Bitwig 6

Official support for: bitwig.com
Post Reply New Topic
RELATED
PRODUCTS

Post

tumface wrote: Wed Sep 17, 2025 4:19 am
limitlesssss wrote: Tue Sep 16, 2025 8:32 pm Can you nudge/edit items in Bitwig Studio 6 at sample-level accuracy?
Yes. Bitwig is subsample-accurate.

Yes, but there's no command to do these, and Bitwig always shows an interpolated waveform without showing the sample points, so it's not easy. Many people have requested better waveform display features and options, but they haven't done it.
stamp wrote: Wed Sep 17, 2025 1:00 am Samples? May I ask what you need all this precision for?
Why do you need to know what they need it for? This is a feature other DAWs have, and there's no need to play gatekeeper. (Sorry if I'm misunderstanding why you're asking.)
Hey tumface, when you say Bitwig always shows an interpolated waveform, does that mean that it (at least visually*) may not necessarily snap to the closest sample position when nudging a waveform? Meaning that even at the highest possible zoom level, the nudging action appears to be infinitely smooth, without any sudden jump from the current sample position to the next?

If that's true, then it makes it very difficult to judge if you even moved a waveform one sample or not. Unless you use the Time Shift tool. This was a problem I had in the past with REAPER, until Justin (the developer) made me aware of the [Snap to project sample rate] option in Snap/Grid Settings. Before that, I would zoom-in completely and move a waveform and it would appear that it was moving in-between samples. But as soon as I activated the [Snap to project sample rate] the waveforms snapped to the next/previous samples when nudged.

* I'm sure in the background it HAS TO snap to the closest sample position since there's no in-between samples in digital audio, that's the whole point of digital audio. It's not infinite.
Last edited by limitlesssss on Tue Sep 23, 2025 11:51 am, edited 5 times in total.

Post

Duplicate. Apologies.

Post

cfernn wrote: Mon Sep 22, 2025 5:04 pm For those who have been using 6 for a couple weeks now...

How we feeling about it compared to 5? Is the automation a lot better? The overall experience better?

I'm surprised to say that I'm not so sure if I'm convinced (yet) ...but I haven't used 6 myself. I just keep watching videos and reading the forums. They did implement a lot of arranger features that I've been wanting, but I can't shake the feeling that it somehow looks worse than before. Maybe it's just an aversion to change on my part.
To me it’s like the love child of Logic and Ableton, drenched in FL vibe, I also think it looks uglier or less pro, but the improvements in the piano roll, automation, etc. make up for it ... Overall it’s positive, and the dark-background arranger was definitely needed.
"Where we're workarounding, we don't NEED features." - powermat

Post

limitlesssss wrote: Tue Sep 23, 2025 11:45 am Hey tumface, when you say Bitwig always shows an interpolated waveform, does that mean that it (at least visually*) may not necessarily snap to the closest sample position when nudging a waveform? Meaning that even at the highest possible zoom level, the nudging action appears to be infinitely smooth, without any sudden jump from the current sample position to the next?
Correct. Bitwig will allow placing waveform with continuous (probably FP64?) precision, but will play back with the timing quantized to single sample precision. It won't visually show you where the waveform will be snapped to during playback. Also, the waveform drawing style that Bitwig has chosen to use (and they don't seem to want to change) makes it difficult to tell where the sample points are positioned in the waveform.
limitlesssss wrote: Tue Sep 23, 2025 11:45 am If that's true, then it makes it very difficult to judge if you even moved a waveform one sample or not. Unless you use the Time Shift tool. This was a problem I had in the past with REAPER, until Justin (the developer) made me aware of the [Snap to project sample rate] option in Snap/Grid Settings. Before that, I would zoom-in completely and move a waveform and it would appear that it was moving in-between samples. But as soon as I activated the [Snap to project sample rate] the waveforms snapped to the next/previous samples when nudged.
The Time Shift device is probably the best choice for one-sample accuracy in Bitwig. REAPER is more sophisticated than Bitwig when it comes to audio editing. If your projects consist mostly or entirely of recording and audio editing, REAPER is a superior choice to Bitwig. I use both professionally. I would not ever choose to use Bitwig for comping, editing, mixing, and mastering recorded instruments from a band or performers, especially not multi-mic drum or cab setups. I don't imagine Bitwig will attempt to become this type of tool any time in the next 10 years. Different focus.
limitlesssss wrote: Tue Sep 23, 2025 11:45 am * I'm sure in the background it HAS TO snap to the closest sample position since there's no in-between samples in digital audio, that's the whole point of digital audio. It's not infinite.
This is not true. While the digital audio is stored and represented in memory as evenly-spaced numeric values, the Nyquist–Shannon sampling theorem shows that it it will be constructed into a continuous waveform at sampling rate/2. The discrete samples become a continuous waveform when converted from digital to analog. [1] There is no per-sample time quantization in the continuous signal, and, using the right math and code, the continuous signal can be manipulated digitally without requiring a conversion to an electrical analog signal.

So, you can adjust waveform positions by less than a whole sample when working digitally, but it requires more processing power than simply offsetting by whole samples.

If you are trying to perfectly phase align two microphones, whole samples is not precise enough. Moving audio clips by 1 sample in the waveform views of a DAW will not give you enough precision to reach the best phase alignment.

There are dedicated tools that can do this for you. Some are automatic. Some can rotate the entire phase of the waveform instead of only offsetting the entire waveform in time. Some are manual. For example, here is a manual one, in plugin form, that allows you to adjust the time position by less than 1 sample: https://www.eventideaudio.com/plug-ins/ ... ime-align/ It doesn't have any features to rotate the phase. Voxengo PHA-979 allows you to rotate the phase manually.

Personally, if I was doing this more than once per month -- which I'm not -- I would just buy Sound Radix Auto-Align, as it gives better results than trying to do it by hand, and is much faster. Doing this by hand, repeatedly, seems like tedious make-work.

[1]: Assuming everything is working correctly. You can find old 50 cent DACs in 1980s hardware with various terrible DAC problems that introduce aliasing. This is not a problem on any hardware in the last 30 years that costs more than $1.

edit: I originally wrote "2/sampling rate" instead of "sampling rate/2" which is a very funny mistake
Last edited by tumface on Tue Sep 23, 2025 6:20 pm, edited 1 time in total.

Post

tumface wrote: Tue Sep 23, 2025 1:12 pm
limitlesssss wrote: Tue Sep 23, 2025 11:45 am Hey tumface, when you say Bitwig always shows an interpolated waveform, does that mean that it (at least visually*) may not necessarily snap to the closest sample position when nudging a waveform? Meaning that even at the highest possible zoom level, the nudging action appears to be infinitely smooth, without any sudden jump from the current sample position to the next?
Correct. Bitwig will allow placing waveform with continuous (probably FP64?) precision, but will play back with the timing quantized to single sample precision. It won't visually show you where the waveform will be snapped to during playback. Also, the waveform drawing style that Bitwig has chosen to use (and they don't seem to want to change) makes it difficult to tell where the sample points are positioned in the waveform.
limitlesssss wrote: Tue Sep 23, 2025 11:45 am If that's true, then it makes it very difficult to judge if you even moved a waveform one sample or not. Unless you use the Time Shift tool. This was a problem I had in the past with REAPER, until Justin (the developer) made me aware of the [Snap to project sample rate] option in Snap/Grid Settings. Before that, I would zoom-in completely and move a waveform and it would appear that it was moving in-between samples. But as soon as I activated the [Snap to project sample rate] the waveforms snapped to the next/previous samples when nudged.
The Time Shift device is probably the best choice for one-sample accuracy in Bitwig. REAPER is more sophisticated than Bitwig when it comes to audio editing. If your projects consist mostly or entirely of recording and audio editing, REAPER is a superior choice to Bitwig. I use both professionally. I would not ever choose to use Bitwig for comping, editing, mixing, and mastering recorded instruments from a band or performers, especially not multi-mic drum or cab setups. I don't imagine Bitwig will attempt to become this type of tool any time in the next 10 years. Different focus.
limitlesssss wrote: Tue Sep 23, 2025 11:45 am * I'm sure in the background it HAS TO snap to the closest sample position since there's no in-between samples in digital audio, that's the whole point of digital audio. It's not infinite.
This is not true. While the digital audio is stored and represented in memory as evenly-spaced numeric values, the Nyquist–Shannon sampling theorem shows that it it will be constructed into a continuous waveform at 2/sampling rate. The discrete samples become a continuous waveform when converted from digital to analog. [1] There is no per-sample time quantization in the continuous signal, and, using the right math and code, the continuous signal can be manipulated digitally without requiring a conversion to an electrical analog signal.

So, you can adjust waveform positions by less than a whole sample when working digitally, but it requires more processing power than simply offsetting by whole samples.

If you are trying to perfectly phase align two microphones, whole samples is not precise enough. Moving audio clips by 1 sample in the waveform views of a DAW will not give you enough precision to reach the best phase alignment.

There are dedicated tools that can do this for you. Some are automatic. Some can rotate the entire phase of the waveform instead of only offsetting the entire waveform in time. Some are manual. For example, here is a manual one, in plugin form, that allows you to adjust the time position by less than 1 sample: https://www.eventideaudio.com/plug-ins/ ... ime-align/ It doesn't have any features to rotate the phase. Voxengo PHA-979 allows you to rotate the phase manually.

Personally, if I was doing this more than once per month -- which I'm not -- I would just buy Sound Radix Auto-Align, as it gives better results than trying to do it by hand, and is much faster. Doing this by hand, repeatedly, seems like tedious make-work.

[1]: Assuming everything is working correctly. You can find old 50 cent DACs in 1980s hardware with various terrible DAC problems that introduce aliasing. This is not a problem on any hardware in the last 30 years that costs more than $1.
Thank you for the Bitwig education tumface. I guess Bitwig cannot be my sole DAW then. I rely heavily on moving stuff sample by sample all the time.

You're right when it comes to analog domain, but what I was talking about had to do with the digital domain only. And of course Eventide Audio Precision Time Align will let you go sub-sample, but I'm sure it does it by oversampling the signal first. So it's still bound by the sampling theorem. It also has to use anti-aliasing filters to do so. I personally stay away from these kinds of tools whenever I can.

But I admit, I never gave Precision Time Align a fair try. Because of my aversion to the use of oversampling whenever unnecessary.

Thank you again tumface.

Post

As usual tumface prioritizes sounding smart.

Ultimately, in the digital realm, audio files are a discrete list of samples, and you need to provide a list of samples to your DAW during playback.

It’s true that you can shift audio by an arbitrary value, and playback the audio at an arbitrary speed. However, you still need to provide the DAW a discrete list of samples, and if those samples don’t line up 1-to-1 then you’re going to need to resample somehow.

So I would argue that shifting an audio waveform sample-by-sample is a different operation than shifting an audio waveform continuously such that samples don’t line up anymore.

Post

coroknight wrote: Tue Sep 23, 2025 6:00 pm As usual tumface prioritizes sounding smart.

Ultimately, in the digital realm, audio files are a discrete list of samples, and you need to provide a list of samples to your DAW during playback.

It’s true that you can shift audio by an arbitrary value, and playback the audio at an arbitrary speed. However, you still need to provide the DAW a discrete list of samples, and if those samples don’t line up 1-to-1 then you’re going to need to resample somehow.

So I would argue that shifting an audio waveform sample-by-sample is a different operation than shifting an audio waveform continuously such that samples don’t line up anymore.
I agree with you. That's why I said (implied) that a sample is the smallest unit of digital audio and there's no such thing as in-between samples in digital. Because in order to be able to for example place the playhead exactly in the middle of a sample, you need to first upsample the signal at 2x, an operation that is NOT lossless and requires anti-aliasing filters.

Post

limitlesssss wrote: Tue Sep 23, 2025 6:19 pm Because in order to be able to for example place the playhead exactly in the middle of a sample, you need to first upsample the signal at 2x, an operation that is NOT lossless and requires anti-aliasing filters.
It doesn't require oversampling. If it did, adjusting by 1/100th of a sample would require using 100x oversampling. You use a sinc window to shift the samples as a fractional delay filter. It's slightly more lossy than adjusting gain, but not as much as changing sample rate. It creates a new sampling pattern, which is a form of resampling, but it's not changing the sample rate. You shouldn't be afraid of it, if it's what you need. If the phase alignment is so important for whatever it is you're doing, the impact of getting it misaligned will be hugely more than a FIR filter or whatever to shift by subsamples.

Post

tumface wrote: Tue Sep 23, 2025 6:25 pm
limitlesssss wrote: Tue Sep 23, 2025 6:19 pm Because in order to be able to for example place the playhead exactly in the middle of a sample, you need to first upsample the signal at 2x, an operation that is NOT lossless and requires anti-aliasing filters.
It doesn't require oversampling. If it did, adjusting by 1/100th of a sample would require using 100x oversampling. You use a sinc window to shift the samples as a fractional delay filter. It's slightly more lossy than adjusting gain, but not as much as changing sample rate. It creates a new sampling pattern, which is a form of resampling, but it's not changing the sample rate. You shouldn't be afraid of it, if it's what you need.
This is very interesting tumface. I've heard of sinc function before. Can you point me to a reference so I can educate myself more? Thank you again.

Post

limitlesssss wrote: Tue Sep 23, 2025 6:28 pm This is very interesting tumface. I've heard of sinc function before. Can you point me to a reference so I can educate myself more? Thank you again.
I looked at a few really quick and this one looks good: https://www.mathworks.com/help/dsphdl/u ... ilter.html I remember seeing a different (cheaper?) one in a DSP textbook a long time ago, but I think it took some shortcuts in the name of performance. I don't even remember which book it was.

Post

coroknight wrote: Tue Sep 23, 2025 6:00 pm As usual tumface prioritizes sounding smart.
You said you blocked me. Please stop engaging with me. Thanks.

Post

tumface wrote: Tue Sep 23, 2025 6:46 pm
coroknight wrote: Tue Sep 23, 2025 6:00 pm As usual tumface prioritizes sounding smart.
You said you blocked me. Please stop engaging with me. Thanks.
Nah. Learn to read and stop being such a nitpicky know-it-all. You should have been able to interpret what they were talking about via context clues and reasonable assumptions.

Shifting a waveform sample-by-sample is different from resampling.

Post

coroknight wrote: Tue Sep 23, 2025 7:27 pm Nah. Learn to read and stop being such a nitpicky know-it-all. You should have been able to interpret what they were talking about via context clues and reasonable assumptions.

Shifting a waveform sample-by-sample is different from resampling.
Please stop harassing me on these forums. You repeatedly insulted me in previous posts, even though I have never insulted you. You said you blocked me, and have now gone on to continue finding posts of mine to reply to (with incorrect information) while adding personal insults.

firefox_6bW0ZfUhCU.png

I do not want to interact with you ever again. Please stop replying to me, quoting me, and insulting me. Thank you.
You do not have the required permissions to view the files attached to this post.

Post

tumface wrote: Tue Sep 23, 2025 7:33 pm (with incorrect information)
Yet you still can't help yourself lol. What exactly did I say that's wrong?

This is exactly what happened last time, you continue to misunderstand and misinterpret what people say, spout off some technical knowledge to make yourself feel smart and then complain when people push back.

Post

Please stop replying to me to trigger notifications. I will not be engaging with you. Do not engage or interact with me. Leave me alone. Thank you.

Post Reply

Return to “Bitwig”