• The forum software that supports hummy.tv has been upgraded to XenForo 2.3!

    Please bear with us as we continue to tweak things, and feel free to post any questions, issues or suggestions in the upgrade thread.

No matching key for decryption

I was under the impression that SD recordings on the Humax were not encrypted by the Humax
Well you're wrong there to start with. Everything is encrypted on recording.
And ask yourself what exactly defines whether a recording is SD or HD, and whether it is the Humax software that is deciding this, or the WebIf etc.
With your modulated feed, the normal 'rules' which off-air stuff mostly complies with don't actually apply.
 
I'm probably wrong but I was under the impression that SD recordings on the Humax were not encrypted by the Humax but this recording is encrypted.
You are indeed wrong, as you would know if you read anything about what you're trying to do instead of diving in head first. Do you not recognise what I meant by "learn to walk before you try to run"? Things Every HDR-FOX Owner Should Know (click), section 5.

" A well known source of the Q variety"
"didn't have this issue before my wife wanted the "Q" version of source box."
Really?
Yes. Many of us are not Sky fanboys, this isn't a Sky forum, and not even a Sky section... so why would we automatically know and why would you want readers to scratch their heads? Why couldn't you have simply said something like "Auto-decrypt doesn't seem to be working for recordings made by broadcasting the HDMI output from a SkyQ box via a DVB-T modulator" – clear, concise, and to the point. It's not SkyQ at issue – that's just obfuscation, the modulator is the key point. However, see next item...

The bottom line is you are trying to defend the indefensible, and demolishing your own credibility in the process. Why write "well known source of the 'Q' variety" when you could have simply and clearly written "SkyQ"?

I've looked at the logs and found that the two logs auto.xxx had no entries since march
Doesn't that sound alarm bells? Can't you work out that if there is no auto-processing activity, nothing will get auto-decrypted (whatever the source)? Fix it, then your "problem" goes away (probably).

barrage of critism from you as it is clear to me that you enjoy doing so
Bollox. I'm just continually frustrated that the hundreds or thousands of hours I have put into documenting this stuff gets totally ignored by the likes of you, who then post a load of crap, are cocky about it, and can't take the resulting criticism. You're insulting the people most likely to be able to help. Primary rules: be clear, use language everyone can understand, don't make unwarranted assumptions. And your spelling could improve too. Newbies' Guide to the Forum (click - but you can't take advice so you won't take that advice either... go on, prove me wrong).

I still don't know where "No matching key for decryption" came from, so far as I can see you have offered no explanation.
 
Last edited:
Thank you for the information.

I've looked at the logs and found that the two logs auto.xxx had no entries since march. the auto log without a "." shows objects being queued for decryption but no other inormation.

Worth a try though thanks.

Cheers,

Bob.

This is looking more like Autodecrypt is not actually running!

Can you look at the Queue page on the webif and post a screen shot!

Can you also change under Auto-processing log level to Debugging information on the Settings page and then try another, short, SkyQ recording and then have another look at Auto.log
 
Tried a manual decrypt using the web interface. This runs smoothly to about 85% and then suddenly jumps to 100% but watching the files via samba I could see that the decryption file was still growing smoothly long after after the web interface said 100%.
When the decrypt file was the same size as the source original it stopped growing but the files did not swap in any way as I would have expected it left both as they were in the directory.
The decrypt file is indeed decrypted as proved when edited by video-redo after transfer to PC via samba.

I can understand the progress indicator might be confused, because the MPEG2 HiDef file will be much bigger than it was expecting.
Thanks BH, that's a much bigger(!) clue.
Standard definition MPEG2 at fixed bitrate best quality, uses 4GB per hour (see the specs for a DVD-R minutes).
Now the DVB-T modulator is encoding HD to MPEG2 - and has the whole mux available for the one input HDMI stream.
We could easily expect TS file sizes exceeding 16GB/hour.
Auto-decrypt (and auto-shrink) fail if there isn't three times the file size in free hard drive space. If the OP doesn't have 50GB available, then one hour won't auto-decrypt (or proportional for whatever his recorded filesize is).
 
Thanks BH, that's a much bigger(!) clue.
Standard definition MPEG2 at fixed bitrate best quality, uses 4GB per hour (see the specs for a DVD-R minutes).
Now the DVB-T modulator is encoding HD to MPEG2 - and has the whole mux available for the one input HDMI stream.
We could easily expect TS file sizes exceeding 16GB/hour.
Auto-decrypt (and auto-shrink) fail if there isn't three times the file size in free hard drive space. If the OP doesn't have 50GB available, then one hour won't auto-decrypt (or proportional for whatever his recorded filesize is).
There have also been timeout issues with the manual webif processes if the host doesn;t respond within the deadline set by the web page
 
I still don't know where "No matching key for decryption" came from, so far as I can see you have offered no explanation.
From /mod/webif/lib/auto/plugin/decrypt/queue.hook:
Code:
      } else {
              log "  Direct decryption" 0

              set key [$ts getkey $mode]
              if {$key eq ""} {
                      return {"FAILED" "No matching key for decryption"}
              }

              ::auto::log "Using key ($key)" 2
              if {[catch {exec /mod/bin/stripts -@ $key $rfile "$tmp/[\
                  file rootname $bfile]" } msg opts]} {
                      ::auto::log "Decrypt error - $msg - $opts"
                      system endop decrypt
                      return {"FAILED" "Decryption failed"}
              }
      }
 
I would be more than happy to make material available but could do with info on how to do this especially baring in mind the size of the files. I can make a 10 minute long file that fails as described but even so, it's still very large.

Auto-decrypt (and auto-shrink) fail if there isn't three times the file size in free hard drive space. If the OP doesn't have 50GB available, then one hour won't auto-decrypt (or proportional for whatever his recorded filesize is).
To satisfy my idle curiosity, could you please go back to the HDR and report the size of file that fails to auto-decrypt and the amount of available hard drive space.
Then I'll back away knowing which tree I was barking up. Cheers ;)
 
" A well known source of the Q variety"
"didn't have this issue before my wife wanted the "Q" version of source box."
Really?
Yes, really. It took me a while to work out what you were probably blathering on about. I didn't bother after that.
I can make a 10 minute long file that fails as described but even so, it's still very large.
So make a 30s recording, that being the minimum the device will do, and upload it (.ts and .hmt) to a file-sharing site of your choice e.g. Dropbox, and tell people where to find it. You will also need to state your encryption key (MAC address and serial number of the box will do, or you can read it off the WebIf).
 
Very good, but I still can't work out what's going on.

If the HDR-FOX is marking it as StDef, then it ought to get treated like an StDef recording regardless. The codecs might get confused if they were set up to expect StDef resolution, but it plays OK so that doesn't seem to be an issue. From an encryption point of view, it's just anonymous data isn't it?

Being marked as StDef should mean the recording isn't protected, and doesn't require unprotecting.
Yes, that is indeed what I thought.

If for some reason the protection flag is set, maybe auto-unprotect isn't being invoked.
Under those circumstances what if any, errors would I see?
It shouldn't be difficult to decrypt the plain TS manually, whether on-box by stripts or on a PC using HFODU, but why isn't auto-decryption working, and why isn't WebIF manual decryption cleaning up? I can understand the progress indicator might be confused, because the MPEG2 HiDef file will be much bigger than it was expecting. Is the problem that everything else is getting confused by these "peculiar" recordings too?
Yes, I'm with you on the thinking there.

No matching key for decryption
How is the title of this thread relevant:
In all other situations where I've reported issues with using software, or asking advice one of the key items requested is the error message I get.
Are you saying the error message isn't a key item in reporting this?

What led you to that conclusion?
What conclusion? I've not concluded anything.

The title coloured my thinking right from the off (and seems to have confused other people as well).
It's the error message I get! Error messages are often used as titles when asking for advice.

If the HDR-FOX encrypted the recording in the first place (which it did, as it does with every recording) then of course it has the matching key for decryption (unless you've done something to change the key in the mean time, or if has forgotten its key until the next reboot).
I've not done anything along those lines that I know of.
You can provide more info to help. The list of installed packages, while trying to be helpful, isn't.
I realised some time ago that nothing I do will ever be right, you clearly enjoy having a go.
At least you can see I tried to give what I thought might be required.
What would be more useful is the WebIF logs. Go to WebIF >> Settings >> Auto-Processing Settings and bump the Auto-processing log level up to "Debugging information". I'm not sure whether a reboot is required to action it, so you might as well. Then you can make another SkyQ transcription recording and see what the auto-decrypt makes of it in WebIF >> Diagnostics >> View Log Files >> auto.log (and maybe other log files will be revealing also).
Thanks for the instructions, I'll have a try at doing that shortly.
Ultimately, as I said before, an upload of an example (unmolested) recording file set will be the gold standard for analysis.
More than willing to do that, would you advise how I should go about it please? I wouldn't expect to attach a video file to post here.

Thank you
Bob.
 
Yes, really. It took me a while to work out what you were probably blathering on about. I didn't bother after that.

So make a 30s recording, that being the minimum the device will do, and upload it (.ts and .hmt) to a file-sharing site of your choice e.g. Dropbox, and tell people where to find it. You will also need to state your encryption key (MAC address and serial number of the box will do, or you can read it off the WebIf).
Before trying that make a short recording from a normal terrestrial channel and check whether auto-decrypt works for that.
Could also try a short SD channel from the SkyQ box
This would identify whether there was a problem with normal auto-decryption and whether it was related to the definition
 
To satisfy my idle curiosity, could you please go back to the HDR and report the size of file that fails to auto-decrypt and the amount of available hard drive space.
Then I'll back away knowing which tree I was barking up. Cheers ;)
Yes of course thank you.

The last recording I did was only 10 minutes long and W10 says it's 1,327,296KB.
Approx 30% of the disc is used and the free sapce is reported as being 603GB.

Bob.
 
Some various points in post order ...
...
Tried a manual decrypt using the web interface. This runs smoothly to about 85% and then suddenly jumps to 100% but watching the files via samba I could see that the decryption file was still growing smoothly long after after the web interface said 100%.
When the decrypt file was the same size as the source original it stopped growing but the files did not swap in any way as I would have expected it left both as they were in the directory.
...
The inaccurate progress completion report is caused by a bug in the WebIf where the progress monitoring JS function is timed out by the web server (IIRC default is s/t like 6 or 10 minutes). It affects any long-running command launched by the WebIf client code (eg, media conversion/extraction). As far as the script logic is concerned the operation was aborted, and the output file doesn't get renamed, but it's actually still running.
...
BTW Qudos for finding a way to archive restricted HD content!
The modulator breaks the protected HDMI chain, at the cost of re-encoding the digital AV stream in MPEG2. A DVB-T USB receiver would allow the Q content to be saved to disk with no encryption.
...
What I don't understand more than anything is why a manual decrypt works and all else fails. Yes, there are issues with the gui saying 100% far too early and the files not swapping at the end the the .decryption file is decrypted?
...
The manual decryption makes you select either DLNA or direct decryption, whereas the queue-based decryption has different logic and tries to select the best method automatically.
Code:
...
              set key [$ts getkey $mode]
              if {$key eq ""} {
                      return {"FAILED" "No matching key for decryption"}
              }
...
As I previously pointed out, this means that no key known to the getkey() method matched the encrypted .ts. If the file isn't corrupt, the testing method must be wrong, but I ran some examples and I still believe it. It does this sort of thing (-q\ just reports 1 or 0 for match or 0, -/ prints a detailed report):
Code:
# stripts -q/ 00000000000000000000000000000000 The\ Taming\ Of\ The\ Shrew_20210718_0909
1
# echo $?
0
#
So this test command either reported an error or its output was not 1, for all the keys tested.
 
Just some thoughts which may or may not help.
Is the SkyQ output same as the previous box? Is it 1080p or is it higher?
Are the settings for the digital modulator still the same? Ie unchanged from the previous SkyHD box.
Are all HD recordings produced on the HDR from the digital modulator missing the HD flag? (Try different channels eg BBC1 HD, ITV HD , Sky1 HD on the SkyQ box).
Are all SD recordings fault free? And do they auto process without issue?

Verify or tweak the output HDMI settings for the SkyQ box.
Verify or tweak the output settings for the digital modulator.
Try some more HD recordings to see if tweaks made any difference.

Finally a side quest (may help diagnosis)
Record test clip1, say, 10 minutes BBC1 HD (101) on the HDR.
Does this auto decrypt ok?
Transfer the decrypted test clip to your laptop.
It should playback fine (eg on VLC) and you should be able to see the media info etc.
Connect HDMI from laptop to digital modulator (in place of the SkyQ source).
Playback the test clip1 and record using your normal SkyQ method (only this time you'll record the test clip1 from laptop via digital modulator to Humax HDR).
Does this test clip2 behave similar to test clip1 in terms of auto processing etc on the HDR?
After decrypt test clip2, transfer to laptop.
Compare file size and media info for the two test clips.

Side quest2 - when you're bored
Replace all SkyQ in this post with 'Q variety'.
Replace all SkyHD in this post with 'H variety'.
Replace all HDR with 'T2 variety'.
Did it make this post better or worst?

_
 
Last edited:
So make a 30s recording, that being the minimum the device will do, and upload it (.ts and .hmt) to a file-sharing site of your choice e.g. Dropbox, and tell people where to find it. You will also need to state your encryption key (MAC address and serial number of the box will do, or you can read it off the WebIf).
I want to do this I really do but I have two issues I need to get past first.
I've been pressing the record button twice and then adjusting the record time down to 10 mins which seems to be the minimum. I've just discovered I can add a 1 min recording to the scheduled list. I can't imagine how you get a 30sec recording.
Secondly, I don't have any on-line storage setup. When I was at work I used theirs but since I retired I've never needed anything like that. Take me a while to find and sign up for something.

Thanks
Bob.
 
The last recording I did was only 10 minutes long and W10 says it's 1,327,296KB.
Approx 30% of the disc is used and the free sapce is reported as being 603GB.
Ten minutes of DVB-T MPEG2 HD 1.3GB
That's approx 8GB per hour.
Freeview HD channels MPEG4 are between 1GB to 2GB per hour (but they are squeezed to the bare minimum to maximise the channel count - including 'stat muxing').

My free space theory is shattered with your 1TB hard drive being only 30% filled :notworthy:.

I can't imagine how you get a 30sec recording.
Start an instant record, monitor a clock with seconds display and press stop - needs to be followed with choosing the recording to stop and confirm via onscreen display.

Just one more question;-how have you set automatic decryption for the recording? I could choose to set it on the MyVideo folder but I've only set it on chosen sub-folders (series).
 
Last edited:
Some various points in post order ...

The inaccurate progress completion report is caused by a bug in the WebIf where the progress monitoring JS function is timed out by the web server (IIRC default is s/t like 6 or 10 minutes). It affects any long-running command launched by the WebIf client code (eg, media conversion/extraction). As far as the script logic is concerned the operation was aborted, and the output file doesn't get renamed, but it's actually still running.
That explains a lot, to me anyway.

The modulator breaks the protected HDMI chain, at the cost of re-encoding the digital AV stream in MPEG2.
Yes, that was my understanding.

A DVB-T USB receiver would allow the Q content to be saved to disk with no encryption.
Really? That hadn't dawned on me if I'm honest but it follows from the previous statement.
Might give that a try, if I can find such a device.
The manual decryption makes you select either DLNA or direct decryption, whereas the queue-based decryption has different logic and tries to select the best method automatically.
I just use the web interface and the opt+ button on the file to decrypt manually. I unaware of making a choice as to DLNA or direct decryption. Which option does my method use?
As I previously pointed out, this means that no key known to the getkey() method matched the encrypted .ts. If the file isn't corrupt, the testing method must be wrong, but I ran some examples and I still believe it.
Interesting thanks.

I need to find out how to do a 30sec recording and find somewhere to store it so the people can see it.

Cheers,

Bob.
 
Before trying that make a short recording from a normal terrestrial channel and check whether auto-decrypt works for that.
Could also try a short SD channel from the SkyQ box
This would identify whether there was a problem with normal auto-decryption and whether it was related to the definition
Just been doing these tests..

A normal recording from a freeview HD channel eg. 114, 101 etc. decrypts automatically just fine.
I've tried two recordings from the SkyQ. One was from BBC2 HD and the other from BBC1 SD and neither decoded and both gave the same error ie. No matching key for decryption.
I am certain where these files were recorded from but they are both 1 minute long but oddly, the SD file is larger than the HD file. I suppose it must be to do with movement or changes in the picture.
The SD file is 70,380KB and the HD is 68,584KB.

Sky Q was in 1080i mode. I've not found any difference in anything changing it to 1080p. I suppose it is only a case of where the de-interlacing is done but I would expect the 1080i file to be smaller, I've not tested that though.

Cheers,

Bob.
 
Back
Top