Patch Phrase!

Well after having my 2500 for about 3 weeks now, I thought that I would share a experience with you guys as well as give a little feedback!! Well I am sorry to say that I am a bit dissappointed with the "Patch Phrase" part of the 2500... It is truly a crap-shoot, trying to get something to come out ok!

Example: I will take my Fantom X and make a 4 bar loop of lets say a guitar, and the BPM will be set at 80, well I then sample that 4 bar loop into the 2500. Next I go into the trim page and make sure everything is still looping tightly, then I go to chop-->Convert. While in that page it shows that my initial BPM is 164 which isnt right, no problem because I already know the tempo of the loop (80).. Once converted to a PP, the loop sounds choppy, and sometimes it don't even finish out the loop!! I think Akai needs to tweek that out a little bit more so that it can truly be effective. Oh and before I forget, not being able to convert a stero sample into a PP is also limiting, I did the same thing as mentioned above with another 4 bar loop made on the Fantom with panned toms and congas and lost the whole effect once I sampled it into the phantom and converted it to a PP!!

On last thing, I think the should do like my Fantom X and have more than on Continuous Sample track, now that would be fire. Well I hope I haven't rambled on to long, peace!

I agree. I too am a new 2500 user(3weeks or so) and also was quite dissapointed with the patch phrase conversion capabilities ,.............at the same time I didn't spend 2000.00 bucks for nothing I will make a sheer effort to get all I can out of this feature. I also have noticed that with a simpler sample (e.g. one instrument no hi-hats or bass kicks..etc) it seems to work a little better.

You can convert your sample to 2 x samples, one sample is left and one sample is right, patch phrase both and then run them parallel in a sequence.


"Once converted to a PP, the loop sounds choppy"
That 164bpm seems to be like the destination bpm for you 80bpm loop.

Seems that the 2500 did exactly what it was supposed to do by stretching your 80 to 164 thus stretching more than 10-15%, what it did was more like 200%. When you stretch for more than 10-15% your stretch will sound unnatural and choppy. You gotta understand the procedure and rules for patched phrase and the look into that 164bpm settings in detail.

I don't have a 2500 but from reading the manual and understanding patched phrase is a timestretch sorta thing, this may be your problem.

http://www.mpc-forums.com/viewtopic.php?t=54844&highlight=

Also check that thread out.

do you double your tempo/time sig? i do, and whenever i detect tempo on my 1k it gives me a reading of twice as much or there abouts. i always figured it was because i double the tempo/time sig and that the machine somehow takes that into consideration.?.

The only reason for that is, the MPC detects tempo based on a 2 Bar loop. So if you record a 4 Bar loop like you said, the tempo will obviously be off by a factor 2. Same with a one bar loop, it will probably have read around 40 then.

Patch Phrase is only good for drum loops with clear spaces between the hits. And then only for moderate tempo changes. All it does is chop the loop, and play the chops in intervals according to the tempo, just like you would with your hands. There's no more magic about that.

Man those are some good replies, see this is what a forum should be about! So what you guys are saying is that when I mad a continuous guitar loop, that it will be less effective in PP because their aren't any defined breaks, as with a drum loop?

Also why does it default to a tempo 2 or 3 times the original with a four bar loop, lets say that I had a four bar loop that I know was 80bpm, and I put it thru the PP conversion, the default tempo that it displays would be around 164, then if I put the tempo at 80 because I know the bpm already, that would make for more than a 10 to 15% change! It seems wierd that they would have it work that way.. Do you have any suggestions on how to work around that.

Peace!

The default is hardcoded in the OS, you can't change it. Although that would be really easy to make adjustable to the user.

Just calculate the correct tempo yourself. If it shows 164, and you know you're on 4 bars instead of two, divide it by two. That would give 82 BPM. Now if you know you recorded it at 80, you would have to set the original tempo to 160 BPM. The difference is probably because the recorded sample is slightly longer or shorter than the four bar sequence on the MPC.

But if you play the loop by yourself already, you could even record it in the correct tempo in the first place. For these kind of samples, I would avoid PP or Timestretch whenever possible.

_Stilo_ wrote:The only reason for that is, the MPC detects tempo based on a 2 Bar loop.

I just skipped through the manual again, and I found this statement about using Detect Tempo on p. 67:

"I. Set the start point and the end point of the sample to match to the length of one bar at 4/4."

So the proposed method is obviously to select one bar of the sample, no matter how long it is, and detect the tempo of that. Not sure where I got the 2 bars from, maybe I was mistaken. But my explanation in general was right, and most of the time you will know if the detected tempo is too low or too high, cause it's always at least by a factor 2.

_Stilo_ wrote:

_Stilo_ wrote:The only reason for that is, the MPC detects tempo based on a 2 Bar loop.

"I. Set the start point and the end point of the sample to match to the length of one bar at 4/4."

quote]


I mean I am no mathematician, but isn't 1 bar at 4/4 going to detect the tempo at 4 times the original? his loop was 4 bars at 80bpm, so it should have read 320 in detect.........Right? Im I missing something here?

You're right. I'm confused myself right now, I just tried to detect a 1 bar sample, I know it loops fine with 81.2 BPM, and the detect tempo gives me double the tempo.

Also, it always gives the same tempo regardless of where I set the start and end point, as opposed to what the manual suggests.

The most reliable method is still, put the sample in a sequence, and adjust the seq's tempo until it loops perfectly. Then you have your tempo. Works everytime.

_Stilo_ wrote:
The most reliable method is still, put the sample in a sequence, and adjust the seq's tempo until it loops perfectly. Then you have your tempo. Works everytime.

Totally!!!!!!!1

Lampdog wrote:You can convert your sample to 2 x samples, one sample is left and one sample is right

Is there a function for that on the 2000XL? Cause I don't see that in my 2500, I have to resample the Main Outs two times to get it split. Or sample the two channels seperately in the first place of course...

_Stilo_ wrote:

Lampdog wrote:You can convert your sample to 2 x samples, one sample is left and one sample is right

Is there a function for that on the 2000XL? Cause I don't see that in my 2500, I have to resample the Main Outs two times to get it split. Or sample the two channels seperately in the first place of course...

make 2 copies of the same sample. assign 1 to the left and 1 to the right.

cre8tive wrote:make 2 copies of the same sample. assign 1 to the left and 1 to the right.

I was talking of splitting the stereo sample into the two parts...which are usually different, or I wouldn't need to retain the stereo. Yes, I could copy them, and silence the left or right channel, then pan it, but that would be a waste of memory. I did not find a Split function however.