Possibly giving up Sony SVHS deck in exchange for JVC 9911

[VideoTechMan]

Hey everyone, im hoping that I have this post in the right section since its dealing with restoration, since it involves the question of the JVC 9911 SVHS unit.

I have read of course that the JVC 9911 model was the top choice for getting good captures from VHS tapes due to the built in DNR and TBC circuitry inside.

My question is, my Sony SLV-R1000 SVHS deck which I have had for several years has done great service (and still does) I have thought about selling it on ebay or something in order to purchase the JVC model in place of it as to get the most out of my VHS tapes for capturing. I also have the Datavideo TBC-1000 which I could also use in conjunction with the JVC unit to get the best that I can out of my VHS captures. From what I read here on the site regarding the JVC and TBC-1000, they are a great combination for doing VHS captures at the best quality possible (according how well the source is of course).

I wanted some thoughts on whether I should give up the Sony deck (which I paid $1,000 for it new several years ago and still in great shape) for the JVC HR9911 which has the additional circuitry to help clean up the video before it gets to the computer. Would hope to grab the JVC unit while its available before they stop producing them if I find it would be more beneficial for my video work.

Any thoughts would be greatly appreciated.

VTM

[FulciLives]

Well using pricebrabber.com I found the JVC S-VHS 9911 at B&H PHOTO for $359.00 which was the lowest price that pricebrabber.com could find.

There are several other websites like pricebrabber.com such as mysimon.com and CNET Shopper etc.

So look for the cheapest place you can buy it brand new.

- John "FulciLives" Coleman

[lordsmurf]

I would buy and test the JVC first before selling the Sony. Make sure you really like it (which you probably will).

[pfh]

Buy the new one. The last couple weeks I've spent looking on ebay the jvc 9900 and 9600 are way over priced due to demand. For the prices they are getting + shipping you end up near the cost of brand new 9911 IMO. B&H is the cheapest. I found an old 6600 from a studio on ebay for $100 that I'm gonna try out but i refuse to buy a used 9900 for $300 for the above reason. If this 6600 + sima color corrector doesn't work out for me then I'll buy a new 9911.

[davideck]

VideoTechMan -

I have a JVC 7600U(with TBC/DNR) and a JVC 6800U (without TBC/DNR).
My preferred choice is to capture with the 6800 because it provides a more detailed image with less artifacts than the 7600. I will use the 7600 if the tape was recorded in 6 hour mode with a VCR other than the 6800. The TBC/DNR processor can make an old noisy tape look much better, but it also introduces its own set of artifacts into the picture. If your tapes are reasonably good quality then your SONY may well provide better results, particulary if the tapes were originally recorded on it.

[lordsmurf]

The TBC/DNR processor can make an old noisy tape look much better, but it also introduces its own set of artifacts into the picture.

I'd like to see some evidence to back up that claim.

Whie I know no VCR is perfect, the likelihood of the JVC "introducing" an artifact is rare, in the many years I've been using the JVC line of equipment.

The "fix" for this is more often to find a VCR that is equally as screwed up as the source tape (comparable misalignment). I actually PURPOSELY have one VCR that is "wrong" in order to play "wrong" tapes.

[gshelley61]

The Sony SLV-R1000 / SVO-2000 (same unit) has a detailed and sharp picture, but it can be a little too grainy. Not a big problem if you're capturing with a DVD recorder that has video input noise reduction (like JVC, Pioneer, etc.)

However, there is no question in my mind that the JVC S-VHS units with the 4MB TBC/DNR processor deliver a smoother, less noisy image than the Sony. I've owned them both at the same time. The JVC's TBC/DNR can be shut off... then you have a picture very similar to the Sony's anyway. It's really nice to have the TBC/DNR feature for most tapes - it does a great job.

I agree with LS... get the JVC and check it out before you get rid of your Sony. If you don't like the JVC, you can re-sell it on eBay for very nearly what you will be paying to get it.

[davideck]

The TBC/DNR processor can make an old noisy tape look much better, but it also introduces its own set of artifacts into the picture.

I'd like to see some evidence to back up that claim.

The JVC DNR function introduces a grain on any surface that is moving.
It is also noticeable at the edges of moving objects as a grainy trail that is temporarily left behind. The amplitude of these artifacts increases with the noise level of the source tape.

For a given tape, I would also claim that my 6800U provides a more detailed picture than my 7600U with DNR/TBC turned on.

I am not suggesting that these artifacts always overshadow the benefits of the noise reduction. But with good quality recordings where DNR is not necessary, a better capture may be obtained by turning the DNR off. With the JVC units, this means turning off the TBC as well. In this situation the SONY VCR in question may be every bit as good, or even better, than a new 9911.

[lordsmurf]

The JVC DNR function introduces a grain on any surface that is moving. It is also noticeable at the edges of moving objects as a grainy trail that is temporarily left behind. The amplitude of these artifacts increases with the noise level of the source tape.

That is totally ridiculous. In the many years I have used these machines, and the many dozens (maybe hundreds) of other uses that I have spoken to in the years, none have ever reported anything of the kind.

The DNR is there to REMOVE GRAIN .. not add it. The image is also not temporally augmented (not by much anyway), so "trails" are not possible.

Maybe yours is broken. Or some other factor elsewhere in your setup.

[davideck]

The image is also not temporally augmented (not by much anyway), so "trails" are not possible.

The JVC DNR is a temporal averaging filter. Trails can result from temporal averaging. They are indeed possible.

I have both a 7600 and a 9600. They both exhibit these artifacts. Take a critical look at the JVC output on a good monitor, and you'll see what I'm talking about.

[lordsmurf]

It's not temporal ... not really. Maybe within a few frames at most. But seeing as there are almost 30 frames per second, your eyes cannot see that. It's intra-frame more than anything else. The RAM buffer cannot hold enough info for true temporal processing, a measley 2-4MB won't cut it for harsh temporal control.

If you want harsh temporal filters, you need a modern computer, or a system that is built like one. Gross amounts of RAM/CPU is needed for temporal filtering like you suggest. Your buffer would have to hold, or have access to, up to 90 frames at any given time, about 3 seconds worth. At minimum, maybe 30-45 full frames for even the most minute trails.

I don't see what you're talking about, nor does anybody else I've ever met.

This is the reason I always have to tell people to not get carried away with NR filters in software. Newbies can get filter-stupid too easily, making flat/plastic images full of trails.

[davideck]

It's not temporal ... not really. Maybe within a few frames at most. But seeing as there are almost 30 frames per second, your eyes cannot see that. It's intra-frame more than anything else. The RAM buffer cannot hold enough info for true temporal processing, a measley 2-4MB won't cut it for harsh temporal control.

If you want harsh temporal filters, you need a modern computer, or a system that is built like one. Gross amounts of RAM/CPU is needed for temporal filtering like you suggest. Your buffer would have to hold, or have access to, up to 90 frames at any given time, about 3 seconds worth. At minimum, maybe 30-45 full frames for even the most minute trails.

I don't see what you're talking about, nor does anybody else I've ever met.

This is the reason I always have to tell people to not get carried away with NR filters in software. Newbies can get filter-stupid too easily, making flat/plastic images full of trails.

A filter that uses information from more than one frame is by definition a temporal filter. I am not talking about harsh temporal control.

If you recirculate the data in a single frame buffer with an input source, you can average across as many frames as you like. 90% buffer and 10% input will hold data from 10 frames. 99% buffer and 1 % input will hold data from 100 frames. All with a single frame buffer.

A 30-45 frame trail would last at least 1 second and be very noticeable. I am talking about artifacts much less objectionable. A trail that lasts for only several frames is noticeable.

[lordsmurf]

A trail that lasts for only several frames is noticeable.

When you have 30 frames per second .... no way. You're imagining a problem that does not exist. "Seeing" a couple frames worth of overlap would mean you have unique eyeballs that no other human in history has ever had, namely that you can perceive all 30 frames.

Human eyes only usually see about 10-20fps. The only reason tv even looks good to us is because our species is genetically stupid when it comes to vision.

This sounds like an expansion of the marketing BS purveyed by video games and game cards, along with all those stupid nerds that insist they can magically see the difference between 45 and 90 fps. What they see is a game that looks like shit when it's played slower than what it was designed for. It's not the fps, or any other factor.

To say you can discern a couple of frames overlapping/trailing is quite honestly ridiculous.

Again, maybe your machine is broken.

A filter that uses information from more than one frame is by definition a temporal filter. I am not talking about harsh temporal control.

To say you can see "trails" ABSOLUTELY means it is harsh filtering. Don't play semantics with me and spout off dictionary definitions.

By standard practice, what is happening on the JVC is intra-frame NR.

Your assumptions about the "recirculation" of data is also totally incorrect. That's not how it works. There is simply not enough RAM for anything you have suggested.

In fact, the difference between the 2MB and 4MB is the 2MB is more susceptible to just not work at all on really bad tapes. The tape can overload the buffer. It will clean some errors, but not all of them, just cannot handle them all.

[davideck]

I have done lots of experiments with recirculating framestores as part of my job.
Perhaps you did not understand my explanation.

Many of us in the US are sensitive to the 25hz flicker that we see with PAL sources.
A two frame trail would approximate a 15 hz flicker; easily noticeable.

[VideoTechMan]

Thanks for your replies and suggesstions. I will look into trying out the JVC before putting the Sony up for sale, but in most cases, I have a good feeling JVC will be a great keeper, not to mention it will take up less room on my AV setup. And most likely be my final SVHS vcr to buy because once I am done getting the material from my tapes I will be done with VHS, unless I need to record some stuff for those who still prefer VHS (like my mom).

Alot of my older VHS tapes are about 14-17 years old, and were recorded in EP mode (back in the days where it didnt matter much on tape speed). Fortunately most of them are still in good condition. Ever since coming to this forum I started doing my recordings in SP mode for best quality possible.

Reading on the 9911's specs, it looks the type of VCR I would need to give my VHS tapes a little boost.

Lordsmurf, how well does your 9911 (i believe you have?) and the TBC-1000 work together with old but reasonably good VHS tapes? I know the majority of the older tapes I have were recorded on a 2 head VCR back in the days.

VTM

[lordsmurf]

I have the 9800 and I recently added an SR-V10U. I also have many hours logged on a 7900. I also have a large handful of ancient mono tapes. This combination of hardware does some pretty amazing things, in terms of quality repairing. It's almost hard to explain. Using the DNR/TBC is like removing a see-through curtain, so you can actually see the truly clean image. The standalone TBC stabilizes any issues the tape may have had, which may or may not show visible improvements, but will always yield better cooperation by the digital equipment, which in the end, is a quality issue in and of itself.

[proxyx99]

From the link that was provided in another thread:

"The prototype we saw was a 36" widescreen set with 720p resolution. Production models (which should begin appearing in very limited numbers by the end of this year) will be in the 50"-55" range, with resolution of 1080p. SED contrast ratio is rated at a mind-boggling 8600:1. The technology is similar to having 2 million tiny picture tubes. An array of electron emitters (one for each pixel) creates images by firing electrons at the phosphor-coated screen.

For the picture quality comparisons, the SED was flanked on the left by a similar-sized plasma panel, and on the right by an LCD. Video material consisted of a series of high-definition clips fed from a prototype HD DVD player. SED's superior color and black level were immediately apparent. Probably the most striking feature was SED's amazing pixel response time (how fast each pixel can switch on and off). When strings of alphabet letters scrolled quickly across the screens, individual letters remained clear and distinct on the SED, while some blurring was visible on both the plasma and LCD. With an incredibly quick claimed response time of 1 millisecond, SED can keep up with sports and other fast-action video, creating a smoother, more natural look.

Toshiba and Canon are positioning SED as the new "high-end" flat-panel TV technology. Based on this demo, it appears to be the real deal! "

As you can see trails may be visible even on a high end displays and to eliminate them you need to go much lower the 1/30 of the sec. in this example it is 1 ms. Just popped into my mind while reading this thread. 1 frame delay would be quite visible I think.

http://www.crutchfieldadvisor.com/S-Zlq2oF0Bj4i/reviews/20050119/ces_TV.html?page=1

[lordsmurf]

That's a phosphor delay, not really the same thing. Those last a lot longer, and are very similar to harsh temporal NR. The "as you can see" statement also has no image to "see".

[davideck]

A trail that lasts for only several frames is noticeable.

When you have 30 frames per second .... no way. You're imagining a problem that does not exist. "Seeing" a couple frames worth of overlap would mean you have unique eyeballs that no other human in history has ever had, namely that you can perceive all 30 frames.

To say you can discern a couple of frames overlapping/trailing is quite honestly ridiculous.

lordsmurf -

Suppose you display a 50-50 mix of the current frame with the previous frame.
That would be equivalent to a one frame trail. Are you claiming that you would not notice the difference between this 50-50 mix and the current frame itself?

Deinterlacing is another good example. It is necessary because the temporal displacement between adjacent fields is enough to be visible. Objects can move significantly across the screen from field to field. Viewing both fields simultaneously (without proper deinterlacing) is equivalent to a one field trail; 1/60 of a second. Even that is visible. Are you claiming otherwise?

[proxyx99]

That's a phosphor delay, not really the same thing. Those last a lot longer, and are very similar to harsh temporal NR. The "as you can see" statement also has no image to "see".

Figure of speach But with your experience... all you need to do is close your eyes to see...

[davideck]

Your assumptions about the "recirculation" of data is also totally incorrect. That's not how it works. There is simply not enough RAM for anything you have suggested.

lordsmurf -

You are misinformed here.

ONE Framestore is all that is required to produce a temporal average over as many frames as you like. During every frame, the output of the framestore is mixed with a new input frame and written back into the SAME Framestore. That is how a recirculating framestore works.

Special Effects like motion blur and image trails are done in precisely this manner.

Your claim that huge amounts of RAM are required is incorrect.

[lordsmurf]

I still think you're seeing things that do not exist, and I don't think your understanding of the tech is correct either. So I guess we'll just have to agree to disagree here.

[davideck]

But lordsmurf

You never replied to my post...

A trail that lasts for only several frames is noticeable.

When you have 30 frames per second .... no way. You're imagining a problem that does not exist. "Seeing" a couple frames worth of overlap would mean you have unique eyeballs that no other human in history has ever had, namely that you can perceive all 30 frames.

To say you can discern a couple of frames overlapping/trailing is quite honestly ridiculous.

lordsmurf -

Suppose you display a 50-50 mix of the current frame with the previous frame.
That would be equivalent to a one frame trail. Are you claiming that you would not notice the difference between this 50-50 mix and the current frame itself?

Deinterlacing is another good example. It is necessary because the temporal displacement between adjacent fields is enough to be visible. Objects can move significantly across the screen from field to field. Viewing both fields simultaneously (without proper deinterlacing) is equivalent to a one field trail; 1/60 of a second. Even that is visible. Are you claiming otherwise?

[davideck]

Hey lordsmurf -

I feel that I am entitled to a reply from you.

You have made bogus claims against my eyes (and my credibility) that demonstrate your lack of understanding on some fundamental video characteristics.

[Scooter_NJ]

I have the same Sony you do and my advice is DO NOT GET RID OF IT!!!! If you can be in a position to afford the JVC as well...then add that but don't sell the Sony! That machine is a workhorse and DAMN reliable!

[davideck]

lordsmurf - I see you posting out there... Still no reply?

Scooter_NJ - I agree completely.

In fact, I would claim that the introduction of the TBC/DNR allowed JVC to reduce the quality of their transports. As a basic VCR with TBC/DNR turned off, my 7600 and 9600 units provide only average playback quality. The transports are vitually identical to my 3600, and many parts are interchangeable. The TBC/DNR helps them out a lot.

I think the units that had the best transports were those just before TBC/DNR.
Its hard to beat an HR-S5800 or an HR-S6800 (and perhaps that SONY you're talking about).

I would claim that for good quality tapes, my 6800 without TBC/DNR makes better captures than my 7600 with TBC/DNR. My 7600 was in great shape when I bought it, but I rarely use it for capturing. I wish the DNR could be turned off separately.

One little known gem is the HR-S5400. I have two of them because I bought a spare! Very stable and responsive transport. GREAT picture quality. Small, lightweight, and really rugged. Neat little remote.

So hang on to that SONY. You may not find a better VCR anywhere...

A trail that lasts for only several frames is noticeable.

When you have 30 frames per second .... no way. You're imagining a problem that does not exist. "Seeing" a couple frames worth of overlap would mean you have unique eyeballs that no other human in history has ever had, namely that you can perceive all 30 frames.

To say you can discern a couple of frames overlapping/trailing is quite honestly ridiculous.

lordsmurf -

Suppose you display a 50-50 mix of the current frame with the previous frame.
That would be equivalent to a one frame trail. Are you claiming that you would not notice the difference between this 50-50 mix and the current frame itself?

Deinterlacing is another good example. It is necessary because the temporal displacement between adjacent fields is enough to be visible. Objects can move significantly across the screen from field to field. Viewing both fields simultaneously (without proper deinterlacing) is equivalent to a one field trail; 1/60 of a second. Even that is visible. Are you claiming otherwise?

[swiego]

This has been an informative topic. I have the 9911U and have been thinking long and hard about the R1000 after having quite a bit of experience with some of its immediate younger brouthers in the SLV range. The 9911U has been a mixed blessing so far. The Digipure does improve picture quality - generally. In some cases it is pretty dramatic. In other cases, it is slight. I am mostly pleased, but not terribly enthusiastic. The Sony puts out a very, very good picture as well. There is no TBC, but I think the electronics are of equal or greater caliber.

The biggest issue with the 9911U is the build quality. I've never been so aghast at what must be the most flimsy consumer product with a $500+ retail pricetag. No wonder B&H and Amazon sell it for nearly half its retail. Like others, I think you should try the JVC but hang onto the Sony in case you are not satisfied. After getting used to upper end Sony build quality for their SVHS decks, the 9911U may feel like a huge step back.

[davideck]

swiego -

I feel the same way about my 7600 and 9600 decks as you do about your 9911. I would really miss my 6800 and 5400 machines. But the Dynamic Drum System is really cool.

If you're aghast at the 9911, you should see the 29XX, 39XX, 59XX series.
It is hard to appreciate the mechanical transport tolerances required to playback a VHS tape when almost eveything you see is plastic. I bought a brand new 2902 for $69 and took it back a week later because I didn't think it was worth it.

Strange paradox how VCR playback quality has become really important at a time when good VCRs are no longer readily available.

[gshelley61]

The Sony SLV-R1000 (SVO-2000) are also cheaply made and full of plastic parts. Your best bet if you want a rugged, precision tape transport is to keep your eyes open for an older high end consumer machine, or a professional broadcast unit (they are very large and heavy, though)

I had a Panasonic AG-7750 on hand for a short while, and it was awesome in the build quality department. Like a friggin' tank. 50 lbs. and huge.

[swiego]

I'm surprised that the R1000 would be cheaply made. I have the SLV-595HF and SLV-780HF, both lower models on the same rung. Yes, they do have lots of plastic, but so does my Miele vacuum cleaner and like the Miele, I'm pretty sure these decks are going to last forever. I don't feel nervous pressing buttons, loading or ejecting tapes, or working the transport.

The JVC is a different affair. It's made of the same materials as my old 595HF, but everything is 1/2 or 1/3 the thickness. Labels are peeling off. The things whirs and whizzes forever while trying to eject a tape (this is brand new, mind you) and I find myself crossing my fingers each time I use it, muttering under breath, "please don't break, please don't break." It really is a cheaply made product, which the older Sonys are not. I can't speak to the SLV-R1000 though, although I would hope that it's of sufficient quality that you wouldn't worry about getting some years' use out of it.