It has been awhile. Looks like the last time I posted here was around January of 2012...little over 1 year ago. Wow!
Last year, I set myself to start transferring my home movies, but then became overwhelmed by how involved a process it is: equipment needed to do a good job (which is heavily emphasized by the pro's here). And so I put it off until I was ready to try and tackle this once again. Never really made real use of being here since then.
My ultimate goal was to archive my VHS tapes to the point where IF needed I could throw the tapes away with little worry that I'd be missing out on the actual real tape. Things always can happen in your life...you move, do something different, don't have room to keep boxes full of VHS tapes. And worst of all, something bad could ruin them where they are then lost forever!
Since then, I've simply popped in now and then to read forums and acquire the hardware I need And, I worked hard at getting the things I wanted. Unfortunately, this has taken some time! But, I want to do a good job so it's important to me. I've had to force myself to be patient!
I'll do my best to keep things as simple as possible, but at the same time ask the tough questions I feel I need to ask to accomplish my goals. So please bare with me! Otherwise, over the last few months I have become a fan of other members (aside from the PRO's) who have shared their hardships and project challenges in transferring their own home movies.
For example, I'm a fan of NJRoadfan and the thread he started:
Other members such as robjv1 and plenty more that have made this such a great place to learn about how to transfer VHS correctly. They share the same struggles I do.
Of course loardsmurf, kpmedia, JPM, admin, and all of the PRO's are what make this place solid! They are our tree trunk, while we are their branches. So keep up the good work!
I guess I've digressed...hehehehe...sorry!
In starting this up again, I think right now I want to start with preparing my VCR and capture card ONLY to start with. To me, this keeps it simple: two devices. Later I can see how other things such as TBC or proc amps change my setup.
I've been going a little crazy about the test patterns and proper calibration techniques. Seems to be a lot of ways to do different things. And, from reading past forums, I've gathered test patterns don't do a whole lot since the source is the ultimate master. If the source is off (ex: like how the original VHS was recorded), then all bets are off right?
So keeping things simple: I have a VCR and capture card (in this case ATI USB 600). My goal here would first be to see how I can prove to myself that the video signals going from the VCR to my capture card are correct in terms of luminance and color? I suppose this is where test patterns come in. Test patterns, from what I gather, help to determine if the hardware is doing what it should (is there noise, is the video blury, too sharp, is the color off, ...etc).
My first question is, what are the test patterns that I should run through my VCR to my capture card to prove that the setup is correct and calibrated? SMPTE NTSC? Maybe SNELL & WILCOX as was mentioned in this interesting forum: http://www.digitalfaq.com/forum/vide...i-600-usb.html. lordsmurf recommends DVE (Digital Video Essentials) or Avid (I've not looked into this one). Yet, jmac698 says to avoid it (I've come to respect his guidance): http://www.digitalfaq.com/forum/vide...html#post14561. Sometimes it gets a bit confusing, so I need to find some clarity here.
I'd like to nail that down. Make sure the path between my VCR and Capture card is GOOD before I introduce other devices and original VHS home movie sources. So I feel I'm at the calibration stage, see if their is any noise in my transfer path, see if I have a bad capture card, ...etc.
Not sure if that's an OK approach? But, I have to start somewhere!
My first question is, what are the test patterns that I should run through my VCR to my capture card to prove that the setup is correct and calibrated?
You can't really. What you would need is test patterns on each tape that you play back (which is how the pros do it) to either ensure that everything is correct, or make adjustments based on the variances that appear. And the real way to do it isn't by looking at it but by measuring the values.
You can do like some of us have done and play back DVDs to check out the performance of the capture system relative to a known starting point, but that doesn't tell you anything about your VCR's levels or the tape's.
Maybe you could also record the test patterns onto a tape and play those back, but that isn't going to tell you very much.
By SMPTE NTSC I will assume you mean the SMPTE version of color bars. Snell & Wilcox isn't the name of a particular test pattern but rather a company (who makes pro devices and tests). Digital Video Essentials carries both SMPTE color bars and the Snell & Wilcox comb filter patterns you see posted about here.
One neat thing about DVE is that the HD DVD version is a flipper: the HD version on side 1 and the older SD version on side 2. Possibly can be found for cheaper than the regular DVD version since the death of the format.
So I thought I would experiment and give this a try today:
ADVC300 Test Pattern -> ATI 600 USB
The ADVC300 is capable of displaying a test pattern from the device itself. My guess, though I could be wrong, is that the test patterns should output the NTSC color bars pretty much to spec? If this is true, then I would imagine I could calibrate my capture card to this? Correct? If not correct, then I'm not sure why Grass Valley would have built in a test pattern into the device if it didn't have some purpose.
Here are the results of my captures:
(Composite):
ADVC300_COMPOSITE_TEST_PATTERN_FEW_FRAMES.avi:
- This is using ADVC300 composite out to ATI 600 USB composite in.
ADVC300_COMPOSITE_TEST_PATTERN_WAVEFORM.jpg:
- This is a waveform of the composite capture.
(S-VIDEO):
ADVC300_SVIDEO_TEST_PATTERN_FEW_FRAMES.avi:
- This is using ADVC300 svideo out to ATI 600 USB svideo in.
ADVC300_SVIDEO_TEST_PATTERN_WAVEFORM.jpg:
- This is a waveform of the svideo capture.
A few comments:
If you compare the waveforms, you'll see the colors captured between S-Video and composite are at the same levels. I would say that seems to be consistent at least. It would have unnerved me if the colors were off between using S-Video or composite. So I am happy to see that if I use S-Video or composite, I get the same colors at the same levels in the waveform.
Again, if you compare waveforms, you'll notice there is more spread per each color for S-Video. This was actually disappointing to see. When you watch the SVIDEO test pattern, you'll see it seems to be much more blurry when compare to the composite capture. I'm not sure why this is? To rectify this, I tried a different S-Video cable and it provided the same result.
My ATI 600 USB's ProcAmp settings that I can set in VirtualDub are set to their defaults for the filter. I'm not sure if I would need to adjust the ProcAmp on the ATI 600 USB filter to compensate in some way for the poor S-Video quality. Or, perhaps I have another problem.
In the wave forms, you'll notice additional white lines. I superimposed these to help show where the correct colors should be. For example, the thin white lines at 0% and 100% is RGB 0 and RGB 256 respectfully. The other white lines show NTSC broadcast level lines which are RGB 16 to RGB 235. Then I have the thin white lines for the correct colors at the various steps.
You'll notice the colors are below the correct color levels. Looks like the black is getting clipped at the bottom. And the white, doesn't exactly reach the 235 mark.
Does this tell me anything though? If the ADVC300 is generating a good test pattern and outputting it correctly, then it would seem my ATI 600 USB's ProcAmp's is off?
What may I be doing wrong here or where is my understanding about something off?
The ATI 600 USB issue with colors in S-Video was brought up in one of the threads you linked in your OP and more-or-less confirmed. But thanks for the further evidence. Disappointing indeed.
One thing you definitely want to do with the Proc Amp settings is turn the sharpness down to 0. The other settings are source-dependent but that one, never budge from 0.
The ATI 600 USB issue with colors in S-Video was brought up in one of the threads you linked in your OP and more-or-less confirmed. But thanks for the further evidence. Disappointing indeed.
One thing you definitely want to do with the Proc Amp settings is turn the sharpness down to 0. The other settings are source-dependent but that one, never budge from 0.
Wow, that is truly disappointing. I do have an additional ATI 600 USB (as well as a non ATI USB capture card/stick). I should perhaps unplug this one and put my second ATI 600 USB in and re-run my same exact capture again? I think I may do that.
What's the point of using the ATI 600 USB if the colors and sharpness is off for S-Video but descent for composite? It seems that the composite looks better than the S-Video. To me the whole point of using S-video over composite is because I would expect it would be a bit better. If anything, not worse than composite.
This is why I am trying to calibrate my stuff as well as possible....to see if my hardware is doing what it should or see where flaws may exist and try to understand them. And, if I didn't attempt to try to run some test patterns, I may have never noticed this issue until today.
However, is the way I am going about this OK or pretty sound? I am assuming the ADVC300 is generating a correct test pattern. And, the test pattern most likely would be correct.
With the S-Video blurry, it's either two things: (1) ATI 600 USB shifts the colors. (2) ADVC300 is blurring the colors on output.
I do have another USB device (non ATI) that will capture the video over S-Video in VirtualDub. Perhaps I should also give that a try. If the S-video is much better on the second 'non' ATI capture card, then I could most likely assume the issue is with the ATI 600 USB. However, if the capture on the second 'non' ATI capture card is ALSO blurry, then I could possibly point to my ADVC300 not putting out a clean signal from the S-Video?
Is that worth trying? Then I can post my captures and letting you help me decide.
Running this back through my DVD player, I notice that color bars seem to become blurry in the S-video captures like my previous posts with the NTSC color bars captured from my ADVC300.
In my lasts posts, I was only looking at color bars. However, these test patterns have color bars as well as text and lines.
What's interesting, is that although the S-video capture color bars look more blurry, the text and lines are more sharp then those captured with composite. I found this a bit intriguing.
For instance, if you compare "ATI600_COMPOSITE_TEST_PATTERN_NO_SHARP_001227.jpg " with "ATI600_SVIDEO_TEST_PATTERN_NO_SHARP_001304.jp g" you'll notice that although the colar bars don't seem as sharp in comparison with the composite ones, when looking at the text and lines, the S-video is more sharp than the composite. For example compare the text "720 x 480 / 4:2:0" found in the pattern between S-Video and composite. The S-video looks sharper, than does the composite. However, the color bars are more blurred in the S-Video than in the composite.
So looking at my past posts, I am wondering if it's normal for the color bars to have a wider spread in the vertical direction on the waveform than those of the composite captures. It seems a bit counter-intuitive though. I would expect good color bars to have less spread vertically in the waveform. So, I'm a bit puzzled by this...if this is how it should be, but why?
By the way, the "no sharp" means that I set the ProcAmp for the ATI USB 600 to 0 for sharpness as suggested by msgohan.
What program are you using to capture the video, and into what format?
I'm using VirtualDub to capture into one of the YUV formats. This is similar to the other .avi's I've uploaded. Those were captured in VirtualDub too. I haven't uploaded the .avi's for the last .jpg's because as you know video is always quite big.
So, I pulled 4 frames each from the captured .avi's and saved them as jpeg's.
What program are you using to take the screenshots? That is what is wrong then.
Ahhhh there is something wrong with them? I am using Vegas to output the frames. I suppose there is an RGB conversion, but I figured that wouldn't really matter? I certainly could be wrong.
Hmmm, maybe I should go back and pull frames from VirtualDub. I've not done that before, but I could try
Your first composite image is on a bad frame (transition between the two stills).
I'm not too sure what you want me to comment on, as these are the same patterns and same device that I used. The only difference is the DVD player.
Sorry about that.
I've attached ATI600_COMPOSITE_TEST_PATTERN_NO_SHARP0170.jpeg to replace ATI600_COMPOSITE_TEST_PATTERN_NO_SHARP0240.jpeg.
The rainbows in the composite captures are what they call dot crawl?
A previous question: I am wondering if it's normal for the color bars to have a wider spread in the vertical direction on the waveform than those of the composite captures. It seems a bit counter-intuitive though. I would expect good color bars to have less spread vertically in the waveform. So, I'm a bit puzzled by this...if this is how it should be, but why?
In earlier posts when I was looking ONLY at color bars, I noticed that the composite color bars looked better than the S-Video color bars. And, when looking at the color bars, I saw a greater vertical spread on the waveform than I did on the composite. My understanding is that good color would have a thinner vertical spread, just like it does in a pure NTSC color bar test pastern.
What I noticed in these test patterns, is the color bars still have a wider vertical spread in the waveform in S-Video then it does in composite. However, if you look at the text in S-Video, it is sharper than that of the composite. That has thrown me off or seems a bit counter-intuitive. So, I'm thinking it's OK to have a wider vertical spread in the waveform for colors over S-video.
But back to original what I'm trying to accomplish: calibration. The ADVC300 outputs a test pattern through S-video and composite. The S-video and composite color levels are practically identical. However, as was said earlier, the S-video color bars have a wider vertical spread. Disregarding the wider vertical spread, I was wondering if it would be correct to adjust the ProcAmp controls of the ATI USB 600 to be in proper alignment with the ADVC300's color bar test pattern? Would this be a proper way of calibrating my ATI USB 600 capture card?
I'm not knowledgeable on this topic, but I'll try to join in on the discussion.
The text is sharper because there's no luma-chroma crosstalk. I don't know what's up with the colour in s-video, but at least there's no dotcrawl like in composite. I can't imagine the chroma resolution, and spread issues, having significance on much other than test patterns. This stick is mostly used for VHS caps, in which case, chroma is so crap that ATI 600's issues should have no impact.
The rainbows in the composite captures are what they call dot crawl?
A previous question: I am wondering if it's normal for the color bars to have a wider spread in the vertical direction on the waveform than those of the composite captures. It seems a bit counter-intuitive though. I would expect good color bars to have less spread vertically in the waveform. So, I'm a bit puzzled by this...if this is how it should be, but why?
Rainbows are called rainbows. Dot crawl looks like it sounds; dots crawling along at certain color boundaries. Both are caused by incomplete Y/C separation. The best explanation I've read was in a book at my library, but I don't remember which it was now... Some sort of "video engineers' handbook".
I don't know about your waveform question; I'm not familiar with how to set up for color, myself. When I "calibrated" my ATI 600 I did the color part by eye, guessing and checking and comparing to the VOB file. (Yes I am vaporeon800, but I didn't create that video -- just re-shared it.)
And, I worked hard at getting the things I wanted. Unfortunately, this has taken some time! But, I want to do a good job so it's important to me. I've had to force myself to be patient!
