Home |
Search |
Today's Posts |
#1
|
|||
|
|||
changing pitch between 44.1k and 48k
I have a 44.1 file that was being told to be 48k. Apparently I didn't
have my settings right when I dumped it in to my DAW. I need to change the pitch up (speed it up) roughly a half step to make it right. What is the pitch formula in cents to achieve the exact pitch change? |
#2
|
|||
|
|||
On 8 Jul 2005 12:52:48 -0700, "jonothon" wrote:
I have a 44.1 file that was being told to be 48k. Apparently I didn't have my settings right when I dumped it in to my DAW. I need to change the pitch up (speed it up) roughly a half step to make it right. What is the pitch formula in cents to achieve the exact pitch change? The last post in this thread derives the formula and gives the exact (well, to 20 or so decimal places) value. It appears to be 147 cents, or 1 semitone (a half-step) and 47 cents (slightly under 1/2 of a half-step): http://groups-beta.google.com/group/...8333 1f222591 ----- http://www.mindspring.com/~benbradley |
#3
|
|||
|
|||
"jonothon" wrote in news:1120852368.922941.234430
@g14g2000cwa.googlegroups.com: I have a 44.1 file that was being told to be 48k. Apparently I didn't have my settings right when I dumped it in to my DAW. I need to change the pitch up (speed it up) roughly a half step to make it right. What is the pitch formula in cents to achieve the exact pitch change? Rather than changing the pitch, why not set it back to 44.1, then do a proper sample rate conversion? Produces no pitch errors. |
#4
|
|||
|
|||
"jonothon" wrote in message
oups.com I have a 44.1 file that was being told to be 48k. Apparently I didn't have my settings right when I dumped it in to my DAW. I need to change the pitch up (speed it up) roughly a half step to make it right. What is the pitch formula in cents to achieve the exact pitch change? Much DAW software has a means for changing the sample rate of a wav file without changing the file's audio data. It just changes a few bits in the header. The file then has the proper pitch, (if it ever did.) ;-) |
#5
|
|||
|
|||
That's how this happened, I think. I was running out of my studio to a
session elsewhere, dumped a 48/24 file into Peak using s/pdif input and clock, resulting in a 48k file. Coverted the sample rate to 44.1 to burn an audio cd. I mut have screwed up somewhere, cuz when i got to the session, the audio cd played back slow. By the way, 147 cents sounds ok, but it's still not right. tempo is still a tiny bit slow. The right amount of pitch shift must be in the cracks. The other thing is I remember fixing files like this before, but can't quite come up with it for some reason. Anyone? |
#6
|
|||
|
|||
"jonothon" wrote in message
ups.com That's how this happened, I think. I was running out of my studio to a session elsewhere, dumped a 48/24 file into Peak using s/pdif input and clock, resulting in a 48k file. Coverted the sample rate to 44.1 to burn an audio cd. I mut have screwed up somewhere, cuz when i got to the session, the audio cd played back slow. By the way, 147 cents sounds ok, but it's still not right. tempo is still a tiny bit slow. The right amount of pitch shift must be in the cracks. The other thing is I remember fixing files like this before, but can't quite come up with it for some reason. Anyone? It's very easy to do in Audition/CEP - the function is called Edit, Adjust Sample Rate. I first encountered this problem with a DAT tape that was sent to me. The various files were recorded at some arbitrary mixture at 44 & 48. The DAT machine was a "consumer" device that recorded analog inputs at 48, but recorded from its digital input at whatever sample rate the digital data was coming in at. In this case the digital source was a CD player, so its sample rate was 44. |
#7
|
|||
|
|||
jonothon wrote: By the way, 147 cents sounds ok, but it's still not right. tempo is still a tiny bit slow. The right amount of pitch shift must be in the cracks. If you your pitch scaling app (I assume that's what you are using because the parameter is in cents) is set to preserve the length, then the tempo will not be affected by the process. If the problem was an SRC with a target rate other than 44.1 (and ending up being labled 44.1 in the file header) then you want to SRC again while keeping the sample rate associated with the file at 44.1 Khz. Turning off the "preserve length" option in your scaling app will do just that and the pitch and tempo will change reciprocally. Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#8
|
|||
|
|||
Use my free utility Header Investigator if you're on XP - or SoundHack if
you're on OS X to change the sample rate in the header to match the true sample rate of the data. Rail -- Recording Engineer/Software Developer Rail Jon Rogut Software http://www.railjonrogut.com "jonothon" wrote in message oups.com... I have a 44.1 file that was being told to be 48k. Apparently I didn't have my settings right when I dumped it in to my DAW. I need to change the pitch up (speed it up) roughly a half step to make it right. What is the pitch formula in cents to achieve the exact pitch change? |
#9
|
|||
|
|||
Bob Cain wrote: Turning off the "preserve length" option in your scaling app will do just that and the pitch and tempo will change reciprocally. Nah, not reciprocally but by the same proportion. D'oh. Bob -- "Things should be described as simply as possible, but no simpler." A. Einstein |
#11
|
|||
|
|||
|
#12
|
|||
|
|||
On Fri, 8 Jul 2005 15:52:48 -0400, jonothon wrote
(in article .com): Rail, Thanks for your generosity. I recently reconfirmed the importance of a master clock. I recorded a board feed to a laptop and also ran a DAT with room mics at a club. When I combined laptop and DAT tracks later in a Pro Tools session, I couldn't get them to sync up. They were the right pitch, but within 30 seconds it was obvious that one was walking away from the other. Fascinating. Ty Ford -- Ty Ford's equipment reviews, audio samples, rates and other audiocentric stuff are at www.tyford.com |
#13
|
|||
|
|||
Ty, please help me a bit here. If the sample rates are different both
the pitch and duration must be off. Were the pitches very close so the initial sample rates only a little different? The timing error between the two tracks would then increase with time and become more and more apparent. Is this what you observed? Thanks jwvm |
#14
|
|||
|
|||
"Rail Jon Rogut" wrote:
Use my free utility Header Investigator if you're on XP First you brought peace to my universe with your utility for allowing other software to use my Pro Tools hardware, now this. You are truly a nifty dood. -- "It CAN'T be too loud... some of the red lights aren't even on yet!" - Lorin David Schultz in the control room making even bad news sound good (Remove spamblock to reply) |
#16
|
|||
|
|||
In article , Ty Ford
wrote: Ty, please help me a bit here. If the sample rates are different both the pitch and duration must be off. Were the pitches very close so the initial sample rates only a little different? The timing error between the two tracks would then increase with time and become more and more apparent. Is this what you observed? Thanks jwvm Yes that's what happened. Pitches were on to my ear, no weird anaomalies when they were mixed together. What I'm saying is that I think 44.1 on one machine might be 44.1003 and 44.1 000275 on the other machine. (numbers are only examples and do not indicate actual operating conditions,) If I had a Vocalign plugin, I could probably realign the timebase of one to another without changing pitch. Don't have it so I can't say what the outcome would be. Regards, Ty Ford Ty, I had this happen in my early days of Protools. I was dumping from multitrack into PT using Digi's SSD Sync box and, after working on some vocals in Autotune, went to dump back to the 24 track, and the sync would drift after 20 - 25 seconds. Hmmmm. I found what happened to me was I forgot to change the PT sync clock to "Internal" for the analog dump in. I had previously been dumping stuff in digitally and it was on Digital in. Not doing that again took care of it. One of the nice things a Digi 192K interface does it continue blinking at you if you try to do this, and PT even shoots a dialog box at you saying there is no digital sync. Good hardware and software. Kinda makes you wonder, tho, how it records supposedly without any sync, because it will still record. David Correia Celebration Sound Warren, Rhode Island www.CelebrationSound.com |
#17
|
|||
|
|||
There are tolerances at play.. unless the clocks are tied together (ie use a
single clock) the tolerance will be different between different digital devices. Never expect 2 (or more) systems running on their internal clocks to be running at the same speed. Rail -- Recording Engineer/Software Developer Rail Jon Rogut Software http://www.railjonrogut.com "spud" wrote in message ... On Sun, 10 Jul 2005 12:46:51 -0400, Ty Ford wrote: On Sat, 9 Jul 2005 23:29:22 -0400, wrote (in article . com): Ty, please help me a bit here. If the sample rates are different both the pitch and duration must be off. Were the pitches very close so the initial sample rates only a little different? The timing error between the two tracks would then increase with time and become more and more apparent. Is this what you observed? Thanks jwvm Yes that's what happened. Pitches were on to my ear, no weird anaomalies when they were mixed together. What I'm saying is that I think 44.1 on one machine might be 44.1003 and 44.1 000275 on the other machine. (numbers are only examples and do not indicate actual operating conditions,) If I had a Vocalign plugin, I could probably realign the timebase of one to another without changing pitch. Don't have it so I can't say what the outcome would be. Regards, Ty Ford -- Ty Ford's equipment reviews, audio samples, rates and other audiocentric stuff are at www.tyford.com Also my experience. In a 4 mic drum situation. kick and snare on a laptop and overheads on a portable dat machine. When I lined them up in a DAW later it was obvious they were recorded at different rates. Roughly 30 secs into the song the visual was off and the sound was weird. No amount of pushing and pulling fixed it and I soon lost track of what sounded right and what didn't. I thought 44.1k was an industry standard and would be the same world over forever but not the case it seems. |