Thanks Jason
I moved this over here, including your answer below, fused to Mareks answer.
> I am honestly just so tired of reading this incorrect claim. I invite you to *test* it yourself. Hook up two ELP (or other well behaved) cameras and actually try it, as I did, repeatedly. As Neils did *while developing the library* you are quoting. It works correctly on all three operating systems. This was the entire reason we undertook the months long development effort and is the reason we released it when it was *tested* and finished.
Oh, believe me, I tried and tried again. And I reported the issue here.
https://github.com/openpnp/openpnp/issues/717#issuecomment-381389641
and here:
https://groups.google.com/d/msg/openpnp/fVv7PvArwiQ/-uqPOGoSCgAJ
and here:
https://github.com/openpnp/openpnp/issues/687#issuecomment-381394138
I just tried again with develop branch built 2018-11-20. Two ELP “6mm 2MP Full HD High Speed MJPEG USB” cameras on a new USB 2.0 external hub (tried various hubs including a USB 3.0 hub – to no avail).
Second Camera stays black. No error messages in Log or stderr:
The same hardware on a separate root hub:
Please note that there are different ELP cameras. Mine are the “6mm 2MP Full HD High Speed MJPEG USB” (ebay listing no longer active).
920 (H) x 1080 (V) pixels MJPEG 30fps YUY2 6fps
1280 (H) x 1024 (V) pixels MJPEG 30fps YUY2 6fps
1280 (H) x 720 (V) pixels MJPEG 60fps YUY2 9fps
1024 (H) x 768 (V) pixels MJPEG 30fps YUY2 9fps
800 (H) x 600 (V) pixels MJPEG 60fps YUY2 21fps
640 (H) x 480 (V) pixels MJPEG 120fps YUY2 30fps
352(H) x 288 (V) pixels MJPEG 120fps YUY2 30fps
320 (H) x 240 (V) pixels MJPEG 120fps YUY2 30fps
There are slower ones like the “ELP 6mm 1.0MP 720P HD MJPEG USB Camera”
1280 (H) x 720 (V) pixels MJPEG 30fps YUY2 10fps
960 (H) x 720 (V) pixels MJPEG 30fps YUY2 15fps
800 (H) x 600 (V) pixels MJPEG 30fps YUY2 20fps
640 (H) x 480 (V) pixels MJPEG 30fps YUY2 30fps
352(H) x 288 (V) pixels MJPEG 30fps YUY2 30fps
320 (H) x 240 (V) pixels MJPEG 30fps YUY2 30fps
176(H) x 144 (V) pixels MJPEG 30fps YUY2 30fps
160 (H) x 120 (V) pixels MJPEG 30fps YUY2 30fps
Doubling the FPS on the same resolution might explain why USB 2.0 is still flooded by the fastest hires cameras. I don’t know which camera Niels used for testing. However in my view it is to be expected that the best cameras will exploit whatever bandwidth they can get.
For me this is no issue except perhaps for the misleading Wiki entry I reported. The bandwidth limitation is in hardware and I don’t see how OpenPNP could solve it universally (true USB 3.0 cameras will in time).
However, not being able to set a custom FPS and therefore reducing the bandwidth might be considered an issue. I opened such an issue in openpnp-capture now:
https://github.com/openpnp/openpnp-capture/issues/29
But like I explained in this thread, I believe that high FPS is actually a good thing for low latency. So I personally rather live with two USB connectors from the machine, saving time on each vision capture. I’m also not so sure that the camera hardware would actually support reducing FPS. So I consider this a low priority issue.
Yes I’m using Windows 7. Maybe that’s the reason. But frankly I doubt it.
For a test of said cameras see here:
https://makr.zone/camera-selection/262/
_Mark
Von: ope...@googlegroups.com [mailto:ope...@googlegroups.com] Im Auftrag von Jason von Nieda
Gesendet: Montag, 3. Dezember 2018 16:47
An: ope...@googlegroups.com
Betreff: Re: [OpenPnP] disable move to safe z
Hi Mark,
The "physical FPS" I am referring to is part of the "format". The formats are enumerated by the camera and one is chosen. The custom FPS field you are referring to as not implemented was added to provide a way to reduce the frame rate from the max, instead of just using the max.
I'm not claiming that there isn't latency - of course data takes time to move through the various systems - but there is no piling up of frames when OpenPnP isn't keeping up with reading them. When OpenPnP calls Camera.capture() for a vision operation the image it will get will be the last one processed by the camera.
In case it's been lost in the thread, the claim I was refuting was:
When 2 can't take the 10 FPS all the time then the data from the camera will
buffer up somewhere on the way and when the vision wants a picture it will
not get a fresh image from the camera, it will gets the oldest picture from
the buffer.
If 2 stalled for some reason, then the "fresh" picture can be as old as the
stall.
Again, this is *very easy* to test and prove if you want to see it yourself. openpnp-capture has a simple interface that you can write a test program for in about 10 minutes. Capture an image, write it to disk, sleep for 10 seconds and during those 10 seconds change what is in view of the camera. Capture an image, write it to disk. Look at the two images.
I'm not talking about a theoretical system - I'm talking about what happens in OpenPnP today.
a) why one USB Root Hub can still only manage one camera even with MJPEG compression (as tested under Windows) and
I am honestly just so tired of reading this incorrect claim. I invite you to *test* it yourself. Hook up two ELP (or other well behaved) cameras and actually try it, as I did, repeatedly. As Neils did *while developing the library* you are quoting. It works correctly on all three operating systems. This was the entire reason we undertook the months long development effort and is the reason we released it when it was *tested* and finished.
If you are unable to get two cameras working on a single hub please file a bug report with details on the cameras, operating system, and USB tree. We will attempt to reproduce and fix it.
Jason
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jw7fG2Euvoiv9_4kyn7_yfWxDZ-Z3ufPXym3n-UhhfO7A%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.
I don't remember the soft name that used to test it instead of Openpnp.
But I remember that using it two cameras 2Mpx/25fps connected to the same controller worked without any problems (at occasion with minimal CPU load, some 3-5%).
Unfortunately I have now at home only webcam and ELP that gives no video (something failed in camera hardware). So cannot reproduce it responsibly on the moment.
I will do this asap and comeback with report.
Downloaded OBS.
Thanks Jason
I moved this over here, including your answer below, fused to Mareks answer.
> I am honestly just so tired of reading this incorrect claim. I invite you to *test* it yourself. Hook up two ELP (or other well behaved) cameras and actually try it, as I did, repeatedly. As Neils did *while developing the library* you are quoting. It works correctly on all three operating systems. This was the entire reason we undertook the months long development effort and is the reason we released it when it was *tested* and finished.
Oh, believe me, I tried and tried again. And I reported the issue here.
https://github.com/openpnp/openpnp/issues/717#issuecomment-381389641
and here:
https://groups.google.com/d/msg/openpnp/fVv7PvArwiQ/-uqPOGoSCgAJ
and here:
https://github.com/openpnp/openpnp/issues/687#issuecomment-381394138
I just tried again with develop branch built 2018-11-20. Two ELP “6mm 2MP Full HD High Speed MJPEG USB” cameras on a new USB 2.0 external hub (tried various hubs including a USB 3.0 hub – to no avail).
Second Camera stays black. No error messages in Log or stderr:
The same hardware on a separate root hub:
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/!%26!AAAAAAAAAAAYAAAAAAAAAIksy9bq9upCo/PrMujuDwDCgAAAEAAAAM0II395t1NKiA1IaCuoRkUBAAAAAA%3D%3D%40makr.zone.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jx%3DLwWDbtHm3gyQ_uN0cbbFhGtyEK7UiVYs1r-EFhe6tQ%40mail.gmail.com.
Hi Jason
I am very sorry that these statements caused massive support issues! I will of course try hard to be more careful in the future.
> The easiest way to determine if this is a openpnp-capture issue is to see if you can get the correct behavior in another app. OBS (Open Broadcast Studio) is a good app for testing webcams
I can confirm I do have the same issue with OBS Studio. One cam OK, both cams not OK. The difference is that OBS won’t even let me activate one cam unless I unplug the other. OpenPNP will at least allow the first one to be activated even if the second one is plugged in. Likewise OBS will run both cams if on separate hubs. Unfortunately OBS can’t configure the FPS either (or I don’t know how).
With VLC the same issues. FPS config does not seem to matter, connecting the second cam will always create an error. Again on separate hubs all is OK.
Like I said previously I don’t think this is an issue with OpenPNP at all. It’s just a “natural” limit of USB 2.0 with high speed hires cams. And after unsuccessfully trying with VLC I’m doubtful that the native FPS can be overridden at all.
> the cameras that are recommended on the OpenPnP site and Wiki
A good point. I must have missed that. Or maybe I thought a better model will surely also work J
> If you'll confirm that this looks like the camera you own I'll buy a set and see how they perform in Windows 10.
It looks slightly different, but it is also two years ago, when I ordered mine, maybe they’re just an older revision:
I suggest we do it the other way around. I will test with my cameras on Windows 10 as soon as I get my hands on one. If it works there, everything is clarified.
If it fails on Windows 10, I will order two of the recommended 720p ones. This way I might be able to confirm that it’s actually the HD vs. Full HD bandwidth step-up that prevents sharing the hub.
If all this is clarified we can help others by mentioning the caveats on the Wiki. And that’s all I truly wanted. Sorry again.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jx%3DLwWDbtHm3gyQ_uN0cbbFhGtyEK7UiVYs1r-EFhe6tQ%40mail.gmail.com.
Hi Jason
I am very sorry that these statements caused massive support issues! I will of course try hard to be more careful in the future.
> The easiest way to determine if this is a openpnp-capture issue is to see if you can get the correct behavior in another app. OBS (Open Broadcast Studio) is a good app for testing webcams
I can confirm I do have the same issue with OBS Studio. One cam OK, both cams not OK. The difference is that OBS won’t even let me activate one cam unless I unplug the other. OpenPNP will at least allow the first one to be activated even if the second one is plugged in. Likewise OBS will run both cams if on separate hubs. Unfortunately OBS can’t configure the FPS either (or I don’t know how).
With VLC the same issues. FPS config does not seem to matter, connecting the second cam will always create an error. Again on separate hubs all is OK.
Like I said previously I don’t think this is an issue with OpenPNP at all. It’s just a “natural” limit of USB 2.0 with high speed hires cams. And after unsuccessfully trying with VLC I’m doubtful that the native FPS can be overridden at all.
> the cameras that are recommended on the OpenPnP site and Wiki
A good point. I must have missed that. Or maybe I thought a better model will surely also work J
> If you'll confirm that this looks like the camera you own I'll buy a set and see how they perform in Windows 10.
It looks slightly different, but it is also two years ago, when I ordered mine, maybe they’re just an older revision:
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/04d301d48b58%24b7ee7cf0%2427cb76d0%24%40makr.zone.
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/60a339ee-d04f-4180-bab2-11f196d7fb14%40googlegroups.com.
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/f91c432e-f059-465b-a22a-e8925a577dd7%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/07284cf1-d239-45b9-a6bc-3b1c3fde086c%40googlegroups.com.
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/c9472cd8-e2f1-4d7b-a17e-b2799b82c15d%40googlegroups.com.
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/2929cbb4-8826-4da0-b1e3-24f0c816bcbd%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/2544ad7e-0b5a-4c21-bd91-d1419ae61d95%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jxBh3O4UqAF_HQSOqOSK7iZXuDCPsG9bZVfhfqShQtbVQ%40mail.gmail.com.
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/0325bce0-0bc2-4f42-bf95-c8ae3487cec9%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/38126d69-59f6-4942-9cc2-1672266619f4%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/7274f1c9-a045-4de8-a182-867005d5fbd0%40googlegroups.com.
Mark has two 1080p I guess.
Hi Jason
> I suspect that this is the camera is that is causing the trouble. I see that the 1280x720 resolution is 60 FPS, but there are lower resolution / FPS combos, so I wonder if one of these would do the trick? I would think 1024x768-MJPEG@30fps should be less bandwidth than the 1280x720-MJPEG@30FPS that I use.
Tested two of my ELP-USBFHD01M with 1024x768-MJPEG@30fps as you suggested. Does not help (believe me, I tried all possible settings back when I setup the machine J). Also tried fiddling with exposure, as Marek hinted, again unsuccessfully. Same for restarting OpenPNP. Tried two different USB hubs.
The first cam will work, the second not (black screen, no error). I can get both running individually depending on order of config.
I also tried to find a way to measure/compare consumed USB bandwidth. Windows actually has a properties dialog in the Device Manager showing bandwidth usage on USB 2.0 host controllers. I had to go to my old computer, because unfortunately this tab is not available on the USB 3.0 controllers I have in the newer computer.
ELP-USBFHD01M with 1024x768-MJPEG@30fps:
HBV-1517 S1.0 (that’s the newer cam Juha sells with the Liteplacer) at 1280x1024, 7FPS, YUV2:
Tried other, lower resolutions, they all take 48%, as the FPS is increased to fill up the bandwidth.
Also tried the built-in webcam 1280x720@30fps MJPG:
While this is not the ELP model this is a good indicator, that in deed the 30 vs. 60 fps do matter.
In conclusion the following facts seem interesting:
1. It is well established that two YUV2 webcams won’t work on one root hub.
2. HBV-1517 S1.0 at 1280x1024, 7FPS, YUV2 is reported at 48%.
3. 2 x 48% < 100% so why does it not work?
4. Guesswork: USB 2.0 is half-duplex (USB 3.0 is full-duplex), so the system might only allocate 50% of the bandwidth for the up-stream. Maybe there is dynamic allocation of the up- vs. down-stream but 2 x 48% is just too close to 100% to be stable, also considering that the reported 48% may be an average over some time. Strangely I found no info on the web.
5. Seeing that even the smallest ELP-USBFHD01M 1024x768-MJPEG@30fps format results in the same 48% bandwidth use as the YUV2 cam, this explains for me, that likewise one USB root hub will not fit two of those cameras.
6. The reported 26% use with the built-in webcam 1280x720@30fps MJPG format might be seen as a confirmation for the opposite case. Seeing that the properties dialog reports 1% base load even for unconnected devices and allowing for some rounding errors it does make sense that 2 x 26% - baseload - rounding errors could roughly fit into the 50% half duplex upstream budget.
I also checked a Logitech C310 and confirmed that formats are either at 48% bandwidth or at 26% bandwidth for the very low resolution formats:
This makes perfect sense to me: a camera will either use ~47% i.e. practically the full upstream bandwidth or ~25% i.e half the upstream bandwidth, allowing one or two cams on one hub.
Jason has already updated the Wiki to suggest to people to buy the recommended 720p models. Perhaps the the Wiki could be more specific, explaining that for two hires cameras to work on the same hub they must be MJPG. However the inverted contraposition (more precisely: “falscher Umkerschluss” in German) is not necessarily true: a hires camera being MJPG does not yet guarantee that you can run two on the same hub. An MJPG camera can still be too fast/too hires!
I will still try to get my hands on a Windows 10 system. Perhaps there is still a difference.
_Mark
P. S. Soon this will all go away with true USB 3 cameras.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jxAdGa%2B%2BQttpHuaV4%2Bhm2sA1fWBc3LM5KJ3%3DvoF5TPa6A%40mail.gmail.com.
Hi Jason
> I suspect that this is the camera is that is causing the trouble. I see that the 1280x720 resolution is 60 FPS, but there are lower resolution / FPS combos, so I wonder if one of these would do the trick? I would think 1024x768-MJPEG@30fps should be less bandwidth than the 1280x720-MJPEG@30FPS that I use.
Tested two of my ELP-USBFHD01M with 1024x768-MJPEG@30fps as you suggested. Does not help (believe me, I tried all possible settings back when I setup the machine J). Also tried fiddling with exposure, as Marek hinted, again unsuccessfully. Same for restarting OpenPNP. Tried two different USB hubs.
The first cam will work, the second not (black screen, no error). I can get both running individually depending on order of config.
I also tried to find a way to measure/compare consumed USB bandwidth. Windows actually has a properties dialog in the Device Manager showing bandwidth usage on USB 2.0 host controllers. I had to go to my old computer, because unfortunately this tab is not available on the USB 3.0 controllers I have in the newer computer.
ELP-USBFHD01M with 1024x768-MJPEG@30fps:
Tried other, lower resolutions, they all take 48%, as the FPS is increased to fill up the bandwidth.
Also tried the built-in webcam 1280x720@30fps MJPG:
While this is not the ELP model this is a good indicator, that in deed the 30 vs. 60 fps do matter.
In conclusion the following facts seem interesting:
1. It is well established that two YUV2 webcams won’t work on one root hub.
2. HBV-1517 S1.0 at 1280x1024, 7FPS, YUV2 is reported at 48%.
3. 2 x 48% < 100% so why does it not work?
4. Guesswork: USB 2.0 is half-duplex (USB 3.0 is full-duplex), so the system might only allocate 50% of the bandwidth for the up-stream. Maybe there is dynamic allocation of the up- vs. down-stream but 2 x 48% is just too close to 100% to be stable, also considering that the reported 48% may be an average over some time. Strangely I found no info on the web.
5. Seeing that even the smallest ELP-USBFHD01M 1024x768-MJPEG@30fps format results in the same 48% bandwidth use as the YUV2 cam, this explains for me, that likewise one USB root hub will not fit two of those cameras.
6. The reported 26% use with the built-in webcam 1280x720@30fps MJPG format might be seen as a confirmation for the opposite case. Seeing that the properties dialog reports 1% base load even for unconnected devices and allowing for some rounding errors it does make sense that 2 x 26% - baseload - rounding errors could roughly fit into the 50% half duplex upstream budget.
I also checked a Logitech C310 and confirmed that formats are either at 48% bandwidth or at 26% bandwidth for the very low resolution formats:
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/003d01d48ca6%24d74462f0%2485cd28d0%24%40makr.zone.
Hi Jason
Nice tool. Attached the ELP-USBFHD01M usbtreeview report.
However: it does not make much sense, clearly USB 2.0 does not support 995’469’360 bps (or 124.4Mbyte/s) as a minimum:
----- Video Streaming MJPEG Frame Type Descriptor -----
---> This is the Default (optimum) Frame index
bLength : 0x1E (30 bytes)
bDescriptorType : 0x24 (Video Streaming Interface)
bDescriptorSubtype : 0x07 (MJPEG Frame Type)
bFrameIndex : 0x01
bmCapabilities : 0x00
wWidth : 0x0780 (1920)
wHeight : 0x0438 (1080)
dwMinBitRate : 0x3B55A830 (995469360 bps -> 124.4 MB/s)
dwMaxBitRate : 0x3B55A830 (995469360 bps -> 124.4 MB/s)
dwMaxVideoFrameBufferSize: 0x003F4A4D (4147789 bytes)
dwDefaultFrameInterval : 0x00051615 (33.3333 ms -> 30.000 fps)
bFrameIntervalType : 0x01 (1 discrete frame interval supported)
adwFrameInterval[1] : 0x00051615 (33.3333 ms -> 30.000 fps)
I think the bandwidth that is actually detected by Windows is more reliable.
I also made a measurement of the ELP-USBFHD01M 1024x768-MJPEG@30fps stream by using a USB sniffer, summing up the payloads over one second from the captured timestamped log. The net sum is 26'170'704 byte or roughly 44% of the USB 2.0 60’000’000 byte per second gross bandwidth. Allowing for some overhead this corresponds nicely to the reported 48%. And this is the actual data stream, not much room for interpretation.
_Mark
Von: ope...@googlegroups.com [mailto:ope...@googlegroups.com] Im Auftrag von Jason von Nieda
Gesendet: Mittwoch, 5. Dezember 2018 16:12
An: ope...@googlegroups.com
Betreff: Re: [OpenPnP] Re: openpnp-capture multiple cameras
Hi Mark,
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/9cd161f6-2440-4b6b-8fb9-0d2035f177e0%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/b69064a1-5efd-414b-b43f-cd9f3bf86d44%40googlegroups.com.
Thanks, will immediately order and then try this hub.
I always thought that the hub being 3.0 does not matter, as it is just marshaling not repackaging data, simply switching the limited USB 2.0 stream/bandwidth through. My USB 3.0 hub (ICY BOX IB-AC6702) does not work.
Strange.
_Mark
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/ac26fa3f-b205-4d6f-832f-a2be92b85cc8%40googlegroups.com.
I think I tried an cheap 2.0 hub one day and it didn't work, so I was surprised by that its working, too...
First I disabled all the unwanted devices from device manager.
Connected all camera and MKS 1.3 to laptop and restarted my laptop.
than started software without connecting any extra devices to laptop. and now both the camera working fine.
I have tested today working openPNP with mentioned 1Mpx(720p) and 5Mpx, both from ELP. Connected to the same controller they work well. CPU load seems be little higher than if the cameras connected separately but not much.
I have also installed python and soft to measure lattency/lagginess of the cameras. I guess you know this.
I have found this 5Mpx camera is very slow and has very long response time reaching 200ms at 4-5fps if resolution is like 1600x1200. Much faster if 640x480 (40ms and 30fps) but no reason to use the camera like that in this mode. So it's not the camera that should be used for fast vision automated system like we do.
The same about this 1080p A4Tech camera. The same lagginess and slow. Not recommended for those whom the vision speed is important parameter.
Issue with auto-exposure occurring at A4Tech is also existing in ELP 5Mpx.
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/df5eae6d-bf52-4ffa-bf6b-7d650624304a%40googlegroups.com.
One thing I found is that I have two USB100W05MT on my machine, not USB100W03M. I'm not sure what the difference is. It may just be a newer version. They have the same sensor and same frame formats.
(...)Jason
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/5064c0fe-b2ba-4a3f-b82a-e443f87c3927%40googlegroups.com.
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/4babc135-8777-45a2-a25f-868a833a7c04%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/72ca9815-7846-4d3a-b781-849258d4fb5f%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/e15b0bd2-fa4e-458e-a68a-e489a857c59c%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/74e3122a-5e0c-4d09-b5b5-8cd146fa17f2%40googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BKNHNyXrHJVGjDtFJTi8XrHpNyqhgSq0QVvcDgMu3tGCQ8TTw%40mail.gmail.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/CA%2BQw0jz%3DpFmkrvLRFz5PbMYt1LqQD%3Do6CiwaoiEiY%2B79Kyhmhw%40mail.gmail.com.
Thanks, will immediately order and then try this hub.
I always thought that the hub being 3.0 does not matter, as it is just marshaling not repackaging data, simply switching the limited USB 2.0 stream/bandwidth through. My USB 3.0 hub (ICY BOX IB-AC6702) does not work.
Strange.
_Mark
Micheal G.> I was always thinking, I can't have two cameras at one port, but now I have tried and can confirm that
@Mark,Hi Mark,Could you pls check again your port report and confirm me whether you really have only two String Descriptors 0 and 1 in the last section "String Descriptors".I have there 9 of them (0-8) and errors visible in descriptors 6-7-8.Take your time.------ String Descriptor 8 ------
bLength : 0x04 (4 bytes)
bDescriptorType : 0x03 (String Descriptor)
Language 0x0409 : "°" *!*ERROR contains 1 NULL character
Data (HexDump) : 04 03 00 00
> Yes I saw it but there are no errors in "String Descriptors" section of your file.
There is another one:
bInterfaceCount : 0x01
*!*ERROR bInterfaceCount must be greater than 1
Are your errors reproducible, each time the same errors? Have you checked cabling? These ELP cables are very thin at the PCB end and not strain relieved, the connectors tiny. Easy to break these and get what in German is aptly called a “Wackelkontakt” (~“wiggle contact” = intermittent contact).
_Mark
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/203b79eb-f293-47cf-a313-ab4e460ad295%40googlegroups.com.
Hi
> Jason wrote:
> As I mentioned, I suspect Windows 7 to be an issue. Microsoft broke MJPEG on at least one occasion so I wouldn't be surprised if they broke it in Windows 7 and then decided not to fix it since that OS is no longer supported.
To exclude the Windows Version as one possible reason, I just tried on a Windows 10 box:
· 2 x USB ELP-USBFHD01M.
· On the same, brand new USB 3.0 hub.
· Latest OpenPNP x64 download (of today), fresh and virgin install.
· Default setup, added the two cameras with FullHD 30fps, deleted the virtual cameras.
· Tried several thing, like restarting OpenPNP etc.
The same result as with Windows 7.
Just one camera at a time works.
Like I said that’s fine with me. It just confirms that MJPEG alone is not a silver bullet in terms of bandwidth limits on USB 2.0
People should either buy the 720p and NOT the 1080p ELP camera as recommended by the OpenPNP Wiki …
… or live with two USB connections to the computer and benefit from better resolution and/or fps à shorter settle times.
_Mark
I had no occasion to connect 2x720. I have tried 720+1200 (my actual setup going to be changed into 720+1080) and it worked, also 2x1080 but not bith ELP.
Interesting is that It didn't work for me still 3 months ago but did ok when I tested it again near 2 weeks ago. While I did not do anything with my computer, no system or java updates, no hardware change. It's like started work better since moment when Jason updated opencv (ok, I remember Jason said it's not connected, but just coincidence).
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/7d2498cd-7529-42a9-a287-4da04164460f%40googlegroups.com.
Hi Jason
I was kinda „hijacking“ someone’s Windows 10 computer, so my time was limited and I just tried one format, as I already established earlier that all the formats use the same bandwidth, including the 1024X768 XGA MJPEG@30fps you suggested to try (then and now) as documented here:
https://groups.google.com/d/msg/openpnp/k7M-Z-P1ivg/b_sPT0WuDQAJ
This makes sense as (M)JPEG compression can also be parametrized. In return the 1024X768 XGA MJPEG format will probably show higher quality/fewer compression artefacts. I assume the camera just consumes the same fixed bandwidth budget (bitrate) in all modes.
A bit like the
option.
I would do it this way if I were to implement a camera :-)
It greatly simplifies the implementation/testing etc., as the transport stream is always the same, even if the user were to switch the format on the fly.
_Mark
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/053701d4a774%246f82b930%244e882b90%24%40makr.zone.
> Success, in my book.
Agree 100%
Von: ope...@googlegroups.com [mailto:ope...@googlegroups.com] Im Auftrag von Jason von Nieda
Gesendet: Dienstag, 8. Januar 2019 18:10
An: ope...@googlegroups.com
Betreff: Re: [OpenPnP] openpnp-capture multiple cameras
Thanks Mark, I had forgotten that you had tested this before. So I guess this camera is just problematic - case closed :)
The FOURCC message may indicate that that camera supports uncompressed RGB, and I don't think we support that mode in the library. We wouldn't want to use it anyway, it would be very high bandwidth.The reticle messages are not important. Just means no reticle preference has been set for that view, so it uses the default.JasonOn Tue, Dec 4, 2018 at 11:25 AM Jason von Nieda <ja...@vonnieda.org> wrote:Hi Marek,Which camera specifically are you referring to with the exposure issue? If you can use sensor numbers or something in this discussion it will help me keep things clear. I am making a list of cameras and the problems or successes people have with them and this will eventually be used to update the Wiki.Alternately, you can use the model number of the camera. For instance, on the ELP the model number is shown on the board. Like this one: http://www.elpcctv.com/720p-usb-camera-module-usb20-omnivision-ov9712-color-sensor-support-yuy-and-mjpeg-with-36mm-lens-p-194.html in the picture and in the item description you see the model number USB100W03M.Thanks,JasonOn Tue, Dec 4, 2018 at 11:15 AM Marek T. <marek.tw...@gmail.com> wrote:Clear now!"Exposure" couldn't be a reason of troubles a half year ago we've seen.But the fact is that everytime I reconnect this camera from USB (but still on the same port) and run Openpnp, the screen is black until check on an auto-exposure. Then get the vision and can set auto-exposure off - in the moment there is no change on the image brightness, it remains "normal".Doing the same on vmcap, the vision is also black but not absolutely total - the most bright objects are visible. When I check a-e image gets normal brightness and when check it off the brightness drops down a little (brightness does not mean slider value of course). Maybe it's only this camera specific but Mark could to check it on his setup.
--
You received this message because you are subscribed to the Google Groups "OpenPnP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to openpnp+u...@googlegroups.com.
To post to this group, send email to ope...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/openpnp/0325bce0-0bc2-4f42-bf95-c8ae3487cec9%40googlegroups.com.