Hello again,

revisiting my old thread to share my more recent lessons-learned and to reply to the registered members who PM'ed me throughout the past years.

No real surprise here, my current S-/VHS setup as follows:
JVC HR-S9600 {S-video/RCA audio} > Panasonic DMR-EH495 {HDMI} > HDMI-Splitter {HDMI} > Intensity Shuttle {Thunderbolt} > Mac

I am using the same setup for Laserdisc, replacing the JVC with a Pioneer CLD-D925, sticking to S-video because the composite video out (RCA, Scart) of the D925 is fed with a merged Y/C signal anyway.
On the HR-S9600 Digital 3R is turned OFF per suggestion in various forum threads, and TBC is left ON. On the D925 the HQ-circuit function is turned OFF as it softens the video unnecessarily.

In addition, I am running the video/audio signal of the EH495 to a small multiformat TV that supports S-video through Scart and also offers seperate S-video and RCA connectors for testing purposes. It's easier to read and configure the device settings that way, too.

So far the Intensity Shuttle has been a hassle-free experience both in terms of the hardware and the software suite containing the apps/drivers (i.e. Desktop Video 12.4.2 running on macOS 12.6.5). The abovementioned S-/VHS setup works for PAL sources whereas the LD setup also allows for NTSC (NTSC 3.58, not PAL-60) because the D925 and the EH495 both support it.

Lessons learned:
- The tv format (PAL/NTSC) must be set accordingly in both software applications, Desktop Video (where you select it in the source settings) and Media Express (where you have to select the tv format in project specs), otherwise you'll end up with a black capture screen.
- Changing the tv format on the EH495 resets the HDMI settings to AUTO so you need to adjust them again to 576i/480i for use with the Intensity Shuttle.
- I couldn't catch a difference between setting the JVC to B.E.S.T. OFF/Picture Control NORM or B.E.S.T. ON/Picture Control AUTO. With the latter option however B.E.S.T. automatically kicked in on some tapes that otherwise failed to display the video for a couple seconds, so that's nice. Your mileage may vary...:wink2:
- The NR function of the EH495 (remote press Display > Video > NR ON) improved the flickering I encountered with various Laserdisc.
- At this point in time, the Intensity Shuttle is apparently discontinued but still supported for Intel based Macs according to Blackmagic Design's support documents.

Well, I guess nowadays there's already an overwhelming ammount of user feedback on the benefits of the abovementioned capture method (DMR+Intensity Shuttle). So, drawbacks? While the Media Express capture software offers the option to "stop capture if dropped frames are detected" I see no way to confirm this since it never stopped on me. Other than that? Cables, adapters... any of which can fail, too... :D

Since LD is composite by nature you may want to check if using composite yields better results, The catch is which one has a better comb filter, Your LD player over S-Video or the EH945 when LD in composite? I would test both under the same conditions and post samples here.

You would want to use EDIT mode to minimize digital processing since you are relying on the EH945 processing, B.E.S.T should be always kept on, It is like the tape bias for analog audio tapes, It checks the level of RF signal in the recordings and adjust head coils current accordingly making up for RF signal loss minimizing artifacts such as comets, streaks and so on.

If you are using it for analog capture it will detect dropped frames, However since you are using it for HDMI ingest anything before the EH945 is already baked in, If the EH945 drops frames there is no way Mediaexpress would know about it because it was already been digitized and routed over HDMI, That's why it never reported dropped frames because the signal is already digital.

Thanks for the advice. I was under the impression that the S-video output of the D925 is the superior source compared to the composite output, because the latter receives the merged Y/C signal by means of the internal video processing design of the D925. As if I connected a S-video to composite adapter to a S-video source, which is why people mod the D925’s internals to avoid splitting and remerging (layman’s terms) the composite signal before it is fed to the output/connectors?

Thanks, I wasn’t aware of that as the manual only mentions that EDIT “minimizes picture degradation during editing”. So when the manual states picture degradation that would be digital processing then?

Thanks, that’s good to know. I was wondering how feasible that option is since I read some user reviews complaining about dropped frames while simultaneously capturing to a PC’s internal capture card. I didn’t find any clue as to the frame dropping by the EH945. Then again, I am well aware that there are always more reliable solutions.

The B.E.S.T function adjust video EQ and various things depending on the output from the tape (haven't found exact specifics for the newer JVCs though) so the degree of difference between it on/off is going to depend a bit on the tape.

It's going to turn down/off some of both digital and analog processing to varying degrees to avoid too much of a compounding effect of all the filtering on successive dubbing. It can be helpful for digitizing as well to avoid the VCR over-processing the output, and leave more of the noise reduction up to much more modern post-processing in software if you want to squeeze out the maximum detail from the tape.

It's more the analog part of the processing that is turned down as the digital noise reduction from the TBC/DNR function is still pretty active. Using the EDIT setting is going have a more pronounced difference on tapes where the video signal is weak.

Thanks again for the advice, and thanks to lordsmurf and team for the possibility to upload/host samples.

Laserdisc sample as follows:

D925 set to NTSC 3.58; EH495 set to NTSC.
Media Express settings: 525i59.94 NTSC, QuickTime Uncompressed 8-bit YUV.

S-VID.mov = S-video
S-VID_NR.mov = S-video, EH495 NR ON
S-VID_NR_HQ.mov = S-video, EH495 NR ON, D925 HQ ON
COMP.mov = Composite
COMP_NR.mov = Composite, EH495 NR ON
COMP_NR_HQ.mov = Composite, EH495 NR ON, D925 HQ ON

(I forgot to mute the samples for composite video.)

That would be a strange design, to take the composite signal from the Laserdisc, split it out to Y/C for the S-Video output, then merge the two signals back again for the Composite output.

You might try using your JVC VCR in passthrough mode to convert the composite signal to S-Video and see if that looks better than using the S-Video from the D925 directly.

With regard to the video processing design choice made by Pioneer for the D925, I learned that is common knowledge ever since the player was reviewed on the site “laserdisc archive” ages ago, and later modded by people on LDDB forums to circumvent this.
I’m totally okay with that because it was the best choice for PAL & NTSC playback in the EU back when I purchased it in the 90s. There are certainly more capable LD players for NTSC, there might be for PAL, but given the fact that I capture my LD library for fun, not preservation, I don’t really mind.

Perhaps I misunderstand your point, but since the composite signal is split to Y/C before even “leaving” the LD player for the sake of the S-video output, only to be merged again when it is actually fed to the composite output, I do not understand how that butchered up composite signal can be improved or why it should be. I was under the impression that video signal clarity lost is, well, lost.
Feel free to check out the samples though. I am interested in everyone’s opinion because I’m certainly not trained to spot slight nuances in colors or sharpness, and so forth.

The 3-D comb filter in the JVC HR-S9600 will be far superior to the 2D adaptive comb filter in the D925. What I wonder is whether the external comb filter improves the recombined composite signal enough to make it better than the D925's internal comb filter or whether you need the composite out mod to see any improvement.

@traal: Yeah, I’ve read through most if not all those LDDB forum threads throughout the years. Even checked out the service manual out of curiously. But back on topic. i.e. capture workflow on Macs, which of the six samples above yields the best quality in your opinion?

The attention span and patience is getting shorter and shorter nowadays, I think the OP miss-understood my point in my previous post, I never asked to capture LD over S-Video and convert it back to composite, that would be totally wrong and un-necessary, I asked for two samples:
1- LD -> Composite -> DMR -> HDMI (testing DMR comb filter)
2- LD -> S-Video -> DMR -> HDMI (testing LD comb filter)
No need to post six samples.
I'm pretty sure the comb filter in the DMR is far superior to the one in LD but I though I would confirm with real samples. So which two out of the six samples are the ones we're looking for? doesn't matter what setting for DNR as long as test conditions and setting are exactly the same for both samples as I outlined in my previous post.
Not sure what a comb filter is? here is a nice write up.


Edit mode in the VCR means it is used as a player only, feeding an editing deck, Capturing is no different than this task.

Take it easy mate, it’s right there… :D

S-VID.mov = LD -> S-Video -> DMR -> HDMI
S-VID_NR.mov = LD -> S-Video -> DMR -> HDMI + EH495 NR ON
S-VID_NR_HQ.mov = LD -> S-Video -> DMR -> HDMI + EH495 NR ON, D925 HQ ON
COMP.mov = LD -> Composite -> DMR -> HDMI
COMP_NR.mov = LD -> Composite -> DMR -> HDMI + EH495 NR ON
COMP_NR_HQ.mov = LD -> Composite -> DMR -> HDMI + EH495 NR ON, D925 HQ ON

And yeah, I threw in additional samples where I also enabled the NR feature on the EH495, and where I also added the HQ circuit setting on the D925. Why? Because this thread is easily found through web searches and I also received PMs regarding this thread over the years. I like to think there is someone out there who might be interested in that comparison.

Me take it easy? I couldn't be more easier than that, I sometimes edit my posts more than once to make sure the message is clearly delivered and yet (generally speaking) some will gloss over it like a todler flipping thru a political magazine pages.
I will examine the samples and get back.

COMP.mov seems to have less rainbowing in the subtitles but otherwise more red/green chroma noise than S-VID.mov.

I thought so too, I don't know what type of comb filters both the LD and the DMR use, line or 2D but it looks like the OP could benifit from one of the modern 3D comb filters found in modern devices like the Eval board, SingMai or similar devices.

https://www.digitalfaq.com/forum/att...1&d=1683490907

Thanks for the info, I felt the same about the subtitles the moment I switched cables.

Thanks, mate, I’m sure that info will come in handy to someone using a similar capture setup/workflow like the one I described earlier, but looking to achieve better results.

LOL! :laugh:

That's a terrible sample image. The 3D is hot as can be, overexposure. The 2D is better. I hope nothing butchers that badly, then is consider to be "better" in anyway. I'd almost rather have color-accurate dot crawl.

That sample picture is for demonstaration purposes only, I think it mainly shows the effictivness of 3D comb filter at getting rid of dot crawl, In the top image with the girl the levels are actually more right in the 3D one than the 2D one if you look at the background behind her, the bottom pictures are not that far apart in terms of levels.

That's what I was thinking. The problem becomes that the 2D vs. 3D aspect is also suspect, not very realistic. I hate fakey samples. (I know you didn't make it.)

Nah, I think the exposure is pumped, regardless of the background.

Manufacturer marketing depts really suck when it came to "simulated" (FAKE!) images. Or cherry picked, just as bad.

I know why you posted it, and we've now pointed out the surreal nature, so that's all it needed. Context. :salute:

@lordsmurf: In case you checked out the aforementioned samples, would you also agree that S-video with the NR function of the EH495 enabled produces the most feasible results for LD in light of the described workflow and the hardware at hand?

To make a more informed decision, we need another sample from the LaserDisc, a short clip with high spatial frequency like a pattern on on a person's shirt, both moving across the screen and not moving.