slow down capture, speed up analysis or increase ring buffer

Forum for questions and support relating to the 1.25.x releases only.
dogmeat206
Posts: 51
Joined: Tue Oct 11, 2011 10:15 pm

Re: slow down capture, speed up analysis or increase ring bu

Post by dogmeat206 »

bb99,

So I attempted to apply that performance patch......tons of errors =\

[root@zoneminder ZoneMinder-1.25.0]# patch -p1 < zm_1_24_4-perfpatch_v1m.patch
patching file config.h.in
Hunk #1 FAILED at 177.
Hunk #2 succeeded at 292 (offset 18 lines).
1 out of 2 hunks FAILED -- saving rejects to file config.h.in.rej
patching file configure
Hunk #1 succeeded at 5787 (offset 1 line).
Hunk #2 succeeded at 5811 (offset 1 line).
Hunk #3 succeeded at 6244 (offset 1 line).
Hunk #4 FAILED at 7542.
1 out of 4 hunks FAILED -- saving rejects to file configure.rej
patching file configure.ac
Hunk #1 FAILED at 246.
Hunk #2 FAILED at 284.
2 out of 2 hunks FAILED -- saving rejects to file configure.ac.rej
patching file db/zm_create.sql.in
Hunk #1 succeeded at 329 (offset 15 lines).
patching file scripts/ZoneMinder/lib/ZoneMinder/ConfigData.pm.in
patching file scripts/ZoneMinder/lib/ZoneMinder/Memory.pm.in
patching file src/Makefile.am
patching file src/Makefile.in
Hunk #1 succeeded at 242 (offset -4 lines).
patching file src/zm_camera.cpp
patching file src/zm_camera.h
patching file src/zm_config.h.in
Hunk #1 succeeded at 38 (offset 3 lines).
patching file src/zm_config_defines.h
Hunk #1 FAILED at 27.
Hunk #2 succeeded at 250 (offset 21 lines).
Hunk #3 succeeded at 470 (offset 42 lines).
1 out of 3 hunks FAILED -- saving rejects to file src/zm_config_defines.h.rej
patching file src/zm_event.cpp
Hunk #1 FAILED at 38.
Hunk #2 FAILED at 596.
Hunk #3 FAILED at 637.
Hunk #4 FAILED at 751.
Hunk #5 FAILED at 1185.
Hunk #6 FAILED at 1196.
Hunk #7 FAILED at 1208.
Hunk #8 FAILED at 1219.
Hunk #9 FAILED at 1240.
Hunk #10 FAILED at 1269.
10 out of 10 hunks FAILED -- saving rejects to file src/zm_event.cpp.rej
patching file src/zm_ffmpeg_camera.cpp
Hunk #3 FAILED at 91.
1 out of 6 hunks FAILED -- saving rejects to file src/zm_ffmpeg_camera.cpp.rej
patching file src/zm_ffmpeg_camera.h
patching file src/zm_file_camera.cpp
patching file src/zm_file_camera.h
Hunk #1 succeeded at 36 (offset 2 lines).
patching file src/zm_image.cpp
patching file src/zm_image.h
can't find file to patch at input line 5121
Perhaps you used the wrong -p or --strip option?
The text leading up to this was:
--------------------------
|diff --git a/src/zm_jpeg.c b/src/zm_jpeg.c
|index a7517c5..996003a 100644
|--- a/src/zm_jpeg.c
|+++ b/src/zm_jpeg.c
--------------------------
File to patch: src/zm_jpeg.cpp
patching file src/zm_jpeg.cpp
Hunk #1 succeeded at 393 with fuzz 2 (offset 4 lines).
patching file src/zm_jpeg.h
Hunk #1 FAILED at 38.
1 out of 1 hunk FAILED -- saving rejects to file src/zm_jpeg.h.rej
patching file src/zm_local_camera.cpp
patching file src/zm_local_camera.h
patching file src/zm_mem_utils.h
patching file src/zm_monitor.cpp
patching file src/zm_monitor.h
patching file src/zm_mpeg.cpp
patching file src/zm_mpeg.h
patching file src/zm_remote_camera.cpp
patching file src/zm_remote_camera_http.cpp
Hunk #1 FAILED at 1096.
Hunk #2 FAILED at 1112.
Hunk #3 FAILED at 1123.
3 out of 3 hunks FAILED -- saving rejects to file src/zm_remote_camera_http.cpp.rej
patching file src/zm_remote_camera_rtsp.cpp
Hunk #4 FAILED at 231.
1 out of 4 hunks FAILED -- saving rejects to file src/zm_remote_camera_rtsp.cpp.rej
patching file src/zm_remote_camera_rtsp.h
patching file src/zm_rgb.h
patching file src/zm_stream.cpp
patching file src/zm_thread.h
patching file src/zm_utils.cpp
Hunk #1 FAILED at 18.
1 out of 2 hunks FAILED -- saving rejects to file src/zm_utils.cpp.rej
patching file src/zm_utils.h
patching file src/zm_zone.cpp
patching file src/zm_zone.h
patching file src/zma.cpp
Hunk #1 succeeded at 91 with fuzz 2 (offset -2 lines).
patching file src/zmc.cpp
Hunk #1 succeeded at 149 with fuzz 2 (offset -2 lines).
patching file src/zmf.cpp
Hunk #1 succeeded at 159 with fuzz 2 (offset -2 lines).
patching file src/zms.cpp
Hunk #1 succeeded at 86 with fuzz 2 (offset -2 lines).
patching file src/zmstreamer.cpp
Hunk #1 FAILED at 146.
1 out of 1 hunk FAILED -- saving rejects to file src/zmstreamer.cpp.rej
patching file web/includes/actions.php
patching file web/lang/big5_big5.php
patching file web/lang/cn_zh.php
patching file web/lang/cs_cz.php
patching file web/lang/de_de.php
patching file web/lang/dk_dk.php
patching file web/lang/en_gb.php
Hunk #1 succeeded at 81 with fuzz 2 (offset 9 lines).
Hunk #2 succeeded at 149 (offset 10 lines).
patching file web/lang/en_us.php
patching file web/lang/es_ar.php
patching file web/lang/et_ee.php
patching file web/lang/fr_fr.php
patching file web/lang/he_il.php
patching file web/lang/hu_hu.php
patching file web/lang/it_it.php
patching file web/lang/ja_jp.php
patching file web/lang/nl_nl.php
patching file web/lang/pl_pl.php
patching file web/lang/pt_br.php
patching file web/lang/ro_ro.php
patching file web/lang/ru_ru.php
patching file web/lang/se_se.php
patching file web/skins/classic/views/js/monitor.js.php
patching file web/skins/classic/views/monitor.php
patching file web/skins/classic/views/monitorprobe.php




Trying your suggestion on the pre-event.......
bb99
Posts: 943
Joined: Wed Apr 02, 2008 12:04 am

Re: slow down capture, speed up analysis or increase ring bu

Post by bb99 »

I'm going to leave the reply to this to Mastertheknife.
dogmeat206
Posts: 51
Joined: Tue Oct 11, 2011 10:15 pm

Re: slow down capture, speed up analysis or increase ring bu

Post by dogmeat206 »

Nope, still doing the same thing. Where it starts recording when I am already in the door. Doesn't show the door open like it did before.

Dunno if its a sensativity thing or not but my co-worker was saying the live few wasn't showing perfectly either. So I dunno if its because of that.
bb99
Posts: 943
Joined: Wed Apr 02, 2008 12:04 am

Re: slow down capture, speed up analysis or increase ring bu

Post by bb99 »

I assume you restarted ZM so I'll post a sample Buffer screen shot for you to compare:
Attachments
Screenshot.png
Screenshot.png (7.1 KiB) Viewed 10200 times
dogmeat206
Posts: 51
Joined: Tue Oct 11, 2011 10:15 pm

Re: slow down capture, speed up analysis or increase ring bu

Post by dogmeat206 »

I did, and I will use your snap shot here.

The thing that its doing now is it creates 6 events for when I am up there, instead of just running a straight event for teh entire time, it closes one, opens a new one, closes one, opens a new one.

Real jutty picture tho. Like I said, now the door opens, you see me take two steps, then I am 10 feet across the floor. Dunno if this is a FPS issue, or what the thing is trying to do, but I had a smooth image before....
dogmeat206
Posts: 51
Joined: Tue Oct 11, 2011 10:15 pm

Re: slow down capture, speed up analysis or increase ring bu

Post by dogmeat206 »

Ok....so I guess the post frames is what was the problem with that.

how many posts do I need to have until I can post a picture?

This time, the event didn't start until I was already in the room and had been standing in it for a while. In other words, it didn't see me in the picture until I had been there for 30 to 45 seconds. It should have triggered as soon as the door opened to the room...

Any idea?

I REALLY appreciate your help. You seem to have got me quite a bit furher then I was before...
bb99
Posts: 943
Joined: Wed Apr 02, 2008 12:04 am

Re: slow down capture, speed up analysis or increase ring bu

Post by bb99 »

I'd like to eliminate a camera issue; a good suggestion would be to reboot them and then test again. Also did you ever enable the motion detection on the Axis cameras? I'm having a hard time figuring the lack of sensitivity in this system. Check the "Reference Image Blend %ge", set it to a value of 7-10 (seven being the more sensitive to motion). Also check zone settings against these:
Attachments
Screenshot-1.png
Screenshot-1.png (18.94 KiB) Viewed 10199 times
bb99
Posts: 943
Joined: Wed Apr 02, 2008 12:04 am

Re: slow down capture, speed up analysis or increase ring bu

Post by bb99 »

Missed the question on posts to get over the new user block; I don't know the answer but apparently links to files hosted remotely (other then this site) happens early on.
dogmeat206
Posts: 51
Joined: Tue Oct 11, 2011 10:15 pm

Re: slow down capture, speed up analysis or increase ring bu

Post by dogmeat206 »

Zones for the first Cam.

If you look, top left is the door coming in. Then Cage doors, and the floor.
Attachments
Screen Shot 2011-10-24 at 1.26.25 PM.png
Screen Shot 2011-10-24 at 1.26.25 PM.png (173.19 KiB) Viewed 10198 times
dogmeat206
Posts: 51
Joined: Tue Oct 11, 2011 10:15 pm

Re: slow down capture, speed up analysis or increase ring bu

Post by dogmeat206 »

This is the door frame, set to 'Best - High sensativity'
Attachments
Screen Shot 2011-10-24 at 1.29.37 PM.png
Screen Shot 2011-10-24 at 1.29.37 PM.png (215.95 KiB) Viewed 10197 times
dogmeat206
Posts: 51
Joined: Tue Oct 11, 2011 10:15 pm

Re: slow down capture, speed up analysis or increase ring bu

Post by dogmeat206 »

I reset the zone sensitivity of the door to what you suggested. Going to run up and try it out.

The cameras seem clean, I can pull the live view from them, and it seems to be fine. Machine's been rebooted and it sunning at a .7 load.
dogmeat206
Posts: 51
Joined: Tue Oct 11, 2011 10:15 pm

Re: slow down capture, speed up analysis or increase ring bu

Post by dogmeat206 »

So this is weird, I reset the zone to your settings and its odd. I walked into the room, did a circle of it all and then walked out the door. Well it shows the door open, me come in, walk to the end of the cage then it turns me around and shows me walking about.

So those settings for sensitivity works well. Question is why didn't it continue to record me walking around the room and such? It seemed to just stop after I made it ot the end of the cage and didn't kick back on untl I was walking out.
bb99
Posts: 943
Joined: Wed Apr 02, 2008 12:04 am

Re: slow down capture, speed up analysis or increase ring bu

Post by bb99 »

This is one of the reasons I abandoned multiple zones versus one zone covering the whole area. If you only changed the one zone (door) the rest weren't picking up your motion and burned up your post buffer frames. Then as the door settings realarmed the recording plus the new prealarm buffer settings kicked in. Setting up multiple zones with the same sensitivity at varying distances change the number of pixels affected forcing you to draw a happy median for settings or doing some insane geometry math or just flat out trial and error for sensitivity. I do realize that this is a megapixel camera (makes what you want to do easier) but the wide angle lens (3mm I'm guessing) makes it more difficult. Am I making any sense? Edit: Each zone is treated as a different camera within one monitor; does that explain it better?
dogmeat206
Posts: 51
Joined: Tue Oct 11, 2011 10:15 pm

Re: slow down capture, speed up analysis or increase ring bu

Post by dogmeat206 »

BB99, that makes tones of sense. Very weird tho.

My only problem is there are lights on the systems that come on and off, which trigger alarms.
caseystone
Posts: 98
Joined: Fri Feb 25, 2005 3:41 am
Location: England

Re: slow down capture, speed up analysis or increase ring bu

Post by caseystone »

Yeah, maybe try just one zone covering the top 1/3 of that image, settings like what worked for you on the door.
SB G620 on Asus P8H67-M LX, 4GB, WD10EURS 'DVR' drive. Ubuntu 10.04.3 server. System load <=1.

8-port 240fps BTTV capture card, 6 cams. Modect 1 zone blobs of mx area 3, mx blobs 1. Some on alarmed pixels mn5/mx80. Mons not linked.
Locked