Gaps in histogram?

[Gary34]

I have gaps on the luma in my Histogram("levels") function and a line that goes up on the right side through the red part and on the blue also.

Also I have 6 lines on one side of my Histogram function (waveform monitor).

Histogram spikes no TBC is with my composite VCR and no TBC in the workflow. I know it's not my VCR and I know it's not my TBC. I am glad about that.

I've seen this https://www.digitalfaq.com/forum/vid...luminence.html but it doesn't really explain everything for me.

I am using the proc amp in my TBC 3000 to control levels except on the one that the TBC is out of the workflow.

[lordsmurf]

Attach the full size in a zip, the forum shrunk the image attachments.

[keaton]

Just a guess, but this reminded me of something I saw years ago when using an ATI 600 USB capture device, which didn't have a proc amp in it. I would see spikes across Vitrualdub's histogram if I didn't use a certain contrast setting in Virtualdub Levels menu. I forget, but it might be that you can only see the histogram in either Overlay or Preview Mode. You may have to switch modes until one of them allows you to see the histogram without capturing.

If you are seeing spikes, then try adjusting Contrast in the Levels menu to see if there's a setting that reduces or eliminates those spikes. I no longer had this problem once i switched to an ATI capture card instead of a USB. With the capture card, it did not introduce any spikes when changing Levels. However, i usually leave them at default 128 and used an external proc amp to adjust brightness/contrast.

[Gary34]

Here's my script crop(0,8,704,0).coloryuv(analyze=true).converttoyv 16(matrix=“rec601”,interlaced=true).Histogram(“Col or2”).Histogram
Histogram(“Levels”,3)

[Feedbucket]

Your histograms are processing the output from each prior step. The 6 lines in the waveform are coming from the ColorYUV output and the lines in the levels graph are coming from the waveform Histogram.

Maybe the proc amp is breaking up the overall waveform? If you're doing something like stretching the Y.

[aramkolt]

No expert here, but I always thought that histogram spikes or persistent dots on say a vectorscope were due to a large percentage of pixels being assigned very specific values, which can be indicative of chroma noise or something else rounding a lot values to the same value which could be something like a proc amp, or a passthrough device that does analog to digital and then back to analog.

I think knowing these things should point you to the source of the issue:

Do you get the same spikes if the line TBC and any digital processing is turned off on the VCR?

Do you get the same spikes using a different cable?

Do you get the same spikes when capturing using S-Video instead of composite?

Do you get the same spikes when using a different VCR?

Do you get the same spikes when using a different capture card?

Do you get the same spikes when playing back a commercially made tape?

Do you get the same spikes when using a different capture program like AmaRecTV?

[Gary34]

That helped. I’ll get back to it later.

[lordsmurf]

Errors seen in an analysis tool don't always mean errors on the item/stuff being analyzed, but rather a flaw in the analysis tool itself. When it comes to missing data from histograms, this has often been why.

[Gary34]

I have my old premier pro uncompressed intensity shuttle AVIs done with the same Pannasonic 1980p ag. I have captures with and without the frame and line TBC. With my regular VCR and with the 1980.

The line through my V channel and partly into my U channel and the 6 lines on the waveform monitor and the lines on the left of my y channel are definitely from Avisynth. I get them in Pmod and hybrid no matter what combination of hardware or capture software I use. I don’t get that in the Vapoursynth waveform monitor in hybrid.

[Gary34]

So i'm trying not to jump to conclusions here but I think it's my capture card. I made the mistake of getting my Pinnacle from eBay. I think that's the problem. I bought it unopened. I broke the seal on it. I think it was a bad card and i'm going to get one off of LS. I might be wrong. It could be something else. IDK. It's really not worth the trouble of all of this trouble shooting just to save a few bucks on a card.

I looked back at my old Intensity Shuttle captures and I recaptured the same tape with the Pinnacle and no TBC and also no adjustments on my capture card. Before I came to this site the guy I followed when I was using the shuttle said the contrast and brightness controls are only a concern if you are creating content for broadcasting over the airwaves. I know it's ridiculous. Anyways this is all clipping like crazy.

I used two different cables to make sure it wasn't the cables. I captured Tekken to make sure it wasn't the tapes. I can't eliminate the VCR as a cause because there's years between these captures. The intensity ones were right after it got refurbished. My Pinnacle works easily with Vdub but I can't get it to work with another software so I can't eliminate that as a cause. I also have looked through these captures and I don't notice any kind of banding. It's just the histograms aren't smooth but I can't see anything wrong with it so maybe it's just the graphs. Maybe it's not banding. Anyways I don't like the uncertainty so i'm getting rid of it.

My guess is that the card is the problem. It can't be my frame TBC because it does it with or without it. same thing with the line TBC. I don't know that. it just seems like the most likely thing.

All the other oddities I was getting are only in my Pmod they aren't in Hybrid.