What is the image like viewed from DVD

Elton said:
Have any of you guys tried downconverting with MPEG Streamclip into a better SD codec like DV50 or PhotoJPEG, animation codec etc.?

Streamclip gives me phenomenal results, and the key is downconverting to a better SD codec than plain-jane DV. When I downconvert to 16x9 SD with the PhotoJPEG codec at 100% quality I get color very close to 4:4:4. THAT is where shooting HDV for SD really pays off.
Click to expand...

Hi Elton. Haven't tried that so far, but I have tried going direct from HDV m2t files on the timeline to SD DVD format MPEG2 files, which does not involve any intermediate codec. The results are fine, but I don't consider them to be that special compared to what I get from the XL2.

The A1 is a great camera package, but its strength is in HD, as it should be. The SD features are very useful for backwards compatibility while we are all waiting for delivery formats to make the transition to HD, but I think it is not accurate for anyone to say that they are an improvement on what the best of the existing SD cameras can do.

Richard
 
I disagree. My results may differ from yours, but the simple principle of oversampling would seem to confirm what my eyes tell me.
 
Elton said:
I disagree. My results may differ from yours, but the simple principle of oversampling would seem to confirm what my eyes tell me.
Click to expand...

Hi Elton. I think in this case the principle is not really so simple. The sensors are quite different so it is not just a case of having more pixels and then downsampling. The quality of data captured by each pixel is also important. For example, my eyes tell me that the A1 is noisier than the XL2 when used in similar lighting situations, and I've seen several posts that indicate others see the same thing (even excluding the ones that are due to AGC being left switched on inadvertently).

Anyway I am not trying to put down the A1, and I believe there are pros and cons in using it for SD. But I also think it should be kept in perspective because it is by no means perfect (which is not a complaint, just a fact).

Richard
 
I've argued this before but I will have to agree with Barlow on this one. I am basing my opinion on examples to my eyes, not theory.
 
jaegersing said:
Hi Elton. I think in this case the principle is not really so simple. The sensors are quite different so it is not just a case of having more pixels and then downsampling. The quality of data captured by each pixel is also important. For example, my eyes tell me that the A1 is noisier than the XL2 when used in similar lighting situations, and I've seen several posts that indicate others see the same thing (even excluding the ones that are due to AGC being left switched on inadvertently).

Anyway I am not trying to put down the A1, and I believe there are pros and cons in using it for SD. But I also think it should be kept in perspective because it is by no means perfect (which is not a complaint, just a fact).

Richard
Click to expand...

Hi Richard.

You make some great points, and I'm not really sure if the A1 is in fact cleaner in SD DV mode than the XL2; by physics alone it shouldn't be...but here's the thing:

When you do a good downsample from HD to SD, noise becomes more finite. From my experiences, the pattern of noise *usually* seems finer and less apparent than the HD version, and hence, looks cleaner.

That's my story and I'm stickin' to it! :)
 
Elton:

Do you downconvert after or before editing? I am using Vegas which is an 8 bit program and does not provide the option for a 4:4:4 colorspace look, where as Primier Pro 2.0 is 32 bit floating decimal. I have both but recently fell in love with the workflow and features of Vegas for HDV.

Thanks,

S Cannon
 
Almost always downcovert after editing. I usually edit HDV, output a Sheer codec HD QT master file, (to avoid rebuilding GOP's) and downconvert via MPEG Streamclip to a better SD codec.

From there I'll make a deliverable DVD.

Unfortunately, this is all on the Mac with FCP. With Vegas, I would think you'd have the ability to export an HDV timeline to any number of SD codecs (QT has more codec options), some of which as I've stated, can contain a higher color sampling in SD.

I've heard Vegas' downconverts are excellent so I would imagine that you would retain some extra color info by converting to a better SD codec.
 
Elton said:
Almost always downcovert after editing. I usually edit HDV, output a Sheer codec HD QT master file, (to avoid rebuilding GOP's) and downconvert via MPEG Streamclip to a better SD codec.

From there I'll make a deliverable DVD.

Unfortunately, this is all on the Mac with FCP. With Vegas, I would think you'd have the ability to export an HDV timeline to any number of SD codecs (QT has more codec options), some of which as I've stated, can contain a higher color sampling in SD.

I've heard Vegas' downconverts are excellent so I would imagine that you would retain some extra color info by converting to a better SD codec.
Click to expand...

Elton:

Thanks for the response.


S Cannon
 
[SIZE=+0]Hi Elton,[/SIZE]
[SIZE=+0][/SIZE]
[SIZE=+0]Thank you for all your input. Can you post some short examples and screen grabs of SD vs HD down converted to SD. Once the videos are played from the DVD is there still a noticeable difference? How would the Cannon compare to the HVX in DVCPro 50, once again compared when viewed from the DVD. I once heard that HDV when down converted is as good or better than Digibata. Is this true once it is played from the DVD? [/SIZE]
 
This is something that still concerns me. The HDV downconversions I've seen do seem softer and "shimmery" (good and accurate description there, Jaegersing) then SD originated material shot on SD only cams. The Canon's seem more shimmery then the Sony stuff. Someone said that the whole -3 gain set up lets more light hit the ccds and the highlight handling is more difficult for these cameras to handle. I don't know if there is any truth to this. Has anybody else heard this?

Jaegersing, Elton, could you guys please post some clips and/or grabs to illustrate your views.

Jaegersing, have you tried After Effects. I've heard some say it's the best for downconversion.

Thank you.
 
-3 dB has nothing to do with how much light hits the CCD's, but rather how much the signal is amplified.
 
You must log in or register to reply here.
Back
Top