Top edge of frame bending on Kramer SP-11D?

[nvrsk]

I am starting to do some captures of old VHS tapes.
since this is a hobby without a budget, I've collected most hardware from friends where it was available.
so, the videosignal goes like this:
Panasonic AG-7500 (or) Sony SLV-SE810K -> Kramer SP-11D -> DV Camera (or) Blackmagic Intensity Shuttle USB

the problem I am facing is that sometimes (unpredictable) the top edge of the frame is bending to the right.
if I exclude SP-11D from the scheme - there is no such a defect.
here's a link to comparison frame: https://screenshotcomparison.com/comparison/25240
here's a link to video showing how it happens: https://www.youtube.com/watch?v=rK5bSHbLJPo

the tested tapes were different: homemade videos, TV recordings, VHS copies.
defect appears on the same tape more often on second, thierd, fourth and so on capture.
swapping composite and S-Video as input does not make the difference.

Kramer SP-11D itself was luckely obtained as brand new and never used before.
model without the rack mounts, during the boot it displays "30" on it's screen - I suppose it is the version of firmware.
playing with it's settings as well as DIP-switches on it's back did not gave any positive result.

so I wonder: is there something I can do about this sort of problem?
may be I have to include another pass-through device in the scheme before Kramer, like DMR-ES10?

I supposed Kramer's TBC function intended to prevent the bending but not causing them, so I'm puzzled...

[lordsmurf]

Those VCRs, and that TBC, is not suggested gear. Why? Problems, including the sort of issues you're seeing. It's not intended for a wide variety of consumer sources. Those ancient editing decks were for in-studio S-VHS work, while the TBC expected clean (or pre-cleaned) input.

I'm not sure why you bought those.

Suggested workflow is
- late 90s / early 00s S-VHS decks with line TBC, JVCs or Panasonics, specific models
- DataVideo/Cypress type TBCs, very few specific others

What you have there is early 90s deck, with broadcast rackmount TBC.

[lollo2]

Quote:

if I exclude SP-11D from the scheme - there is no such a defect.

Are you sure of that? The problem you are facing is because you're missing a lineTBC correction, but AFAIK this is on the tape, it is generally not introduced by an external element. But I am not familiar with SP-11D device, so YMMV.

[lordsmurf]

Quote:

Originally Posted by lollo2

Are you sure of that? The problem you are facing is because you're missing a lineTBC correction, but AFAIK this is on the tape, it is generally not introduced by an external element. But I am not familiar with SP-11D device, so YMMV.

TBCs and VCRs can introduce tearing. And certain genres of decKs/TBCs, brands, and models are known for it. In this case, old editing decks, and rackmount TBCs. That stuff just was not engineered with this low end source in mind.

[nvrsk]

Quote:

Originally Posted by lollo2

Quote:

Are you sure of that?

absolutely.

Quote:

Originally Posted by lordsmurf

TBC expected clean (or pre-cleaned) input.

so, is there any way "to clean" the signal right to avoid the top edge dancing in the Kramer?
all in all, partially I became pretty happy with V/H shifting and other editorial results from Kramer.
I am also in the future will need to do SECAM transcoding, hoping Kramer to do that task as well.

Quote:

Originally Posted by lordsmurf

I'm not sure why you bought those.

as I said before, this is a hobby without a budget, and hardware was collected as is here/there/everywhere. =)

Quote:

Originally Posted by lordsmurf

DataVideo

impossible to get at my location. especially nowadays. =(
and the the price is a rocket in the sky...

Quote:

Originally Posted by lordsmurf

Cypress

is Cypress CTB-100G may be suitable for the task?
I've seen you mention that model sometimes on forum as a AVT-8710 clone which you consider good.
I may try to obtain one, but that's not clear and will take a long, long time...

[hodgey]

The avt/cypress and datavideos things don't really help all that much with this issue afaik, might be better off with a ES10 or other dvd recorder for tapes which exhibit this, or some other device with line-TBC/jitter correction.

Quote:

Originally Posted by lollo2

AFAIK this is on the tape, it is generally not introduced by an external element. .

It's more that it's a type of error not all video decoders/devices will handle equally as well. It's something to do with the gap/offset at the head switch point being too large/unstable for the device to handle. If you look at some older manuals for mixers and such it's noted as something that can occur. Tape tension differences between recording/playback is one cause I've seen noted, or head alignment issues or other problems (likely during recording if it occurs with both vcrs) .

Judging by how well the area below head switch is converted in the image without the kramer it looks like whatever is used to capture has a solid line tbc function though, presume it's a dv device rather than the blackmagic.

[lordsmurf]

Quote:

Originally Posted by lollo2

Are you sure of that? The problem you are facing is because you're missing a lineTBC correction, but AFAIK this is on the tape, it is generally not introduced by an external element. But I am not familiar with SP-11D device, so YMMV.

It can absolutely be introduced by workflow gear. Anything of it. VCR, TBC, capture card. Skew is an issue with top of frame timing, and can present in multiple ways, including being ignored, corrected, made better, or made worse. The suggested VCRs and TBCs will generally NOT add issues,

But it is a line timing issue, first lines in frame.

While frame TBC mostly cleans the signal, tearing can be prevented if the TBC quality is good, and the post-VCR output didn't have the issue. And in this case, it sounds as if the not-suggested rackmount Kramer is causing it. Every now and again, somebody argues with me about rackmount TBCs not being the proper tool for conversions tasks, and this is a perfect example of why it is (and the arguers are wrong).

Quote:

Originally Posted by nvrsk

so, is there any way "to clean" the signal right to avoid the top edge dancing in the Kramer?

Because of this:

Quote:

if I exclude SP-11D from the scheme - there is no such a defect.

Probably not.

Quote:

as I said before, this is a hobby without a budget, and hardware was collected as is here/there/everywhere. =)

Lack of a proper gear budget unfortunately usually gives you problem gear, or inferior gear, or no gear. Random VCRs, random TBCs, and random capture cards net random quality.

Quote:

is Cypress CTB-100G may be suitable for the task?
I've seen you mention that model sometimes on forum as a AVT-8710 clone which you consider good.
I may try to obtain one, but that's not clear and will take a long, long time...

Only if it's a non-defective unit, and a non-defective unit hasn't been sold new in 10+ years. What yuo see now is the "black" generation of TBCs,

Quote:

Originally Posted by hodgey

The avt/cypress and datavideos things don't really help all that much with this issue afaik, might be better off with a ES10 or other dvd recorder for tapes which exhibit this, or some other device with line-TBC/jitter correction.

ES10/15 type would pre-clean, but that exact Kramer can still be an issue. Some members here have tried that TBC, and found it lacking, though I forget who. It's similar to other Kramers, which can vary from marginally acceptable to avoidable.

Quote:

It's more that it's a type of error not all video decoders/devices will handle equally as well. It's something to do with the gap/offset at the head switch point being too large/unstable for the device to handle. If you look at some older manuals for mixers and such it's noted as something that can occur. Tape tension differences between recording/playback is one cause I've seen noted, or head alignment issues or other problems (likely during recording if it occurs with both vcrs) .

^ This. The more "pro" the gear is (editor, VTR; broadcaster TBC), and the older the gear is, the more this is likely to occur.

Quote:

solid line tbc function

Hmmm... "solid". Don't know if I agree with that. At best, it's trying, but not succeeding. All the layman jitter, the skew, and even some wiggly in that structure (bridge?) at the end. The sample submitted is pretty crappy for judging TBC. Line TBC testing need far more stable areas, not handheld water shots.

[latreche34]

OP have you tried removing the TBC from the chain? Did the flagging disappear? You can't troubleshoot a defect with stacked components, that's not how troubleshooting works, If you want help you're going to have to go onto the process of elimination and provide samples.

[Hushpower]

Quote:

Originally Posted by nvrsk

may be I have to include another pass-through device in the scheme before Kramer, like DMR-ES10?

I agree. Looks to me like classic a ES-10/15-fixable video, or ES-35, EZ-48, perhaps Sony HXD-890 or similar.

@Latreche
OP have you tried removing the TBC from the chain? Did the flagging disappear?:

Quote:

Originally Posted by OP

if I exclude SP-11D from the scheme - there is no such a defect.

[nvrsk]

I was able to locate and get hands on DMR ES-10 from a friend.
I was also able to get for testing another SP-11D, which is an old one (with firmware v.1.5) and was higly used backdays.

now I've had some time for this my hobby and made a comparison video with 6 variants of gear showing the results of VHS capture via S-Video in PAL-B:

1. Panasonic AG-7500 → DV Camera → VirtualDub
2. Panasonic AG-7500 → Panasonic DMR-ES10 → DV Camera → VirtualDub
3. Panasonic AG-7500 → new Kramer SP-11D v3.0 → DV Camera → VirtualDub
4. Panasonic AG-7500 → Panasonic DMR-ES10 → new Kramer SP-11D v3.0 → DV Camera → VirtualDub
5. Panasonic AG-7500 → old Kramer SP-11D v1.5 → DV Camera → VirtualDub
6. Panasonic AG-7500 → Panasonic DMR-ES10 → old Kramer SP-11D v1.5 → DV Camera → VirtualDub

https://youtu.be/xS9zy2CAXg8

the Kramers behave differently:
the new one (v.3.0) is stable but has the bending appear sometime.
the old one (v.1.5) has frequent frame jumps, sometimes loosing half-frames, and has framedrops.

so I am wondering: is there any sort of "service manual" available for Kramer SP-11D ?
there are lots of variable resistors/capacitors on it's motherboard that may be placed there for a fine tuning.

Quote:

Originally Posted by Hushpower

Looks to me like classic a ES-10/15-fixable video

anyway, if I pass the video through DMR-ES10 prior to any SP-11D - most of the problems disappear.

[latreche34]

If you are using a DV camcorder as a digital convertor shouldn't you be using WinDV or SClive instead of VDub since the signal is DV? From the samples in the link I don't see the Kramer doing you any favor, just ditch that P.O.S.

[lordsmurf]

Quote:

Originally Posted by nvrsk

the Kramers behave differently:
the new one (v.3.0) is stable but has the bending appear sometime.
the old one (v.1.5) has frequent frame jumps, sometimes loosing half-frames, and has framedrops.

Again, that model, regardless of firmware, is not made for consumer sources like VHS. In fact, it often just makes the quality worse. Exactly as I've been stating all along, as you're seeing in samples.

The ES10 is pre-cleaning with line TBC and non-TBC frame sync. The sync (frame timing) can still suck, cause drops, so that mild correction is being handled by the Kramer.

But you also now have the damage from ES10, especially luma hotness for PAL. There's another recent thread here when a user is trying to make a custom device to fight the luma gain. Not sure if he's successful yet.

[nvrsk]

Quote:

Originally Posted by latreche34

shouldn't you be using WinDV or SClive instead of VDub

I've started to capture from Vegas, as my friend who supplied me with a DV-Cam showed me, but have switched to VDub very fast as I use it frequently since early 2000th and know it much better.

I noticed that since DV-Cam provides straight (uneditable) digital stream - there is no difference in result depending on capture software.

Quote:

Originally Posted by lordsmurf

that model, regardless of firmware, is not made for consumer sources like VHS.

I've got your point.
but from the other side: there were lot's of VHS tapes backdays that were brought to TV stations/studios and played flawlessly on the air or edited into the final cuts.
I was working in the videostudio in middle 90th and we've had a Panasonic TBC in a rack (don't remember the model as it's gone ~25 years ago on studio closure), and I perfectly remember that we had sometime to fine-tune it's controllers on the back in case smth goes wrong with framesync on the output.
since it is professional equipment - it has to handle this sort of problems.

Quote:

Originally Posted by lordsmurf

The sync (frame timing) can still suck, cause drops

I do have an external sync input on AG-7500. if it is there, as well as external sync switch on it's pull-out panel, there is possibility to use it. the problem is that I don't remember how we did it backdays and what was used as a syncrogenerator... =(

Quote:

Originally Posted by lordsmurf

luma hotness for PAL

I did not dig deep for that yet since the primary problem is picture stabilizing. this would be my next step and I suppose Kramer is giving opportunity to fine-tune this right from it's front panel.

[latreche34]

I don't think vdub captures DV the same way sclive or WinDV streams it, Streaming is 1:1 digital transfer, Capturing is converting DV to whatever parameters you set and often destructive or lossy.

[Hushpower]

Quote:

I've started to capture from Vegas, as my friend who supplied me with a DV-Cam showed me, but have switched to VDub very fast as I use it frequently since early 2000th and know it much better.

I noticed that since DV-Cam provides straight (uneditable) digital stream - there is no difference in result depending on capture software.

I also don't think that's right. What codec are you using to capture your DV with in Virtual Dub?

[lordsmurf]

Quote:

Originally Posted by nvrsk

I've started to capture from Vegas, as my friend who supplied me with a DV-Cam showed me,

NLEs are horrible with capturing, never do it.

Quote:

I noticed that since DV-Cam provides straight (uneditable) digital stream - there is no difference in result depending on capture software.

DV cameras/boxes only transfer DV.

Quote:

but from the other side: there were lot's of VHS tapes backdays that were brought to TV stations/studios and played flawlessly on the air or edited into the final cuts.

I'm not swayed. Lots of garbage was aired, and still is, but should have been corrected. The analog era at least "hid" some of the errors, but not the digital era whatsoever.

Quote:

we've had a Panasonic TBC in a rack (don't remember the model as it's gone ~25 years ago on studio closure),

VTP/VCR/VCP also heavily mattered.

Quote:

Originally Posted by latreche34

I don't think vdub captures DV the same way sclive or WinDV streams it, Streaming is 1:1 digital transfer, Capturing is converting DV to whatever parameters you set and often destructive or lossy.

I'm no longer sure what's going on here. You don't want to hardware convert to DV, then capture analog, and back to the inferior DV codec in software. That will really screw things up. Not as much in PAL as NTSC, but still not great.

The whole workflow is messy.

[Hushpower]

Quote:

But you also now have the damage from ES10, especially luma hotness for PAL. There's another recent thread here when a user is trying to make a custom device to fight the luma gain. Not sure if he's successful yet.

The trimpot (100Ω) is working well to reduce the luma signal from my PAL ES-15. I haven't done any detailed tests yet on whether the whole contraption produces as good an image (interference-wise), but it certainly allows fine control of the luma channel before it gets to the capture card/stick.

http://www.digitalfaq.com/forum/vide...ntiometer.html

[nvrsk]

Quote:

Originally Posted by lordsmurf

NLEs are horrible with capturing, never do it.

I have that same opinion.
there was Adobe Premier as well, which I dropped also. =)

Quote:

Originally Posted by latreche34

I don't think vdub captures DV the same way sclive or WinDV streams it

I don't know how WinDV does it, but I did not see any settings for codec setup that are available for making a change to it in any tested capture software, it's just "dvsd", as DVCam's internal ADC hardware produces it and transfers digitally to PC via FireWire.

Quote:

Originally Posted by lordsmurf

DV cameras/boxes only transfer DV.

and it looks equally in mediainfo for any software that I've tested:

ID                          : 0
Format                      : DV
Commercial name             : DVCAM
Codec ID                    : dvsd
Codec ID/Hint               : Sony
Bit rate mode               : Constant
Bit rate                    : 24.4 Mb/s
Width                       : 720 pixels
Height                      : 576 pixels
Display aspect ratio        : 4:3
Frame rate mode             : Constant
Frame rate                  : 25.000 FPS
Standard                    : PAL
Color space                 : YUV
Chroma subsampling          : 4:2:0
Bit depth                   : 8 bits
Scan type                   : Interlaced
Scan order                  : Bottom Field First
Compression mode            : Lossy
Bits/(Pixel*Frame)          : 2.357

Quote:

Originally Posted by latreche34

Capturing is converting DV to whatever parameters you set and often destructive or lossy.

why is that?! you are mistaken - capturing is recording of the stream to the disk.
it could be right for the RAW stream which comes from BlackMagic (for the purpose of disk space saving), but in current case I see that the only stream that comes out from the source is properly packed DV. so I save it AS-IS for later long term editorial with AVS. the load of processor during capture is close to zero.
It is the same as, for example, capturing IPTV/Satellite, or even "fishing" Sat-internet - the stream has to be recorded unchanged, and there is a loss and mess otherwise.

[latreche34]

Unfortunately that's not how Vdub works, it is not a DV stream software, its a capture software, unless you install the DV codec in vdub, it is not going to output DV, even if it has a DV codec I wouldn't trust to perform as WinDV or SClive.

[lordsmurf]

Quote:

Originally Posted by nvrsk

I have tht same opinion.
there was Adobe Premier as well, which I dropped also.

Yep. NLEs are a tool to edit, not capture. Use the tool for the intended reason.

Quote:

and transfers digitally to PC via FireWire.
you are mistaken - capturing is recording of the stream to the disk.

Not necessarily. You can re-encode DV to whatever (including a new DV) on the fly. You must be careful. NLEs can do this, cheap/crappy ones especially (Pinnacle Studio, etc).

Quote:

Originally Posted by latreche34

Unfortunately that's not how Vdub works, it is not a DV stream software, its a capture software, unless you install the DV codec in vdub, it is not going to output DV

To add: VirtualDub with DV codec encodes to new DV, not DV stream from the tape.

Otherwise correct.