146 replies to this topic

[OldDeuteronomy]

I thought it might make sense to open a common thread for all kind of bugs that we few ET9000 owners stumble upon. Maybe some or most of these things need to be fixed in the drivers, but I think there are also things that can be fixed from your side.

I won't list any softcam-related issues here, because I think that these kind of issues need to be fixed by the softcam makers.


[list]
[*]GUI transparency slider is missing. Applies to both OE1.0 and 2.0.
[*]"Mute" does not mute audio (at least not via HDMI, haven't tried SPDIF - will most likely work fine via all kind of analogue audio outputs). Applies to OE1.0 and 2.0.
[*]The "Information->About" screen states the "image version" as "unavailable" and lists the tuners as "AVL2108". This is for OE2.0. In OE1.0, the image version was displayed correctly, and the tuners was showed correctly as "BCM4xxx" (don't recall the exact model number).
[*]Autoresolution Plugin can be installed but causes a GS once it's being enabled. Applies to OE1.0 and 2.0.
[*]Some keys of the original remote control are not working as they should, most likely because they are not present on a DMM remote control and therefore not supported in your image. I haven't tested if these keys were working in the original image (which is also an OpenPLi built from October 2010). These keys are: Filelist, Playlist, Timeshift, Timer, Mark, PiP. However, the appropriate functions can still be used by using the DMM alternatives (for instance Yellow for timeshift, 2xBlue for PiP and so on).
[*]PiP is not working for HD channels. If enabled while watching a HD channel (no matter if scrambled or not), the PiP window appears but remains black. With a SD channel however, PiP works fine. Applies to OE1.0 and 2.0. Could this be a HW limitation? Or driver issue?[/list]I will update this list as I find new issues. Those few (at the moment at least ) other ET9000 users are welcome to append their findings to this thread.

One question to the OE team: I regulary check http://openpli.git.s...itweb-index.cgi for changes. However, it seems that not all changes are listed there. For instance, I believe that today's update on the ET9000 included a new driver (wasn't fast enough to realize everything on the screen ), but I can not find anything about this in any of the sections on the site mentioned above. Do I need to look elsewhere, or is it just that not everything finds its way into the changelog?


Many thanks in advance for your support!

[hemertje]

Mute does work via SPDIF

[malakudi]

Most of the above mentioned bugs are driver realted.

Newer drivers from 16/12 report tuners as Avalink, older drivers reported tuners as BCM4506. Maybe Avalink is the correct name.

PiP only works with 2 SD or 1 HD and one SD channel (HD to the primary main video decoder only). This is by design, accoring to capabilities of BCM4705. PiP doesn't work at all with latest drivers from 16/12 (only applicable to OpenPLI 2.0 because they are built with gcc 4.4)

Autoresolution crashes because drivers don't export /proc/stb/vmpeg/deinterlace* controls.

[MiLo]

This is what I already reported to clarke-tech so far:

- Attempting DVD playback (from file structure on disk) will completely hang enigma, no chance of recovery other than killing it.

- The loopthough only functions when the device is supplying power to the LNB. There's no signal to the next box in the chain when the et9000is in standby. This renders the second loopthrough port useless in mostcases, because one has to control both the et9000 and the "slave" box so that they tune to the same satellite, polarization and band in order toget a signal on the slave.

- On first boot, there's no signal on the composite or scart connector. One has to wait for the E2 wizzard to start switching outputs.

- Letterbox and Pan&scan options are "reversed", when set to "letterbox"in Enigma, the display will be pan&scan, and vice versa.

- No combined composite + HDMI output. The combination is very practicalin combination with a video transmitter and such.

- SCART loopthrough doesn't work.

- Analog audio is very, very soft (at ~1% of normal levels orso).

- When I leave an 8GB USB stick in the device, it fails to boot. Itdoesn't do that with all sticks though. I've attached the serial portoutput when the box halts during boot.

[OldDeuteronomy]

Thanks for that information. Do you have any past-experience with CT, and if yes, would you say they are, well..., helpful?

BTW, just found another bug, which is, unlike the ones from my first post, really annoying. I'm using a media player as streaming client in my office room. The player is Internet-TV-capable, so I can use it to stream every TV station from the receiver using adresses like hxxp://192.168.1.10/web/stream.m3u?ref=1:0:19:EF11:421:1:C00000:0:0:0:

I am using this for quite a while now, so far with a VU+, so I know for sure that the issue can only be with the ET9000. The issue is that the streams sometimes work forever without any problems, and some other times it's being interrupted after a few minutes. I can then simply restart the stream, and if I'm lucky it will work "forever", or it will drop again after 10 minutes or so. It happens with HD and SD channels, scrambled or FTA. On the TV which is hooked up to the receiver, there are no dropouts at all, so it only affects the streams.

Is there any way to find out what causes this issue? I have seen this with the latest OE1.0 build as well as with the 2.0 build from today.

[malakudi]

New drivers released (26/12)

20101226
-OSD alpha setting added
-Allow simultaneous hdmi/scart output ('simul_hdmi_scart' module option, default enabled)
-Scart output fixes
-RCA audio fixes
-Autoresolution handling improvements
-I2c adapter improvements

[malakudi]

@Milo: Does PiP work for you with gcc 4.4 drivers and OpenPLI 2.0 beta? PiP only worked for me with last gcc 4.1.1 drivers on OpenPLI 1.0 beta.

[OldDeuteronomy]

@Milo: Well, after reading what you've told them and what they fixed, I would say they ARE helpful, aren't they?

Will try the new drivers immediately. Thanks malakudi for the headsup!


edit: Where can I find the new driver?

[swiffer]

good idea this subject.

I am a newcomer in linuxbox world. I bought the ET9000 from clarke-tech because it seems good from white paper & internet informations I read. And also, because I am the owner of a CT5000HD since 4 years without any problems.
I am confident that clarke-tech will do its best concerning the drivers.

I use OpenPli v1.0 20101712 to understand this Enigma new world.
With this version :
- BCM4509 is reported within the tuner configuration menu
- mute works well for HDMI
- I confirm for the keys

- I did not succeed to work (or configured correctly) with only one satellite dish and the little cable given (I guess to set it on the output of LNBA to input of the LNBB)?!? As soon as a record is active, I can zap on another channel of the same tp, but as soon as I go on another TP, when it tries to used the second tuner, I got an error message
- I got some strange behaviour before I installed an internal HDD. The menu "Hard disk drive" of the Magic skin was not listed. When I selected the default skin, the menu was there. I can by this way format an external USB disk (ext2). But a lot of error & crash when I tried to setup to records...

These are my current problems

But from now, its current state allows me to use it daily. So it is not a so bad result for a young receiver as this one.

[swiffer]

I saw in software management menu that I can save and restore my current settings on the HDD.
Does it means that If I set a new image (for instance OpenPli v1.0 xxxxxx), I can restore my settings from v1.0 yyyyy ?
Does this work also between a v1.0 and v2.0 ?

If it is the case, I will test as soon as possible the new versions to participate to the debugging effort of the v2.0.

If not, how sould I do to save my current v1.0 state to switch easily between the two versions.

[OldDeuteronomy]

Found the new drivers now and attached them here. Can confirm that GUI transparency setting and Autoresolution PlugIn now work fine. I'm quite impressed that they fixed it over XMas.

Attached Files

•  et9000-drivers-gcc44-20101226.zip   3.73MB   1 downloads

[malakudi]

http://www.et-view.c...ownload/drivers , as it is mentoned in et9000-dvb-modules.bb

[swiffer]

If I copy them in /lib/modules/2.6.18-7.4/extra, will these new drivers work with my v1.0 current version ?

[malakudi]

If I copy them in /lib/modules/2.6.18-7.4/extra, will these new drivers work with my v1.0 current version ?

No, these are only for OpenPLI 2.0

[OldDeuteronomy]

OK, now this is a real issue: All of my recordings are incomplete! No matter from which channel or if it's a timer or manual recording - none of my tries led to a complete recording. I noticed that a recording from 3 days ago (when I still had the last OE1.0 image installed) was incomplete as well, so the issue is not related to OE1.0 or 2.0. Things I tried to far: Reformatting the HDD, replacing the HDD. No change.

Since it wouldn't surprise me if others did not really tested recordings so far: Can anyone with a ET9000 confirm that recordings work fine for him?

Any suggestions to solve the problem are of course highly appretiated.

[swiffer]

what do you mean by incomplete ?
Since now, I do some records (but only few minutes) and all work fine. I will set some long records this morning, to see what happens.

[swiffer]

perhaps you can precise too, which kind of HDD (USB, internal Sata) and which format used ?

[jedeje]

I just checked and experience the same behavior with (Timer) recordings, the recent ones are all incomplete. I record onto an external USB 2.0 HD running the 2.0 BETA from December 19th.

My older recordings done with 1.0 images build before the 13th are OK. New(er) driver issue?

Gr
JeDeJe

[OldDeuteronomy]

First of all, I'm really glad that I'm not the only one with this problem. If I would have been, I would probably had return the box. But now, knowing that it's a common bug, I'm quite confident that it will be fixed. If it wouldn't, I could still return it because it's faulty, even after our 14-day-by-law return right.

OK, some more details. I'm using an internal SATA drive. I have tested with 2 different drives so far, a 2.5" 320GB WD (which is my default drive that I also already used in the VU+), and a Samsung 128MB SSD, which also works fine in the VU+. Reformatted both drives (in the box). Since jedeje sees the same issue with an external USB disk, the HDD can be closed out from being responsible.

I am currently using the OE2.0 with the online-update from today morning and the new drivers from yesterday. I have just tested another recording (manually via REC key), which stopped after approx. 30 minutes. The first recording I did was last thuresday. At this point, I still used the OE1.0 build from 17/12, so for me the issue was also present with OE1.0 and with the older drivers.

One interesting thing that I noticed yesterday with one of the interrupted recordings: When I accessed the ET9000 HDD via Windows Explorer, the size of the .ts file was 180MB, and when I streamed this recording via Webinterface, it ran for approx 90 seconds, which would mache those 180MB. But then I connected to the box via FTP, and in the FTP browser, the very same .ts was displayed with about 4GB. I could download this 4GB file, and when I played it back on the computer, it ran for about 45 minutes, which again matched with the 4GB filesize. I just don't understand why the same file is different in size depending on the way how I access the box. If the recording would have been complete, it would have run for 90 minutes, so the filesize would have been around 8GB.

There is no "system" in how long a recording runs before it interrupts. I had recordings that stopped after 5 minutes, other run for one hour or longer.

Next thing I will try is to do a recording via LAN to my server's drive.

[MiLo]

@Milo: Does PiP work for you with gcc 4.4 drivers and OpenPLI 2.0 beta?

Yes, but only for SD channels. Appears to be a hardware limitation.