jagalactic

Right - not the whole story, for sure (re: sequential, etc. - although sequential is less important on SSDs than spinning disks). A 16/44 *stereo* sample is 4 bytes (2 x 2 x 16 bits) because it's actually a left and a right sample.

Most disks and SSDs are fast enough for audio. Here is some math. Each sample of 16/44.1 audio is 4 bytes (2x16 bits = 32 bits = 4 bytes) Each sample of 24/96 audio is 8 bytes (because 24 bit samples are stored in the 3 low-order bytes of a 32 bit 'word', and there is one 'word' for left and one for right; If I'm wrong and it's 6 bytes, the MB/s requirements go down by 1/4 - but I don't think I'm wrong; X86 systems don't handle any 3 byte packed types well). So how much data is in 1 second of stereo 16/44? 4 x 44100 = 176400 = 176K How much data in 1 second of 16 stereo channels of 16/44? 16 x 4 x 44100 = 2822400 = 2.8MB One second of 24/96 stereo audio is: 8 x 96000 = 768000 = 768K One second of 16 stereo channels of 24/96 = 16 x 8 x 96000 = 12288000 = 12.2MB I'm not suggesting this is the whole story, but it's important perspective. If I'm recording 16 simultaneous 24/96 STEREO tracks (which is 32 interface channels), the *audio* data is less than 13MB per second. Either thunderbolt or USB3 (or even USB2) is fast enough to move that audio and do a few other things at the same time (though doing a lot is asking for trouble). If your disk or SSD is a lot faster than that - and it's on USB3 or TB - you should be fine. YMMV

Thanks Simon. You're not the only person who recommended dynamic eq and multi-band compression for singling out the problem frequencies. But I'm pretty motivated not to process the whole track(s). Some of my tracks have notes in the frequency range where the noises are. And, it turns out there is a way to get access to the Izotope RX editor without buying. I just discovered this today, but slice.com has "rent to own" on a bunch of audio software, plugins and samples - including RX 7. So I rented it for $15.99/mo. I would own it after 25 payments (which is full price, but not more), but I'm not required to keep renting it. Apparently they allow stopping and restarting payments, provided they are still supporting the software package in question. So far it's working great. I can't completely vouch for slice.com yet, but so far it's solving a real problem for me. Cheers, John

I'm mixing a jazz trio album - acoustic piano, drums and bass. There is a lot of lovely leakage between the drums & piano, which is a feature and not a bug. Throughout the piano recording, the pedal noise is audible here and there. No squeaks, but just a bit of a "boom" when the pedal is pressed and the mutes lift. I am only interested in attenuating it during a handful of quiet passages, where it could distract from the subtlety of the music. Otherwise it's a natural sound that should remain. I won't EQ the entire tracks in this range. I suppose I could automate a linear phase EQ to briefly attenuate targeted frequencies, but I have two concerns: 1) I don't want to introduce audio artifacts as the eq settings are changed, and 2) I find that EQ can have an audible effect even when flat - and I want to avoid that. I found this demo of the iZotope Spectral Repair tool (https://www.izotope.com/en/learn/products/rx/how-to-remove-piano-pedal-thuds-in-rx-6.html) - doing exactly what I want to do. But it looks like $399 for the minimum package containing that tool, which isn't in the budget. Can anybody suggest other techniques or tools that might be effective? Recording is finished, so I'm not in a position to re-do it. And it is my judgement that the piano player did not abuse the pedals (or any other part of the piano). And you can be sure this is true because I am the piano player Thanks! John

You are The Man, David. Thank you!!

I'm working on a project where we overdubbed a handful of cleanup takes a few days after original tracks (which we're mostly pretty clean single-takes). This turned the original track(s) into composite tracks, but the volume is not identical - the setup had been torn down and then re-assembled. Ideally I would like to have a per-take gain adjustment. In this case the new takes are about 3db quieter than the original ones. Although I could apply automation to compensate, that would be hard to work with in case I change up any of the composite takes. Ideally I would just make those quieter takes louder. Is there a (straightforward) way to do this? Thanks!

Arnaud: thanks for the detailed clarification, and yes I understand that plugins, eq, will interfere with the nice simple dB math - because it's hard to know the exact and perceived gain effects. In my original post, I realized that the comment that "...the kick drum mic is messing with the otherwise-working (very subtle) compression on the overheads..." indicates that the submix bus gain is not the whole story due to the compression on the submix bus. But it's "very subtle" Thanks for all the excellent info! John

Excellent, thanks very much guys.

Say I have a track recorded, and in my mix I have -6db on the Logic mixer fader. Say that channel is routed to a bus with +3db gain on its fader (and that bus goes to the stereo out). Will that track be at the identical volume if I send it to the stereo out with -3db on the track fader? (i.e. -3 = -6 + 3) ...or is it more complicated than that? I'm mixing drums on an acoustic jazz recording through a bus, but the kick drum mic is messing with the otherwise-working (very subtle) compression on the overheads. I'm using my ears, but I'd like to know if the technical answer is yes or no. And I add this with LOVE: I may not read answers that begin with "I don't know, but..." Thanks, John

I think you're right, JakobP. I did two re-tracking sessions on the piano part in a 10-piece jazz ensemble recording (originally recorded live, but one track had problems in the piano performance). So upon re-tracking piano, there was apparently playback latency (the existing tracks played slightly late) that caused the piano performance to be a little late. On one of the sessions, I put Logic in low-latency mode, and there was no problem. On the other session, I failed to do this and I had a sync problem as a result. Of course all recordings need to be in-time and in-sync, but this track is a fast samba; the rhythmic piano comping sounded awful when it was just a few milliseconds late (but great once I got it shifted into sync). Glad it didn't need to be redone... I was tracking on a late 2017 MBP, and few if any plugins were active, but... somehow it happened. Cheers...

Ah, mystery solved - it's basically operator error. First I selected region(s) by clicking on them with the pointer tool, but the region inspector looked "empty" to me (although it wasn't - the sole adjustable parameter is the Delay - which is what I need). Since it looked like a selected region had nothing in the region inspector, I tried the marquee tool to select a portion of a region - and noted that there was lots of stuff in the region inspector. But changing the delay parameter with a partially selected region didn't have any audible effect. When I revisited selecting regions with the pointer tool, and adjusted the Delay, it worked. Thanks for reading and responding...

Apparently not a common issue. Let me elaborate on what's important, just in case. The track in question appears to be a solid performance (it's a piano part on a Samba - a fast latin jazz tune, fyi, and it must be played and played back "in the pocket") . The performance was in the pocket, but the audio in logic plays back slightly late. I was able to zoom way in on the track view, and drag a take forward "slightly" - and it landed in the pocket. But there was a half-time breakdown in the middle of the tune, and the piano track was separately recorded for the first and last verses. By dragging the take, I got it into the pocket but I have no idea how many milliseconds forward I dragged it - meaning I don't know how to drag the first/last verse takes the exact same amount (although I is clear that the exact same amount for both is the correct answer). More generally, any time I have takes that are slightly out of time like this, the correct answer would be to move ALL of them forward the exact same amount - but I don't see how to do that. Surely it exists. If there was a dead-simple delay plugin (like a subgroup delay on a console for the purpose of time-aligning PA speakers to account for the speed of sound), AND/BUT capable of negative delays, this would be ideal. I'd just put that in an insert, dial in the right number of milliseconds, and then bounce a corrected copy of the track. Bueller?! (Thanks community)

I have some composite piano tracks (audio) that are a few milliseconds late. They sounded in the pocket when tracking, but they're consistency behind the pocket on playback. I have found that setting logic in low-latency mode when adding overdubs seems to avoid this, but I forgot in a session yesterday. It looks like the region inspector should allow me to set a delay (or time offset), so that a region should play a bit early (or late, but I need to shift it early). However, selecting a region and setting the delay parameter does not have any audible effect. Should the region inspector delay parameter be usable for audio tracks? Is there another way to accomplish this? Thanks, John

Having thought on this for a while, I think I'm going to treat it like a pedal board - I'll mount the hard drive, USB hub, and possibly the audio interface on a board with velcro. Route the cables nicely, and rig it with a (probably hinged) platform on top for the macbook. I can set it up where just the necessary cables emerge, and the whole thing can travel as a unit. When (and if) I "git 'er done" I'll post pics...assuming I'm not the only one who cares. Meanwhile, I'm still interested if others have found a good solution... John

I've used a Yamaha S90 as a Mainstage controller - it's fairly similar. These Yamaha boards are bare bones as midi controllers, but they work. You'll need to tell it to transmit midi (utility menu, midi switch). If it's like my S90, it can control midi via either a standard midi cable or via USB; however, the USB connection was not usable for me because I had ground loop hum problems (YMMV, possibly a defect in my S90). Since I wanted to use the S90's patch selection buttons to send midi program changes to Mainstage, I just put the S90 in voice mode on bank A, and then the 16 patch selector buttons sent program 1-16 (with bank B sending 17-32, etc). This works well if you don't need to mix Yamaha sounds with Mainstage sounds (I mute the S90 audio output while controlling Mainstage). If you need to mix 'em, you'll need to use a different method for selecting patches (unless you can live with voice-mode-only patches on the MO, and put them at the same patch numbers you're using in Mainstage). Hope this helps, John