|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
65.19.76.104
In Reply to: RE: High Resolution - Fact or Fiction? posted by Inmate51 on April 09, 2014 at 11:03:05
I've done these tests on a number of occasions, starting with 176/24 and downsampling to 44/16. Then I upsample back to 176/24 and compare the original and the processed file. That way, the entire real-time playback chain is identical throughout the comparison and the only difference in the complete record - playback chain is the two passes through the sample rate converter. I use the 64 bit SRC software in iZotope RX 2 Advanced.
I listen for tonal balance, musical transients and soundstage. There are a lot of controls available with the iZotope SRC and you can set them so that 44/16 sounds pretty good, but no setting will allow all of the original tonal balance, musical transients and soundstage to make it through the 44/16 "knothole", but the differences are subtle, and not a reason to avoid buying 44/16 recordings, unless they are not available in higher resolution.
I digitize cassette tapes at 88/24. I have found that these digital transfers sound better when played at 88/24 than after downsampling to 44/16. In a sense, 44/16 isn't as good as a good cassette tape.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Follow Ups:
Tony, that all makes sense. I wonder, though, if you're gaining anything useful by using 24 bits, given the limited dynamic range of the cassette (typically). You could save some disk space by using 16 bits, and possibly not hear a difference.I record live concerts at 88/24 typically with just a pair of Schoeps condensers, using 24 bits mostly so I have less noise for subsequent processing (EQ, etc.).
:)
Edits: 04/11/14
When I record cassettes I like to make only one pass over them, especially when they are old. So I don't know how loud they are recorded. I tend to keep the gain down, thereby wasting one or two bits, which are precious with a 16 bit format, but not so with 24 bits where the limit is more likely to be headroom in analog circuitry, etc... Some of the cassettes are original recordings made on metal tape using Dolby C. These have up to 70 dB of S/N ratio, which is roughly comparable to what you get with 16 bits where you have left headroom to ensure a clean recording. So I think it useful to capture the analog signal at 24 bits.
In the end, I release this material in 44/16. The original 88/24 recording gets converted to 88/32fp for editing and mastering. The levels get normalized at this point so there are no more wasted bits. Then I downsample to 44/32fp and finally dither to 44/16. There is usually an audible difference between the 88/32fp and the 44/32fp due to the lower sampling rate. There is usually a further audible difference between 44/32. Some cassettes were professionally recorded and I have access to duplicating cassettes. Others are production copies, but not high speed duplicated. Still others are not so professionally recorded and can need all manner of work. There are also the occasional dropout and glitches that need digital repair. It is a delight when I get a clean tape that needs nothing more than a flat transfer and splitting into tracks, but this is not the norm.
In this work I am not concerned with disk space. When I am done I keep the original 88/24 transfer, the 88/32fp master and the 44/16 final. Sometimes I keep additional 88/32fp versions if I think that some rework may be necessary. The cost of disk space is insignificant compared to the value of my time.
Tony Lauck
"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: