Gear / Technical Help > Post-Processing, Computer / Streaming / Internet Devices & Related Activity
24/48 to 16/44 conversion introducing clipping...
jes1982:
--- Quote from: vanark on March 10, 2024, 10:05:25 AM ---My solution - don't convert to 16/44. It has been 7 or 8 years since I made 16/44 files. I simply circulate my 24/48 files.
--- End quote ---
Thanks for the input. Just to clarify: I also archive 24/48 FLACs. My standard practice now is to save (1) my 24/48 "raw" file; (2) edited 24/48 FLACs (I won't say "mastered" because I don't know what I'm doing); and (3) 16/44.1 FLACs. Overkill, I guess. I'm in the process of trying to clean out my parents' house and cursing the things that weren't thrown away. When my son has to deal with my hard drives, if he looks at what's there, he may wonder why his father had so many damn copies of the same show.
But, yes, the audience for the music I tape prefers the redbook version and I'm going to go along with that for now. I suspect it's because the files take up less space, not a desire to burn to CD-R.
jes1982:
--- Quote from: vanark on March 10, 2024, 11:57:32 AM ---
--- Quote from: EmRR on March 10, 2024, 11:53:02 AM ---The recommendation for files that will become mp3’s is to normalize or master to -1dBFS. No higher. I’ve noted in mastering things for release that it’s common to see peak increases of +0.2 from sample rate conversion, occasionally as much as +0.8.
--- End quote ---
I've never had an issue at -0.2 dB
--- End quote ---
Thanks for the comments. This is interesting and gave me an idea. For one of the files I was having trouble with where I had initially set the maximum amplitude at -0.1db, I increased the hard limiter by 0.1db increments. I didn't get a clean conversion until the maximum amplitude was set at -0.5db. This means the sample type conversion boosted the the left channel by +0.41db and the right channel by +0.29db.
Now I wonder: if I can find or create a "hotter" file for testing purposes, will the sample type conversion boost the gain even more than +0.41db?
nulldogmas:
--- Quote from: jes1982 on March 10, 2024, 01:02:44 PM ---
Now I wonder: if I can find or create a "hotter" file for testing purposes, will the sample type conversion boost the gain even more than +0.41db?
--- End quote ---
Huh, interesting. I've long been normalizing to -0.5dB without issue, but I'm now wondering if it wouldn't hurt to go a bit lower.
AbbyTaper:
I use Sound Forge and normalize to 98.14% (not sure what the dB value of that is), and then dither with iZotope's mbit process. I never noticed any clipping after dithering.
Like Rory, I've now stopped offering 16 bit files (unless the source was 16 bit). I haven't had any complaints so far, or requests for a 16 bit version. So either most people are accepting 24 bit, are dithering the files themselves, or just don't bother downloading.
nulldogmas:
--- Quote from: AbbyTaper on March 10, 2024, 03:38:17 PM ---I use Sound Forge and normalize to 98.14% (not sure what the dB value of that is), and then dither with iZotope's mbit process. I never noticed any clipping after dithering.
--- End quote ---
Looks like 98.14% is about -0.16dB.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version