Author Topic: Dither/Resample ? (Read 12475 times)

Electric Cowgirl · « **on:** November 11, 2003, 04:12:12 PM »

Hey all-

Got a question....I am recording in 24/96. I want to dither to 16 and resample to 44.1. What order should I do this in??

Sorry for the dumb question.

dmonterisi · « **Reply #1 on:** November 11, 2003, 04:14:02 PM »

resample then dither. all digital editing (resampling, fades, etc) should be done before dithering to 16 bit.

Electric Cowgirl · « **Reply #2 on:** November 11, 2003, 04:39:23 PM »

Thanks!!! +T

Kindguy · « **Reply #3 on:** November 11, 2003, 04:56:10 PM »

T+ to both of ya. I was just about to ask this question.

nic · « **Reply #4 on:** November 12, 2003, 09:41:43 AM »

all digital editing should be done before you resample/dither(not just dithering)

Willie T. · « **Reply #5 on:** November 14, 2003, 02:14:13 PM »

Since we are on this editing question should it be tracked out before dithering?

MattD · « **Reply #6 on:** November 14, 2003, 03:38:24 PM »

I don't think it matters, but tracking will be faster AFTER dithering because you'll be writing smaller files.

Chris K · « **Reply #7 on:** November 14, 2003, 04:00:36 PM »

Quote from: luvean on November 12, 2003, 09:41:43 AM

all digital editing should be done before you resample/dither(not just dithering)

whyis this so important...i have bee tracking a show, resampling, then splitting the files at the track markers.

then i take the 1st track and do a fade in, and the last track of the set/show and do the fade out

is this incorrect?? should i apply the fades before chopping up the set??

since it is a fade i thought this would be ok???

chris k

nic · « **Reply #8 on:** November 14, 2003, 04:32:56 PM »

you should do the fades before you resample/dither.
as far as tracking, technically I guess it should be done before resample/dither, but most do it afterwards.

say you have 48 and a 44.1 files from the same show, (on a pc) load 1 in cdwave and create a cue sheet....then just load the other file and the same cue sheet and its instantly tracked(granted you might have to rename the file in the cue sheet)!

Phil · « **Reply #9 on:** November 15, 2003, 12:05:03 AM »

If you using Wavlab, just do a "save as" on your 24/96 file and there is a option to save it as 16/44.1.

Simp-Dawg · « **Reply #10 on:** November 15, 2003, 11:42:08 AM »

i like to do all my tracking at the very end of everything. i record at 24/48, and my process is like this:
1. apply any necessary audio adjustments (for example one of my channels got turned down by a beeyotch standing on my case at mule the other night, so my first step is to apply gain over the section on that channel so it matches the rest of the recording)
2. resample to 44.1 (using sound forge at highest accuracy setting with anti-alias filter)
3. bit-depth conversion (again in sf, currently using highpass triangular dither and high-pass contour noise shaping, though i don't know the difference between that and other settings)
4. save new 16/44.1 file with new name (i keep all the 24bit recordings for archival)
5. open 16/44.1 file in cdwave, track, save cue file, open it on the 24/48 file and track that too
6. flac

tadow.

rabhan · « **Reply #11 on:** November 28, 2003, 08:17:42 PM »

well, when i have over 2 minutes of crowd before the roar and over 2 minutes of crowd after the last instrument sound, i choose to add my 5sec set fades after i cut the tracks with CDWav.....

rabhan · « **Reply #12 on:** November 28, 2003, 08:20:36 PM »

not to mention, how/why does it matter that a "fade" is affected in a sonic realm bc it has been resampled and dithered? how is it any different than having the adc(V3 for example) using ANSR vs using CEP/WL after the fact? where is the evidence that fading 5 secs of audio before or after these processes are applied makes any difference?

dklein · « **Reply #13 on:** November 28, 2003, 10:40:16 PM »

Quote from: Phil on November 15, 2003, 12:05:03 AM

If you using Wavlab, just do a "save as" on your 24/96 file and there is a option to save it as 16/44.1.

as long as you realize you are both resampling and dithering with the internal algorithms at default settings. Lots of folks like to choose their quality settings and dither type.

dklein · « **Reply #14 on:** November 28, 2003, 10:52:27 PM »

Quote from: luvean on November 14, 2003, 04:32:56 PM

you should do the fades before you resample/dither.
as far as tracking, technically I guess it should be done before resample/dither, but most do it afterwards.

I agree, but only for convenience reasons - edit it once instead of twice. While fades are 'destructive' editing, and therefore should be done prior to dithering, most probably don't give a crap about getting the absolute highest quality audio for their fade points

The rest of the file will be the same when it's re-saved (after the fade has been done).

When you track is irrelavent, but if you're going to edit after you've actually split the tracks, you may not be adhering to sector boundaries. Again, kind of an moot point from a listener's perpsective - if you join a fade-out to a fade-in, there won't be any pops or clicks from zero filling the sector because the last sample of a fade-out and first sample of a fade-in are zero. But the etree police may come looking for you

BC · « **Reply #15 on:** November 29, 2003, 04:26:35 PM »

I don't think it is a good idea to just "save as" for downsampling your 24 bit files. I am guessing that just saving it with different sampling rate and bit depth does not resample or dither, it just cuts off the extra information.

In wavelab you can add in the resampler and dither to the processing chain then "render" the file to obtain a correctly resampled and dithered file.

Ben

Quote from: Phil on November 15, 2003, 12:05:03 AM

If you using Wavlab, just do a "save as" on your 24/96 file and there is a option to save it as 16/44.1.

Electric Cowgirl · « **Reply #16 on:** December 03, 2003, 04:50:07 PM »

Ok heres what I ended up doing to get my 24/96 YMSB show down to 16/44.1.

1- Resampled to 44.1 using wavelabs resample function.

2- Dithered using the UVH22 algorithm in wavelabs.

3- I clicked Render in the masters section of wavelabs to create a new file that has the above changes.

4- Then I joined the "split" files using the add wave to end function...this was seamless.

5- I put in the track marks.

6- Used the create new cd menu to export each track as a separate file.

7- Used Nero to add faders and burn cds.

Something I discovered that I will throw out to hopefully save someone else time is TRACK AFTER YOU RESAMPLE/DITHER. After resampling the track marks were not in the right place in wavelabs, so do everything else first.

Also I can clearly hear the difference between 16 and 24 bit (blind test) on my SHITTY stereo. Was it worth the effort/frustration of switching from DAT to Laptop? HELL YA!

Brian Skalinder · « **Reply #17 on:** December 03, 2003, 05:16:19 PM »

Quote from: BeckyT on December 03, 2003, 04:50:07 PM

Also I can clearly hear the difference between 16 and 24 bit (blind test) on my SHITTY stereo. Was it worth the effort/frustration of switching from DAT to Laptop? HELL YA!

One of these days I'll get around to going 24-bit, but I just won't do it on a laptop. C'monnnnnnn hard-drive recorders! I figure I'm about a year away from making the jump. Great post, Becky!

rabhan · « **Reply #18 on:** December 03, 2003, 08:46:15 PM »

just remember that hdds can and will fail too! i lost one on this past tour, thankfully i ran dat backups.

just my $.02

Electric Cowgirl · « **Reply #19 on:** December 04, 2003, 11:15:01 AM »

Nothings perfect.

dmonterisi · « **Reply #20 on:** December 04, 2003, 11:23:43 AM »

Quote from: BeckyT on December 04, 2003, 11:15:01 AM

Nothings perfect.

now, that's a defeatist attitude

F.O.Bean · « **Reply #21 on:** December 05, 2003, 12:09:28 AM »

Quote from: dmonterisi on December 04, 2003, 11:23:43 AM

Quote from: BeckyT on December 04, 2003, 11:15:01 AM
Nothings perfect.

now, that's a defeatist attitude

Electric Cowgirl · « **Reply #22 on:** December 05, 2003, 12:55:07 PM »

No way! That is a positive attitude. If I was hell bent on everything being absolutly perfect I would be a depressed, pissed off, and generally dissatisfied person.

FWIW-I was attempting to say that no system is perfect. There is always a chance for something to go wrong and ruin the recording. DATs misload, can have dirty heads, or eat your tape. My laptop can decide it dosen't want to record for a zillion different reasons....so ya there are plenty of imperfections.

Author Topic: Dither/Resample ? (Read 12475 times)

rabhan

rabhan

rabhan