Summing in T..few questions..

[stash98]

I have noticed some things when I mix in T. First off, I always have 64 bit mixing on.

First Question is ...why is the deafult setting on the 2 bus -3 instead of zero?

Now the next thing.. I have made some observations..like when I mix in T2, I get the peaks LOUD..like at -3 to 0. I then have to use very light compression to catch peaks.

Now a good mixer would say.."hey man, you mix too hot." And I'd agree...but then I go to my next step..

I use tapeit2 instead of render..I do this because I use a lot of programs that dont like to render all the time (battery, kontakt..you know..). So I am capturing a 32 bit floating file.

So after this. I'm not expecting square wavs (because of the comp) but I'm expecting a pretty hot mix..but I'm pleasantly suprised w/ this waveform:




So I am just trying to figure out the nuances of the 2 bus in tracktion..how loud is loud? I am pushing zero in this puppy each mix.

I ask this because I notice that when I mix according to thr "rules" (peaks at around -9) I get a very soft mix that requires way too much compression/limiting to get to an acceptable volume.

So I am just trying to learn all I can about the engine.

To end all of this.these mixes are the best I have made, and are incredibly consistent. I am very happy with the consistency..this is more of me just making sure I know all I can about how the 2 bus works in Tracktion.

[pHz]

looks (sounds ??? ) like you have tapeits automatic gain reduction switched on ... it checks for peaks then drops the recording level to give you 3db headroom over those peaks (if i understand how it works correctly myself) ...

TapeIt will measure the recording level for you and reduce gain accordingly, until there is no clipping anymore (with a safety headroom of 3 dB).

slainte rob

[djsubject]

i mix hot as f**k all the time!

maby because i am a hot head

i allways say if it sounds good its good!

-3db on the master is because........ummm well i dont know but when you mix hot it sure peaks easy

do as you wish & trust your eres

PS. i enjoy maxing the hell out of a track after its been exported!!!! thats how i can tell how good my mix is! if i cant get it that loud before it sounds squashed then it needs doing again


Subz



DISCLAMER: i am only a bum who makes music in his bedroom!!

[rpc9943]

so what i record music and bands and mixdowns for a small business on my own in my bedroom.

RonC

[stash98]

phz...how do i turn this off..I think you just nailed the issue!

[djsubject]

so what i record music and bands and mixdowns for a small business on my own in my bedroom.

RonC

me aswell! but it seems i spend most my time here not working in a pro studio getting paid pro cash

what i have lernt is different music needs to be mixed different!

EG: if i push my uad-1 compressor so much that it peaks the shit out of the peak meter on that track (on my kik's) i get a much rounder kik than if i mix it not so hot then gain it up somewhere else

Subz

[stash98]

yeah..different styles have totaly different mixes..for dance you want that kick to be huge compared to a rock track.

[stash98]

pHz...

I went back into that mix..I noticed that tapeit will show you if the gain reduction is pushed on..so I watched, and I saw none..I took off the compressor and the song was peaking at -1.2, so I didnt need it! Now -1.2 should be LOUD..but here is the mix..BTW it sounds great




so, I'm learning by my ears and waveforms like this, that T2 was meant to be pushed hard and loud to sound really good...what do you think?

[valley]

Are you ever zooming far into the waveform? 'cos the image you are showing above tells you nothing very useful about peak levels.

A typical waveview of an audio file longer than a few ms in length will be greatly compressed on the x axis. Depending on the method used to render the wave file as an image, the compression will yield greatly different looking results. At the most basic the waveview is simply an avearge of teh level for a give number of samples, IE:

lets say each pixel represents 1000 samples at the current zoom level - set the y value to be equal to the average of those y values. If there is a huge peak but the average level is -15dB, you wont even know the peak is there unless you zoom in.

Your wave editor should be able to give you the peak and RMS levels for the wave file. They'll give you a better idea of what is going on that just eyeballing the wave image.

[stash98]

yes.. I can do those things..but from a quick look, I can tell what is louder then something else. Like many commercial tracks zoomed out into this view almost look like complete square wav.


Basically, my whole point is that I'm trying to find the "sweet spot" in tracktion. It would make sense to get RMS levels though.

do you mean RMS power maybe? That says it is -19. under statistics in Sound Forge. if you are familiar w/ soundforge you may know exactly what to do.

[valley]

yes.. I can do those things..but from a quick look, I can tell what is louder then something else. Like many commercial tracks zoomed out into this view almost look like complete square wav.

But you are still confusing peak with power.

For example:

I took off the compressor and the song was peaking at -1.2, so I didnt need it! Now -1.2 should be LOUD..but here is the mix..BTW it sounds great

Peak levels just show you how loud your track gets for some period of time. What you really care about if you are trying to mix hot, is how loud it is *most* of the time.

RMS gives a much better idea of how hot a track is. Peak just tells you that your quiet track has a few loud transients.

Basically, my whole point is that I'm trying to find the "sweet spot" in tracktion. It would make sense to get RMS levels though.

There is no sweet spot in Tracktion. Ideally your mix should be rendered such that it never clips, but peaks somewhere between -3dB and 0dB (the closer to 0 you can get, the better).

If you want to raise the overall amplitude you need to use compression, or even limiting if you want it really hot.

A compressor effectively allows you to raise the average level without raising the level of peaks, which is what you are trying to achieve.

A good demonstration, albeit a bad mixing technique, would be to place something like Kjaerhaus' classic limiter set to -6dB in the master filter section. Render the mix with and without the plugin, and compare the results.

do you mean RMS power maybe? That says it is -19. under statistics in Sound Forge

Yup, that's the one.

[stash98]

ok cool..yeah you are right.the peak thing is not the way to base it..BUT..let me tell you why I usually end up doing it.

I have mixed for long enough (8 years) to know what my peaks will be (usually vocals or snares in rock..snares or kicks in dance). So I guess that is why I do it.the RMS does make much more sense.

I totally see what you mean though. What is a good RMS level for a premastered track such as mine?

[freeztar]

I totally see what you mean though. What is a good RMS level for a premastered track such as mine?

Depends on how loud you want it. The 16-18 range seems to be good for most of my stuff, but for dance music and hiphop you might want even more loudness than that. Use your trusty ears.

[valley]

Use your trusty ears.

Or, if you really aren't sure you trust yourself, then rip a track that is broadly of the style and sound that you want, load the wav into soundforge, and:

a) check the RMS value for the track
b) take a quick look at the spectral content, and note how balanced or otherwise the frequency spread is.

Now, compare that to your own work.

The spectral comparison can be particularly telling, because very often what is perceived as a quiet recording, can actually be simply too much bass, or sub-bass.