Hi.
I have a Panasonic nv-hs 1000, connected to Datavideo tbc-1000.
As a capture device, I have:
- Canopus advc 110
- VC500 (older model)
I will be capturing PAL, and since I have about 70 home tapes, I will
- Rather not be doing manual restorations / corrections per tape
- No editing, besides cropping bad recordings
- I will probably want to save original "semi-archive" quality just in case
- I will want a copy which is non-interlaced for watching on PC.
I started comparing the two capture devices, and I have a tendency to like the ADVC workflow, however I'm not sure if it's results are better.
Also, I got very mixed up in terms of how to set capture levels:
- On VC500 (colors are too bright, but if I don't, they show up in "red" in virtualdub)
- On ADVC colors just seem a bit saturated, but I don't know if it's even possible to control saturation.
I don't think either version is acceptable. There's no excuse for ruining a capture by not controlling input levels on the VC500 caps and blowing away brights. Even the black levels are too high. Plenty of info exists on how to do it with VirtualDub, and I've done it myself with that card.
The DV versions have the opposite problem with badly crushed darks and overall murky output. They have a characteristic analog-to-DV look. They're on the verge of appearing denuded due to missing inner detail removed by DV compression, and the color looks a little cooked. Color on all of them, really, is seriously haywire.
Your line tbc is doing a poor job -- look at all the noisy buzz and mice teeth on edges. Was the line tbc turned on? These look as if they play with a lot of annoying noise. Ugly DCT ringing and sharpening halos are all too obvious. Why would you want any of these immortalized in binary format?
Quote:
Originally Posted by ido2
I will be capturing PAL, and since I have about 70 home tapes, I will
- Rather not be doing manual restorations / corrections per tape
- No editing, besides cropping bad recordings
That's a shame. but so be it. Why do you need a tech forum about video quality to "help" you do nothing with low-quality captures? Anybody can make bad captures. No skill involved.
Quote:
Originally Posted by ido2
- I will probably want to save original "semi-archive" quality just in case
- I will want a copy which is non-interlaced for watching on PC.
A "copy"? You can't copy as deinterlaced. You have to deinterlace and re-encode. Considering the poor quality, it seems like a waste of time. It will only look worse, since even QTGMC costs something when deinterlacing, but I doubt you'd get as far as using QTGMC. Perhaps it would be easier to find a PC media player that deinterlaces as it plays. Most of them do. I like the "semi-archive" phrase -- a neat euphemism for not knowing how much damage you're looking at.
Quote:
Originally Posted by ido2
I started comparing the two capture devices, and I have a tendency to like the ADVC workflow, however I'm not sure if it's results are better.
Yeah, it does look a little "different" after DV compression throws away a lot of the original data and many objects can't even be seen under the crushed blacks. It's the sort of bad video that most forums teach users to either improve or avoid.
Seriously, if you'd like help on getting better captures and improving the results with a little cleanup, many here can offer plenty of tips. But you appear to be asking us to approve of poor practice that doesn't even achieve basic standards.
These images, however, are useful and informative demos of typically bad VHS transfer. Surely you want better.
I'm very happy you replied to my question.
My responses to your notes:
Quote:
Originally Posted by sanlyn
Makes me sad to have to say this, but....
I don't think either version is acceptable. There's no excuse for ruining a capture by not controlling input levels on the VC500 caps and blowing away brights. Even the black levels are too high. Plenty of info exists on how to do it with VirtualDub, and I've done it myself with that card.
A. What would be the correct way to calibrate levels in such that:
1. Would not have to be done per cassette;
2. I tried manually controlling contrast/brightness in such that it would not fill the red section in virtualdub levels histogram, however, this is how I ended up with what I attached.
Quote:
Originally Posted by sanlyn
The DV versions have the opposite problem with badly crushed darks and overall murky output. They have a characteristic analog-to-DV look. They're on the verge of appearing denuded due to missing inner detail removed by DV compression, and the color looks a little cooked. Color on all of them, really, is seriously haywire.
B. I have no idea if it's even possible to control levels on the device itself (I'm not talking about post-processing). Is this even possible ? Or should I throw the Canopus to the garbage and use only the VC500 ?
Quote:
Originally Posted by sanlyn
Your line tbc is doing a poor job -- look at all the noisy buzz and mice teeth on edges. Was the line tbc turned on? These look as if they play with a lot of annoying noise. Ugly DCT ringing and sharpening halos are all too obvious. Why would you want any of these immortalized in binary format?
C. Both TBC's were turned on. What should I check / change to fix this ?
Quote:
Originally Posted by sanlyn
Yeah, it does look a little "different" after DV compression throws away a lot of the original data and many objects can't even be seen under the crushed blacks. It's the sort of bad video that most forums teach users to either improve or avoid.
I guess again this means throwing the Canopus..
Quote:
Originally Posted by sanlyn
Seriously, if you'd like help on getting better captures and improving the results with a little cleanup, many here can offer plenty of tips. But you appear to be asking us to approve of poor practice that doesn't even achieve basic standards.
These images, however, are useful and informative demos of typically bad VHS transfer. Surely you want better.
My point is, I'd like to define a streamline workflow, in which I can transfer the videos to digital form, with small effort (ie. I don't have time to process each cassette's defects, willing to sacrifice in quality), hoping good gear and a smart flow would be good enough.
This is exactly why I am here looking for feedback on how I should improve my workflow;
This is also why I purchased the VC500 (I actually bought the Canopus long ago).
I've read countless threads, and this is how I ended up with the gear and thought the workflow would be simpler, but without posting, it's hard to understand if the output from my gear is workable or not, and what should be changed.
Looking over your images, the brighter VC500 shows that when making your histogram readings before capture, you apparently didn't crop away black borders or head switching noise. The black borders in your VC500 images have values of RGB30 to RGB45. which aren't even close to black. Gamma is too high and your midtones are wiped out.
Remember that black borders or bottom-border head-switching noise affects the histogram. These can be temporarily masked with the "Cropping..." dialog (see next post) during capture setup
Cropping can be used to mask off areas that affect the levels histogram, such as black or white borders and head-switching noise. You can crop left, right, top, and bottom.
The scenes themselves also look as if darks are underexposed on tape. Some of the lighting in your images indicate that the luma values have a wider range of dark values than digital video can accept. So, after you crop off unwanted areas when reading the histogram,you'll simply have to let a small amount of overrun go into the left-hand red area.
Slight overruns into unsafe areas can be corrected in post-processing and some detail can be rescued. But you don't want to do any post-capture cleanup, so you'll have to live with what you get. I'd suggest that you always use the lower quality DV captures. DV is destroying detail and wiping out fine textures, but you seem to prefer lower detail and lots of compression noise, so I guess you'll have to live with it. It's absurd to think that you can use the same level settings for all of your tapes. Why you chose to use a capture card and lossless capture with no intention of repairing obvious and extensive VHS defects is beyond me anyway. For the tapes I don't see why you didn't just use a DVD recorder. At least you could play your videos on something besides a PC. DV is PC-only -- no TV playback, no external media players, no web posting. Windows doesn't even come with a DV codec any more, the codec is in the media players.
The images tell us that your VCR's line tbc is pretty weak, giving you lots of noisy ragged edges, and your player is imposing ugly DCT ringing and edge halos by oversharpening the image. These effects are obvious in all your images.
What you're asking, basically, is how to do nothing with your tape captures. That's what you're already doing.
So first of all, I'd need to calibrate the capture cards;
I've found out that the advc can generate colorbars which can probably be used for calibration, so I've done the following tests:
A. Captured bars from advc out to vc500 in (attachment colorbar_ire0.avi)
B. Recorded bars onto tape:
B.1. Captured bars using advc
B.2. Captured bars using vc500
* I've set the jumper on the advc to PAL + IRE0, I understand this should be set for PAL.
** I used svideo everwhere
*** When using VCR as passthrough (ie. connecting advc out to VCR and then VC500, image looks almost the same as it looks when connecting in (A) above.
Capture (A)'s output looks good to me (is it ?) ;
I assume levels will not be correct for recordings from VCR tape (am I right?) but I did this to put levels on VC500 in a known working ballpark. (attachment bars_nre0_vc500_through_vid.png).
I also attached (vc500direct_histogram.png; vc500direct_waveform.png) chroma levels for reference.
The first capture I made from tape look horrid.
https://takeafile.com/?f=sakubetuje (Sorry for the large file; vc500_recorded_is_this_bad_tape_or_bad_vcr.avi)
The recording was made onto a very very old VHS tape. Just for my knowledgfe, is the playback quality like this because of the tape deteriorated ? Or a problem with the recording heads on the VCR ?
The "better", but still bad tape, output capture recordings: bars_advc_from_tape.avi and vc500_recorded_is_this_bad_tape_or_bad_vcr.avi.
The levels look very different from the capture made in (A), but I don't know if this is expected or just very very bad.
What should be my next step ?
- Is the VCR bad ?
- Are the tapes bad and not good for calibration ? (other tapes did not show these artifacts).
- Should I try record avdc output on a different VCR/TAPE and use that for calibration of chroma ?
Thanks for all that. But you don't understand what's happening and you're wasting your ti8me. You re-adjust the input signal for each tape, not for standard colorbars. What works for the colorbars or a recording on one tape won't work for another tape. TYou also fail to understand that once clipping occurs during a recording or capture, it can't be corrected later.
I better understand what you are saying, However, my work could still provide useful information on (I think I'm correct by saying these):
A. Checking that VC500 does not crop extreme blacks/whites
B. Understand why ADVC is saturating colors (by recording something it has produced)
C. Also, I'd still like to hear what's on your mind regarding the bad tape recording. I need to know if my VCR is bad.
Would you mind looking at the capture and commenting on the above points ?
Regardless, could you please point me to a step-by-step procedure of how chroma should be calibrated (per-tape), I haven't found one which is clear to me.
I am misunderstanding something important here obviously:
A. I understand now that each tape is different, but if recorded from same camera, doesn't it mean the chroma should be more or less the same ?
- Otherwise doesn't it mean that colors on different tapes should be different (so if I calibrate them to min-max, chroma would have more spread out, but would be offset than what it was supposed to be ?)
B. I am still missing how it can be calibrated per tape, as different scenes in throughout the tape (or even the same scene) will have different extreme readings. back to the question - there is not 'black'/'white' reference it can be calibrated to.
I better understand what you are saying, However, my work could still provide useful information on (I think I'm correct by saying these):
A. Checking that VC500 does not crop extreme blacks/whites
B. Understand why ADVC is saturating colors (by recording something it has produced)
It's usually the player that oversaturates, but consumer DV compression does tend to make subtle but visible changes in color purity and gives a lot of "hot" red that's difficult to correct later..
Quote:
Originally Posted by ido2
C. Also, I'd still like to hear what's on your mind regarding the bad tape recording. I need to know if my VCR is bad.
That looks like a multi-generation tape that was recorded or played without a line tbc.
Quote:
Originally Posted by ido2
Regardless, could you please point me to a step-by-step procedure of how chroma should be calibrated (per-tape), I haven't found one which is clear to me.
On a t5ape by tape bnasis, that isn't possible. Trying to correct analog color during capture is an exercise in utter futility. Analog color, especially VHS, changes hue and balance every few minutes and often scene-by-scene. The secret to correcting it is what everyone uses after they get a decent capture: post-processing.
Quote:
Originally Posted by ido2
I am misunderstanding something important here obviously:
A. I understand now that each tape is different, but if recorded from same camera, doesn't it mean the chroma should be more or less the same ?
No.Color accuracy is largely determined by exposure and by the nature of the lighting. If you have a camera that used an auto-white balance feature, the color balance changes with changes in the predominant color of objects in the frame. It is very difficult to correct in post6 processing, and often impossible.
Quote:
Originally Posted by ido2
- Otherwise doesn't it mean that colors on different tapes should be different (so if I calibrate them to min-max, chroma would have more spread out, but would be offset than what it was supposed to be ?)
What you're missing is that calibrating for a color bars recording will give you perfect colorbars if done properly. It has nothing to do with the color changes in tapes.
Quote:
Originally Posted by ido2
B. I am still missing how it can be calibrated per tape, as different scenes in throughout the tape (or even the same scene) will have different extreme readings. back to the question - there is not 'black'/'white' reference it can be calibrated to.
Look, it's pretty simple. Play some section of tape and adjust the histogram as well as possible during a few short samples of tape segments. Do you want to make it perfect for the entire tape? You can't. Diversions from the safe zones you've set up have to be cleaned up in post processing. Since you want quickie push-button results and refuse to go through the repair\correction/restoration phase where it's required, you're stuck with what you've captured.
