Hi guys
Has anyone else had this problem? When I play AR through Ableton it has lovely sound quality. Then I record it, and play it back. And when I play it back it is tinny and horrrible. I’ve checked and there don’t seem to be any reasons for this, no effects on the audio or master tracks, etc. Any ideas? Thanks alot!
I’ve found that the AR is usually slightly behind in time, even if you have it synced via USB. This results in Ableton shifting the transients around to get it to sync with abletons clock. You can turn this off by disabling warping, but if you do that you’ll have to set 1.1 manually.
Try turning off warping first. Beyond that, what are you recording with? Gain staging etc?
Unfortunately you’ll find that digital recordings of analog synths and drum machines will always be missing a little something. But the difference shouldn’t be that drastic.
What sample and bit rates are you using in Live? Did you change either of those inadvertently? Does your system have sample or bit rate settings that would supersede your settings in Live? Do you have the same problem when you make digital recordings of other analog devices?
Recording in Live should not sound “tinny and horrible”.
Try recording directly in to another app such as an audio editor. Ocenaudio and Audacity are both free.
You could also try Windows sound recorder, or possibly Quicktime on Mac.
The answer will help pin down if it’s caused by a setting in Live or something else.
@barmyarmy: If it sounds really tinny you’ve probably got some monitoring problems: Check if the incoming sound gets accidentally recorded alongside the monitor signal thus resulting in troubles with the phase of the signal making it sound “tinny”.
@sempervirent: digital technology is not the problem, the reason why recordings of your analog gear may sound different is, because there’s gain staging and all kinds of other errors one can make resulting in quality loss. AD/DA converters nowadays work perfectly. Signal loss due to digital conversion is a big myth and simply BS.
More on this from the guy with the weird beard:
Could you give us a recorded audio example?
If you’re monitoring via Ableton (signal is going through the engine), there shouldn’t be a difference if it’s a recorded signal.
Agreed. I could probably record a loop of audio passing out of and back into the analog inputs of my soundcard 20-30 times before hearing any sort of signal degradation and serious noise floor build up assuming I had my gains set correctly, and even then any artifacts would probably only be heard in really busy parts with lots of HF content. Might be a different story if you are recording via the integrated sound card on an 8 year old laptop…
Sounds like a phasing issue.
Agreed. I could probably record a loop of audio passing out of and back into the analog inputs of my soundcard 20-30 times before hearing any sort of signal degradation and serious noise floor build up assuming I had my gains set correctly, and even then any artifacts would probably only be heard in really busy parts with lots of HF content. Might be a different story if you are recording via the integrated sound card on an 8 year old laptop…[/quote]
I used to think the same thing until I visited the recording studio that does most of the SACD remasters in the world. They have a proprietary system that records at 2.8MhZ using a technology that is different from PCM (which is what the rest of us are using). After a few hours of listening tests, comparing these recordings to 192kHz PCM recordings of the same live sessions or the same master tapes, it was obvious that PCM is a fairly rough approximation of what we actually hear. There’s no myth or BS to that statement, just basic physics.
Mmm, I think that’s more cognitive bias actually. The physics argument doesn’t hold wter. You’re sampling at 2.8Mhz, but it’s 1-bit sampling so you’re recording a pulse train instead of pulse code modulation. If you’re normally recording at 24 bit to measure the amplitude at a given moment, that information has to be made up somewhere, hence the need for super-fast sampling rates. 24 bits by 96Khz is equivalent to 1 bit sampling at 2.3 Mhz. Extensive blind listening tests in a population of professional musicians don’t find any statistically significant difference in quality between the two formats: http://old.hfm-detmold.de/eti/projekte/diplomarbeiten/dsdvspcm/aes_paper_6086.pdf