Is 24-bit Resolution a Prerequisit These Days?

[Telefunkin]

Hi All,
I read somewhere yesterday that library companies would expect nothing less than 24-bit versions of anything that they sign. Not only that, but they would also ask for different mixes, different length versions and would probably want run and other component stems.

That set me worrying. We all submit mp3 files to Taxi of up to 192kbps, but I've assume that the most they might ask for after that would be a .wav file. Hmmmm, no less than 24-bit though? What about 48khz v 44.1kHz? Does it matter? Any other demands?

I guess what I'm asking is - what is the pro spec that we are likely to be asked for if any of us newbies is lucky enough for a 'forward' to turn into a contract? I'd like to be prepared and avoid being an under-informed idiot, should I ever be so fortunate.

No doubt this topic has been discussed before, and I suppose I really should have searched for it. However, I hope you'll forgive me for not doing so as I believe the answers will be significant interest to other new(ish) members.

Thanks for any advice,
Graham.
P.S. Anyone else from the UK? I'm near Sheffield.

[hummingbird]

It varies.

I have been asked for:
48khz, 16 bit - AIFF
48khz, 24 bit - AIFF
44.1kHz, 24 bit - WAV
44.1kHz, 16 bit - WAV
48khz, 16 bit - WAV

As for stems and submixes, that varies too. For some I just send the final mix of the track as AIFF or WAV by whatever means they specify (data cd/dvd ,FTP, or whatever service they are using), that's all they want.

For others, two or three submixes as well as the final.

Rarely have I provided stems although I have seen them asked for in listings, again rarely.

But it's easy enough, once finished with the Final Mix of the track, to run it off as a 16 bit AIFF and a 24 bit WAV.

You might also take the time to create two or three submixes then... percussion/bass (no string/synth), bass/strings (no drums), that sort of thing, use your composers hat to create submixes that you like... making sure something carries through (no silence) in each one. Figure these out by simply muting various tracks within the track and finding elements that work together.

As for stems I create them for myself as a backup... when I have finished designing a track I will run a 48khz, 24 bit of each instrument, VI's especially (in case the VI is lost at some point I will have the melody/counterpoint/sound as a wav) and put copies of everything in a folder. I use this as the source for my final mix and I make a text file with notes about my mix or effects etc to go in the same folder. In my experience when I look back at something I worked on 3 years ago I no longer remember what the heck I was doing, so some notes can be helpful.

In backing up I suggest the 48k, 16bit AIFF and the 44k, 24 bit WAV final; plus 2 - 3 submixes if wished; plus the back-up stems and your notes. If the track is signed I put a text file with the information in the same folder. I also run a 192k and 128k mp3 for uploading or pitching... they could always be converted from the aiff or wav but if it's there already it's much easier to grab.

I have literally hundreds of tracks much of which is grist for the mill, when I'm looking for ideas often I'll go pick a few folders and listen to drafts or finals of stuff. [It's kinda like being a photographer who takes 100 photos to find one good one.]

Have I answered your question? Hehe

Hummin'bird

[Telefunkin]

Wow Hummingbird,
You certainly did answer, and thank you very much for the comprehensive run-down. I need to get with the program! Up to new, with a small bag of forwards but nothing signed, I've been dreading letting myself down by not being able to provide whatever is asked for, but I now know that I'd better be prepared. I might also need to upgrade some kit! Work, work work [well OK, its really play, play, play ]

Many thanks indeed, and I'm sure there will be plenty of other folk who would benefit from your advice here.

Best regards,
Graham.

[andygabrys]

Hi Graham,

I don't really feel there is much to be worried about honestly....but it might depend on what you are doing right now.

First a couple of hotly contested "facts" (choose whichever side of each one works for you and you can find somebody that supports your viewpoint):

A) For mp3 files: increasing the size of the mp3 by choosing 320 kb/s or similar always means better audio / other issues such as encoding for streaming are equalizers and the extra high rate mp3's don't necessarily help / most people can't tell the difference between 192 kb/s and 256 kb/s or 320 kb/s (but a lot of people CAN distinguish 96 or 128 kb/s vs. all the higher resolution mp3s)

B) For WAVE and AIFF files: Increasing sample rate sounds way better / most people can't tell the difference past 44.1 khz.

C) For WAVE and AIFF files: increasing bit depth to 24 bit or even 32 bit obviously sounds better / a lot of people can't tell the difference / we are only using the top 2 or 3 bits for dynamic range on a hot master so why bother with anything greater than 16 bit?

Forgetting about all that:

1) the "standard" for audio used in Film and TV is a PCM file (WAV or AIFF) at 48khz, 16 bit.

2) Any DAW available today can recode WAV or AIFF files at nearly any sample rate and bit depth.

3) The "standard" for what publishers are asking for is either as above or in some cases 44.1 kHz 16 bit, and as you mentioned sometimes either 44.1 k or 48 k at 24 bit - 24 bit is often requested if the publisher is then going to master the music themselves.

4) to cover yourself for most eventualities that happen today, you operate your session at 48 k 24 bit depth. You bounce your final mixes at 48k 16. And then you convert the files to whatever other format might be needed. If you know that your publisher is going to be asking for 24 bit files to master themselves, it would be prudent to leave off the mastering plugins and just bounce at 48 k 24 bit. Easy. Most of the time its not voodoo.

5) as far as quality of mp3 files to submit - whatever is asked for and as long as you bounce both the WAV or AIFF and mp3 file at the same time, you have saved tons of work. I am sure most DAW's do this these days (I know Logic and Pro Tools do). Since TAXI spec's 192 kb/s, this is what I use everywhere else too. See point A) above.


That's the first part of the question. the second part (alt mixes, stems etc) varies again by publisher and by use.

Here are some common ones:

1) Full Mix
2) Instrumental mix (if its a vocal song)
3) TV mix (if its a vocal song - this is the song with BG vocals but no lead)
4) Underscore / narrative / no-lead mix (taking out some of the most attention grabbing parts of the mix i.e. melody so the music will be more like wallpaper underneath dialogue
5) various "Minus" mixes - i.e. no drums or no percussion, or no guitars, or no keyboards or maybe minus everything except for guitars. Usually most easily done by muting various tracks in your full mix session and re-bouncing with a distinct name

timed edits - usually used for things like commercials

60 sec
30 sec
15 sec

Stinger (a couple seconds of the song then a shot and fade or shot and quick ending usually used to go out to or back from a commercial during a TV program.)


just my 2,364 cents.

[Telefunkin]

Wow again Andy,
Even more comprehensive advice, that is gratefully received.

At present, I do set all my mp3's to 192 kbps for Taxi Submissions, and I also save a wav file of the same mix, so I've got that bit right.

I have been working at 16-bit and 44.1kHz though, and from your comments (and those from Hummin'bird) I guess I should make sure I can't be caught short and start working at 24-bit and 48kHz. Maybe it sounds dumb to have stuck with 16-bit but there are reasons, and I expect that I'll be able to do better in the near future.

I'll have to hope that, should I be fortunate enough for any of my existing forwards to get picked up, they don't demand something I can't provide.

Once again, thanks for the great tips.

Cheers,
Graham,
UK (where its been cool and rainy again, just for a change!)

[TheElement]

Great info Andy. I work at 48khz and 32 bit float and also mix down at the same to 48khz and 32 bit float wav. this is ok?

Thanks

[andygabrys]

at some point everybody has got to do some personal research and do some critical listening, as there are few black and white absolutes (thats why the first part of my original post is phrased as questions).

If mixing at 48 kHz 32 bit float sounds better, then go for it. That is what I do in both Logic and Pro Tools. Most DAW's these days have 32 bit float mix engines with amazing tons of headroom.

If mixing down at 48 kHz 32 bit sounds better, then go for it. Myself I don't hear the difference, and so i bounce at 48 k 16 bit with dithering. Like I said above if somebody wants 24 bit files, I bounce out at that. Its just whatever is easiest really. Seriously, its convenience.

and going with the theory that capturing and mixing at 24 bit and 44.1 or 48khz is going to give a good quality start, and I down sample and dither down to shorter word length (bit depth) from there. Or down to mp3. But starting with a good quality product to make that mp3.

In reality, you can up and down sample from 44.1 to 48 kHz and not too many people are ever going to pick it out in a blind test.

for kicks, google information about who can actually tell the difference between a decent quality mp3 and a wav hi res file. George massenberg is one of the few. its interesting. There are also blind listening tests out there on the internet.

[orest]

Thought I would chime in as well.

Yes, it all depends on the libraries you are working with and as mentioned before: if you have a decent sound card and a full version of a DAW you can convert your songs into pretty much everything you need.
You can also use other programs to convert files.

I usually record in 16 bit 48 kHz

About stems, something I recently learned is to do build up sting endings with a natural fade out (no drawn in fade out curves).
Editors and producers use these endings a lot in TV-shows and commercials.

I can post some examples if anyone is interested. Sometimes it's easier to listen.