Topic: resampling question 48>44.1 for a 44.1 file

[terrapinj]

I gave a patch out at Kottke/Gordon in LA on Sep 30 from my UA-5 at 44.1. The guy asked about uploading to etree, I said sure no problem I don't have a steady highspeed connection to do it and it should keep people from bugging me about copies. Anyhoo, he just emailed me and said that there is a pitch problem with the version on etree. He thought I had sampled at 48 and resampled from 48 > 44.1 using SF. I have never resampled before, but is it possible to tell the program to resample from 48 > 44.1 if the file is already 44.1? If so would this cause the pitch to be too high, some have commented that mike sounds like a chipmunk or had a bunch of helium. The funny part is that some people have commented on how good the show sounds and even rated it at 4.7/5 with 3 votes.

http://bt.etree.org/details.php?id=18759

Thanks for any input. If you've got this show already any feedback is greatly appreciated.

[pfife]

Down in the right hand corner of the window, there should be a spot where it says "44.1" or something.  You should be able to right click on that, and then select 48 - which would effectively change the header.  Then, try to resample to 44.1.

hth

[dnsacks]

sounds like you patcher simply told the computer to assume his dat was recorded at 48k rather than 44.1k and the computer thus indicated in the header that the samples it was fed from the dat were recorded at 48k.  The patcher then resampled the "48k" wav to 44.1k, creating a pitch problem since the original digital data was in fact recorded at 44.1k. 

While the patcher could re-resample back to the proper sample rate, the CORRECT way to handle this is for the patcher to re-transfer the dat to the pc and "tell" the pc that the wav being transferred is a 44.1k dat (and to not resample anything)

[terrapinj]

sounds like you patcher simply told the computer to assume his dat was recorded at 48k rather than 44.1k and the computer thus indicated in the header that the samples it was fed from the dat were recorded at 48k.  The patcher then resampled the "48k" wav to 44.1k, creating a pitch problem since the original digital data was in fact recorded at 44.1k. 

thanks a lot - just wanted to make sure that it was feasible to tell the computer to resample to 44.1 if the file was in fact 44.1

While the patcher could re-resample back to the proper sample rate, the CORRECT way to handle this is for the patcher to re-transfer the dat to the pc and "tell" the pc that the wav being transferred is a 44.1k dat (and to not resample anything)

that's what is doing I believe

just had to make sure that resampling would cause the pitch to be off

thanks for the input guys

[pfife]

resampling DEFINATELY would cause the pitch to be off.

[F.O.Bean]

sounds like you patcher simply told the computer to assume his dat was recorded at 48k rather than 44.1k and the computer thus indicated in the header that the samples it was fed from the dat were recorded at 48k.  The patcher then resampled the "48k" wav to 44.1k, creating a pitch problem since the original digital data was in fact recorded at 44.1k. 

thanks a lot - just wanted to make sure that it was feasible to tell the computer to resample to 44.1 if the file was in fact 44.1

While the patcher could re-resample back to the proper sample rate, the CORRECT way to handle this is for the patcher to re-transfer the dat to the pc and "tell" the pc that the wav being transferred is a 44.1k dat (and to not resample anything)

that's what is doing I believe

just had to make sure that resampling would cause the pitch to be off

thanks for the input guys

I'd do the same

[terrapinj]

show was retransferred and has been tracked and is now up on etree if anyone is interested, the resampling was definately the problem, thanks for all the feedback everyone

http://bt.etree.org/details.php?id=18972


edited to add a few smiley's for my 420th post