Google Doubles Size of Youtube Videos

[braincell]

New plan, record the piano part again with more velocity and a noise gate, then try and raise the gain and compress the mix. I can't seem to get the kind of compression I see on commercial recordings with a flat top.

[valis]

Try Waves L2/L3, iZotope Ozone (multiband comp & limiter functions together), Sonnox Oxford Inflator & Oxford Limiter, Sonic Timeworks Mastering Comp, Voxengo Elephant or even the free TLS Pocket Limiter (or here from the dev's site).

Really though getting a 'flat top' is obviously not the ideal for 'mastering', it's unfortunate that high levels of lossy (data) compression benefit from it. Whatever tool you choose keep a listen for what happens to the transients & decay of your notes.

[braincell]

I will try it thanks! I've been using less or no compression these days. This one is said to be a clone of Waves L1:

http://www.yohng.com/w1limit.html

[braincell]

Your advice was very good Valis but that wasn't the problem. I have noise in the piano part which for some reason isn't apparent on my studio monitors. I never noticed this before. I applied fairly heavy noise reduction to the piano parts and am rendering the video again with h64 video and mp4 audio.

[valis]

Perhaps it's low frequency content? It's really easy to eat up headroom with subsonic information if you don't know it's there. Try highpassing 20-25hz perhaps, or post up a sample of the noise to listen to?

[braincell]

I think you are right. I'll try that and post something if it doesn't work. It does have a lot of low frequencies. I didn't know that low frequencies cause noise which seems to be a high frequency but I guess this is because of overtones.

[braincell]

I decided this is too much trouble for just a few measures so I switched it to a more treble brand of piano. I think that will solve the problem. It's interesting that this piece will have two different types of pianos but they are separated so I don't think anyone will notice. I'm still going to use the massive compression like you said.

[valis]

It's all about results, if you think the end result is improved. I think there's a line between feeding the encoder more signal to retain more bandwidth & killing attack transients & making decay too 'reverby', one that you'll have to define for yourself.

Your encoder really should be hipassing at 20hz though, if you're not already doing that during the 'mastering' you're doing.

[braincell]

I cut back on the compression because it sounded distorted. It's too reverby but at this point I am going to settle with pretty good and giving up on perfect. I always hated mastering anyway. By the time I get to that, I am already sick of it.