I was reading a tutorial and it mentioned that you should use dithering when exporting a track which I didn't know about. I don't know if he meant dithering when exporting a pre-mastered track, a mastered track or both.

I read about dithering and even exported a blank song with and without dithering. I could hear the background noise when I exported with dithering when I turned it up quite a bit and the non-dithered one was silent.

I then read this article http://productionadvice.co.uk/when-to-dither/ and it said you should dither your pre-mastered track before you send it off for mastering and then the mastering engineer should dither it a second time. I think the article is also saying that you only need to dither a 32bit float file once when exporting the mastered copy but not when exporting your pre-mastering copy.

Next, I tried to determine which format should be used when sending a track off to get mastered, 16bit, 24bit and 32bit float. It seemed like 24bit was the most common recommendation. Then I found out that FL Studio won't dither a 24bit file (or a 32bit float file) only 16bit. So, out of this I have 3 questions:

1. When sending a pre-mastered track to get mastered, is sending a 32bit float file ok?

2. If so, does the 32bit float file need to be dithered?

3. If 32bit float is not good to send to get mastered, can 24bit be sent for mastering without being dithered?

I hope I explained that well. It might not even matter, I don't really know.

___________________
Songwriter/artist and reluctant producer.

My Soundcloud

From memory, Reason only allows dithering for 16 bit exports as well. It appears to point towards the extra bit depth of 24 bit audio (not to mention 32 bit) meaning that the resolution is sufficient (for example, smooth reverb tails etc) that dithering makes no audible difference.

The mastering engineer should be able to handle either 24 bit or 32 bit, and likely won't care much which you provide, if it's being mastered down to 16 bit. If you're able to hear a difference between a track which has been mastered to 16 bits down from 24, vs the same track mastered to 16 down from 32, you have some of the best ears in the world.

Thanks for the reply! Guess I will export to 32 bit float in the future.

___________________
Songwriter/artist and reluctant producer.

My Soundcloud

I usually export my pre-mastered tracks at 24bit 44.1KHz

quote:

Originally posted by paulazizeh I usually export my pre-mastered tracks at 24bit 44.1KHz

Do you add dithering?

___________________
Songwriter/artist and reluctant producer.

My Soundcloud

Dithering will be applied by the mastering engineer if they're mastering to a bit depth which will benefit from dithering (for example, 16 bit).

There's no reason to apply dithering to 24 bit audio - I doubt that there's a combination of playback system and human ears which could hear the difference between a dithered and undithered 24 bit file.

Yes, you can dither, but there's absolutely no real-life benefit, and I'd find it interesting on a technical level whether there'd be reasons to dither a 24 bit file prior to the mastering engineer applying the 16 bit dithering.

You can safely provide the 24 bit (undithered) version of your song to a mastering engineer - it'll produce exactly (to any human ears) the same result as providing the 32 bit (undithered) version to them, if the end result is a 16 bit file.

quote:

Originally posted by derail Dithering will be applied by the mastering engineer if they're mastering to a bit depth which will benefit from dithering (for example, 16 bit). There's no reason to apply dithering to 24 bit audio - I doubt that there's a combination of playback system and human ears which could hear the difference between a dithered and undithered 24 bit file. Yes, you can dither, but there's absolutely no real-life benefit, and I'd find it interesting on a technical level whether there'd be reasons to dither a 24 bit file prior to the mastering engineer applying the 16 bit dithering. You can safely provide the 24 bit (undithered) version of your song to a mastering engineer - it'll produce exactly (to any human ears) the same result as providing the 32 bit (undithered) version to them, if the end result is a 16 bit file.

I'm no mastering engineer but this pretty much sums up my understanding of when and why you dither. The word length of 24 bit or higher encoded audio is sufficient to accurately represent the arithmetic involved, keeping the detail in the sound, you normally add dither when truncating back to 16 bit to fix truncation errors which happen without dithering. I think whenever you process audio in the DAW, it expands the wordlength, so the only time you would need to dither twice is where you encode an analogue source to digital, then truncate that back to 16 bit too.

Perhaps one of the engineers or mastering gurus here can chime in and correct me if I'm wrong.

___________________
Youtube
Mixcloud
Twitch

quote:

Originally posted by derail Dithering will be applied by the mastering engineer if they're mastering to a bit depth which will benefit from dithering (for example, 16 bit). There's no reason to apply dithering to 24 bit audio - I doubt that there's a combination of playback system and human ears which could hear the difference between a dithered and undithered 24 bit file. Yes, you can dither, but there's absolutely no real-life benefit, and I'd find it interesting on a technical level whether there'd be reasons to dither a 24 bit file prior to the mastering engineer applying the 16 bit dithering. You can safely provide the 24 bit (undithered) version of your song to a mastering engineer - it'll produce exactly (to any human ears) the same result as providing the 32 bit (undithered) version to them, if the end result is a 16 bit file.

quote:

Originally posted by Teezdalien I'm no mastering engineer but this pretty much sums up my understanding of when and why you dither. The word length of 24 bit or higher encoded audio is sufficient to accurately represent the arithmetic involved, keeping the detail in the sound, you normally add dither when truncating back to 16 bit to fix truncation errors which happen without dithering. I think whenever you process audio in the DAW, it expands the wordlength, so the only time you would need to dither twice is where you encode an analogue source to digital, then truncate that back to 16 bit too. Perhaps one of the engineers or mastering gurus here can chime in and correct me if I'm wrong.

Thanks for the responses. Just making sure I wasn't missing an important step in production.

___________________
Songwriter/artist and reluctant producer.

My Soundcloud