Go Back    Forum > Featured > General Discussion

Reply
 
LinkBack Thread Tools
  #41  
01-28-2016, 08:42 PM
sanlyn sanlyn is offline
Premium Member
 
Join Date: Aug 2009
Location: N. Carolina and NY, USA
Posts: 3,648
Thanked 1,307 Times in 982 Posts
With my PV-S4670 and PV-4665, the usual amount of head switching noise at the bottom of an image is 2 to 4 pixels. With the AG-1980 the usual is 8 pixels, although today I captured a VHS tape made in 1992 with a 1991 Panasonic whose image played with only 2 pixels of noise at the bottom. The usual noise at the bottom of images with my SLV-585HF is 8 pixels, although a couple of times it played 12 pixels of bottom border noise with a recording made on an old RCA vcr.

These values remain the same whether or not I have the PA-100 in circuit.
Reply With Quote
Someday, 12:01 PM
admin's Avatar
Ads / Sponsors
 
Join Date: ∞
Posts: 42
Thanks: ∞
Thanked 42 Times in 42 Posts
  #42  
01-28-2016, 08:58 PM
msgohan msgohan is offline
Free Member
 
Join Date: Feb 2011
Location: Vancouver, Canada
Posts: 1,323
Thanked 334 Times in 276 Posts
But it appears to be saying that if you turn down the gain, for example, the bottom of the video will remain as contrasty as the input.
Reply With Quote
  #43  
01-28-2016, 09:20 PM
sanlyn sanlyn is offline
Premium Member
 
Join Date: Aug 2009
Location: N. Carolina and NY, USA
Posts: 3,648
Thanked 1,307 Times in 982 Posts
You mean the "contrast" in the head switching noise? It will be cropped and the image re-centered anyway as much as needed by replacing the crop with new borders. (as opposed to crop and resize the image, which changes the original image proportions. But some people don't care about the original image).
Reply With Quote
  #44  
01-28-2016, 09:25 PM
msgohan msgohan is offline
Free Member
 
Join Date: Feb 2011
Location: Vancouver, Canada
Posts: 1,323
Thanked 334 Times in 276 Posts
My concern is that the ES10 could use it as part of its AGC calculation, as I said.
Reply With Quote
  #45  
01-28-2016, 09:52 PM
sanlyn sanlyn is offline
Premium Member
 
Join Date: Aug 2009
Location: N. Carolina and NY, USA
Posts: 3,648
Thanked 1,307 Times in 982 Posts
In VirtualDub capture I monitor the incoming image with VDub's histogram. Borders are cropped from the image to measure contrast/brightness, etc., and adjusted for the cropped image. Then the crop is disabled before I start the capture. Whether or not the Panny's "AGC" does anything I don't know. (It has AGC that's this active? That's new. I know the Toshiba has some mild AGC activity to protect itself and your TV, but I have no figures on how much AGC the Panny applies). It can't be much, because I get the same image whether I have the Panny in circuit or not.

All of these pass-thru recorders have IRE settings. But I guess you've checked the settings in yours already.

I'd be more concerned about another fault the PA-100 displays with some cards: the bottom 16 to 24 pixels display a gray overlay band when the black level is increased or decreased. I think TGrant's repair package fixes that, too.

-- merged --

After some research it seems you're correct about AGC, although this link deals with the ES15 -- so similar to the ES10 that I'm making the assumption that AGC acts the same way in both, just as AGC behaves this way in the less powerful Toshiba I often use: http://forum.videohelp.com/threads/3...=1#post1980760. In summary the Panny AGC, like that in the Toshiba, attempts to tame "hot" brights from some VCR's. How much the head switching noise affects AGC is anyone's guess, but since I crop that off when measuring output from the pass-thru unit (cropping black or white side borders as well), the measurements are for the image content only after the pass-thru unit has done whatever it does, and without those areas affecting the histogram. Those unwanted areas are removed in Avisynth and replaced with black borders to maintain the frame size. The same cropping technique is used when doing color correction.
Reply With Quote
  #46  
01-30-2016, 03:30 AM
Goldwingfahrer's Avatar
Goldwingfahrer Goldwingfahrer is offline
Remembered
 
Join Date: Aug 2013
Location: Switzerland
Posts: 453
Thanked 84 Times in 74 Posts
why not just connect a voltmeter?
16-235 = 0 - 700 mV

NTSC 525

White levels = 100 IRE, 714 mV
Maximum signal level = 120 IRE
Minimum signal level = -20 IRE

https://en.wikipedia.org/wiki/Broadcast-safe
Reply With Quote
  #47  
01-30-2016, 10:13 AM
sanlyn sanlyn is offline
Premium Member
 
Join Date: Aug 2009
Location: N. Carolina and NY, USA
Posts: 3,648
Thanked 1,307 Times in 982 Posts
One can have 16-235 input but still have clipping at either or both ends. I've seen many examples.
Reply With Quote
  #48  
01-30-2016, 11:35 AM
Goldwingfahrer's Avatar
Goldwingfahrer Goldwingfahrer is offline
Remembered
 
Join Date: Aug 2013
Location: Switzerland
Posts: 453
Thanked 84 Times in 74 Posts
Yes me too.

Here are some excerpts
I have white hair

in Postpro one must also have something to work

before capturing the image I look at the interlaced monitor

This is not the ES10 the problem
either the feeder or I


Attached Files
File Type: zip Sample images.zip (503.6 KB, 6 downloads)
Reply With Quote
  #49  
01-31-2016, 08:43 AM
Jarvis Jarvis is offline
Premium Member
 
Join Date: Aug 2011
Location: Australia
Posts: 81
Thanked 15 Times in 13 Posts
Back for a brief post. Interesting info about SignVideo and AGC etc. Thanks again to you guys for your continued research. I'm almost positive the Vidicraft proc amp I'm trying to get right now is PAL-compatible, but if I can't obtain that one then I'll be getting the Vivanco. Either way I'll have something to test with on my own.

I remember something Sanlyn said about my JVC VCR levels being too high. Of course they are too high for the ES10, but it didn't really occur to me that the JVC may not be sending the full range to begin with. So I looked back on some captures and it turns out it's true. My NV-FS200 actually outputs a better luminance range than the JVC (HR-S9600). Not by much but it is noticeable in some scenes. Unfortunately the picture is worse in almost every other aspect.

There is a compounding effect in my chain right now with each device mishandling levels, obviously ES10 being the main culprit. They need to be pulled down at the first stage. Maybe even a potentiometer adjustment like Goldwingfahrer mentioned. Certainly, for home video with exposure problems, all this luma clipping is devastating. It does irreparable damage to the dynamic range.
Reply With Quote
  #50  
01-31-2016, 03:04 PM
msgohan msgohan is offline
Free Member
 
Join Date: Feb 2011
Location: Vancouver, Canada
Posts: 1,323
Thanked 334 Times in 276 Posts
Quote:
Originally Posted by Jarvis View Post
it didn't really occur to me that the JVC may not be sending the full range to begin with.
All VCRs with internal ADC can internally clip brights and crush blacks, depending on the levels from the tape. I've seen this with posted samples from both JVC and Panasonic VCRs.

VCRs with ADC = all VCRs with DNR&TBC, as well as other models that may do some sort of unknown digital processing. (Ex: Mitsubishi HS-U748 doesn't advertise DNR feature, but service manual shows "YCS/DNR" chip that does YNR after ADC.)

I've also noticed clipping & crushing with my own JVC HM-DH5U, which has an odd AGC that bases its decisions on picture content rather than just signal properties (i.e. sync pulses or color burst, according to that jagabo post). At times, it lowers whites down to light-grey in order to try to bring the superwhites down to white.

This is an HDMI capture, so you can see see exactly where the digital clipping occurs. Captures from the S-Video output are just the same thing sent through a DAC (+ADC on cap card), so they show the same clipping but it looks a bit rounder and the added noise makes the precise level at which it's hitting the hard limit less clear.

If you watch the attached video, you can see the timestamp starts out clipping as overly-bright white. Then a bright light reflecting off the glass enters the frame, and the entire picture including timestamp is lowered to accommodate the over-exposed reflection. After the reflection exits frame, the picture level rises again. These transitions happen fairly quickly, but at least they don't just flicker from one level straight to the other.

Screenshots, resized by the forum software to fit the page. Originals are attached. The purple arrows are just indicating the bright head-switching noise (and corresponding histogram peaks, only visible in the attached version).
fish-datestamp-bright.jpg

You must be logged in to view this content; either login or register for the forum. The attached screen shots, before/after images, photos and graphics are created/posted for the benefit of site members. And you are invited to join our digital media community.


The highest pixel level in the sample is Y=253, but this is due to head-switching noise. The actual levels in the active picture area are around 247-248 with a few spikes to 249. So there is a little headroom going unused, for no reason. After the reflection exits frame, the highest pixel value is suddenly only 229 before it rises back up over the next 6 frames.


Reply With Quote
  #51  
01-31-2016, 09:24 PM
Jarvis Jarvis is offline
Premium Member
 
Join Date: Aug 2011
Location: Australia
Posts: 81
Thanked 15 Times in 13 Posts
msgohan,
I always assumed this kind of AGC was caused by the camcorder at time of recording, rather than how the VCR is reading it. Commonplace in my experience, actually the tape I posted screen caps of here has constant activity like that in certain scenes. I've played it on a handful of VCRs and the exposure fluctuations are the same across the board. Otherwise, being a 2nd-gen tape, the problem may have happened at the time of dubbing (there seems to be interference noise in places too).

On a side note - I wonder if there's an Avisynth script or NLE plugin that can automatically (to some degree) map such exposure changes over the course of a scene, and balance them out on its own. Would be nice to resolve without manually plotting envelopes. I suppose it'd be like an anti-flicker or something.

I've attached a clip showing the exposure problem in my case.
JVC HR-S9600EU (TBC on) > ADVC-110. Tested with other VCRs and capture devices with same result. Losslessly cut from original DV file, no adjustments.


Attached Files
File Type: avi scene0004-AGC.avi (54.05 MB, 9 downloads)
Reply With Quote
  #52  
01-31-2016, 10:13 PM
Jarvis Jarvis is offline
Premium Member
 
Join Date: Aug 2011
Location: Australia
Posts: 81
Thanked 15 Times in 13 Posts
Quote:
Originally Posted by msgohan View Post
I've also noticed clipping & crushing with my own JVC HM-DH5U, which has an odd AGC that bases its decisions on picture content rather than just signal properties (i.e. sync pulses or color burst, according to that jagabo post). At times, it lowers whites down to light-grey in order to try to bring the superwhites down to white.
Based on my own JVCs, I find they all work just about if not the same in regards to picture and levels (tracking differs though). Lowering whites down to light-grey to accommodate superwhites, as you say, is the same that I've noticed. This is good behaviour. The levels are still above safe range but can be recovered as long as the capture device doesn't legalize values on input.
The ES10 on the other hand just pushes the whole waveform up in comparison, clipping that information.

PS: Still no luck in obtaining the Vidicraft PRC-100 I wanted, the seller hasn't responded lately. Does anyone here own one or a similar model with luminance metering? I'll probably have to put up an ad in the marketplace forum soon.
Reply With Quote
  #53  
01-31-2016, 10:16 PM
msgohan msgohan is offline
Free Member
 
Join Date: Feb 2011
Location: Vancouver, Canada
Posts: 1,323
Thanked 334 Times in 276 Posts
Quote:
Originally Posted by Jarvis View Post
I always assumed this kind of AGC was caused by the camcorder at time of recording, rather than how the VCR is reading it.
I typed that these same tapes of mine (camcorder originals) play in other VCRs without this fluctuation, but apparently I deleted that line before posting.

The camcorder itself has AGC for its auto-exposure control that causes similar variations, but that doesn't affect the timestamp.
Reply With Quote
  #54  
01-31-2016, 11:12 PM
Jarvis Jarvis is offline
Premium Member
 
Join Date: Aug 2011
Location: Australia
Posts: 81
Thanked 15 Times in 13 Posts
Quote:
Originally Posted by msgohan View Post
I typed that these same tapes of mine (camcorder originals) play in other VCRs without this fluctuation, but apparently I deleted that line before posting.

The camcorder itself has AGC for its auto-exposure control that causes similar variations, but that doesn't affect the timestamp.
Understood, indeed the timestamp shouldn't be affected if it was a camcorder issue. Can't say I've noticed anything like that on my JVC VCRs but doesn't necessarily mean it's not there. Panasonic (NV-FS200 at least) seems to be most accurate towards levels, haven't seen AGC issues on it either but perhaps further testing is needed.
Reply With Quote
  #55  
02-04-2016, 05:03 AM
Jarvis Jarvis is offline
Premium Member
 
Join Date: Aug 2011
Location: Australia
Posts: 81
Thanked 15 Times in 13 Posts
Update:
Managed to purchase the Vidicraft PRC-100, waiting for it to arrive then I'll post some more tests.
This has the luma meter which will come in handy.

Also, some questions for those in the know:
1. I'm considering getting the DMR-EH65. If one of you guys here own both EH65 (or similar model) and ES10, would you be willing to post a TBC comparison between the two? A short clip (or even picture) and some comments would suffice. I know the ES10 is supposedly better, it just isn't clear by what degree.

2. Also interested in trying a HyperDeck Shuttle 2 to replace the PC for capture. Anyone have experience with it or heard anything? Captures via HDMI to uncompressed, which I would later convert to lossless. Presumably there should be no problems.
Reply With Quote
  #56  
02-04-2016, 12:14 PM
msgohan msgohan is offline
Free Member
 
Join Date: Feb 2011
Location: Vancouver, Canada
Posts: 1,323
Thanked 334 Times in 276 Posts
One thing that sucks about the HyperDeck Shuttle 2 is that it always includes 8 channels of uncompressed audio, regardless of the number of channels in the source. This takes a huge amount of space on the SSD.
Reply With Quote
  #57  
02-04-2016, 07:26 PM
Jarvis Jarvis is offline
Premium Member
 
Join Date: Aug 2011
Location: Australia
Posts: 81
Thanked 15 Times in 13 Posts
Quote:
Originally Posted by msgohan View Post
One thing that sucks about the HyperDeck Shuttle 2 is that it always includes 8 channels of uncompressed audio, regardless of the number of channels in the source. This takes a huge amount of space on the SSD.
Not ideal, but for 1 hour of capture, the difference only amounts to 3GB when compared to 2 channel audio.
What concerns me more is the 10-bit pixel format.
Reply With Quote
  #58  
04-04-2016, 10:44 AM
Jarvis Jarvis is offline
Premium Member
 
Join Date: Aug 2011
Location: Australia
Posts: 81
Thanked 15 Times in 13 Posts
Hey guys, finally back for an update.

After acquiring the Vidicraft PRC-100 proc amp, I am happy to report that I'm able to turn down VCR output levels and passthrough the DMR-ES10 without any clipping, preserving all information. The ES10 makes no attempt to adjust the levels it receives from the Vidicraft, and is able to perform all its line TBC and frame sync operations on the signal. This is excellent news and just the result I had hoped for.

I'm looking to make a detailed post in the coming days, full of info and examples covering a lot more than the above. I've been making a lot of tests in regard to TBC and levels performance between devices, bear with me while I prepare it all.

PS: I've also attached the PRC-100 manual that I managed to find. Although the manual only mentions NTSC compatibility, I am confirming that it works with PAL. It can also be powered with a DC adapter which negates the need for a step down transformer. The luma metering on it is excellent. The model is composite only.


Attached Files
File Type: pdf VidicraftProcAmp.pdf (2.34 MB, 17 downloads)
Reply With Quote
  #59  
04-04-2016, 11:17 AM
sanlyn sanlyn is offline
Premium Member
 
Join Date: Aug 2009
Location: N. Carolina and NY, USA
Posts: 3,648
Thanked 1,307 Times in 982 Posts
Congrats on getting your proc amp. Vidicraft became SignVideo and changed their product line. The old PRC-100 became the SignVideo PA-100, which I've been using for years.

Below: Not too much changed with the PA-100, mainly the analog luma meter replaced with an LED display and some differences in front panel control function.

Look familiar?

VidiCraft PRC-100


SignVideo PA-100


Attached Images
File Type: png VidiCraft PRC-100.png (289.4 KB, 110 downloads)
File Type: png SignVideo PA-100.png (93.9 KB, 108 downloads)
Reply With Quote
  #60  
04-04-2016, 11:26 AM
Jarvis Jarvis is offline
Premium Member
 
Join Date: Aug 2011
Location: Australia
Posts: 81
Thanked 15 Times in 13 Posts
Quote:
Originally Posted by sanlyn View Post
Congrats on getting your proc amp. Vidicraft became SignVideo and changed their product line. The old PRC-100 became the SignVideo PA-100, which I've been using for years.

Below: Not too much changed with the PA-100, mainly the analog luma meter replaced with an LED display and some differences in front panel control function.

Look familiar?
Indeed, I was actually looking at Signvideo and Studio 1 (also before Signvideo) proc amps today, there's a couple on eBay. Wondering if they would be PAL-compatible also? Would be ideal as they also have s-video connections, compared to composite-only Vidicrafts (though I think there were Vidicraft BNC versions too). They fetch exorbitant prices however.
Reply With Quote
Reply




Similar Threads
Thread Thread Starter Forum Replies Last Post
Elite Video BVP4 vs. SignVideo PA-100, proc amps for PAL? vaccastracca Restore, Filter, Improve Quality 10 05-18-2021 06:42 PM
Sign Video detailer and proc amps discontinued Cyclone82 Restore, Filter, Improve Quality 6 06-09-2012 10:12 PM
Composite vs s-video for detailers and proc amps? Acden Restore, Filter, Improve Quality 3 04-02-2012 06:51 AM
Internal TBC VCR or HQ Non-TBC VCR + Panasonic DMR-ES10? LukeS Restore, Filter, Improve Quality 1 10-25-2011 04:06 AM
Proc Amps and Detailers PAL Compatible Juan-ito Capture, Record, Transfer 4 10-18-2010 11:57 PM

Thread Tools



 
All times are GMT -5. The time now is 05:59 AM