|
Post by Charlene Adzima on Jan 20, 2021 15:29:54 GMT
Hi all,
I've got what I think is a playback resolution issue and I'm not sure how to fix it. I made a video and my friend recorded his part with mine, but upon playback with my video, he starts much sooner and plays in a higher key than I do. What setting can I check on my end and on my collaborator's end?
Thanks!
Charlene
|
|
|
Post by lizrad on Jan 20, 2021 18:56:57 GMT
Hello Charlene, Here are some troubleshooting questions: 1) Regarding the higher pitch your collaborator recorded: Find the individual video file in your Drive. Playback that individual file. Does the pitch in the individual file match the pitch in the in-app playback? 2) Does your collaborator's recording play at a faster tempo than your recording, or does he just start sooner, but the actual tempo is not any faster than your own? If this is a syncing issue, here is a tutorial how to use the nudge feature to adjust. If nudge's 350 millisecond correction-window is not sufficient, first confirm with your collaborator that wired headphones where used, set-up calibration was performed in a reasonably silent room, and the latency measurement applied at set-up is within 50 milliseconds (0.050) of his average latency measurements. For more information, you can look at the troubleshooting page of the website "Why am I not with the metronome and/or other recordings when I listen back to my recording?". Please note that 'nudge' will correct syncing issues (starting recordings at the correct time) but will not correct tempo issues (recordings going at different speeds).
|
|
|
Post by Charlene Adzima on Jan 20, 2021 21:43:00 GMT
Hi,
Thanks for the help. I did some testing and it turns out my collaborator's recording is actually going slower than my own. I was using the metronome at 100 and confirmed with the tap tempo feature on my metronome that I was playing at 100 bpm. I used the tap tempo feature again to see what my friend was playing at and his was at exactly 92 BPM. I also did a more thorough check of the pitch differential and although I'm playing in D, he is playing in ~C, so a bit lower than mine. I also noted that I started at 12:1 and he starts at 9:1. I end at 44:1 and he ends at 45:1 (2 beats per measure, for reference). We double checked our tuning over the phone and he was in tune with me, but the recording was not. I downloaded both parts and they both came out the same as they did in CL. The bit rates were both 48 kHz, if that means anything. Could my friend's have been recorded at a different bitrate?
Let me know if anything jumps out at you. Otherwise, I'm just going to have him try again so maybe we'll hit it or I'll have more data on what's going on. Also for reference, I've collaborated successfully with other folks so I think it has something to do with my current collaborators set up.
thanks in advance,
Charlene
|
|
|
Post by lunchbox on Jan 20, 2021 23:09:14 GMT
Could you clarify what you mean by "I downloaded both parts". Do you mean you downloaded directly from Google Drive, or did you use one of the export features to download them?
|
|
|
Post by Charlene Adzima on Jan 20, 2021 23:28:02 GMT
I downloaded both parts from Drive separately.
|
|
|
Post by lunchbox on Jan 21, 2021 0:55:48 GMT
Thank you for the information. I agree with your suspicion that your friend's part could have been recorded at a different bitrate; the ratio of the differing recording lengths definitely seems to point to something like a 44.1kHz vs 48kHz mismatch. However, Cyborg Llama currently doesn't process recordings before they are stored on user's Google Drives, so the bitrate information within the files, in user's drives, should be correct, or at least have correctness independent of Cyborg Llama. The fact that the files both report 48kHz as their bitrates would suggest that they were recorded at that bitrate. Since that doesn't seem to be the case, it might be possibly an underlying computer/reporting problem i.e. the computer is incorrectly reporting the bitrate to have been recorded at 48kHz. To figure out if there is a computer/reporting problem, could your friend check if the stretching occurs when recording into both of the following: 1)A native program, such as Audacity www.audacityteam.org/2)Another in-browser recorder, such as the one at the following site: virtualspeech.com/voice-recorderAdditionally, I'm curious that your friend's part is slower; since 48kHz is typically the upper end of sample rates, any mismatching that would occur would typically be to a lower sample rate. In other words, the mismatched file would be the one that appears faster, which in this case, would be your recording. Could it be that your recording is the problematic one? I guess it probably isn't since, as you said, you have collaborated with others before, but could this be a newly emerged problem? To check, you can follow the same troubleshooting steps as your friend. Another question: what type of project are you using? (Metronome, Clicktrack, or Conductor)
|
|
|
Post by Charlene Adzima on Jan 21, 2021 13:55:21 GMT
Thank you for your insights :-) I will check these out and let you know what happens.
|
|
|
Post by Charlene Adzima on Jan 24, 2021 18:34:46 GMT
Ok, so I did some digging and testing and it seems that when you take the audio from my friend's recording, resample it at 48000 and play it back at 44100, it came out fine. I have no idea how this happened and we've decided to go a different route on a recording project. Just more information to add to the knowledgebase.
Thanks all!
Charlene
|
|
|
Post by lizrad on Jan 25, 2021 2:57:03 GMT
Thanks for checking back in. We're glad you've found a solution that'll work for you! If you still have the project and would feel comfortable doing so, please share the project with cyborg.llama.app@gmail.com; we would love to look into the issue further to see if we can better understand the issue and perhaps clarify this further.
|
|