Hi,
I'm learning a lot while using Reaper but this piece of soft is so rich that mistakes are unavoidable by a newbie that I am. And I'm almost ashamed to say that I though the volume level slider of the track, controls the recording level!
And I'm here to ask you if there is any way to save/fix my guitar and vocal recorded tracks without having to record them again.
I somehow managed to record all my 4 tracks drum, bass, guitar, vocal without any of them going into the red zone while keeping them often in the orange zone. I think I just adjusted (wrongly) the tracks' levels to be in the green and orange zones. So, I suppose that I should have done those adjustments with the input level coming out of my Line6 UX2 sound USB card(used for both guitar and mic input).
After recording, I played the whole 4 tracks in Reaper and saw some clipping on different parts of the guitar and vocal tracks. So I lowered the level via the take's volume level to fix them.
Question 1: can this be considered as a fix ? I ask this because I didn't change anything in the real "recorded" sound, but just lowered the level afterward so that when playing back (or rendering), the level never goes to the red zone. That is, clipping actually occurred, I just tried to fix it afterward.
Question 2: My drum and bass tracks are MIDI virtual instruments (VSTi). When opening their FX windows, I can control the sound level of each virtual drum and bass within those windows. Are those levels considered as "input" levels that I should control to not clip ?
Question 3: It is easy for virtual MIDI instruments to find the best volume levels to not clip and in the same time to get them somewhere between green and orange zone (and leave them always in those positions). This is harder with guitar (because I can change the volume on guitar accidentally) and vocals (because I haven't any volume level in my throat!). Anyway, this makes me think that each track should be recorded to not clip at recording time without any consideration on their relative levels between each other at this step. Then, at mixing time (just before rendering) I should use the tracks' sliders to control the level of each track relative to the others. This time, I should watch the master control to avoid clipping at this step (when mixing/rendering) by too much rising a track'slider level. Is this the correct way to do things ?
If a recorded wave is making the master track go over 0db then it is clipping and should be redone. This would be with the master fader set to 0db.
If you happen to have a great performance that has portions that have clipped according to the meters, but you don't hear any distortion or general signal degradation, then you might want to use that take for the sake of a great performance. Otherwise, do it right.
The midi tracks are not a problem. You can adjust then in the Vst or with the track fader. I would tend to keep the level under control in the vst if it is going over 0db on the track meter with the track fader at 0db.
The main reason for that is because some effects that you might place after the vst could be sensitive to input levels and could overload if the vst is putting out a too high level.
When you record waves with vocal, guitar, etc then a good level is around -18db RMS. Keep peaks under 0db and preferably under about -6db.
It sounds like that would be too low, but 0db RMS in the digital world is like 0db on a tape deck's meters.
Thank you PAPT for your answer (and also on my other post on compression - I'll answer to that about channel splitting).
Quote:
Originally Posted by PAPT
If a recorded wave is making the master track go over 0db then it is clipping and should be redone. This would be with the master fader set to 0db.
Not sure to understand. Should I always set the master fader to 0db, then watch it to see if it's clipping, in which case I must redo a take after lowering the volume ?
Quote:
Originally Posted by PAPT
If you happen to have a great performance that has portions that have clipped according to the meters, but you don't hear any distortion or general signal degradation, then you might want to use that take for the sake of a great performance. Otherwise, do it right.
This is exactly the case. I can see some portions going to red zone but overall I hear no distortion. However you didn't answer to my 1st question regarding fixing with the take's volume level. I don't think this can fix anything but it's better than nothing because I'll not clip twice (again at the mixing time). No ?
Quote:
Originally Posted by PAPT
The midi tracks are not a problem. You can adjust then in the Vst or with the track fader. I would tend to keep the level under control in the vst if it is going over 0db on the track meter with the track fader at 0db.
The main reason for that is because some effects that you might place after the vst could be sensitive to input levels and could overload if the vst is putting out a too high level.
I understand it's better to control it in the VSTi. But why this time it is also possible to control the recorded volume with the track fader while it is not possible with guitar and vocal for which the recorded volume level can only be controlled within the device itself ?
Quote:
Originally Posted by PAPT
When you record waves with vocal, guitar, etc then a good level is around -18db RMS. Keep peaks under 0db and preferably under about -6db.
It sounds like that would be too low, but 0db RMS in the digital world is like 0db on a tape deck's meters.
Thank you very much for this numerical advice. Do you have some similar advice/recipe for my other post regarding compression setting ?
THe difference between an audio recorded as a wave and the audio from a vst is that the vst has not been recorded as a wave yet. So, if you turn down the volume it will no longer clip.
A recording in wave form that is clipping contains the clipping in the recording. If you turn down the volume of the wave you will have a clipped sound playing back at lower volume.
The problem with giving numbers for compression is that the setting depend entirely on the source material.
The threshold you use depends on the volume of the signal coming into the compressor.
If one track has peaks at -6db and you set the threshold to -8db the comp will start to work whenever the level exceeds -8db so it will work on the peaks that go to -6.
If you use -8db for the threshold for a track that peaks at -10db the comp will never start working.
You need to do some reading about compression or you will never understand.
Most people on this forum are glad to help people when they have sticking points, but most of us have no interest in running a school and teaching all of the basics.
THe difference between an audio recorded as a wave and the audio from a vst is that the vst has not been recorded as a wave yet. So, if you turn down the volume it will no longer clip.
A recording in wave form that is clipping contains the clipping in the recording. If you turn down the volume of the wave you will have a clipped sound playing back at lower volume.
The problem with giving numbers for compression is that the setting depend entirely on the source material.
The threshold you use depends on the volume of the signal coming into the compressor.
If one track has peaks at -6db and you set the threshold to -8db the comp will start to work whenever the level exceeds -8db so it will work on the peaks that go to -6.
If you use -8db for the threshold for a track that peaks at -10db the comp will never start working.
You need to do some reading about compression or you will never understand.
Most people on this forum are glad to help people when they have sticking points, but most of us have no interest in running a school and teaching all of the basics.
If Reaper would have somebcustomer service we would jot ne bithering you geeks with the simple +×÷$
If you clip an audio track then it's clipped. Nothing you do afterwards can "unclip" it.
Well, nothing in Reaper at least. If you can't re-record you can see if a tool like Izotope RX's de-clipper could redraw the waveforms based on what it thinks they should have been, as a last resort. See https://www.izotope.com/en/learn/pro...ed-vocals.html
I mean, if you zoom in far enough, you could probably use the take volume envelope to "draw" on the waveform, too. You still have to kind of guess what it's supposed to look like, or just try to draw something that looks (and hopefully sounds) smoother.
Sometimes, depending on the material, you can just EQ out some of the higher frequencies where the worst part of the clipping distortion lives.
But it only actually matters if it matters. If you play the mix and don't hear anything that bothers you, then there's not really anything to fix. A lot of times, if the signal is already distorted, a little extra clipping doesn't hurt so much. Similarly, if you're just going to distort it some more - whether that means amp sim or console/tape emulation - a couple of clipped peaks will sometimes kind of be masked and maybe even rounded off a little bit.
Honestly, that's one of the best ways to "fix" clipped signals: distort them more so it sounds like it was on purpose.
Hi, I've been recording for a few years but I feel like a beginner especially after this issue. I recently sent my files to a friend for mixing. I just sent him the clean D.I. signal without any effects. He said the files had clipping yet I remember when I recorded it there was no bad sound or anything in the red. Could it be I raised the volume on some recorded items(seems like the logical) or maybe there are some other settings causing this. O.k. Thanks for any feedback.
I really enjoy learning.
Honestly, that's one of the best ways to "fix" clipped signals: distort them more so it sounds like it was on purpose.
I've done this a few times even with vocals. I had this one where the client recorded on one of the digital versions of the Porta Studio (post cassette, pre DAW - for him anyway). The vocal had some fully blown out parts. Not just a few clicks or crunches. Sections of full on phone recording sound! I went full radiophonic and it ended up being the perfect thing for the song!
Fun stories aside...
Clipped just a little?
Can you hear the clip?
If yes, is it over the top terrible (phone recording territory)?
If no, move on. (Use caution in the future of course!)
If you really did make some distortion sounds but it's also a really important recording that can't be redone easily, iZotopeRX has a clip repair function that works better than something like this should ever be able to work. That has its limits though too. Crackling and a few crunches from that level of clipping can often be transparently repaired. Full on phone recording sound cannot.
With most converters, 0 is 0, and you’re not going to get anything bigger into Reaper no matter what format you record. If the converter is clipping, then it’s clipping, and recording more bits is literally just wasting bits.
It IS completely possible that you changed the gain in Reaper before rendering stems. There’s about 700 ways to do that. This is where floating point files can help. Any time you render(freeze, bounce, whatever) anything that’s not meant to be your final destination file, go to a floating point format. Then you know that whatever you did the gain can’t cause actual clipping in the file, nor can it really compromise the noise floor. Unless you do something seriously absurd, your friend can just turn them up or down and move on.
The hardware input and AD converter is really truly a hard limit @H3o. If you clip the input, that data is sliced off and left on the cutting room floor as it were.
You may have read about products like some of the portable recorders used in broadcast/field work that use two sets of preamps and AD converters and stitch the result together in a 32 bit file. This is a clever way to work with hardware with limited range to increase it.
IF you are applying input FX (or recording output of FX in general) for whatever reason, THEN yes I agree you might get some benefit going to floating point. There is no good reason to add an input FX just so you can smash your interface, though. It’s fixed point coming in, and if you’re not doing anything else to it, you might as well just record the fixed point and worry about the rest down the road.
Yeah... I think it's time to stop feeding the troll.
Yeah I’ve been down this road before, and really want trying to start the same argument. Just trying to make sure the OP and other folks asking for help didn’t get sidetracked.
Hi,
I'm learning a lot while using Reaper but this piece of soft is so rich that mistakes are unavoidable by a newbie that I am. And I'm almost ashamed to say that I though the volume level slider of the track, controls the recording level!
And I'm here to ask you if there is any way to save/fix my guitar and vocal recorded tracks without having to record them again.
I somehow managed to record all my 4 tracks drum, bass, guitar, vocal without any of them going into the red zone while keeping them often in the orange zone. I think I just adjusted (wrongly) the tracks' levels to be in the green and orange zones. So, I suppose that I should have done those adjustments with the input level coming out of my Line6 UX2 sound USB card(used for both guitar and mic input).
After recording, I played the whole 4 tracks in Reaper and saw some clipping on different parts of the guitar and vocal tracks. So I lowered the level via the take's volume level to fix them.
Question 1: can this be considered as a fix ? I ask this because I didn't change anything in the real "recorded" sound, but just lowered the level afterward so that when playing back (or rendering), the level never goes to the red zone. That is, clipping actually occurred, I just tried to fix it afterward.
Question 2: My drum and bass tracks are MIDI virtual instruments (VSTi). When opening their FX windows, I can control the sound level of each virtual drum and bass within those windows. Are those levels considered as "input" levels that I should control to not clip ?
Question 3: It is easy for virtual MIDI instruments to find the best volume levels to not clip and in the same time to get them somewhere between green and orange zone (and leave them always in those positions). This is harder with guitar (because I can change the volume on guitar accidentally) and vocals (because I haven't any volume level in my throat!). Anyway, this makes me think that each track should be recorded to not clip at recording time without any consideration on their relative levels between each other at this step. Then, at mixing time (just before rendering) I should use the tracks' sliders to control the level of each track relative to the others. This time, I should watch the master control to avoid clipping at this step (when mixing/rendering) by too much rising a track'slider level. Is this the correct way to do things ?
thanks
nalooti
you need to 'set up' b4 u can sta recording, and START W yr ROOM
I could 'fix it' ($350)
so its prob quickeer to do the job the right way the first time