IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Question about low-lantency, test for low-lantency, find the blocksize
pi_jom
post Aug 13 2009, 13:11
Post #1





Group: Members
Posts: 4
Joined: 13-August 09
Member No.: 72268



Hello!
I have downloaded the low-lantency ,while i set the blocksize=2000, and sub-blocksize=32 ,

i think the encoder will run like this : first compress 2000 /frame and then 32 /frame;

--------------------------------------------- but the fact is , it still need to pack 2000 / frame.
| 2000 | 32 | 32 | 32 |.......
--------------------------------------------- am i misunderstanding it? sad.gif

And the command is "-b320 --blocksize=2000 --sub-blocks=32 1.wav 1.wv"
Go to the top of the page
+Quote Post
bryant
post Aug 13 2009, 21:27
Post #2


WavPack Developer


Group: Developer (Donating)
Posts: 1292
Joined: 3-January 02
From: San Francisco CA
Member No.: 900



Yes, you are misunderstanding it a little. The parameters that you are using will create blocks with 64000 samples per block (2000 * 32)! However, the encoder will compress 2000 samples at a time and provide those to the application, so the effective latency is 2000 samples. Every 64000 samples (32 of these “sub blocks”) the encoder will terminate the large “super block” and start over.

I suspect that is not what you want, because even the regular encoder is efficient with 2000 samples per block. If you want the sample latency to be 32 samples, try --blocksize=32 and --sub-blocks=32. This will process the samples 32 at a time, and every 1024 samples will start a new “super block”.

Hope this makes more sense... smile.gif
Go to the top of the page
+Quote Post
pi_jom
post Aug 14 2009, 02:03
Post #3





Group: Members
Posts: 4
Joined: 13-August 09
Member No.: 72268



QUOTE (bryant @ Aug 14 2009, 04:27) *
Yes, you are misunderstanding it a little. The parameters that you are using will create blocks with 64000 samples per block (2000 * 32)! However, the encoder will compress 2000 samples at a time and provide those to the application, so the effective latency is 2000 samples. Every 64000 samples (32 of these “sub blocks”) the encoder will terminate the large “super block” and start over.

I suspect that is not what you want, because even the regular encoder is efficient with 2000 samples per block. If you want the sample latency to be 32 samples, try --blocksize=32 and --sub-blocks=32. This will process the samples 32 at a time, and every 1024 samples will start a new “super block”.

Hope this makes more sense... smile.gif


O i see, Thanks very much ! rolleyes.gif It would really help a lot.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 23rd October 2014 - 19:15