Old Motifator threads are available in the Archive.
Michael Trigoboff
Total Posts: 647
Joined 09-05-2011 status: Guru |
My problem is that the audio files (MP3, WAV) that I produce play at extremely low volume compared to every other audio file that I have (music, podcasts, etc). I’ve uploaded before normalization and after normalization images of what my audio looks like. Normalization is the only way I know of to fix that problem.
How would I do that? That’s exactly what I’m trying to find out. So far, all I know how to do is use Cubase to boost the velocity of my notes, or normalize the audio file. And boosting the velocity doesn’t actually give me that much of an effect on audio output levels. I keep thinking there must be some setting that, when I export audio from Cubase, would allow me to say how loud I want that audio to be. But I haven’t found it yet.
Exactly.
Originally I was working in pattern mode. When I got the pattern chain the way I wanted it, I made a song out of it. Then I saved it in an “all” file and imported the song into Cubase. Since then (a few months ago), I’ve been working in Cubase exclusively, except when I need to play and record something.
:-) In the early 1970s, I bought myself a Fender Rhodes. It turned out I couldn’t use it because the attack time was too slow for how I play unless I hit the keys way too hard. So I sold it to a guy who showed up wearing a wrist brace. I asked him what happened, and he said he hurt his hand playing the piano. Maybe he ended up being Metallica’s keyboard player…
I just want to be able to play my audio files without having to turn up the volume on my stereo to 11.
I don’t really want to use a compressor if I don’t have to. I’m just grasping at straws at the moment… Image Attachments
|
Michael Trigoboff
Total Posts: 647
Joined 09-05-2011 status: Guru |
Maximum of ~76. Average of ~55.
Isn’t there some sort of gain adjustment for the audio output anywhere in the Cubase/Motif XF system? |
5pinDIN
Total Posts: 11891
Joined 09-16-2010 status: Legend |
A possible drawback to such routing is that it might introduce more latency than you find acceptable. If so, once the velocity issue is resolved, you could go back to direct monitoring. |
5pinDIN
Total Posts: 11891
Joined 09-16-2010 status: Legend |
Quite a few dB can often be gained by getting the velocities higher. In addition, several of the Voices have a lot more “life” to them at higher velocities. The XF’s “audio output” going to Cubase is actually in the digital domain (it’s “digital audio” - before all the analog audio circuits), so there’s no adjustment from the XF.
For the heck of it, in case this was changed from the default 127…
EDIT: One more possibility, when using an FW16E…
Just to make sure that I understand your workflow…
|
5pinDIN
Total Posts: 11891
Joined 09-16-2010 status: Legend |
Here’s one more idea, to help determine what might make a difference…
|
philwoodmusic
Total Posts: 1055
Joined 07-01-2013 status: Guru |
When I said that there are ways of making softly played music loud, I did state that it depends on good recording levels and the quality of the recording itself. It’s really not worth talking about those until you’ve managed to get good input levels to Cubase in the first place. So, at this point, you really need to go through everything 5pinDIN has said. |
Michael Trigoboff
Total Posts: 647
Joined 09-05-2011 status: Guru |
It was set to 51. I had done that in the VST Editor to adjust the volume in my headphones.
I’m currently only using the L&R;output, and it’s set to +6dB.
Yes. Following your suggestions, I set the tone generator volume to 127, and set velocity curve to “fixed” and velocity value to 127. Here’s an image of the audio that resulted. Still pretty low volume. :-( Image Attachments
|
5pinDIN
Total Posts: 11891
Joined 09-16-2010 status: Legend |
According to the Reference Manual pages I previously mentioned, the L+R Gain and Assign L+R Gain settings on the XF affect only the output at the rear panel jacks, and it’s the FW Gain settings that determine the output levels via the IEEE1394 (FW) connector. I thought that you added an FW16E to your XF, and that’s how it’s interfaced to your computer/Cubase. If so, you might try increasing those FW levels to +6dB. (I apologize in advance if that doesn’t help, although I blame the manual writers. Â :-) Â )
You might find this helpful:
Â
I see a several dB increase in peak level. Of course, don’t leave the Velocity Curve set to “fixed” - return it to “soft” and store the change. Then work with the Input Transformer in Cubase. |
Michael Trigoboff
Total Posts: 647
Joined 09-05-2011 status: Guru |
I do have an FW16E now. I raised the FW levels to +6dB, but didn’t get an extremely significant effect.
I had seen that already. Interesting information, but I didn’t find anything that fixed my problem.
The problem with the Input Transformer is, as both you and Phil mentioned, there are Voices which will sound very different at the higher velocities. Back when I was in my amateur rock band in the 1970s, I had a Wurlitzer electric piano plugged into an amp. Sometimes I would turn the volume on the piano down and hit the keys pretty hard (which I could do back then). Doing that gave me a particular sound—after a while all the metal reeds in the piano would start resonating and as I played I’d be feeding energy into that cloud of resonance. It made for an interestingly spooky “ghost” dancing behind what I was playing. Other times, I would turn the volume on the piano way up and hit the keys very softly. This produced a very different sound, but courtesy of what I did with the volume knob, it came out at exactly the same loudness. I can do this kind of thing playing “live” on my XF7. It amazes me that this is apparently not possible when recording digital audio either directly from the XF7 or by using Cubase/FW. But I appreciate all the attention you and Phil have paid to this thread. |
5pinDIN
Total Posts: 11891
Joined 09-16-2010 status: Legend |
For your purposes, normalization of what was originally recorded might indeed be the best approach, although it might result in somewhat increased noise levels. It also may be worthwhile bringing your situation to the attention of Steinberg at their own forum.
Â
I think both of us want to see users get as much as they can from their Motifs. I always hope that Motifator members will find value in my posts, even if I don’t always have “the” answer. |
Michael Trigoboff
Total Posts: 647
Joined 09-05-2011 status: Guru |
Where would this noise level come from? I can see it could happen if I was using an analog device like a microphone. But my current song is all MIDI. Is there some noise that could occur in the exclusively digital realm? |
5pinDIN
Total Posts: 11891
Joined 09-16-2010 status: Legend |
Digital has advantages, but it’s not perfect. It has a limited dynamic range, depending on bit depth, and therefore a limited signal-to-noise ratio (S/N). For example, with 16-bit audio the maximum dynamic range is 96 db (16 bits times 6 dB per bit). That translates to a theoretical maximum 96 dB S/N - assuming the peak level attained is 0 dBFS, the “floor” will be -96dBFS.
The term “normalization” typically refers to peak normalization, in which the original peak is increased to something near but (hopefully) under 0 dBFS, and everything else (including the lowest level program material and noise) undergoes the same increase. And then there’s dithering…
If peak levels approach 0 dBFS without normalization, the result for 16-bit could be a recording with a S/N of about 90 dB. Most people would find that to be relatively “noiseless”. However, if pre-normalization levels are sufficiently low, bringing the peaks up can also bring the noise floor to an audible level. S/N of 50 dB would be noticeable by almost anyone, and even 70 dB S/N can be detected by many. I’m not saying that normalization should be avoided at all costs, just that it isn’t without consequences. |