25fps to 24fps software that fixes pitch shift

[deleted user]

Bumping this one up,

what speed slow down % do I use in audacity, to get 24fps audio down to 23.976?

Thanks if anyone can help.

I don't know how Audacity's slow down works, but I think the factor is always 1000/1001 or 1001/1000 (depending on whether we're talking about speed or stretch factor). Multiply result with 100 for percent. For example 1001/1000 is 1.001. Times 100 is 100.1%. Or 1000/1001 is ~0.9990009. So result would be ~99.90009% (but the number is not 100% precise, it goes on).

Problem is sometimes you can't set it that precisely. I like to do it via resampling personally, as that can have perfect precision.

For example to get 23.976fps audio to 24fps audio, you need a speed up of exactly 100.1%. If the base sample rate is 48000 Hz, the sped up sample rate is exactly 48048 Hz. The simplest way to have this to be a round number is of course to have three zeros at the end in the base sample rate, which is why it doesn't work well with 44.1 kHz, as the result would be 44144.1 Hz, which you can't set as a file's sample rate afaik. Anyway, I simply set the sample rate to 48048 (no resampling, just setting it). That sets the speed. Then from that I resample back to 48000 Hz (or any other target sample rate) with a good resampler and it's done.

Reverse would be, resample 23.976fps audio to 48048Hz, then simply set (interpret) it as 48000 Hz, and you got the correct speed. If the base audio is 44100 Hz, this would require kind of an awkward extra step after resampling to 48048 Hz and interpreting as 48000 Hz: resampling back to 44100 Hz. But you could always opt to just leave it at 48kHz ofc.

Likewise, speeding up 44100 Hz from 23.976 to 24 fps would require first resampling to 48000 Hz, then setting to 48048 Hz, then resampling back to whatever the target sample rate is.

Dunno if that helps.

[Stamper]

A bit.

The audio I have is an italian Blu-ray, set of course at 24fps / 48000KHz.

So I downconvert the video to 23.976, no problem.

Audio needs to be slowed down to ~99.90009% I guess?

Then I check if the duration perfectly match.

[deleted user]

Yup, that sounds about right. But I would wager that these speed-changing algorithms don't use the best resampling out there, so there may be more quality loss than the other way. Your choice tho.

By matching duration, I assume you mean checking against your calculation of how the new duration should be, not against another transfer? As another transfer obviously could have other factors changing the length, like extra content at the end/beginning.

[Stamper]

Checking against the video converted to 23.976 duration.

Then there is this thing where I could do it someone suggested

from 24 to 25 = +4,167
Then from 25 to 23,976 = -4,097

[Falcon]

I have a question for the final export after slow down a pal audio track: Can I up the bit-rate for keep the quality ? 
if I have a track encoded on 192-224 (mono or stereo) or /384/448 (5.1) kbps, after slow down, the music are more grave, and I get a music a little different, If I up the bit-rate, maybe the encoder apply less filter for compressed the audio (and all encoder don't make the strictly same algorithm).

I think convert DD 384/448 to 640 after slow down, or DD 192-224 in DD 320

What do you mean by the music sounding "more grave" after slowdown? Unless you're encoding in a basic profile at a low bitrate under 96kbs, there won't be much of a difference in pitch. It's very common for things like music and audio to sound different during slowdown, unless your source happened to be sped up to 25fps but kept it's 24fps pitch. Are you doing both pitch and tempo correction or just one of them during conversion? I've seen instances where a 24fps source will have it's audio converted from a 25fps master but erroneously retaining it's higher pitch and sounding all "distorted and scratchy" and vice versa with some 25fps material.

sorry to answer so late, but better late than never

The audio becomes more grave when I slowed it from 25 to 23.976fps.
This allows me to find the original pitch/tone 95% of the time, because 25fps source often have the problem of pal speed-up.
I think that re-encoding at the same bitrate as at the beginning is not a good idea because:

-The encoder is different (maybe he haven't the same efficiency)
-Even if one uses the same original encoder, the setting is maybe different (and maybe the encoder has been updated)
-Slowdown my track changes the audio spectrum, because the sound is more grave. I may be refound some frequencies with the slowdown. I don't want to destroy this frequencies because of an inappropriate compression.

I think finalize all AC3 192 PAL tracks, in AC3 256 (with a Dolby Pro Lab encoder) after synchronization done.
Tracks 448, I re-encode them in 640 after synchronization done.

[Pomologe]

There are two more tools doing time modification und pitch-correction in one step:
1. MeGUI
2. TFM Audio Tool
I prefer TFM Audio Tool because the results have better quality and another goodie is:
TFM Audio Tool has a built in DTS-Encoder, so you can convert very comfortable between DTS and DD.
The whole process is very fast and takes only a few minutes...