Poor audio quality when recording (again)

[andrepires]

Hi.
For those who asked for my PC specs. If there is something missing, please let me know.

It's a laptop with:
Processor: Intel Core i7-8700K Coffee Lake 3.7 GHz, 6 Cores
RAM: 64 GB-DDR4 @2400 MHZ
The drive where the OS and DAWs are installed is a 1TB SSD M2 writing at 500mb/s
The drive where I save my audio clips is a 2TB disk at 5400 RPM
I don't know if it is relevant, but the video cards are: 2x SLi NVIDIA GeForce GTX 1070 (8GB GDDR5).
The OS is currently Windows 8.1. The computer came with Windows 10, and I got the same results as now. I had to downgrade because my Fantom G doesn't work fine with Windows 10 (no drivers).

I have the latest version of Ableton installed.
I will now record some samples using different DAWs and publish so you can listen to it.

[andrepires]

Hello again.
I recorded two short samples using a guitar patch from the Montage.
They are named after the DAW I used to record.
In both Cubase and Ableton I just created a black new project and started recording.
No audio effects applied.
The settings for recording are 16bits 44.1 on both Ableton and Cubase AI. The same settings in the Montage as per instructions in the manual.
I didn't use export functions of the DAWs. I just collected the raw wave files from the recording folders.
Here is the link to soundcloud:
https://soundcloud.com/andre-pi/sets/re ... vs-ableton

You can notice that the Ableton version sounds like it is in a different ambient/room/box, don't know, while the Cubase sample sounds exactly as it is in the Montage.
Maybe the right word to say here is not poor quality but different quality.
All I want to accomplish is having in Ableton the exact sound I have in the Montage with no alterations of any kind. Just as I get in any other application.

Thanks in advance for those who are trying to help.

[fishmonkey]

sorry, that's not a very good way to do a comparison. it's much better if you record exactly the same performance. if using a MIDI instrument, then this would be from pre-recorded MIDI notes.

apart from the fact that it's two different performances, the next obvious difference is that the samples are not level-matched—the Cubase sample is about 4dB louder than the Ableton sample.

you may already know this, but anything played louder will usually sound "better", because of the way our ears and brain work...

also, generally you are better off recording in 24-bit, not 16-bit.

[andrepires]

sorry, that's not a very good way to do a comparison. it's much better if you record exactly the same performance. if using a MIDI instrument, then this would be from pre-recorded MIDI notes.

apart from the fact that it's two different performances, the next obvious difference is that the samples are not level-matched—the Cubase sample is about 4dB louder than the Ableton sample.

you may already know this, but anything played louder will usually sound "better", because of the way our ears and brain work...

Alright.
Thanks for helping.
The input levels are the same. Could it be Ableton lowering down the levels?
If I equalize the levels wouldn't be altering the results from what the defaults are?
Anyways. I will do what you said and record exactly the same sound and equalize the levels so we can compare again.
Thanks.

[fishmonkey]

you gotta stop thinking in terms of defaults. if you want to do comparisons, either adjust the input levels so that the signal peaks at the same level, or when making comparisons, level-match the samples in whatever app you are using to play them.

level-matching is a really important concept in music production. it's extremely difficult to make a fair comparison between two similar pieces of music if they are played at different levels. for example, if you are trying to work out if a track was better before or after a change in compression or EQ, you need to level-match. if the compression or EQ makes the track louder, and you don't compensate for this by level-matching, then it will usually sound better. in reality though, it might actually sound worse if you reduce the level and do a fair listening comparison...

[andrepires]

you gotta stop thinking in terms of defaults. if you want to do comparisons, either adjust the input levels so that the signal peaks at the same level, or when making comparisons, level-match the samples in whatever app you are using to play them.

level-matching is a really important concept in music production. it's extremely difficult to make a fair comparison between two similar pieces of music if they are played at different levels. for example, if you are trying to work out if a track was better before or after a change in compression or EQ, you need to level-match. for example, if the compression or EQ makes the track louder, and you don't compensate for this by level-matching, then it will usually sound better. in reality though, it might actually sound worse if you reduce the level and do a fair listening comparison...

Alright. Thanks again.
I'm sure the input levels are the same, but I will record again following your instructions and adjusting the playback levels.

[andrepires]

Hi again.
Following instructions now I recorded again using the same MIDI track so there is no differences in what is being played.

My steps in Ableton:

- Checked the bit rate and frequency settings: 16Bits 44.1
- Checked the input level: -10.3
- Checked the output level: -10.3
- Created a new Audio Track and started recording.
- Switched to the output folder and grabbed the WAV file.

My steps in Cubase AI:
- Exactly the same above plus:
- Imported the sample recorded in Ableton
- Checked if the levels still match and they did
- Exported both samples to MP3 files using High Quality Mode, bit rate 320 (max) 44.1kHz, so I could upload fast to soundcloud. If I just playback the WAV files the result is the same.

Other checks:
- The tracks are being recorded using ASIO Steinberg driver that comes with Yamaha Montage in both Ableton and Cubase AI.
- The connection between the computer and the synth is USB L/R.
URL's to the samples:
https://soundcloud.com/andre-pi/sets/te ... vs-ableton

Thanks for helping

[Da hand]

Apart from the levels, the Ableton recording is Mono and the Cubase is Stereo - this is why you are hearing the big difference. You will need to check your input setting in Ableton to make sure you are recording in stereo.

[andrepires]

Apart from the levels, the Ableton recording is Mono and the Cubase is Stereo - this is why you are hearing the big difference. You will need to check your input setting in Ableton to make sure you are recording in stereo.

Da Hand! This is awesome!
You cannot imagine how happy I am man...
I cannot believe it was just that and I missed it all the time! I feel happy and stupid at the same time
Man you saved it!

@fishmonkey
Thank you too man! You also helped a lot with your instructions.

This is awesome!

[Da hand]

Glad to help andrepires! And to bring some happiness as well haha

[fishmonkey]

Apart from the levels, the Ableton recording is Mono and the Cubase is Stereo - this is why you are hearing the big difference. You will need to check your input setting in Ableton to make sure you are recording in stereo.

hey, good catch!

[Da hand]

Apart from the levels, the Ableton recording is Mono and the Cubase is Stereo - this is why you are hearing the big difference. You will need to check your input setting in Ableton to make sure you are recording in stereo.

hey, good catch!

Thank you! But props to you for all the info and instructions beforehand which led to the comparison files - without which it would have been hard to make this deduction

[lapieuvre]

Glad you found the problem.

But know that there is a difference of audio quality between DAW's. Not huge but it's there.

My suggestion is when you export a mix from Ableton, export in 32 bits. I did some tests and there is a difference in the sound.

The test I made was this one:

Take a song separated in around 100 stems, import it in Logic and import it in Live.

The files were 48Khz, 24 bits, I exported a stereo mix in 48/24.

In logic the sound was much clearer, crispier in the high end, so I tried to export the stems in 48Khz/32 bits in Live, and then the sound quality was equal to Logic.

Another interesting thing is that with the headphone output of a MBP, I could not hear a difference between the 2 48/24 files, it's when I used an Antelope sound card that I could hear the difference.

[mikb]

In logic the sound was much clearer, crispier in the high end, so I tried to export the stems in 48Khz/32 bits in Live, and then the sound quality was equal to Logic.

[fishmonkey]

100 stems? what kind of music do you work on?