Noisy Analogue

[EoinHo]

Hi,

I have an live analogue video feed that is experiencing various noise artefacts.

My apologies if this is the wrong forum but I could not find anything more suitable.

In the first instance I want to find the correct terms for these bit of noise, then obviously try and find the source of the noise and resolve it. The attached cropped image is a screenshot of part of the noise - showing blue bands that pop up for a frame or two. They run fully horizontally but vary in height. I have to date only noticed them in the bottom third of each frame, though they may not be as noticeable in the upper, lighter portion of the frame.
There is an audio feed with this that has high frequency crackling - I would guess it coincides with the banding.

From my electronics background I suspect physical damage to the conductors: dirty contacts, dry solder joints, water in the coax.

The setup is unfortunately over complicated and inaccessible for the next few months due to nesting birds. An SD analogue camera feeds 60 foot coax which connects to an active splitter (which also has unused converted outputs). The output from this then goes to a fibre optic converter, into an extended run of fibre to another converter bringing it back to analogue. It is then digitised in a Windows driven capture card and streamed in flv to an adobe media server.
The screen capture is unfortunately the streamed output captured over the web. At this time there is no possibility of capturing the image earlier in the chain - though I hope to get on site this coming week.

Does anyone recognise this noise?

[smrpix]

Very likely your diagnosis is correct, intermittent interruption to the physical signal.

If you want to post a short sample <10mb there may be some other telling artifacts. A still of a partial frame can only get you so far.

[EoinHo]

Thanks for your speedy reply.

If I have worked this forum correctly hopefully there is an attached video clip to this reply. It has the blue line noise in the frames and crackling in the audio. I forgot to mention the interference becomes more frequent during as the day progresses, a colleague suspected it was heat related.

[smrpix]

a colleague suspected it was heat related.

Very possible. (And that would fall under the category of intermittent interruption to the physical signal.) Is this a ruggedized camera or is it a consumer camcorder in a box? They aren't really built to run 24/7.

You do have some digital artifacting in your file, but my best guess it is as a result of your weak analog signal. You obviously can't check the physical connections to the camera, but obviously you can troubleshoot the rest of the chain by switching out components.

[EoinHo]

The camera is a professional steerable ball camera designed for outdoor use, though it is aged. The manufacturer reckons it is doing well for its age.

The picture is marred by digital artefacts like pixelation, motion blur, but I believe that is a separate issue to do with the digitisation being screwed down as the on site upload bandwidth is poor. But yes if there is noise in the analogue signal then it stands to reason this may be misinterpreted by the digitisation.

So you would suggest the analogue signal is weak as opposed to being dirty? Or both? I just find it strange and curious we see the same artefacts appearing in the same area of the frame. Why blue?
Is there a term to describe this particular kind of coloured banding noise?

Someone else on site swapped "a cable" which resolved the issue for two days. Which cable and where it was in the chain I am unsure. I suspect it may have been co-incidence that the swap appeared to fix the problem, especially as the noise is intermittent.

[jagabo]

Looks like some lose or broken wires or cables/connectors. Maybe a bad solder joint.

[smrpix]

It's not blue. It looks like typical "rainbow."

[jagabo]

It's not rainbow noise (chroma luma crosstalk). If you look at the video it's obviously a bad connection between the analog camera and the encoder board. From from flexing of the internal wiring over the years.

[smrpix]

jagabo, I was using "rainbow" descriptively -- as opposed to "blue." Didn't mean to confuse the issue. Completely agree it looks like a physical problem.

[EoinHo]

Hi,

Here is the latest offering of the noisy analogue. We are experiencing a heatwave in the UK and the heat is almost certainly making the issue worse.

However in the evening the video noise is different and appearing in the upper part of the frame this time. If there was any other way to technically describe this noise that would be helpful. Also you will note the accompanying soundtrack has crackling - the kind of audio noise I would more associate with 100V audio circuits - or does anyone recognise it as something else?

Many thanks for your help

[jagabo]

Is that MKV file what the camera produced? Or did you re-encode the video? If the latter, it would be much better for you to upload what the camera put out as lossless YUY2 AVI or something similar.

[EoinHo]

Yes this is the encoded stream.

Unfortunately due to site security - protected bird species - it is difficult to get direct access. We are fairly confident the line noise is in the analogue side, before encoding. Elements of pixelation and and motion corruption are from the digitisation / streaming which is a separate issue and one we will look at later in the project.

The key thing I am trying to do is describe the noise and suggestions of typical things that could cause it.