Skip to main content

This part of a series of music I had restored from tape written and recorded a few years back. So you could say it's Hybrid recording. This came from a project that was never really finished and with all the time down since March 2020, I had time to finally restore the 16 tracks and drop them into Pro Tools sessions for enhancement and sweetening. I was surprised at the recording quality on most of the tracks. We had a decent little studio at the time although we lacked a few really great pieces in the outboard.

I put this up as a sample of the hybrid way of production but I'll listen to critique as well. You guys have great ears and this isn't quite finished.

The vocals and drums are from the original tracks and all the guitars, Keys, and a new bass track were added this year. (2020). The vocals have been tuned and timed with Melodyne 5. I'm still sorting out the compressor. Right now there's an LA-2A on the track and a multi across the 2 buss. Every element in the mix has a buss. The guitars are an actual stem I had built to send to the keyboardist, Howard Helm, who does all my key work. Anyone needing real keyboard production should hire him! Seriously it's what he does. PM me.

As always, I'll answer any questions about what you hear and why. Thanks and enjoy "If My Heart Don't Fail Me" It's mine and the singers composition from 1987.
 

Attached files If My Heart Dont Fail Me print.dup1.06_02.mp3 (9.9 MB) 

Comments

Boswell Sat, 01/16/2021 - 10:59

It's good, Dave - you like mixing these really hot, don't you!

I can't prove this, but I just have the feeling that one or both of the kick or bass is causing the whole track to duck in the 2-bus compressor. Could I suggest a separate mono mix put through an HP filter to use as the comp's side chain, rather than have it use the full-range mix as the compression source? A different approach would be to have the bass at a lower level in the sidechain mix, but either way, it would probably mean bringing the bass down a dB or so in the overall mix.

Davedog Sat, 01/16/2021 - 11:41

Boswell, post: 466799, member: 29034 wrote:
It's good, Dave - you like mixing these really hot, don't you!

I can't prove this, but I just have the feeling that one or both of the kick or bass is causing the whole track to duck in the 2-bus compressor. Could I suggest a separate mono mix put through an HP filter to use as the comp's side chain, rather than have it use the full-range mix as the compression source? A different approach would be to have the bass at a lower level in the sidechain mix, but either way, it would probably mean bringing the bass down a dB or so in the overall mix.

LOL. Thanks Bos. I'm not shy. LOL. I have the Abbey Road Mastering chain with the CLA Mixdown on the aux master before my print and I'm beginning to think that the 2-buss comp on the bass stem is not playing well with the stuff at the end. I will certainly look at that. I think without these plugs on the aux master it's hitting the meters at around -16LUFS.

I said it wasn't mastered but that's not entirely true. What it is is "semi-mastered" like if a client asked me for a daily to live with for a week. And thats why the Abbey Road and CLA across the aux for the print. If I take those off then I'm sure the effect thats popping up will disappear but then I would have to worry about it at mastering. Although my mastering engineer uses GREAT hardware and he wouldn't let this happen. He'd also send it back with the same instructions!

Thanks. I'll put up another one without the "sorta-mastered" stuff on it. See how that sounds to y'all. Stay tuned.

Davedog Sat, 01/16/2021 - 23:36

Reworked a few things.
Not to make excuses but I'm down my main power amp for my bigger monitors for a while and I have a temporary amp in place. It's a Crown but not as much headroom and power as my regular custom built thingy....."so what!" yeah really.......

Here's something different than the other one. I think I was not paying attention to the low-end when mixing on my Neumann KH120's. I should know better by now I've had em for a few years. But now and then when I'm mixing at a low volume I keep trying to "hear" things that they are actually reproducing but not so much at the low levels.

Funny. I ALWAYS check my finals on them before sending things out to mastering.

I put this 'splaining in for those who maybe don't quite get it yet. Even someone who has mixed a large number of things throughout the years can get complacent or lazy and forget where they are.

Not to say this a great mix yet but I'll wager you guys can hear it better now. Comments appreciated.
 

Attached files IfMyHeartDontFailMev4_03.mp3 (9.8 MB) 

Davedog Sun, 01/17/2021 - 11:58

bouldersound, post: 466822, member: 38959 wrote:
It's not ducking with the kick so much now, but it's still so heavily compressed it's hard to evaluate. Unless that compression is intended as part of the mix, then I'd say it's too compressed.

Interesting. The whole mix (2 buss)doesn't have very much compression at all. Maybe a touch over 1.2:1. There's a small amount of limiting but the levels aren't even hitting it. I'll make one more just to see. I worry when people use terms like "heavily" or "Too".....To the point where they say it's impossible to evaluate something.

Davedog Sun, 01/17/2021 - 13:45

bouldersound, post: 466824, member: 38959 wrote:
It might be compressed at the track or submix level. I don't know how else you would get it to 10.5 LUFS integrated with a Loudness Range of 1.7 LU.

Could you put up a screen shot of these figures please? There just isn't that much compression on anything. The guitars have their fair share but the keys have none, the bass was backed off a LOT. The drum buss has some but not what I would call a "lot" ...There's an SSL comp across that 2-buss (of course). I print my drum sounds once I get them ironed out since I use Slate triggers and don't like how cpu heavy they can be with multiple incidents. As I explained in the post, the drums and vocals came from tape. So there is some of the compression you get from that. Other than that I can't find a point that would be over compressed or would generate those kinds of output levels

Boswell Sun, 01/17/2021 - 14:08

This may be the sort of effect that is pushing up the LUFS figures. It's one beat taken from about 1:02:40 to 1:02:60 in your track showing the bass guitar note lasting the whole beat and the other material riding on top of it. This is a simple screen shot from Audacity with no adjustments to the amplitude, so the top and bottom screen extent of each channel is 0dBFS. However, it may be that the RO MP3 is normalising everything.

 

Boulder and I are not attacking you Dave - I certainly am just as keen to find out how the LUFS figures represent different types of material, and why this track is pulsing my sub unusually heavily. The overall Audacity display shows only very light compression at the 2-bus level.

Attached files

Davedog Sun, 01/17/2021 - 14:27

Thanks Bos. NOT feeling an attack at all. Just want to sort this out as best I can. I'm very curious as to the why and the how. My hearing is still intact and as you have observed I like to 'fill it up' a bit. I always am unimpressed with MP3 encoding. I've started simply making the MP3s I present through the ProTools ...It actually seems to sound better than say the MP3 generated through iTunes or some other devices. I don't know if RO changes anything. Chris would know.

Davedog Sun, 01/17/2021 - 14:40

Kurt Foster, post: 466828, member: 7836 wrote:
were the source tracks recorded with dbX?

Yes. Tascam MS16 w/dbx and at the time I had it biased at the recommended +9 for the 456 tape which was brand new. The digital transfer was done on an Otari MX70 w/dbx and if I'm not mistaken that dbx is encode/decode simply to protect the tape.

Davedog Sun, 01/17/2021 - 14:43

audiokid, post: 466832, member: 1 wrote:
For whats it worth to the community, I've done many test with mp3's here. I've uploaded an mp3 and then downloaded it and it matched to my master mp3.

Thats what I am finding too. Good to know without a doubt. Some of my earlier posts I made the MP3 through other programs rather than PT. The PT MP3s are all @ 320kbps

KurtFoster Sun, 01/17/2021 - 14:53

Davedog, post: 466831, member: 4495 wrote:
Yes. Tascam MS16 w/dbx and at the time I had it biased at the recommended +9 for the 456 tape which was brand new. The digital transfer was done on an Otari MX70 w/dbx and if I'm not mistaken that dbx is encode/decode simply to protect the tape.

i think that is what's going on. dbX will double freq response errors on tape by a factor of 2 as well as dull transients .... so if the response is + or - 3 db with the dbx it becomes + or - 6 and it gets worse if it is printed over 0 dB. then it went through 2 encode decodes ..... also in my expierence i never needed to compress anything if i was using dbX.

Davedog Sun, 01/17/2021 - 14:54

eternalsound, post: 466838, member: 48927 wrote:
Sounds better but yea ..the kick. The whole set is still poofy though.

I think I got it about here. Maybe it's your speakers?

No. Not speakers or mix position or playback quality of any sort. I'm years past that sort of thing.....I don't know "poofy" LOL

Davedog Sun, 01/17/2021 - 14:58

Kurt Foster, post: 466842, member: 7836 wrote:
i think that is what's going on. dbX will double freq response errors on tape by a factor of 2 as well as dull transients .... so if the response is + or - 3 db with the dbx it becomes + or - 6 and it gets worse if it is printed over 0 dB. then it went through 2 encode decodes ..... also in my expierence i never needed to compress anything if i was using dbX.

Interesting.

Davedog Sun, 01/17/2021 - 15:03

audiokid, post: 466844, member: 1 wrote:
Did you lift anything above 12K more than 6db? I just listened again and it sounds like compounding HF oversampling and then your mp3 master crushing it, Hope that makes sense.

Thanks Chris. This is great info. I generally don't EQ anything above 8-12K. I can still hear that stuff and if it's tracked right those frequencies should be where they need to be. I don't tend to use EQ very much except on a 'feature' part. There's a Harrison 32 on her voice at the track itself. The lead guitar has something on it for it's short little part. A bit of API and a Fairchild but that's not going to affect the whole mix. Right now I have the track up and I'm going through all the subs to determine if there is something banging on something else. Thanks.

Boswell Sun, 01/17/2021 - 15:06

Kurt Foster, post: 466842, member: 7836 wrote:
i think that is what's going on. dbX will double freq response errors on tape by a factor of 2 as well as dull transients .... so if the response is + or - 3 db with the dbx it becomes + or - 6 and it gets worse if it is printed over 0 dB. then it went through 2 encode decodes ..... also in my expierence i never needed to compress anything if i was using dbX.

Kurt's point is interesting, but wasn't the bass guitar one the re-recorded tracks, so what we are hearing is the kick but not the bass from the tape?

bouldersound Sun, 01/17/2021 - 15:07

Davedog, post: 466840, member: 4495 wrote:
Thanks! Now if you could give me a simple explanation about what is being shown here I'd appreciate it. Bos put his figures up and also notated their points of interest. I'm very curious about this.

Yep. Integrated is the overall loudness of the track from start to finish. 10.5 LUFS is fairly loud, but not out of the normal range for a CD master. It's a few dB hotter than typical streaming service specs. Loudness range is the overall dynamic variation of the track. 1.6 means it has fairly low dynamic variation. Momentary and Short Term Max values are peak and short term loudness. True Peak Max is the estimated peak of the reconstructed analog signal at the output of a DAC, accounting for inter-sample peaks.

"Loudness" refers to the perceived loudness as determined by carefully designed algorithms. It's designed for comparing the perceived loudness of different source files as they would sound played through a system at the same volume setting. If a -10.5 LUFS integrated song followed a -14 LUFS integrated song, it would generally sound louder. If it also had a lower Loudness Range value, it would sound more constant, with a flatter dynamic profile.

Davedog Sun, 01/17/2021 - 15:11

Boswell, post: 466850, member: 29034 wrote:
Kurt's point is interesting, but wasn't the bass guitar one the re-recorded tracks, so what we are hearing is the kick but not the bass from the tape?

Yes. And the drum tracks have been treated with Slate Triggers to cut back on the mic bleeds. So while the performances are the originals, the sounds are a reasonable approximation of the original kik/snr/tom sounds. The Overheads are unaltered other than timed with the rest of the kit. So actually the ONLY original tracks are the overheads and the vocal

Davedog Sun, 01/17/2021 - 15:31

I gotta thank all you community members for giving a listen and a suggestion or two. It's good for me.

And in retrospect, I find myself guilty of one of my Producer sales points......Because this material is MINE I find I'm trying too hard to make it perfect instead of adhering to tried and true techniques I would normally use and have used on the last few records I've made with others.

I do mix 'hot'. No doubt. I have excellent gear and a very usable mix position that I really know well. Chuck's post of the EQ change opened it up for me and I'll be revisiting things that I normally do these kind of cuts on at track levels. Why they aren't already done refers back to the being enamored of my own creation.

Bad Davedog.:LOL:

kmetal Sun, 01/17/2021 - 18:06

Interesting stuff. The only thing i can add is that on my phone, which is my primary listening device right now, i can't hear any of the issues described. Things sound fine generally, with the upfront vocal sound being the thing i picked up on. I haven't listened on my bt speaker which i would use if sitting down to listen for a while, it has deeper bass.

I don't really have much of a point other than to illustrate how much can be lost/unnoticed by the end user's listening system. Im sure we all know this, just found it interesting in this case.

Im not convinced yet the LUFS is a very important metric. Its a moving target since each streaming platform has a different (proprietary) algorithm, and they are routinely tweaking them. I used to have the mindset of mix/master for each platform so the mix sounds best, ive now switched to the Andrew Scheps, Ken Lewis notion of mix it till it sounds good and nevermind the lufs/destination. I think its interesting to see the lufs readout, but i don't know that id alter anything i liked because of it.

Also Dave, you may be interested in the Fraunhofer Codec software from Sonnox. I haven't used mine yet, but bought it years ago when i realized not all encodes were the same. It does batch conversion which can be a time saver.