somebody at Warners screwed up (apparnetly Wounded Bird just gets digital copies from the record companies).
Odd...