PDA

View Full Version : A Capture Favor to ask


signal
12th September 2007, 03:52
I'm trying to compare a few things and make sure I'm not crazy (probably too late for that).
It would help tremendously if I could get about a 5 to 10 second capture of the following image.

Here's the image in 640x480 png:
http://64.202.175.38/examples/Luma5_640x480_601_7.5.png
720x480 png version here (http://64.202.175.38/examples/Luma5_640x480_601_7.5.png)

If it helps, here's a small DVD iso of the image that also includes two rotated views:
Luma5_DVD.zip (571K) (23MB uncompressed) (http://64.202.175.38/examples/Luma5_DVD.zip)

I need this in a non-mpeg, un-filtered (i.e. no hardware/software "clean up" options used), and preferably in a Huffyuv/Lagarith or other losless codec format.

I know it's odd to be asking for a clip instead of submitting one, but please bear with me on this one.

Guest
12th September 2007, 05:46
Why don't you make the clip yourself?

signal
12th September 2007, 06:40
Why don't you make the clip yourself?

Well.. I did.
And tried to explain but I guess I didn't do a good job.

I'm trying to compare a few things and make sure I'm not crazy (probably too late for that).

I know it's odd to be asking for a clip instead of submitting one, but please bear with me on this one.

Guest
12th September 2007, 15:01
It doesn't make any sense. How can I "capture" that? I could make a clip out of the PNG, or transcode the MPEG, but what does that have to do with capturing?

And what do you want to compare?

If we don't know what you are trying to do, we can waste our time doing the wrong thing for you!

signal
12th September 2007, 18:13
It doesn't make any sense. How can I "capture" that? I could make a clip out of the PNG, or transcode the MPEG And is also why I provided a DVD of it if it helped some in creating another source of this test pattern to be captured. Some can output it through the TV out of their graphics card, use the DVD, create your own method, whatever your choice is to capture that test pattern.

but what does that have to do with capturing?
Point of it is that it's a source unaffected by doing a analog capture of it. So I'm providing the source to use for a test clip to be captured.

And what do you want to compare?
To my own captures of the test pattern.

If we don't know what you are trying to do, we can waste our time doing the wrong thing for you!

What part of It would help tremendously if I could get about a 5 to 10 second capture of the following image. I need this in a non-mpeg, un-filtered (i.e. no hardware/software "clean up" options used), and preferably in a Huffyuv/Lagarith or other losless codec format. I know it's odd to be asking for a clip instead of submitting one, but please bear with me on this one. isn't explaining what I am trying to do?

If there are still questions on the format, how to, or other parameters, then I really apologize.
Please let me know.

Guest
12th September 2007, 19:19
To my own captures of the test pattern. Thanks, but I already knew that. I meant what features of the clip are you comparing and for what purpose?

Your results are going to be affected by the TV out method used. How will you know that the differences you see are due to TV out versus capture?

signal
13th September 2007, 10:24
I'm looking at the Luma side of things only (hence the test pattern) and while yes, what someone captures is going to vary from what I have, I expect that.

I'm not trying to match up and do an apples and apples comparison in that regard.

FlimsyFeet
13th September 2007, 13:53
I can understand what you're asking people to do but I don't see what you are expecting to find or achieve from it.

Say I burn that DVD and play on a DVD player connected up to the s-video input on my capture card. The luma levels in the captured file will obviously depend on the brightness and contrast (and maybe gamma) adjustments for my capture card in virtualdub. I'd adjust the brightness so that the black ends up being captured as black (0,0,0 - when displayed as RGB) because that's good practice, and you should adjust these things before any capture. Your top bar is not white, it's R=G=B=235 - should I be adjusting the contrast to make it 255, or 235, or whatever value your capture was?

Guest
13th September 2007, 14:25
I'm not trying to match up and do an apples and apples comparison in that regard.
OK, then again, what ARE you trying to compare and why?

signal
15th September 2007, 09:54
Folks.. again apologies.

From what I'm reading here, it is still expected that somehow in order to just see how someone else might capture this and for me to make use of it, it just about must be a pixel by pixel, exact configuration capture in order to do so. And again, apologize if I'm misinterpreting that.

But just to be sure...
I'm looking at the Luma side of things only (hence the test pattern) and while yes, what someone captures is going to vary from what I have, I expect that. I'm not trying to match up and do an apples and apples comparison in that regard.

For sanity sake, I was asking to see how someone elses unsolicited random configuration might handle it. YES.. again.. I fully understand I'm not going to back what I have and the many various parameters that can affect a capture will not match mine. I expect that.

Getting back that sample in any capture in the format I asked for, I can truly get the info I'm looking for (as far fetched as it may seem).

And yes, I'm being evasive on the exact nature of that test/comparison for two reasons.

1. I didn't want it to influence the capture

2. I seriously didn't want this to turn into a discussion about what "issue" I might be seeing on my side and the guesses/theories behind it.
Again, I really just wanted a sanity check here before pursuing anything else.

@FlimsyFeet - Thank you for bringing that up. I hope my post here explains, but to answer your question even if the range differs from mine, that is ok. Capture it however you feel best and I can still make use of it.

I had started this thread to ask a favor and to please be patient as I know asking for a clip rather than submitting one was "odd".
I really understand it is being favor (big favor) and knew it was out of the ordinary to ask, but it is a serious request not made half heartedly and wasn't made without thinking about it ahead of time.

GrofLuigi
15th September 2007, 23:12
Here it is:

http://rapidshare.com/files/55984476/capture.rar

Notes:

- The DVD was played on a cheap standalone and passed through S-VHS deck; sorry I couldn't connect directly - too many entangled/short cables - but maybe that's what you're looking for? :) That's my usual procedure - I capture everything from that S-VHS deck. The two cables signal passed through were S-Video.

- There were dropped frames - I don't usualy get them, but changed drivers recently and underclocked this old Athlon XP to lower temperatures that were sky high.

- Cut 5 seconds in VirtualDub (that's what was captured with, 1.7.3) in direct stream copy mode, no audio.

- Capture card: Pinnacle PCTV Rave with btwincap drivers.

- Oh, and Extend Luma White Point was on; on normal captures, that's how I get the histogram to 'fill up'; all other driver settings were at default.

GL