Do not watch HVX files on Quicktime Player!!

S

SPZ

Veteran
Someone posted a comment concerning this in another thread, but I tihink this deserves a separate thread, since the noise issue is pretty much exagerated by this:

If you play the HVX files in Quicktime player, the picture will be brighter and the colors washed out due to some codec interpretation error. If you see the same files on Final Cut Pro, the noise will disappear, and the true beauty of the dinamic range and color will stand out. Judge the picture from the Final Cut output, not the Quicktime playback!
 
If you get bored, see if this is affected by using the "High Quality" setting in Quicktime Player...
 
kenn michael said:
It's not a High Quality setting issue - that's mainly for watching DV clips... It's actually a gamma issue.
Click to expand...

Kenn, sorry, I should have quoted you on the initial post. It was your post that made me think about the diferences when seing the footage in the quicktime player and FCP. And the difference really exists!
 
So there really is no "noise " issue with the camera itself?
I always thought the Karate footage looked damn good myself.
Nice detail in shadows, unlike other cameras in this price range, but I don't use quicktime viewer.
 
Green Hornet said:
So there really is no "noise " issue with the camera itself?
I always thought the Karate footage looked damn good myself.
Nice detail in shadows, unlike other cameras in this price range, but I don't use quicktime viewer.
Click to expand...

No - there's a REAL noise issue. I've got FCP and have seen the MXF footage. I downloaded those raw P2 clips in another thread here and the noise totally popped out at me while other people were chanting "we're not worthy" so maybe I'm just anal or maybe if it says "Panasonic" it looks better to them now matter what it looks like.

I've also shot an HVX with a friend who has the footage. To quote his initial reaction: "Great camera if the noise could be cleaner". This is a guy with a Sony ZR1 saying that. But he also shoots a lot of F900, so he could be jaded by that.

Barry has noted Cine-Like-D is noisier and that's what we shot. So... if you don't think Cine-Like-D is a good shooting mode on the DVX100, problem solved.

I'm waiting to see how this shootout goes before I really decide. Barry did offer the BEST piece of advice. Rent each of these cameras and THEN decide.
 
a swag at why this is:

Keep in mind that FCP is applying gamma correction to compensate for the noise and knock it down (hence the image is more saturated). It basically prefers chroma. QT player isn't. If you change your FCP prefs to approximate (from the default accurate---bad choice of names but another topic as to why FCP does this), you should see the noise return a bit as a result of the desaturation.

The noise is there in the luma it's just whether or not it's above or below the threshold of visibility given the processing applied to it by the application. QT player doesn't deploy as heavy a gamma correction and thus you see it more.

This isn't a knock against the footage orthe camera. All cameras and codecs involve these issues. You see this with much HD footage (the differences).
 
Brad Johnson said:
Keep in mind that FCP is applying gamma correction to compensate for the noise and knock it down (hence the image is more saturated).
Click to expand...

So you're saying view it in QT to get a BETTER indication of the real noise because FCP will artificially try to fix it?
 
what I'm saying is that your viewing conditions (not just monitor but also application) affect the amount of noise you see in any codec. We are used to thinking that different monitors will show the footage different (e.g., HDV footage looks great on a 50" screen at the local pub but less so on the cinema display, but better on an SD CRT).

We need to apply that same understanding to the SW and include those same kinds of discriminations when we view the images. It's not "use this, don't use that" application (as though there is a flaw in the app) but rather understand that by using X application versus Y application, the results will vary in a particular way.

FCP's "accurate gamma" setting will replicate an SD CRT delivery environment better and will be more forgiving to the footage (in my experience).

sorry that I can't just say "yes, QT is more accurate" -- the noise is in the footage, FCP just handles it more graciously. But throw a bleach bypass on it, or grade it out and see what's lurking below in the luma to really tell.
 
The DVX is a very noisy camera... the 100b helped quite a bit but still, much noisier than an XL2... that didnt stop lots and lots of people from creating great images, shorts, features, etc. Noise is the price you pay for better colors in the 1/3" CCD world. I suspect the HVX/XLH debate will be similar to the DVX/XL2... one has more res, the other more pleasing gamma...


ash =o)
 
FCP making IMMEDIATE correction from the HVX footage? I don't buy it. FCP is a PROFESSIONAL application and it should represent acuratelly what the image is, without any correction of any kind. This is up to the user. I viewed my files on FCP 4.5 and there really wasn't any HVX specs or footage for Apple to try to interpret its footage avoiding the weaknesses of the camera...

I've seen the footage in FCP and it really is nowhere near as noisy as in quicktime. The colors look a lot better as well. Quicktime DVCPROHD codec was meant to be used on FCP, no on the player. I believe something went wrong with the integration in the player.

The noise is barelly noticeable. I haven't seen anything on a plasma, tough. However, don't forget Michael Pappas saying he saw dvx footage on a 50" Plasma and it looked GREAT. And Michael is know for is Xl love... :)
 
If we're talking about a gamma issue, which I believe we are, then the issue isn't FCP "covering up" noise to make the image look better, rather it's where the noise falls on the IRE scale when it's being shown. The noise is still there, it is less obvious because the image is being shown correctly. QT is artificially boosting the noise present in the footage, because it is designed for watching video on a computer, and FCP is designed to present video in a way that will make it possible to make color correction decisions. Granted, the only way to truly make a decision about noise in the footage is to look at the image on a CALIBRATED HD monitor. If you can put a pluge on it, you can set the brightness and contrast controls so that you can make a fairly clear estimate of if the gamma is being represented accurately. That said, lacking an HD production monitor, I'm more inclined to trust the FCP output on the LCD rather than the QT. Anyone got an HD editing suite? Still, the noise does remain in the realm of subjectivity, not that it exists or not, but how offensive it is. In FCP, it is far less offensive than in QT.
 
I just took Kaku's night footage that has been floating around and played it in FCP for the first time at 100% and it looked a TON better. The noise is not nearly as noticeable. When I first saw this footage in QT Player I had a pit in my stomach because I thought the camera has a noise "issue." I'm beginning to think that QT Player has been messing things up for a lot of us.
 
doccutter,

I don't believe I'm saying anything different than you. Perhaps "covering up" is too imprecise, but the result is the same (it displays differently--better--than the player but that doesn't mean the footage is cleaner. Agreed on the HD Monitor of course).

In FCP, the result can change since it's a preference--you can choose to have it display more for color correction for a CRT television (switch it to "accurate") or more like the QT player (change the pref to "approximate"). In other words I can make it look like the FCP that SPZ mentions above or like my QT player version of it. It's a choice in FCP.

So, to SPZ's rebuttal, you don't need to buy it from me--go into your system settings prefs and change the gamma correction and you'll notice a shift similar to the way it displays in QT (which only uses the gamma values that FCP calls "approximate"). Or, read on page 517 of your FCP manual which states:

"However, when the video from your camcorder is shown
on your computer display, you often see the image brighter than expected, because
the image was intended for playback on a video monitor. Gamma correction must be
applied to the video so it appears as intended."

My point is simply that of course FCP looks better and the take away is that we need to consider the differences in software like we do the hardware--we always have, but it was perhaps less noticeable for those only working in SD.

This is fairly reminescent of the DV codec wars of the 90s....Promax's versus Apple's etc.
 
I downloaded the MXF files that had the green screen and the water dropping in slow mo into the hand. i thought they looked beautiful. I even encoded them to DVD and played them on my 34" widescreen HDTV (Sony CRT), and they looked awesome.

I'm not worried about any issues.
 
Someone said that the wmv's files of the dvcprohd footage look a lot better than the dvcprohd footage viewed on the quicktime player. Can someone confirm this? If this is indeed true, than I really can only interpret this has some kind of codec interpretation error from the quicktime player. I've seen the wmv 60p overcrank and undercrank footage from the japanesese website (with the old man and with the Audi A6, if I'm not wrong), and it looked really good. Haven't seen the qt counterpart, due to not having quicktime instaled in the office PC, so i can't really judge the differences. Can someone compare both files or take a snapshot from them? This could help alot! (Sorry for not mentioning the guy's name. I really like his footage, tough .He's not Kaku Ito- who also does a great job in presenting hands-on impressions!)
 
SPZ said:
I've seen the wmv 60p overcrank and undercrank footage from the japanesese website (with the old man and with the Audi A6, if I'm not wrong)
Click to expand...

Do you have a link for that?
 
You must log in or register to reply here.
Back
Top