|
|
Post by mrholmes on Jan 13, 2016 21:08:55 GMT -6
It means that if the RME has a maximum input of 13dBu.... And nominal line level is +4dBu...... The 13 minus 4 is 9. Therefore, if you send a signal at line level , it would register at -9dBFS on your daw or in total mix. Which would mean you only had MAX 9 dB headroom above line level. Make sense? Cheer Wiz Got it you are right, makes sense but using lower levels just sound better to me.... |
|
|
|
Post by wiz on Jan 13, 2016 22:37:13 GMT -6
I ain't suggesting you track at that 8)
You need to be hitting your converters so the peaks are roughly no more than -12 dBfs...
Whatever that is in dBu is converter dependant..
Cheers
Wiz
|
|
|
|
Post by ericbradley on Jan 16, 2016 6:54:03 GMT -6
Logic applies TPFD dither automatically when bouncing to 24-bit. There is no need to use Logic's additional dithering algorithms unless you are truncating to 16-bit.Sorry for the confusion on this matter. The above statement was plain wrong. I had run a couple of tests myself before and had found that this wasn't the case but after a talk with a fellow Logic user I have confirmed that this isn't the case. Just wanted to correct it. /Eric |
|
|
|
Post by mrholmes on Jan 16, 2016 15:47:00 GMT -6
Logic applies TPFD dither automatically when bouncing to 24-bit. There is no need to use Logic's additional dithering algorithms unless you are truncating to 16-bit.Sorry for the confusion on this matter. The above statement was plain wrong. I had run a couple of tests myself before and had found that this wasn't the case but after a talk with a fellow Logic user I have confirmed that this isn't the case. Just wanted to correct it. /Eric But with this you can..... and its for freeeeeeeeeeeee www.airwindows.com/ditherto/ |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 16, 2016 20:59:19 GMT -6
Huh? I thought it was a joke when i read "boutique" and "gloss" and "vinyl" dither in DitherBox. But i think they mean it, right?
It is dither. TPDF is A-ok and i just wanted to get no sound at all from dither, i mean this is the point of dither for me. To not get *any* sound from dither *at all*, but preventing quantization noise.
Maybe i should sell a "mellow crisp" dither plugin. (just doing no dither at all for this *rich* modern noisy sound...)
|
|
|
|
Post by mrholmes on Jan 17, 2016 18:18:21 GMT -6
Huh? I thought it was a joke when i read "boutique" and "gloss" and "vinyl" dither in DitherBox. But i think they mean it, right? It is dither. TPDF is A-ok and i just wanted to get no sound at all from dither, i mean this is the point of dither for me. To not get *any* sound from dither *at all*, but preventing quantization noise. Maybe i should sell a "mellow crisp" dither plugin. (just doing no dither at all for this *rich* modern noisy sound...) A ME demonstrated me diffrences I was surprissed too and I have no idea why we can hear it. |
|
|
|
Post by Ward on Jan 17, 2016 20:46:50 GMT -6
We're still debating this? Lord Thundering Savior! (I apologize for the Pentecostal swearing) Can't we end this and just have a resident expert write a tutorial already?
|
|
|
|
Post by tonycamphd on Jan 17, 2016 23:31:26 GMT -6
I'm certainly not that guy haha, but as I understand it,
Dither doesn't add pleasing sound to anything, it masks unpleasing sound, there is a difference. Check out this video, i agree with the vast majority of what he's saying, but I have a few problems with some of his implications...
First, his implied equivalence of his gear to human hearing is something that i just don't buy, measuring devices are very useful, same with null tests, but they are crude in comparison to the biological sophistication of the human hearing experience, and NOT on par with it, i've proven this to myself time and again. Second thing he said that bothers me is.. "no one ever ruined a great recording by not dithering" really? hmmm..., He's trying to prove his point with a single sine wave, even modulated, but when it comes to mixing complex audio waveforms, using many plugins and automation moves ITB that ensure a bazzilion calculations of quantization correction, and since quantization noise is input dependant, and whatever the input is will create nasty fixed quantization harmonics, then stacking track after track of this will certainly be bringing up the level of this nastiness, in my mind this is the #1 suspect to what is causing the annoying digital hash I hear on a lot of recordings, I personally refer to it as "digititus".
I hope that makes sense? haha, and all that said, I fully admit I am no expert on this stuff, my understanding of it has a lot of holes in it, it's hard to find info to fill them, please someone esplain tu me how i'm messed up here, I would appreciate it.
|
|
|
|
Post by mrholmes on Jan 18, 2016 2:55:13 GMT -6
I'm certainly not that guy haha, but as I understand it, Dither doesn't add pleasing sound to anything, it masks unpleasing sound, there is a difference. Check out this video, i agree with the vast majority of what he's saying, but I have a few problems with some of his implications... First, his implied equivalence of his gear to human hearing is something that i just don't buy, measuring devices are very useful, same with null tests, but they are crude in comparison to the biological sophistication of the human hearing experience, and NOT on par with it, i've proven this to myself time and again. Second thing he said that bothers me is.. "no one ever ruined a great recording by not dithering" really? hmmm..., He's trying to prove his point with a single sine wave, even modulated, but when it comes to mixing complex audio waveforms, using many plugins and automation moves ITB that ensure a bazzilion calculations of quantization correction, and since quantization noise is input dependant, and whatever the input is will create nasty fixed quantization harmonics, then stacking track after track of this will certainly be bringing up the level of this nastiness, in my mind this is the #1 suspect to what is causing the annoying digital hash I hear on a lot of recordings, I personally refer to it as "digititus". I hope that makes sense? haha, and all that said, I fully admit I am no expert on this stuff, my understanding of it has a lot of holes in it, it's hard to find info to fill them, please someone esplain tu me how i'm messed up here, I would appreciate it. I am not an expert too, all I said I was able to hear a difference in the mastering studio. In my thinking dithering is a noise that is so low down in the floor it should not make a difference... One thing I hear within the first bar of a tune, if I did forget to enable the dither boutten before going to lower bit depth. You hear it on the verbs, on depth and dimension its damaged. If its an Album I leave the dither decision always to the ME. Done.... |
|
|
|
Post by jazznoise on Jan 18, 2016 6:10:41 GMT -6
I don't really feel anyone should ever care about dither other than it be applied when reducing the bitrate. There's always something better to be worrying about than what is happening at -100dBFS.
|
|
|
|
Post by Ward on Jan 18, 2016 6:25:22 GMT -6
I don't really feel anyone should ever care about dither other than it be applied when reducing the bitrate. There's always something better to be worrying about than what is happening at -100dBFS. Minus a hundred DBFs? Ain't nobody got time fo dat! |
|
|
|
Post by tonycamphd on Jan 18, 2016 8:52:21 GMT -6
I don't really feel anyone should ever care about dither other than it be applied when reducing the bitrate. There's always something better to be worrying about than what is happening at -100dBFS. Minus a hundred DBFs? Ain't nobody got time fo dat! if you are feeding processed ITB tracks to a console via multiple da's, and you are NOT using dither on a per channel output basis, it's a fail, and your audio is suffering for it. The lynx software routing mixer has automatic dither on any channel that has the slightest fader movement ITB. Noise shaping is also very useful as you can send dither energy to specific areas to be filtered off later OTB. If Bob O says it's important with multiple thoughtful answers, you guys should rethink your one word GS style pop shots, i find them quite annoying to be honest. realgearonline.com/post/60200/ |
|
|
|
Post by jazznoise on Jan 18, 2016 9:17:31 GMT -6
Any audio DSP that upsamples and downsamples and doesn't employ filtering and dithering is faulty DSP. Dither is something that was a big deal in the 90's that has been taken out of the hands of engineers in 90% of cases because it's frankly more of a statistical issue than taste. Some people just weighed ones are worse for errors, some suggest that the frequency profile begins to shape the noise floor, some people claim the shaped noise floors sound better.
Why are you putting dither on every channel? Are you downsampling? If not you're just adding noise, even if it's just a tiny irrelevant amount of it. The gain changes in your DAW should be linear and if the engine truncates then, again, it also downsamples. I can't see why it would matter, especially in the stuff we talk about here - which is compressed 2 fuk styles of pop music. We don't use enough dynamic range in most things to worry about bits 16-24 in most situations. In some modern Pop I'd say even bits 12-16 are irrelevant.
|
|
|
|
Post by Bob Olhsson on Jan 18, 2016 9:29:20 GMT -6
Dither doesn't mask anything. It keeps the bottom but from chattering like a noise gate which is far more audible and, more important, builds up twice as fast in each processing stage as the hiss-like sound of randomizing the bottom bit, aka dithering. Noise is not dither even though dither sounds like noise. While I thought noise shaping sounded better back in the '90s, I think in most cases it sounds worse with modern converters. I do try it. Lack of twenty four bit dither is audible as a loss of depth in recordings made with microphones. If the audio has ever been truncated, the distortion masks the effect of the dither which leads people to believe dither doesn't matter.
The only reason we even talk about dither is because the early consumer audio manufacturers and developers were some combination of ignorant and lazy. They left it out of their software creating a horrible reputation for digital audio. Unfortunately there are only a handful of people in the world who know audio, programming, math and DSP. Dither has been a part of digital telephone communications from the very beginning in the 1950s.
|
|
|
|
Post by Martin John Butler on Jan 18, 2016 10:06:49 GMT -6
Great posts guys. Now, I've recorded my tracks in Logic X, I want to bounce. Usually I've added all sorts of plugs in the 2 bus to faux master instead of creating a clean mix and then doing a mastering mix session. i'm beginning to see what I need to do, keep my inputs and outputs to between -20 and -12, and will try to limit my 2 bus plugs to ones that are only for tone, and not mainly for volume, maybe the Ampex ATR-102, maybe one of the Slate VBC compressors.
Can someone please suggest what settings to use at that first stereo mixdown in Logic regarding dither, and what dither setting, if any to use for the "mastering" session. Thanks.
|
|
|
|
Post by Bob Olhsson on Jan 18, 2016 10:30:10 GMT -6
Compressors affect balance so you never want to remove that. Logic always gave me a headache so I'm not very familiar with it. I think the best is to bounce to 32 bit float if it is capable of that. Otherwise, 24 bit with no noise shaping.
|
|
|
|
Post by tonycamphd on Jan 18, 2016 11:13:04 GMT -6
Any audio DSP that upsamples and downsamples and doesn't employ filtering and dithering is faulty DSP. Dither is something that was a big deal in the 90's that has been taken out of the hands of engineers in 90% of cases because it's frankly more of a statistical issue than taste. Some people just weighed ones are worse for errors, some suggest that the frequency profile begins to shape the noise floor, some people claim the shaped noise floors sound better. Why are you putting dither on every channel? Are you downsampling? If not you're just adding noise, even if it's just a tiny irrelevant amount of it. The gain changes in your DAW should be linear and if the engine truncates then, again, it also downsamples. I can't see why it would matter, especially in the stuff we talk about here - which is compressed 2 fuk styles of pop music. We don't use enough dynamic range in most things to worry about bits 16-24 in most situations. In some modern Pop I'd say even bits 12-16 are irrelevant. truncation happens with a single ITB process, that is why the lynx mixer enables dither upon a .01db movement of a fader, per channel feeding a DA, they know the quantization noise will be there, the idea that that ugly harmonic noise is 100db down is all but laughable in these times of hyper crush compression bringing all the uglies in that range, up into the audible range, less we forget that every instance of compression is a multiplication, not an addition, ultimately if i heard nothing but beautiful audio coming from myself and everyone else, i would probably be in the "it doesn't matter" camp, alas this clearly isn't the case. |
|
|
|
Post by Bob Olhsson on Jan 18, 2016 16:32:20 GMT -6
Distortion 100 dB. down is audible when nothing is in the same range to mask it or if the listener has hearing damage near the same frequency and can't mask it. Guitar shredders are especially sensitive to truncation and other digital artifacts.
|
|
|
|
Post by popmann on Jan 18, 2016 17:21:32 GMT -6
Posting any link to that Monty dip shits video might break the number one rule on the site. Just a friendly warning.  ? |
|
|
|
Post by M57 on Jan 18, 2016 17:21:38 GMT -6
Stop blithering you dithering GearNerds!
Getting back on topic.. I'm doing a pretty good job of tracking below -10 dbfs in Logic X on all tracks - and usually well below that; many don't hit -18 with the fader at unity. So what should I be looking for as the mix hits the 2-bus if I'm creating a mix just to post. I usually keep it pretty simple and bump things up 3 or 4 db with a little more compression, then apply some gain if necessary into a brick wall, usually knocking off another 1 to 3 db to get it as loud as I can without sacrificing dynamics. I've been reading (somewhere in these forums no doubt) that there are transients that fool my meters, so I've been setting the ceiling between -3 and -5 db; Logic normalizes at mix down anyway, so..
Actually I have questions about dithering too, but I'm just going to assume that unless I'm sending something out for mastering, I should dither, not blither ..right?
|
|
|
|
Post by popmann on Jan 18, 2016 17:40:24 GMT -6
Said it before, say it again--one of the few advantages of mixing in software is that you can actually EQ/compress/saturate/whatever and maintain no additional PEAK level..
"With the fader at unity" IS the level minus pan law. I don't even use Cubase's channel meters--I use the INPUT channel meters during tracking and a virtual VU to set trim levels at the start of a client mix.
Do the first part, not adding/subtracting peak level on each channel and you really don't have to worry--because it will end up a little higher than your calibration level, IME. For my -13 calibration, my final mixes tend to peak somewhere between -12 and -7dbfs on digital ballistics....but, a -18dbfs calibrated Slate VCC meter looks like a hot'ish analog mix--kick/snare usually going into the red and every now and again briefly pinning the needle for a Tom fill or something.
|
|
|
|
Post by Martin John Butler on Jan 18, 2016 19:13:56 GMT -6
Just had a meeting with my friend at Apple who's a Logic expert, and showed him this thread. Among other things, he said use the input meters on the Apollo when tracking, which sounds like Popmann's comment in a way.
|
|
?
