A Real Close Call

[Lou Bruno]

Just got back from a shoot comparing the SONY FX-1 with the CANON XL-2 in widescreen 16:9. I set the SONY in HDV.

I played back both tapes on a SAMSUNG 32" LCD WIDESCREEN.


I first placed the CANON tape in a minidv player using a S-VHS as an output cable.

I placed the HDV tape into one of my SONY HD decks.

When I compared the quality in widescreen the HD quality was slightly more vivid compared to the S-VHS output of the Canon XL-2.

Now here is where it got interesting.

I placed the Canon "footage" into the SONY HD deck. The output is COMPONENT. The picture quality of the two cameras were so close that only an A&B comparison could actually reveal the very SLIGHT difference.

I can only imagine if a HDMI cable was used to output the CANON picture how even better it would look as it is in a true digital realm. The CANON 16:9 will definitely bridge the gap in the next few years before a complete turnover to HD/HDV etc.

NOTE:
I did notice, that I will have to adjust the CORING a little in the CANON "footage." Must be the nature of the RGB isolation. My settings were Green 0 RED and BLUE 2 and Saturation 3. Setup -2.

ALSO.....the SD picture is more defined in the Canon than my Sony.

[Mark Sasahara]

The S-video is a lower quality signal. While it does separate Luminance and chrominance, the inherent quality of the picture ain't good, it's an analog signal.

You're better off going through the BNC video out, or the RCA video out on the camera.

S-VHS is Super VHS format. SVHS was developed by JVC and gives 400 horizontal lines of resolution, an improvement over the normal 250 lines of VHS, but still less than broadcast TV's 500 lines. Sorry, seeing "S-Video" called "SVHS" is a pet peeve.

I'm confused, did you actually play the cassette with the Canon XL2 video in the Sony HD deck? Isn't the pitch for MiniDV different than for HDV?

[Lou Bruno]

I should have written YC connector rather than S-VHS. Most of my clients do not know what a YC CABLE is but rather relate to a S-VHS cable or output.

The RCA composite is actually lesser quality by 10%. Then YC out, Component, then finally HDMI.

Both RCA and YC OUT are analog as is COMPONENT albeit seperate RGB.

HDMI and DVI are true digital cables with HDMI being able to transfer a digital audio signal as well as a video signal......much like a 1394 cable.

BNC is nothing more than a connection that allows the RCA jack to be locked into a professional player/recorder unit with a male and female type lock terminal.

Any HD deck will recognize the format introduced to it. If a 1080 i or p/720 etc signal is played within the unit, the player/recorder will recognize the HDV format as GOPS. It will either play the HIDEF signal through the component or firewire cable either to another deck or NLE system. Of course, if the deck or NLE system is set up to INPUT HDV, then the signal will NOT be downconverted.

If SD is played in a HD unit, that will be recognized as standard definition which requires no down conversion as is the case with HDV.
The pitch of a tape can vary from deck to deck regardless of tape format.
Luckily, the SONY will play the MINIDV from CANON as the pitch is the same.
The problem arises in HIDEF when JVC vs. SONY and most other camera HIDEF recordings. JVC uses a unique form of GOPS (Groups of Pictures) which can only be read in a JVC unit.....go figure :-(

Tape pitch is mostly a universal standard in SP mode. With LP Mode, just like in the old analog world, time base errors occur.

Due to the seperation of colors in Component, I assume the picture surpassed that of both RCA and YC connections. The HDV deck must actually convert the digital signal to analog to use the component out of a HDV DECK.

In closing, the picture quality (to my eye) looks great from the Canon and only a side by side comparison would actually reveal a slightly better picture using my two SONY HDV cameras.......

Originally Posted by Mark Sasahara

The S-video is a lower quality signal. While it does separate Luminance and chrominance, the inherent quality of the picture ain't good, it's an analog signal.

You're better off going through the BNC video out, or the RCA video out on the camera.

S-VHS is Super VHS format. SVHS was developed by JVC and gives 400 horizontal lines of resolution, an improvement over the normal 250 lines of VHS, but still less than broadcast TV's 500 lines. Sorry, seeing "S-Video" called "SVHS" is a pet peeve.

I'm confused, did you actually play the cassette with the Canon XL2 video in the Sony HD deck? Isn't the pitch for MiniDV different than for HDV?

[Mark Sasahara]

Thanks Lou.

[Greg Boston]

Originally Posted by Mark Sasahara

The S-video is a lower quality signal. While it does separate Luminance and chrominance, the inherent quality of the picture ain't good, it's an analog signal.

You're better off going through the BNC video out, or the RCA video out on the camera.

I beg to differ Mark. S-Video has the chroma and luma signals separated already and avoids having to have it done in the display as would be the case with composite video where everything has to be separated out before it can be used. This separation step of composite video introduces degradation.

Now here is where it gets really interesting. I saw rumors a year ago or so that the S-Video output of the XL2 was pre-compression much like the component output of the Sony Z1. I mentioned this in another thread recently, and another DVINFO member actually emailed me video samples of a shot that was captured via firewire, and the same shot captured with a Kona card via S-Video. There was a VISIBLE quality difference of the color gradation on the S-Video signal. It looked superior to the digital firewire acquisition. I have to qualify that by saying the S-Video was a live capture and not coming off the tape. But, so was the firewire version.

There are some variables to consider in this method but the end result was a much smoother picture with virtually no color banding.

-gb-

[Mark Sasahara]

Greg, Thanks. I had always thought that S-Video was not the way to go. But as you say, I guess it depends on what's done, or not done before it goes out the camera. According to Billups' Digital Moviemaking book S-Video is an inferior signal with lower resolution. I guess, it ain't necessarily so.

The Firewire vs S-Video out is a little test I'll have to try.

[Greg Boston]

Originally Posted by Mark Sasahara

Greg, Thanks. I had always thought that S-Video was not the way to go. But as you say, I guess it depends on what's done, or not done before it goes out the camera. According to Billups' Digital Moviemaking book S-Video is an inferior signal with lower resolution. I guess, it ain't necessarily so.

The Firewire vs S-Video out is a little test I'll have to try.

Creating the composite signal in a 3 chip camera requires mixing the red, green, and blue discrete colors of each ccd into a chroma, luma, and sync pulse signal train. The chroma is actually color difference called Y-R and Y-B There is no Y-G because if you know the percentage of red and blue, then the rest has to be green. The predominant color in most scenes is green (now you know why they pixel shift the green to increase 'apparent resolution) so they don't include it to conserve bandwidth. In doing all this conversion and mixing, you have degraded the original signal somewhat at the composite RCA and BNC jacks before it ever gets out of the camera.

Now, it gets to the other end and whatever is attached has to split all that stuff back apart creating another quality hit. This is why proc amps are used because the first thing to get lost in these conversions is higher frequencies which equals detail. Among other things, the proc amp boosts the high frequencies to help combat the quality loss before being input to the next device in the chain.

I haven't read the book you are referring to, but what he likely meant is that digital is a direct one for one copy so the generational loss I described above is virtually non-existent. The digital version is going to be higher resolution pixel wise, but the chroma sub-sampling at 4:1:1 for mini-dv and dvcam will degrade the chroma resolution somewhat and that's where the s-video helps. It smooths out the 'digital edge' of subtle color variations. That's the thing I was talking about in my earlier post.

Most everything will have a composite video input. Somewhat fewer items will have S-Video input. Even fewer will have component input cause it takes more circuitry to employ but yields higher quality because nothing has to go through the mix and separate stages that induce quality losses.

Sorry to be so long winded. ;-)

-gb-

p.s. Some finer technical points have been omitted for brevity.

[Mark Sasahara]

Greg, thanks again. Your explanations are great, no apology neccessary. Looks like I need to do a lot of re-reading.

[Lou Bruno]

You're Welcome..

Originally Posted by Mark Sasahara

Thanks Lou.

[Chris Barcellos]

So have you three determined that the high definition in HDV is only a myth and that we really don't need it because we have S Video out from the XL2 ?

[Lou Bruno]

Component out from the HD deck gives a clearer picture than S-Video.
HDV is not really a myth...it is awesome. The XL-2 can bridge the gap for the next few years as the native 16:9 chip do an excellent job in regards to widescreen.

Originally Posted by Chris Barcellos

So have you three determined that the high definition in HDV is only a myth and that we really don't need it because we have S Video out from the XL2 ?

[Bruce S. Yarock]

I have an XL2 and an FX1. Viewing the FX1 via component and the XL2 via rca or s-video on my 30" Sony Vega hdtv, there's is NO comparison. I love my XL2, and it has a great image, but the increased resoltion of hdv is quite impressive.
Bruce S. Yarock
www.yarock.com

[Richard Hunter]

I think the display device is critical for this type of comparison. First, the panel's native resolution will determine whether it can display the 1080i image directly, or whether the video has to be resampled. Second, if resampling is required, the method used can affect the outcome a lot. Examples below are the extreme cases.

Example 1: the LCD panel has a vertical resolution of 768 pixels, this means it cannot display the FX1's 1080i without resampling. Further, let's say the panel resamples 1080i by discarding alternate fields (to give 540p) and then upsampling to 720p. This would not provide much better resolution than the XL2's 480p when it is also resampled to 720p.

Example 2: the LCD panel has a vertical resolution of 1080 (there are a few of these on the market, although 1024 is much more likely at the moment). The FX1's 1080i can be displayed directly (ignoring interlacing artifacts), but the XL2's 480p has to be resampled up to 1080. In this example, the FX1's output should show significantly higher resolution than the XL2's, unless the lens on the FX1 is total rubbish or smeared with Vaseline.

So my question to Lou is, what is the resolution of the Samsung display? (Not challenging you, just very interested!)

Richard

[Lou Bruno]

1366(H) x 768(V) pixel resolution......4000:1 Dynamic Contrast.
What do you think?

Originally Posted by Richard Hunter

I think the display device is critical for this type of comparison. First, the panel's native resolution will determine whether it can display the 1080i image directly, or whether the video has to be resampled. Second, if resampling is required, the method used can affect the outcome a lot. Examples below are the extreme cases.

Example 1: the LCD panel has a vertical resolution of 768 pixels, this means it cannot display the FX1's 1080i without resampling. Further, let's say the panel resamples 1080i by discarding alternate fields (to give 540p) and then upsampling to 720p. This would not provide much better resolution than the XL2's 480p when it is also resampled to 720p.

Example 2: the LCD panel has a vertical resolution of 1080 (there are a few of these on the market, although 1024 is much more likely at the moment). The FX1's 1080i can be displayed directly (ignoring interlacing artifacts), but the XL2's 480p has to be resampled up to 1080. In this example, the FX1's output should show significantly higher resolution than the XL2's, unless the lens on the FX1 is total rubbish or smeared with Vaseline.

So my question to Lou is, what is the resolution of the Samsung display? (Not challenging you, just very interested!)

Richard

[Richard Hunter]

Hi Lou. Thanks for the info. I would expect that a different display, especially a higher res one, would show the Sony in a better light if it didn't have to resample the video.

But I also think that your main point about the XL2 bridging the gap to HD is totally valid. We cannot always control what our videos are going to be shown on, so it may not make a lot of sense to obsess about the highest resolution camera if the display is going to be the great leveller. Hopefully this situation will change when 1080 panels become commonplace.

Richard